Dell Technologies Aktie

Good morning, everyone. Asiya Merchant here. Citi -- I lead Citi's tech hardware and tech supply chain, day 3 of Citi's TMT Conference. Absolutely delighted to have David Kennedy here. He is the CFO of Dell Technologies here, and we also have Paul in the audience from IR. So I'm going to start off with just some safe harbor statement here that Dell would like me to read, so please bear with me here.

This presentation contains forward-looking statements based on Dell Technologies' current expectations. These statements involve risks, uncertainties that could cause actual results to differ materially. Factors that could cause results to differ are discussed in Dell's Technologies' periodic reports on Forms 10-K or 10-Q filed with the SEC. Any forward-looking statements made today are based on assumptions as of today, and Dell Technologies undertakes no obligation to update them.

With that, first question is going to be about outlook.

All right. Good morning.

So clearly, just results that are just staggeringly amazing, right? I mean you guys just reported last quarter -- last week, I should say, and your current year fiscal '27 guide is up 70% year-on-year tracking. Of course, we are at an investor conference with a lot of investors being very skeptical, like is there a cliff ahead after these like growth, which are so much higher than, let's say, your analyst event, right, and where you guys thought about. So just given there is pipeline, you have a very solid pipeline, you have a very solid backlog, there's supply constraints at the same time. Like what gives you the confidence that this momentum continues not just through this year, but as you kind of think about the outer years? And there's not a big demand cliff that we should be aware of?

Yes, sure. Good morning, everybody. You're right. Look, what we see is this accelerating appetite for infrastructure. And I think we've seen these building blocks emerge over the last 9 months, 6 months, 3 months and it builds on each other. So obviously, the tailwind of -- from an AI server GPU perspective, we booked $61 billion of orders in Q2. We had booked $61 billion of orders in the previous 3 quarters, so $132 billion over 12 months. So you see this accelerating appetite. It's also pipelines being multiples of that backlog.

So while the bookings keep coming in, the backlog continues to get bigger, and it's across all the segmentation you can think of neoclouds, sovereigns, enterprises. Enterprise is growing the fastest from a rate perspective. Obviously, the dollars are still pretty big on the Neocloud side. You then put another layer on top, which is from an inferencing workload perspective. So your traditional data center, that had a modernization refresh cycle happening anyway. You now put inferencing and AI workloads on top of that. And an emerging theme we've seen in the last 3 months around security and resiliency. So these -- all these components are driving -- if you look at our Q2 results, we had 122% growth in traditional server. Demand was faster than the P&L. And it's broad.

So if I look at that by geography, every geography growing triple digit. If I look at it by customer segment, if I look at it by customer vertical, all these boats are rising. It's not one corner or one niche piece that's kind of floating it. So you build that. And now you introduce in our Q2 results, the need as you use more tokens, do more inferencing, need more compute power for the data, you now need somewhere to store it. So our Storage growth of 26%, and we've guided to double-digit growth for the full year. We're watching all these signals, and we're watching the customer count still be fairly early in the adoption curve and look at the pipelines. And now all of a sudden, we see that confidence that allowed us to guide what we've guided and allow us, I think, to think about beyond into next year or 2.

Yes. And then you marry that with CPU or I should say, constraints, not just in the CPU side, but obviously, we've all heard about memory constraints, HDD constraints. Is there any line of sight to maybe when these constraints sort of ease or normalize? And so -- and could there be more upside here even in the back half of calendar '26?

Yes, we'll go piece by piece. So if I look at the back half of this year, I would -- in one word, I would call our guide a realistic guide. I think I would have positioned our second half guide 90 days ago to be a prudent guide. I think we see way more identification of parts to customer needs of data center readiness. So all of that picture is guiding to the $192 billion. Demand is still faster than supply, though, for traditional server and AI server. So those two buckets continue to have demand that was really robust.

As we look into next year, we see the supply differential only getting worse, not getting better. We see -- yes, there is a small bit of supply growth, but it's pretty minimal on the basis of the demand signals that we're seeing. So it's actually tighter next year than it is this year in relation to that as we look out. And then we'll see as we go beyond from that perspective.

Okay. And within that supply constraints, is it just memory, CPUs, HDD the usual? Are there any other pockets that are getting worse?

Obviously, dominated DRAM and NAND is by far the big brother in that equation. Given the growth in CPUs, obviously, that now comes into the equation also. But look, ultimately, any part associated within an AI box is short to some degree. So optical, transceivers, T-glass, any component that's in there is constrained to some degree. And that's -- our job then is to try and figure out how we can put all these real golden parts, if you like, together into fully configured units to maximize the P&L. And that's part of the operational execution that we've talked about in our guide and to differentiate versus the competition.

Great. And then Jeff, on the call did talk about this earning -- greater longer-term visibility with large sophisticated customers. Is that a function of these supply constraints that you're talking about and Dell kind of executing at its best where -- like you just mentioned?

Yes. I think the conversation in the market is around supply access. So if you go back to -- let's go back 6-months ago, was there a few savvy customers who foresaw some of this and did some advanced purchasing to kind of get ahead of that game? Yes, there was.

But what we're now seeing is, if I pick an enterprise customer, they come forward I know they're buying history. If all of a sudden, they're looking to buy 4-years worth of equipment from me, I'm not in a position to fulfill that. Our job is to try and make sure we diversify it to as many customers as we can and solve as many customer needs as we can. But what we're doing now is entering into conversations about what are your supply needs and your infrastructure needs going out into year 2, year 3 and year 4. And our commitment that for those who want to partner with us, we'll give them that supply access and provide some guarantees that we can position that for them as we look out. Guarantee supply does not guarantee price. That's a specific differentiation. But that's what the market is looking for. It's that supply access.

That's fantastic. ISG margins, like eye-popping 15%. And this is in the face of AI servers, which within that segment carry lower margins, well above 3x almost relative to what you were looking at year-on-year in first half. Pricing, probably a big driver here of those margins, just given the pricing pass-through of some of these components. But what's changed here? And why do you think that could sustain?

Yes. So if you think of ISG rate, there's two tailwinds and one headwind, right? So by far, the biggest tailwind is scale. I might touch on it later. But when you think of our OpEx as a percentage of revenue for the company, we've just guided to 8% of revenue. That's the most efficient we've been ever in the 42-year history of the company. And to give some kind of marker, that was 20% of revenue 6 years ago. So that's one piece.

Part two is the Storage portfolio. So as we've launched and drive our Dell IP storage portfolio, we've now had 6 consecutive quarters of demand growth, almost 2x the market. Obviously, a lot of software and service components to that. So the margin rate enhancement from our storage portfolio will lift ISG profit. And then to your point, we're using those two levers to be able to absorb the demand in AI. So almost 3x growth to $74 billion for the year and mid-single-digit margins, we can absorb that with the two other tailwinds and give the outcome that you see, which is record outcomes.

Okay. And within that, you also have servers, right, traditional servers, I should say. Just because these are built on next-generation infrastructure, next-generation core processors, are you thinking about higher margins even for these products?

No, I think from our portfolio, if you think of commercial PCs and traditional servers, it's about more a sustained margin rate. Look, we're trying to make sure we can give as many customers access to the compute power that's there. Our margin lever is storage. That's our margin lever and then to lean on the scale of the company as we do that.

I think that's pretty impressive, too, just given the memory inflation that we have seen that you're able to sustain those margins in your core compute business. Okay. AI backlog, I think you guys talked about $95 billion, but you do have a growing presence of enterprise customers in there, almost greater than 6,500 now. You have talked about both sovereigns and enterprises growing faster. Obviously, the Neocloud dollar is much bigger. But how does -- as you start to kind of drive the -- sorry, the sovereigns and enterprises to revenues? What is the attach rate to those AI? So could that AI pipeline and as it becomes revenue have higher margins associated with them?

Yes. I guess there's a few things in that. I think our value proposition in terms of the wheelhouse will be that on-prem enterprise customer base demand, which is going beyond the GPU. So traditional server CPU-based. So those tokens generating compute power, which generates the data, which generates more tokens, would you then need storage. So that inflection point, in particular, of the storage growth, that enterprise lift that we see, which we still think is very early in its adoption curve right now. I mean you mentioned the 6,500 customers. That's up over 60% in the last 6 months. But it's also -- you could argue the infancy of -- we serve tens of thousands of customers, right?

We would argue likely a low percentage pick a number between 10% and 15% of enterprises have only done meaningful AI, agentic modernization within their company. The rest have either haven't started or on the early adoption and kind of proof-of-concept kind of stage. So that element of an end-to-end portfolio from an edge PC up to your storage device, that's the wheelhouse. Are there attached motion tied to Neocouds also from a storage and a CPU server perspective? Absolutely, and we'll look to play there. I think we'll see a better conversion rate for sure on the enterprise.

And do you think that the storage opportunity could be as big as the AI opportunity, AI compute opportunity for those enterprises?

Yes. Look, I think you're not going to see in terms of the same dollar value, but the opportunity, if you look at our guide for this year, we're adding $2.5 billion of Storage revenue to our P&L, double-digit growth. So that's something we -- you haven't seen us do. This will be a record Storage quarter for the company's history. I think the more and more you see these inferencing workloads and enterprises adopting these use cases, I think we feel it's pretty durable and robust that it kind of lives on as we kind of go multiyear.

Okay. And when we talk about that enterprise adoption, there's always been this debate like what is it that enterprises -- what is it from a TCO perspective that's getting these enterprises to do it on-prem. Clearly, you have -- they have access to the cloud. So what is it -- like what is -- how is Dell quantifying those economics to these 6,500-plus enterprise customers that's getting them to deploy this stuff on-prem?

Yes. Look, the best example I can give is to use our own company as customer zero, if you like, right? So one of the big "ahas" we get from enterprises when they come into our executive briefing center and they want to have a conversation about AI, it's not as dumbed down as this, but it's almost, hey, help us turn it on and the world is better, if you like, right?

Whereas the reality is there's quarters and years of work to do before you can supercharge the AI in the first place. So we spent the last 3 years in our company standardizing and automating and simplifying all the different processes and workflows and cleaning up the data in the company. So -- and many enterprises are the same. We're 42 years old, through nobody's fault, we've had acquisitions that we've done. We've empowered different teams in different geographies to create their own ERP systems. And -- but ultimately, what that does is the data doesn't talk to each other within an enterprise. And so if I apply AI to that, it can't access corners of the company to kind of fully supercharge an end-to-end workflow. So you actually spend and it's enduring work. It takes multiyear, and it takes a cultural shift in the company.

We were very lucky in, a founder like Michael, who stood up in front of the leadership team 3 years ago and pitched this concept of modernization. In other words, if we don't do it, somebody else will do it to us. And so you see that. And where you now see the inflection point is every enterprise customer needs to identify where their value comes from. So for us, it's our supply chain, our products, our services team and our sales team. So within that, AI can be used as a power of productivity, whether that's growth, whether that's product road maps, pick your idea. Every other function should be using AI from a scale and efficiency perspective then, including my own function in finance.

And ultimately, then you find both the decoupling effect of I can grow faster, I can scale my business through OpEx spend, and they're not linked anymore. So when we talk about our 8% of revenue, that is now structurally based, and it's only going to get better. It's not going to go back to -- well, if I need to grow, I need to add some more variable cost in. We've divorced those two things. They will stay permanently divorced. And I think enterprises that we showcase and we bring them in, they're seeing some of that. But I think they'll all be on different curves. Some will go faster, some will still resist. So I think it will take time for the broader enterprise to get there.

Okay. And when we think about workloads, like the ones that reside on-prem, private clouds versus those companies are probably still using some sort of public cloud maybe for certain workloads. Just as you think about that, any kind of -- how you think about maybe the split between that? Like do you have any kind of forecast that you think about across the board?

Yes. I mean if you listen to the Amazon earnings call from a few weeks ago, they talked about 85% of the data is still on-prem, if you look at that perspective. Look, ultimately, it's going to be about the cheapest token you can have is the one that's closest to the data. So for a big -- take a finance function like is doing linear planning, decision-making and processing. You do not need a frontier model to do that. You can work off N minus 1, N minus 2 models. You can do anything from power of AI up to the lowest form of AI. You can do all that on-prem on your edge device, in your own data center on your powerful PC.

Obviously, there are also proprietary information that drive value for your company that you want to keep in a sandbox and keep away from everybody else. And then there are legitimate workloads you need to send to the cloud in a hybrid world to get supercharged compute power from a large language model perspective and look at that. Finding that right equation and balance will find you the right cost structure and efficiency. And that's why we think there's a strong value prop for the on-prem piece of it. And again, because we sell across the portfolio, it leads into our storage portfolio, in particular, which is where the value creation can be for us.

Okay. Wonderful. We have seen a transition from full rack solutions. We started this AI journey. It was all about just GPU chips. And now we're starting to see full rack solutions come out from some of the largest chip makers. How do you think about Dell's addressable market opportunity in this space? Like as we're getting more and more complex. And in some cases, there's disaggregated architectures as well.

Yes. I would say complexity is our friend. We're pretty agnostic in terms of if a customer in any segment is looking for something they need help with to build, we're available to build it with them and have that conversation. Obviously, we have a deep and fantastic relationship with NVIDIA today. The majority of our business is with NVIDIA today. But ultimately, when I say complexity is our friend, we bring and we've defined the value we're bringing to the conversation. So this idea of time-to-token for any multibillion-dollar deployment, we're on average, 4, 5, 6 weeks advantaged in that differentiation to the customer.

So that time-to-token, the economics benefit they can get from that because it's monetizing is worth all the conversations. So when we secure the inventory to build it, to ship it, to install it, to deploy it, when we turn it on, it stays on 99.9% of the time. We use far less parts than everybody else in terms of that uplift. And we white glove that service. So you get an engineering team and a pod that helps the design of a Neocloud design T minus 3 months, goes through 30, 40, 50 iterations. They are also the ones who will turn it on and make sure it stays on and almost live on-site there as a stint to with us in Dell. And so you have that end-to-end kind of relationship.

The other piece is because we sell across the portfolio, we can be more fungible with the supply that we've got. So one tangible example, we would have taken a decision in the springtime for seeing the PC market, particularly in consumer, was going to get squeezed a little bit. It's projected to be minus 18% to minus 20% on units. So we purposely sent some of our wafer supply for consumer PCs and reconfigured it for infrastructure for the second half. Now you can't do that quickly. It takes 5, 6, 7 months to do that. But we made some of those bets then, which allows us then to enhance some of our guide for the second half of the year. So that fungibility is a good asset for us, too.

Okay. And it's often lost like investors will often ask me, but -- and this moat that you've talked about, this time to deployment, which is critical in terms of monetization, often lost on investors because often they'll say, well, why can't this business just easily flow through to ODMs? And why is it -- why is Dell getting more than their fair share? So I think that's a very valid point when you're talking about it. Do you think that, that moat that you just talked about time to deployment helps with the underlying AI profitability as you think about the AI compute profitability as you think further down?

Yes. Look, I think we expect that to stay fairly stable from a rate perspective. We're in that mid-single-digit range. Obviously, it's a competitive market. So all those players you just referenced, they turn up and bid just like we do in all of these big deployments. We feel that's the right sweet spot because obviously, in a multibillion-dollar deal, mid-single digits, it's still a lot of margin dollars for them. But that's the value point we found for sure.

Okay. And then just shifting to the traditional servers. Obviously, very strong growth here. The growth is frankly, greater, I mean, on a year-on-year basis relative to the GPU side of the things. So how big could this traditional server opportunity be, especially when you think about many years that you've seen this business and it does go through a boom and a bust cycle?

Yes. And that's where -- I kind of referenced it at the start, but I think we're seeing a new TAM emerge and what it looks like. So I think it's more durable and more broad-based than just the cycle of the traditional data center that we've always seen. That's the foundation there for sure. But inferencing and then security are the next two where I think, like you said, we feel enterprises are only at the start of their enterprise journey from an AI perspective. And I think that gives us strong confidence that it's a new TAM that's a multiyear TAM that kind of develops out.

Right. And could you double-click a little bit more on it? Because clearly, this year, given supply constraints, given the pricing of memory that you've had to pass through as well, it's more pricing versus units. And then as you kind of think further out, what's the algorithm for growth there between units and pricing?

Yes. I would introduce a piece in the middle also, which is the density and core growth within the unit. As an example, there are over -- a significant number of the installed base are 13G, 14G. Ultimately, I can replace 5 or 6 of those with one of these, the new ones. As we go to 18G, that density even gets even higher. So the unit count is important, but it's only one small piece of the equation. I think because of the amount of memory you need, the amount of core growth that's in there, I think the ASP forever will stay much higher from a traditional server perspective, stands the traditional inflation or deflation cycle that could be there also.

Okay. All right. Let's talk a little bit about Storage, right? You already referenced it. The growth there was phenomenal. You've been outperforming the market. And again, pricing here has obviously been a driver because all the OEMs have raised prices. As you see these deployments mature and enterprise adoption matures a little bit and mix shifts also towards Dell's own IP and away from third-party IP, how should we think about the longevity here of Storage revenue growth? And more importantly, then on margins as well because you're still benefiting from the shift towards your own IP.

Yes, for sure. Look, as we embarked on our Dell IP Storage portfolio, go back, let's say, 3 years ago, that started with our mid-range PowerStore product. Q2 was the 10th consecutive quarter of double-digit growth that we've seen there. But as we've built out the portfolio, more in mind for AI workloads, so PowerScale now 5 quarters in a row of growth. ObjectScale, 4 quarters in a row of growth, Unstructured is obviously growing also double digit. All of these elements are allowing us at higher ASPs because it can command a higher price with software and service in there, too. So we get the ASP increase, but also the margin thresholds are significantly different, obviously, given it's Dell IP.

So you build those elements together, link it to that CPU growth we talked about earlier, and we think from a product perspective and then the underlying demand that's coming, I think we're positioned really strongly as we head into next year for sure.

Okay. And then you did introduce Project Lightning as well. I guess it's still early in the market, it's probably not reflecting...

Yes, its still a very small piece. I mean it's out there. It's been tested by a lot of our biggest customers, but more to come there. It's not a material element in terms of the numbers today.

Right. And this is more towards the Neocloud versus your traditional enterprise customers?

Yes. And then you get the broader base of the products I just referenced will give us complete coverage across the customer segment stack.

Okay. On the flip side, AI tends to be fairly lumpy. I think we've talked about -- you guys have talked about it before, where there is customer readiness really is mainly driving when you recognize revenue. But when you think about all the various constraints that you already talked about from a working capital perspective, how do you kind of manage that across? Is it just the fact that you just have so many customers now that it's smoother? How are you thinking about that AI data set readiness?

We have built out good muscle. So when we talk about our backlog at $95 billion or when we take orders in $61 billion in Q2, that's gone through a pretty rigorous setting process on two avenues. One, we've completed all the durability tests around credit assessments, financing, making sure the cash is good, et cetera, et cetera, and also on the design piece. So we will have worked particularly on the bigger Neocloud designs, like I said earlier, 20 to 50 iterations of that. So it's a solid customer, credit is aligned, financing is aligned. And we have, I'll call it, an 80-20, 90-10 alignment of data center readiness in terms of timing. So we have approximate ship dates we're going to do that.

Once we take the order and it comes in, obviously, you might have one or two that might slip out. We can easily -- that's the value of our supply chain. We'll move a few things around. But the majority, when we talk about our backlog, it's solid, it's locked in. It's noncancelable. It will endure as kind of part of that. So all those steps have kind of gone through the process.

Okay. All right. A little bit on commercial PCs here, then I can open it up. You are gaining share. I think you've identified that as one of your areas. You obviously have a leading position in commercial PCs, but there were other pockets that you were strategically interested in gaining share as well there. There are supply constraints at the same time. So where are you seeing this greatest traction that you have? And again, the margins on the PC segment were pretty strong.

Yes. Look, ultimately, the market is compressed for the rest of this year, like I said, minus double digit. On the flip side, there are still 400 million-plus PCs over 4 years old. And ultimately, if you go back to the conversation about the cheapest token being closest to the data, none of those installed base PCs will have the battery life or the power to do AI workloads on them. So there is a pent-up demand that we believe will come once it stabilizes and once IT budgets get a little bit more settled and expanded.

The reality is most enterprises this year for their IT budgets are sweating those assets out a little bit more and what dollars they have, they're trying to build out their infrastructure side as they do that. As we go forward, I think you'll see it a little bit more robust. Our main focus will be on those higher ASP commercial boxes as part of the AI storyboard that's there. But obviously, like you said, we have the portfolio across the platform, but very mindful that you're not going to see much elasticity on the consumer side, et cetera, as we go through the next 2, 3 quarters.

Okay. Fair enough. Let me just turn it to the audience. Any quick questions? Please raise your hand so we can bring the mic to you. I see one hand there.

Thanks very much for the discussion here. Yesterday, S&P published report said that they're allocating $21 billion of your debt to Dell Financial Services, up from $10 billion a year ago. Can you just talk a little bit about customer financing, kind of what the opportunity is, just given the pipeline and the backlog in the AI and data center customer base? And would you be willing to broaden your customer financing beyond your kind of traditional into residual value support backstops and the other types of things that we've seen in the market very recently.

Yes, sure. A few things in there. One, just to correct the numbers, it's $14 billion to $20 billion, not $10 billion to $21 billion. So hopefully, it will kind of correct itself in the storyboards. It's roughly an indication of the growth that we've seen, particularly on AI. Look, what we will do, we're not a traditional big bank. So in most cases, the rates I can offer would not be competitive versus the rates they can get in the open market, whether it's the banking or private sector.

We will continue to offer and be selective on certain accounts. It will be short term in nature, though. So we will not -- you will not see us doing long-term Neocloud financing. I have no desire to use Dell's balance sheet to do that. So you're going to keep a fairly prudent approach in relation to that. And we're very active with those big banks and those conversations around those long-term financing options. You've seen it all in the press the last few weeks. We're helping position and part of those conversations so we can help those folks kind of be part of that dialogue.

All right. Anybody else here in the audience?

Okay. Let's talk a little bit about OpEx. You already mentioned that how you divorce that from the revenue growth rate here. Could that operating leverage when you think about it maybe next year and into the following year, could that be even a bigger driver of EPS versus just the top line?

Yes. I think it's probably one of our most misunderstood elements of the P&L. We have -- I'm pleased with the tools and the toolbox I have from an EPS perspective. It's anchored in revenue, obviously. Storage margin rate enhancement gives us margin discussion. Then you look at OpEx scale, like this year, our P&L will find about 4.5 points of scale. It's going to continue to do that. So it's absolutely a viable lever from an EPS perspective. And then our fourth, obviously, from a cash -- as we've developed more cash, given our negative cash conversion cycle, the more you grow, the more you grow profitably, the more cash we'll create, and we're going to hand that back to our shareholders.

So we're pretty pleased from Q1 to Q2, while we saw our amazing share price value increase, we also in unison, increased our share buyback program from $1.6 billion to $3.9 billion in 90 days. That's an indication, hopefully, the commitment that we can see as the boats rise, we'll continue to reward the shareholder base.

Okay. And then just given there's so much growth and there's so much need for all this capital. I mean, obviously, you talked about share buybacks, you talked about dividends. Just when you think about your product portfolio, are there areas where you feel like there is opportunity to maybe double down, beef-up the product offerings that you have?

I think we've if you ask me that question 3, 4 years ago, I would have talked about Storage in particular and also the ecosystem around AI server. So as we continue to see chip generation, we're building out more and more engineering labs to be ready for those engineering hurdles and the next wave of products, and we'll continue to build out that storage portfolio. That's probably the sweet spot of where you're going to see our R&D investment as we kind of move forward.

It's still pretty organic.

Yes, correct.

Okay. All right. And then just as we wrap it up here, if you compare today's environment, David, relative to where we first saw you at your analyst event, right? It wasn't too long ago and where you guys laid out some targets and clearly, you're surpassing that significantly. What do you think is part of the story that's most underestimated by investors? Is it your AI opportunity? Is it the storage attach? Is it the exposure to sovereign? And just maybe help investors understand...

Yes. I think as we wrap it, there's probably two elements that are fundamentally different. One is that scale and leverage conversation we just had. The other piece is, okay, AI GPU server has exploded, and we've seen that revenue. But that traditional server CPU inferencing TAM I think that is new 12 months on versus where we were previously. And I think that's where you're seeing the duality of that revenue growth kind of really pop up. And we're positioned -- we sell across the portfolio. So I think as you see that circular conversation of tokens to data, to compute, to storage drives more tokens to data, to compute, to storage. As that gets bigger and bigger, we're well positioned given the only one who provides end-to-end those products. We're excited about the next few years ahead for sure.

Awesome. Well, I'd like to thank Dell and David, thank you very much and to Paul as well for coming to our tech conference and looking forward to more stellar results.

Thank you, team. Thank you.

Hi, everybody, and welcome to the Dell Technologies keynote at the Goldman Sachs Communacopia and Technology Conference. I have the privilege of hosting Michael Dell, CEO and Founder of Dell Technologies. My name is Kat Murphy. I cover IT, hardware and Dell here at Goldman Sachs. We have about 35 minutes for today's discussion. I'll kick it off with the Safe Harbor, and then we can dive right into it.

This presentation contains forward-looking statements based on Dell Technologies' current expectations. These statements involve risks and uncertainties that could cause actual results to differ materially. Factors that could cause results to differ are discussed in Dell Technologies' periodic reports on Forms 10-K and 10-Q filed with the SEC. Any forward-looking statements made today are based on assumptions as of today, and Dell Technologies undertakes no obligation to update them. So with that, Michael, thank you very much for joining us here today.

Thanks for having me. Great to be here.

To kick it off, Dell has been at the forefront of multiple technology eras with leading portfolios in PCs, servers, enterprise storage and most recently, in AI rack scale infrastructure. You've talked in the past about how the data center, in particular, is moving from being thought of as a cost center to a center for value creation. I was wondering to start, if you can talk about where you think we are in the role of infrastructure for an enterprise. And are we at the beginning of a new enterprise virtualization cycle or enterprise infrastructure cycle? And how might that compare to prior transitions?

I think we're in this sort of gradual process where more and more organizations are figuring out that AI is incredibly powerful in terms of reimagining how they achieve outcomes in their business. And it's a multispeed world. Not everybody is going to go at the same speed. But it is quite evident that this is very different than the technology of the past. We had sort of calculating and computing and now we have intelligence. And the data that exists inside companies is being unlocked and unleashed in a new way.

And they're sort of in this process of figuring out that they have to reimagine their companies. And we started this about 3 years ago. I would say maybe 10% or 15% of companies understand that this is a really important thing, and they have to go do it. A lot of them are very confused about what to do, how to do it. There's an easy first phase of this, which is you just give individual people better tools and they can be more productive in doing whatever they were doing. Maybe they could do it 10% better or 20% better.

That's kind of extraordinary in and of itself, but that's not really the plot. The real opportunity is reimagining the whole workflows. And what you have to do is map out every single process in the company and then simplify, standardize the process, get all the data together and then kind of the last step is you're applying AI. And also understanding that the AI today is pretty extraordinary, but it's also the worst it will ever be. It's improving.

And you can apply agentic technology to replace a lot of the workflows that used to be done by humans, and so it could be done way faster. So you're more innovative, you get speed. And I would say we're really at the very beginning of that in most companies. They don't know how to do it. It's hard. It's also the barrier to adopt technology, it's not technology. It's people, it's cultures, it's regulation, it's courage, and it will take time. But we are seeing with the 6,500 Dell AI Factories in enterprise, we're kind of in this S-curve adoption.

Maybe let's talk more about the 6,500 AI Factory customers. Can you talk about how Dell's AI Factory model helps customers bring AI closer to their own data while also getting the benefits of the broader AI ecosystem?

Right. So as customers go through this, they begin to understand that it's not one size fits all and you really want to have the AI as close to the data as possible. So if you're in advanced manufacturing, you want your AI in the factory, okay? If you've got a restaurant chain, you want your AI inside the restaurant, retail stores, logistics. And you've got this wide variety of small models, large models, open models, closed models. We're bringing the frontier models on-prem.

And for your most valuable data, your proprietary data, your hard-fought knowledge that only you have inside your company, you don't want to send that off to somebody else. And so we're seeing tons of this infrastructure being built around solving that problem. But again, it takes some time to organize all that, pick the right models. And this is what we help our customers do. We have hundreds of reference platforms, starter kits, blueprints and a variety of partners in all sorts of verticals.

And so we can help customers kind of get going here. But it takes time. There's an instinct to want to just say, okay, I press a button now I have AI, right? And it's way more complicated than that. It just takes time.

Talking more about this idea of moving on-prem as a function of protecting your hard-earned data. What are some other considerations that are driving unit economics and considerations to repatriate data, move workloads on-prem, a hybrid model as opposed to that cloud model?

Well, by various estimates, 85% of data is still on-prem, right? And so yes, the cloud is a real thing, no denying that. But it's really more of a hybrid multi-cloud. And if you think about it, most of the data in the world is going to be created in the future. And again, as we adopt AI and agentify, we create way more software, we're creating more data, and it's all happening in real time. And again, you want the AI as close to the data as possible.

So this is kind of what we're building out. And of course, with enterprise, we have more opportunity to architect the full solution where you need more CPUs because of agentic behavior, obviously, data storage where we have a leading position. Networking has to be re-architected. And the frontier models also want to come on-prem with us.

So how much does the ROI of these investments drive some of these decisions as well?

Yes. I would say this is part of also this change that's been happening inside companies where it used to be we were all talking about the IT budget, okay? What's the IT budget? Is it going up? Is it going down? Is it this? And so now companies sort of understand that this is how you become more efficient, how you become more innovative. If you had 200 people doing something, now you can have 50 people doing the same thing with $10 million of AI spend and it costs half as much. And it actually works better.

And so what's happening is the CEOs, the business line executives, they're all figuring out, we need AI to be more competitive, to be more innovative, to be faster. And so the conversation is sort of leaving the traditional budget discussion because this is actually how you create value inside companies.

On the point of IT budgets, there's a lot of questions today from investors around the durability of demand for hardware spend and maybe looking through Dell's very strong recent results and trying to understand how much of that might be attributed to pull forward or otherwise pricing increases parsing that out from the uptick in demand. Can you talk about, from your vantage point, what gives you confidence that this is a durable demand vector and not some of these more idiosyncratic issues that are happening right now in the market?

Yes. I think this goes back to sort of the fundamental root element here, which is the creation of intelligence and just the demand for that inside organizations. And we actually see demand being deferred in many cases. I mean, right now, we have, there are 400 million PCs that are over 4 years old, right? And because the cost of memory has gone up so much, you have people sweating assets. And so they're just deferring future demand that we know is going to come.

And the PC is now 5 years old. It's 6 years old. It's got rust around the edges. This is no good, right? If you're a knowledge worker, you need modern tools. And so those will all be replaced. And then in servers, we have 1.3 million servers that are 13th generation or older. Some customers have definitely not refreshed because they thought the price went up too much or wasn't going to keep going up or whatever.

What I see and I step back and think about this is the ROI that customers have and that the NeoClouds have and that exists around compute is pretty extraordinary, and that's what's driving the demand. And all of the improvements in the models, particularly from basic LLMs to reasoning to agents has occurred well within the time frame required to build a new semiconductor fab. And so you just have a structural shortage probably worse in '27 than in '26 from everything that we see.

And so yes, costs are going up. Yes, we price all that through. But customers sort of go through 5 stages of grief. It's like, okay, price went up. I don't really like that. You know what's worse than that is they don't have availability because they can't run their business, and that's a disaster. And so they tend to go to, okay, are you going to be able to supply us with what we need over the next year, 2 years. And so that's what we work on with our customers. But we watch closely for any signs of double ordering or things that are not being utilized, and we don't see that.

To talk more about the ways in which Dell's supply chain execution has helped in many instances, gain share in some products and navigate through a lot of the constraints that you've talked to? You've talked in the past about the nimbleness of your operations, the strength of your relationship with suppliers. Why is that a moat more now so than ever? And how should that compound over time?

I think the tighter the supply environment, we're probably going to do better. I mean that's been heuristic for a long time. We have incredible relationships with our supplier partners. We also have this broad portfolio. We can sort of move supply around to where the demand is. They also know that we are a reliable counterparty that will continue to require the output of their high CapEx investments through all the cycles.

And we have the broadest set of customers, and they want to know that there's going to be demand in 2032 and 2038. And the relationships go back many, many decades. And so that's incredibly important. And also, the volatility of our demand signal and the integrity of our demand signal is, we believe, the best in the industry because we're more directly working with customers of all types. And we see what the actual demand is from the end customers.

That's very helpful. Switching gears, I want to talk more about the AI server opportunity. You exited the quarter with $95 billion in AI server backlog even after converting $130 billion in orders in the last year. NeoClouds is a big part of that story still today. Investors are naturally comparing the AI server opportunity within NeoClouds to that of the x86 server cloud build-out of the last decade. We touched on why this is a different technology transition than the cloud. But can you help us better understand why this NeoCloud customer may be a more durable customer than the cloud customers of a decade ago and thinking about how sticky that type of opportunity might be?

The NeoClouds have done a very good job finding powered land. And ultimately, we look through them to understand where the real demand is coming from. And on the other side, it's Anthropic and it's OpenAI, it's Microsoft, it's Google, it's Meta. And so we want to see those contracts. And if they don't have one of those contracts, not interested. So we're really understanding what's going on and why this demand is occurring.

I think it's also maybe not completely appreciated what's going on with sovereign AI and demand outside the United States where there is power also and lots of customers. And so there are lots and lots of these companies. They're not all the same. And when this started many years ago, I think a lot of the sources of demand maybe thought that they didn't need them. And it turns out the demand has been way more than anybody thought. And so they do need them.

And I think a number of them have been super clever in how they've built their businesses. And we have given them an incredibly fast time to first token. And if we're delivering a massive cluster of these in a super reliable way, and we're 6 weeks ahead of a competitor, well, that 6 weeks is worth more to them than whatever premium that we're charging. And so it's been a great business. And certainly, the pipeline and the set of demand opportunities we see is quite strong there. But enterprise is growing faster, not as big. And you've got this sovereign, which is sort of a bit of a hybrid of the 2 and also has public sector in there.

Can you talk more about how the complexity and the rate of change in some of these AI platforms and chips is also creating a competitive advantage for Dell?

Yes. The complexity is our friend because it's complicated to install, deploy these things, make them operate reliably with incredibly high uptime, and orchestrating all the supply chain. It's not just the engineering and innovation, it's all the other aspects, including our kind of short-term financing that we bridge to all the sort of longer-term sources of capital that want to finance these projects.

So, and of course, many of them want access to the enterprise, which we can help them with. And so customers come to us, they're like, oh, I need 5,000 Vera Rubins. Well, you're not ready to build your own data center, but we have this partner over here. And so we help all that happen.

Can you also talk more about the sovereign opportunity and maybe the opportunity within the public sector as well that you talked about?

Right. So if you take the top 50 countries in the world by GDP, you can find 2 or 3 sovereign AI projects in pretty much everyone. And these are quite important. Nobody talks about sovereign AI in the United States because it's sort of like the non-U.S. AI thing. But in these countries, it's quite important. Sometimes they're government sponsored, sometimes it's a telco, sometimes it's just private industry.

But there are lots of reasons why these are occurring. And I would say, in some countries, the adoption has been a little slower than the U.S., but it's coming. They might be 12 or 18 months behind, but sizable economic activity, maybe half the opportunities outside the United States. So it's substantial.

Over the last 6 months, there has been much more of a focus across the industry around the role of general purpose compute, which is really core to Dell's offering. As enterprises continue to modernize, you talked about those 1.3 million servers that are 13th generation or older. How do you see the opportunity for continued momentum in traditional compute? The company is guiding for 100% growth in your traditional compute business for this fiscal year. But how do we think about the runway and durability of demand beyond fiscal '27?

Well, one reason why you have durability of demand is you have this massive installed base of old things that have to be replaced. And if you're disassociating your company, you're sort of preparing for all this onslaught of cyber threats and even beyond that, sort of the post-quantum encryption challenge. The stuff that's 4, 5, 6 years old, forget about it. You're going to be replacing a lot of that.

But you still have a lot of traditional server demand that has been there for a long time and keeps growing. And then, of course, you have the agentic behavior, which drives tons and tons of CPU usage. And we're just kind of at the beginning of that. And so we see tons of demand. Now in 5 years, are we going to be talking about traditional servers and AI servers? No, it's just all going to be servers, right?

So it's a fuzzy line that's going to get blurrier all the time. But ultimately, we are the scaled infrastructure player with incredible operating leverage across our scale in servers, storage and increasingly data center networking. And with the customer relationships, the supply chain, the services, the financing to be able to continue to lead in the sectors we compete. And we still make PCs.

We'll talk about PCs. But ultimately, AI is running on data. You've touched on this. And we've talked a lot about the compute side, and you've alluded to the role of data and storage as part of the data platform, but the enterprise storage TAM, we'll talk about, hasn't seen the same inflection as compute has. From your view, what do we need to see to build confidence that the storage opportunity is going to be a significant one for the enterprise? And is this more about simply storing more data? Is that what's happening here? Or do we need to be more thoughtful about how enterprises are moving and protecting and extracting data from these platforms?

So there's always been a lot of data and data has been at the center of everything sort of important and interesting. If you want to do AI, you need data, right? No data, no AI, bad data, bad AI. And actually, as these models get more powerful, you're creating more and more data. And so we have the broadest portfolio of enterprise data storage. Increasingly, these AI Factories have our PowerScale and ObjectScale storage platforms attached to them.

You've seen the sort of acceleration in the storage growth, right, with our pivot to Dell IP Storage, 26% growth in the last quarter. And again, we're largest there, larger than #2, #3, often #4 combined together. And yes, I mean, I think AI is making better AI. That's creating more data and then that's making better AI. And so it's sort of a self-improving process that is driving demand.

And again, most of the data will be in the real physical world. And by the way, we're talking about AI on the screen, right? We're not even talking about AI in the physical world, which is just starting to come with robotics and that will be another massive demand driver.

For enterprise storage?

For enterprise storage, for sure, multi-robot agent coordination in factories and transportation, everything generates data and requires data to make it more intelligent and more effective. And we have leading position, leading capability there.

I'll ask my PC question now. But we've talked a lot about the data center. Intelligence is also moving closer to the edge, mostly through endpoints like the personal computer. How does this change your outlook for the PC opportunity overall? Obviously, a very significant leading enterprise portfolio there, but how does this change your TAM outlook as well as Dell's strategy within the PC category, especially as you navigate these input cost increases that the broader industry is digesting?

So as you know, the device market has been challenged, right, because the cost of memory has gone up so much. So if you look at our PC business, last quarter, revenues grew 20% and units that didn't do so well, right? So average selling price is going up. So what's happening is you're actually growing the installed base of older machines. And so you're deferring this future demand into whatever, right?

It's either some breaking point where it's like my computer is too old now. You expect me to work at this company, you need to give me a computer that actually works, right? And if I want to use AI, I've got to have a modern computer. I need better tools or it just breaks, stops working. It's too old. So, and again, I think the most cost-effective token is going to be the one generated closest to the data. The small models are operating locally and it will continue to be a hybrid architecture. And the PC is still the productivity device that people use in the modern world to get things done.

Across the breadth of your portfolio, PCs at the endpoint edge, all the way into the AI Factory with your full stack solutions there, how has the breadth of your portfolio helped with the sales motion and bringing to market to your customers a variety of solutions that might help them in different operating environments, cost constraints, things like that?

Right. So some of you will remember almost exactly 10 years ago, we bought EMC, combined with EMC. And our presentation to our customers was we're #1 in everything all in one place. And it's worked out quite well, right? We've continued to gain relative share and scale. And so that's basically still the same story with customers, right? Let's say there's 10 different areas or 20 different areas. We are #1 in pretty much every single one.

So you can buy 20 things from us or you can buy 20 things from the #2 company, but it's 20 companies, right? And so nobody wants that. And so we continue to gain share really in all these categories, and we have this compounding advantage of scale, cost structure, the aggressive application of modernization and AI inside our own business to deliver great value to customers.

Does that create a stickiness in your customer relationships as you think about your enterprise buyers potentially rearchitecting or rethinking their broader infrastructure stack, the fact that you can be one vendor who provides all of those types of engagements and what is sometimes considered to be a commoditized hardware industry.

Yes, I believe it has. And because we have everything and you sort of wrap it with services and with Dell Financial Services as well, and you just make it easy for them. And so it doesn't mean we have a license to just do whatever we want or price it ridiculously high, right? We have to be competitive and deliver great value to customers. But that's absolutely what we do.

And we do what we say we're going to do. We do it reliably. Customers can count on us. And particularly now when there are these supply chain challenges, they know from prior periods, we deliver when we say we're going to deliver, right? And not everybody does that.

You talked about all the ways in which Dell is helping your customers implement AI, but you're also a large enterprise yourself. On earnings last week, the company talked about record low OpEx rate levels. You're being very efficient in driving operating leverage. What are the applications of AI that you're seeing the most successful internally within your organization? And how does that help in the sales motion to customers?

So we started this modernization a little over 3 years ago. And essentially, I stood up in front of our leadership team and said, we were going to have a new competitor 5 years from now, this competitor was going to be in every business that we're in, that they're going to be more innovative and faster and lower cost, and they're going to put us out of business. And the only way we're going to prevent that is we're going to become that company.

And so you look at our OpEx as a percent of sales, we're down to roughly 8%, pretty extraordinary operating leverage. We've totally decoupled revenue and cost. And how have we done that? Well, it is back to reimagining the outcomes and mapping out every process in the workflows, identifying everything that we possibly can, and we're still going. There's still a long way to go here.

And so yes, that's super helpful. Customers want to learn about that. They want to know how we've done that. And they need a lot of help doing it. It's not an easy thing to do. Not everybody has the appetite to do it. But it has put us in a great position to be able to, obviously, with an 8% OpEx, we can sell things at a given price and earn a very nice operating margin.

There are many companies here today at this conference overall that are participating in AI, but you have Dell combination of portfolio breadth, customer relationships and global reach importantly. When you think about these sorts of advantages over the course of the next 5, 10 years, what stands out to you as most differentiated for Dell?

I think the breadth of technology portfolio is incredibly difficult to replicate. Plus now we have this scale advantage and operating leverage that's just extraordinary. And we have an execution-oriented machine. I mean that's kind of what we do every day. The supply chain relationships are, they've always been important. They're more important in times like this, but it's always been a durable advantage for us.

And ultimately, we have a process and a set of systems that allow us to understand what the unsolved problems are for customers and then quickly go create the innovations with our own teams and with lots of partners to be able to go solve those problems. And it's an enormous overall market that we operate in, and we're well positioned to deliver value to our customers and ultimately, our shareholders through that. And obviously, we look at the financial model, operating leverage, negative cash conversion cycle, how we deal with capital allocation with our share repurchase and dividends. It all works.

We have a few more minutes here, but maybe just in closing, can you talk to what your vision is for the future of Dell across these 4 major product line items that you participate in and what you think investors should be paying most attention to in the years to come?

I really do see that we're at the beginning of an S-curve adoption in infrastructure related to this takeoff in intelligence capability driven by data. And that is just going to require enormous amounts of infrastructure. And it will be everywhere. Actually, it will be at the edge, it will be in the data centers. It will be the NeoClouds, it will be the hyperscalers. It will be all over.

And I think we're -- we've also opened up enormous amount of development capacity because of AI changing the way we develop products. And so the speed at which we're introducing new features and capability is massively accelerated. And so put all that together with what I said previously, I think a great set of opportunities. I think the operating leverage might be the part that is not completely understood by everyone because they're just not used to it. We're just kind of different.

Great. Well, Michael, thank you very much for being here and joining us today at the conference.

Thank you.

Good afternoon, and welcome to the Fiscal Year 2027 Second Quarter Financial Results Conference Call for Dell Technologies Inc. I'd like to inform all participants this call is being recorded at the request of Dell Technologies. This broadcast is the copyrighted property of Dell Technologies Inc. Any rebroadcast of this information in whole or part without the prior written permission of Dell Technologies is prohibited. [Operator Instructions]

I'd like to turn the call over to Paul Frantz, Head of Investor Relations. Mr. Frantz, you may begin.

Thanks, everyone, for joining us. With me today are Jeff Clarke, David Kennedy and Tyler Johnson. Our earnings materials are available on our IR website, and I encourage you to review these materials. Also, please take some time to review the presentation, which includes additional content to complement our discussion this afternoon.

During this call, unless otherwise indicated, all references to financial measures refer to non-GAAP financial measures including non-GAAP gross margin, operating expenses, operating income, net income, diluted earnings per share, free cash flow and adjusted free cash flow. A reconciliation of these measures to their most directly comparable GAAP measures can be found in our web deck and our press release. Growth percentages refer to year-over-year change unless otherwise specified. Statements made during this call that relate to future results and events are forward-looking statements based on current expectations. Actual results and events differ materially from those projected due to a number of risks and uncertainties, which are discussed in our web deck and our SEC filings. We assume no obligation to update our forward-looking statements.

Now I'll turn it over to Jeff.

Thanks, Paul, and thanks, everyone, for joining us. Another outstanding quarter. I'm proud of how our team executed across the business delivering record revenue and record earnings per share. Revenue was $47 billion, up 58% and earnings per share was $7.04, up 203%. These results reflect the pounding benefits of our competitive advantages, the breadth of our portfolio and the strength of our operating model. Our modernization efforts are driving greater efficiency and significant operating leverage enabling us to grow earnings faster than revenue.

Customers no longer see IT environment simply as cost centers, but as value drivers that enable growth, productivity and competitive advantage. As a result, they are expanding and reallocating budgets to support continued investment. This is creating opportunities across our portfolio from infrastructure to client devices, our world-class supply chain and ability to serve customers across their IT environments are helping us meet more of their needs and gain share. Our deployment and service capabilities are helping customers integrate solutions across their IT environment and capture more value quickly. The proof is in our results.

Over the past 12 months, we have booked more than $130 billion in AI server orders. In just the past 2 quarters, we have generated almost as much revenue from traditional servers and networking as we have in any prior full year in company history. Storage returned to growth and share gain with strong demand for Dell IP storage products and CSG revenue is growing at the fastest rate in 5 years. It is clear why demand for our solutions is succeeding available supply, our results and guidance demonstrate the strength of our position as customers enter a new era of infrastructure modernization.

Customers are modernizing their data centers for both AI and non-AI workloads and the benefits are meaningful. AI is an important catalyst, but the opportunity extends well beyond AI optimized infrastructure. AI requires modern disaggregated architectures that keep data accessible and in motion across compute, storage and networking. It is also accelerating investment across traditional IT environments as customers see greater performance, efficiency and resiliency. Our AI server momentum continues to accelerate. We booked $60.9 billion of AI orders in this quarter than most in our history.

We are also seeing AI-related tailwinds in traditional servers and networking along with early signs of increased storage demand as customers prepare, manage and protect growing volumes of data. Deployment methods are evolving as well. On-prem and edge infrastructure offers attractive token economics for the right workloads while giving customers greater control over their data and intellectual property. Our portfolio, global reach and deep customer relationships position us to help customers design, deploy their right solutions for their performance, cost and security requirements. Together, these trends are expanding our addressable market and driving demand across compute, networking, storage and PCs. This represents a significant long-term opportunity for us. It plays directly to our strengths and expands the value we can deliver across the entire IT environment.

Now on to the results. Starting with ISG. Revenue increased 89% to a record $31.8 billion with operating income of $4.8 billion and an operating income rate of 15%. In AI, demand continues to accelerate. In Q2, we booked a record $60.9 billion in AI orders and recognized $16.4 billion in AI server revenue. We exited the quarter with a record $95 billion of AI backlog, and our pipeline continue to grow sequentially and remains multiples of our backlog even after converting $131.7 billion into orders over the past 12 months. Demand is broadening across neoclouds, sovereigns and enterprise customers and our customer count has surpassed 6,500. The scale and complexity of these deployments reinforce why customers choose us.

AI infrastructure requires much more than assembling and delivering components. These opportunities demand significant engineering, design and deployment expertise with some engagements requiring upwards of 50 unique designs as customers optimize for workload performance, power, cooling and the data center environment. This complexity plays to our strength. Our engineering capabilities, broad portfolio, global supply chain and ability to deploy and support infrastructure at scale globally differentiate us, enable customers to move from design to production more quickly.

We demonstrated those capabilities again by becoming the first to ship rack systems engineered on the NVIDIA Vera Rubin platform. The AI market is evolving rapidly, and we are focused on expanding our platforms and capabilities solving increasingly complex customer challenges and innovating across the infrastructure stack with accelerating demand and a growing pipeline and differentiated capabilities, we are well positioned to capture the opportunity ahead.

Moving to traditional servers. Revenue was up 122% as demand remains exceptionally strong, supported by multiple vectors of growth. First, a majority of our growth is coming from existing customers as they continue to refresh and modernize their data centers to support traditional workloads. Heightened security and resiliency requirements are also creating incremental demand as customers modernize their infrastructure. Second, we are seeing a growing trend of customers that require meaningful CPU compute capacity to support AI and agentive workflows. These workloads are creating incremental demand for traditional servers. We are executing very well against both opportunities and gaining share.

Over the past 2 quarters, we have gained more than 10 points of traditional server share, and we expect to gain share again this quarter. With the majority of the installed base still on 14th generation or older servers, we see a significant and durable refresh opportunity ahead. The strength and breadth of demand, combined with our continued share gains demonstrate the competitiveness of our portfolio and the consistency of our execution.

Turning to storage. Revenue was up 26% as strong demand for our Dell IP portfolio translated into revenue growth and improved storage profitability. Dell IP delivered another record demand growth quarter making this our sixth consecutive quarter of demand growth above market. Demand remains broad-based, enterprises continue to modernize their storage environment as data growth increases the importance of keeping data available and secure. At the same time, we are beginning to see incremental demand from AI workloads, which require customers to prepare, manage and move increasingly large volumes of data.

We saw strong growth across PowerFlex, PowerStore, PowerProtect and PowerVault, with PowerStore posting double-digit demand growth for the ninth consecutive quarter. PowerScale and ObjectScale also drove another exceptional quarter in unstructured storage which has now grown at double digit or better for 3 consecutive quarters. Storage is becoming a more meaningful contributor to our growth and profitability. Dell IP continues to increase as a percentage of our storage mix and margins continued to improve, supporting overall ISG profitability. Our share gains expanding Dell IP mix and accelerating pace of product development give us confidence in the opportunity ahead.

Turning to CSG. Revenue grew 20% with demand growth across all regions and verticals. Commercial revenue grew 22%, our eighth consecutive quarter of growth with demand up for the tenth quarter. Large enterprise customers continue to refresh their PC installed base, driving double-digit growth across all regions. More cost-sensitive customers are extending their upgrade cycles. This is increasing the number of older devices in the installed base and expanding the long-term refresh opportunity for CSG. Consumer revenue was up 7%, the fourth consecutive quarter of demand growth. CSG profitability remained strong, benefiting from price discipline and greater scale.

In closing, we delivered record revenue and EPS with continued strong cash flow and record capital returned to shareholders. Our results reflect several reinforcing factors. First, infrastructure demand is growing structurally driven by data center modernization, AI adoption and attractive economics of deploying workloads on-prem. Second, our broad-based portfolio across AI infrastructure, traditional servers and networking, storage and PCs enable us to serve the full range of our customers' needs. And lastly, we delivered value at scale through our engineering and deployment expertise, supply chain scale and fast disciplined operating model. Our full year operating expense rate guidance of approximately 8% of revenue is the lowest in our company's 42-year history demonstrates the operating leverage this model can deliver.

These advantages reinforce one another. They are driving growth, share gains, profitability and cash generation. By creating more value for our customers, we compound our advantages and create durable cash flow and long-term value for our shareholders. I am proud of our team's performance. We entered the second half with strong momentum and confidence in our position.

With that, let me turn it over to David to walk through the financials and our outlook.

Thanks, Jeff. We delivered another record quarter, capping a very strong first half of the year. The team executed exceptionally well, driving record revenue, record EPS and record shareholder returns. Total revenue was up 58% to $47 billion. Gross margin dollars grew 78% to $9.9 billion. Gross margin rate was 21.1% driven by an improvement in ISG margin rate and a higher mix of ISG revenue. Operating expenses were up 22% to $4 billion, primarily from variable compensation tied to our outperformance.

Building on last quarter, we continue to drive significant scale in the P&L with OpEx down 250 basis points to 8.5% of revenue. Operating income grew 160% to $5.9 billion or 12.6% of revenue, driven by higher revenue, scale and price discipline across servers, storage and CSG. Net income was up 189% to $4.6 billion, primarily driven by strong operating income. Diluted EPS increased 203% to $7.04, a record.

Moving to ISG. ISG delivered record revenue of $31.8 billion, up 89%, marking the tenth consecutive quarter of double-digit or better revenue growth. AI server momentum accelerated, and we set records across the board, including $60.9 billion in orders, $16.4 billion in revenue, and $95 billion in ending backlog. Traditional server and networking revenue was $10.5 billion, up 122% as demand continued to outpace supply. Storage revenue was $4.9 billion, up 26%, with strong demand across the Dell IP portfolio, driving revenue growth and significant margin contribution. Dell IP storage demand has grown above market for 6 consecutive quarters. Unstructured storage remain one of our fastest-growing solutions with broader strength across the rest of the portfolio.

ISG operating income was a record $4.8 billion, up 225%, marking the ninth consecutive quarter of double-digit or better growth, primarily driven by higher revenue across the business. Operating margin was 15%, up 620 basis points. Looking at the key drivers of margin performance, a number of factors came together and went our way this quarter. The demand environment was strong, mix and rates were favorable and the team executed with discipline. While we would not expect every benefit to continue at this level, the quarter also reflects meaningful structural improvements in the business, which is reflected in our second half guidance.

Looking more closely at the drivers. First, we are realizing the benefits of our multiyear modernization journey. That work is driving greater efficiency and strong operating leverage, resulting in significant scale. Second, storage profitability was up with a higher mix of Dell IP and rate expansion across the solutions. And third, we maintained strong operational and price discipline in the dynamic environment reflecting our team's strong execution and continued focus on supporting our customers.

Turning to CSG. CSG revenue was up 20% to $15 billion. Commercial revenue grew for the eighth consecutive quarter, up 22% to $13.2 billion, and consumer revenue increased 7% to $1.8 billion. CSG operating income was $1.1 billion or 7.6% of revenue, driven by pricing discipline and the benefits of scale in the P&L. We will continue to balance customer demand with availability of supply to drive profitable share gain. CSG remains an integral part of the business. It provides scale across our supply chain and manufacturing, completes our end-to-end portfolio with the essential productivity device and is our most capitally efficient business. Together, these strengths make CSG a significant source of cash generation and helps fund growth across Dell and capital returns to our shareholders.

Moving to cash on the balance sheet. We delivered another strong cash quarter with cash flow from operations of $2.2 billion and adjusted free cash flow of $8.1 billion. This is primarily driven by sequential revenue growth and higher profitability. We returned an all-time record $4.3 billion to shareholders this quarter, including repurchasing 9.5 million shares at an average price of $401 per share and paying a dividend of approximately $0.63 per share. This acceleration in shareholder return, up $2.2 billion quarter-on-quarter reflects our agility and commitment to capital deployment as we generate more significant adjusted free cash flow as well as our confidence in our long-term value creation. We ended the quarter with $14.2 billion in cash and investments of $0.2 billion sequentially and our core leverage ratio is at 0.8x.

Overall, our strong cash generation and healthy balance sheet, further validated by positive credit rating actions during the quarter provides significant flexibility to invest in the business and continue returning capital to shareholders.

Turning to guidance. We've had a strong first half of the year, and we expect the second half to be stronger. The momentum we've seen continues, and we are raising our expectations about every line of business. Our second half gross margin rate outlook has improved over the past 90 days and we continue to drive significant operating leverage and scale. For Q3, we expect revenue to be $49 billion at the midpoint, up roughly 80% year-on-year. We expect ISG to grow roughly 145%, supported by $19 billion in AI server revenue. CSG revenue is expected to be up roughly 15%.

Operating expenses are expected to be down low single digits sequentially. Operating income is expected to grow roughly 120%. We expect ISG operating income rate to be up just over 1 point year-over-year, even as AI server revenue more than triples year-over-year. We expect CSG operating income rate to moderate to roughly 6% as we balance demand, share and profitability. We anticipate a diluted share count of approximately 651 million shares. Diluted non-GAAP earnings per share is expected to be $6.50, up over 150% at the midpoint. For the full year, we are raising our revenue guide by $25 billion to $192 billion at the midpoint, up roughly 70%, with diluted non-GAAP EPS of $25.50 up approximately 150%. We expect ISG to grow roughly 120%, driven by AI server revenue up 3x year-over-year to $74 billion. We expect traditional servers to grow just over 100%, storage up in the mid-teens and CSG revenue to grow in the mid-teens.

Excluding the mix impact of AI servers, gross margin rates are up year-over-year. Our modernization efforts are paying off, simplifying, standardizing, automating and enhancing our operating model with AI delivering significant operating leverage with operating expenses to be approximately 8% of revenue, the lowest level in the company's 42-year history. With gross margin improvement and the benefits of significant scale, operating income is expected to grow approximately 120% with over 2 points of rate improvement year-over-year. I&O is expected to be between $1.4 billion and $1.5 billion. Diluted non-GAAP earnings per share is expected to be $25.50, up approximately 150% at the midpoint.

In closing, we've delivered another exceptional quarter, capping a record first half of the year. Over the past 2 quarters, revenue was $90.8 billion, up 71%, EPS grew 208% to $11.90. We generated record cash flow from operations of $6.3 billion and returned an all-time record $6.3 billion to shareholders. The team executed exceptionally well across the business. The second quarter provided further evidence that AI momentum is accelerating with $60.9 billion in orders $16.4 billion in revenue and a backlog approaching $100 billion. At the same time, traditional servers, storage and CSG all contributed, reinforcing the breadth and balance of our portfolio.

Beyond the numbers, I would highlight the operating discipline. The modernization work we've invested in over several years is showing up in scale, in margin structure and in our ability to execute in a dynamic supply environment. We're entering the second half from a position of strength, and we'll continue to balance growth with discipline to drive long-term shareholder value. You are seeing the compounding benefits of our durable competitive advantages, differentiated operating model and operational discipline. We're excited about the second half and confident in our long-term value creation.

Thank you to the team for their execution, and thank you all for your time today. Now I'll turn it back to Paul to begin Q&A.

[Operator Instructions] Let's go with the first question.

Our first question comes from Amit Daryanani with Evercore.

I want to spend some time on the non-AI part of ISG? And if I look at a traditional server growth of 122% was actually faster than AI compute and storage grew 26% as well. I think a worry folks will have is this driven by a combination of pricing and prebuys rather than real demand. So I don't know if you can spend some time just talking about what do you think is driving this demand? And is there a way to think about pricing versus demand versus share gains? And really, any color on what workloads or use cases are you seeing this infrastructure going into and durability effect would be helpful.

Sure, Amit. Let me try a little bit. So if you look at traditional servers and what we're seeing, which is the vast majority of the growth that we saw in the quarter. It's a consistent theme that I think we talked about last quarter. One, there's a modernization in the data center. That modernization continues to drive consolidation. It is increasing space, driving power efficiency and cooling, and it's obviously driving demand. And demand for new servers that have more cores, new servers that have more DRAM, the new servers that have more storage in them as we consolidate an aged installed base.

Secondly, that's probably the next big opportunity for us. As much as we've modernized and to give you a sense that it's not an end near or it's a onetime thing, we still have 1.2 million assets that are 14G or older in the installed base. They have to be upgraded. They're going to have to be consolidated with new technology, whether it's our 17G and the consolidation ratios 6:8:1 or a new 18G, they will begin shipping next month, where we see consolidation rates in the 12 to 14 servers per new 18G server. That is going to happen and a forcing function is going to be the security environment that we live into them. So we think about what's happening in the world of security and driving increased resilience and new requirements like post-quantum photography coming online, old infrastructure has to be updated.

And then increasingly, we're seeing enterprises drive AI workloads, specifically agentic workloads. I know your question was specifically the non ones, but it's complemented by growth there. In storage, we see a very similar dynamic. We have the dynamic of our products are very, very competitive in the marketplace. Data continues to grow. So regardless of the inflationary environment that exists, more data is being created on the planet at the edge and data centers in the cloud, and that data has to be stored. It has to be encrypted and protected. And those are the opportunities that we see, which is why we believe our Dell IP portfolio has a pretty significant tailwind. Think about it.

I think if we blend Q1 and Q2 together, we grew storage 17% in the first half of the year. We continue to see our Dell IP storage growing ahead of the market for 6 consecutive quarters on a demand basis. We could run off a bunch of fund numbers. PowerStores now grown 10 consecutive quarters in a row. We have PowerScale 5 quarters in a row, ObjectScale 4 quarters in a row. Data Domain, 3 quarters in a row, our all-flash arrays have grown now 10 quarters in a row. So there is inherent demand. Our products are more competitive. And we're seeing that play out in the marketplace from the largest enterprise customers down to small and medium-sized businesses. And then there, there's also the opportunity to grow with AI, which is driven by agents and KB Cash and new techniques in the AI world. I hope that helps.

And maybe to add, Jeff, I think it's part of the durability of that growth and demand. Again, we see as part of our guide, our second half growth rates maintaining what you've seen in the first half. So Jeff mentioned the 17% growth in storage, pretty similar to mid-teens for the second half. We'll continue to guide to traditional server growing triple digits again for the second half as we drive that through. So we continue to see pipelines build. You continue to see the use cases that Jeff mentioned, and it all points to a more broad-based, more durable ecosystem.

And the next question will come from Ben Reitzes with Melius Research.

I'll echo pretty impressive quarter and guide there. wanted to ask about a little longer term, your partner in AI servers talked about growing 70% next year in overall revenue. You guys are growing faster than that. Your backlog just surged. I was -- and you also have these CPU racks that are new, adding to traditional servers. So would you be willing to -- should you grow kind of in line with NVIDIA for next year? You guys are really part of the [ ACIE ] segment they have. Do you see that kind of growth rate in your future or anything you want to kind of say about your long-term growth rate, given it's so much better than expected, would be appreciated.

Thanks, Ben. Look, I think if you anchor in on our second half trajectory building on the last question, you can see we like the position in relation to the durability that we see in the demand. We see it across the portfolio, and that's giving us tremendous leverage. As we continue to grow that scale that we get in the P&L, again, offers us the opportunity to continue to find scale and growth in the business.

As the second half growth, which is 68% is pretty much a mirror image to the first half, 71%. And it's obvious we're seeing signs where the data center is turning in from this cost center approach to a value creator. And the ecosystem and the enterprise customers that we're seeing are starting to embrace that. There's lots of complexity and execution. I think right now, really keen to execute a strong second half continue that great momentum as we go through the second half of the year, I think we'll be in a great position at that point, and we'll continue to look for the growth going forward.

And then maybe some more context around that. Our 5-quarter pipeline grew sequentially. That's after booking $131.7 billion of orders over the past 4 quarters. I think that gives you a sense of what's happening today. And then if I look at the longer-term trends, I know you're a believer of this, but as we see it, agentic demand is reshaping the data center and the underlying infrastructure. Inferences past training and it's pure demand in our industry. We think the tokens that inference drives is going to grow 87x to 3,600 trillion tokens by 2030. Training demand grows 5x to 850 Zettaflops by 2030.

Enterprise agentic is expected to be the single largest workload by 2028. We're expecting AI to be 75% of all data center demand by 2030, adding 200 gigawatts of power over that same time frame. And half of that, we believe, is right in our sweet spot with our customers, the neocloud, sovereigns and enterprises. And if you look at that math, we think the opportunity in front of us is more than $1 trillion over that time frame. And we believe we're well positioned. We believe that our model is differentiated, that our engineering is differentiating ourselves with every customer that we interact with. The scale of our deployment capabilities is unmatched globally.

We believe what we're doing on the support side is equally important, helping customers ramp getting to that first token faster than anyone else and then keeping it running. And then the DFS component that we have to help customers in that bridge point from an order to that first token is something that we believe is differentiating us, and we're going to continue to focus on that. And then if you believe that demand is there, it drives more servers, the agentic workload, and it drives more data around that agentic workload growing each of those areas for us as well.

And our next question will come from Mark Newman with Bernstein.

A few more details on the huge strength you're seeing in both traditional and AI servers. First of all, for traditional service, this has been traditionally almost all enterprise customers. And I believe you're lumping in the CPU racks, the agentic AI servers that are CPU racks in there, I believe. Is this traditional server category still almost all enterprise? Or are you seeing a portion of that from, say, neoclouds and then similarly, for the AI server customer mix, both revenue and orders, I know majority in the past has been new clouds or Tier 2 CSPs.

Is that still the same? I wondered if you could give us any hints in terms of the relative growth rate between enterprise versus other larger customers in the AI server mix because previously, you said enterprise had been growing faster. And I just wondered if that is still the case given the huge step-up, particularly in the orders.

You bet. Mark, traditional server, the 122% growth. It's primarily our historical enterprise customers. I'd stress demand outstrip supply. Demand was even greater than the results that we published there. We are supply constrained but demand is from our traditional enterprise customers. That's where the vast majority of the workloads are. That's where the modernization is occurring. That's where the aged installed base is. That's where the heightened awareness around security and resiliency has been driving demand. I introduced last quarter that we are beginning to see an AI servers and when I talked about that and it would be the same that happened this past quarter that our neoclouds buying that. So of our high-frequency trader customers are buying those types of servers as well as very advanced in their AI deployments, our largest and most sophisticated enterprise customers.

So vast majority of that 122% growth are traditional customers across all segments, all geos. AI servers are beginning to show up with that set of customers which is exciting to see that grew quarter-over-quarter. It grew across neoclouds. It grew across our HFT customers as well as our enterprise customers. So that's exciting to see. And the mix inside our traditional AI business is exciting and something that we've talked about, and I think we mentioned in our remarks, we now have more than 6,500 customers buying Dell AI factory. 3,300 of them have happened in the last 3 quarters that took us 8 quarters to get to the first $3,200. That acceleration is enterprise.

Enterprise customers grew quarter-over-quarter year-over-year. Repeat buyers grew quarter-over-quarter and year-over-year. Enterprise revenue grew quarter-over-quarter and year-over-year, and the pipeline of enterprise customers grew sequentially as well. So we are seeing more enterprise customers. The mix didn't necessarily change because we are still winning on the sovereign side as well as the large neocloud side but the momentum with enterprise that's measured by a number of customers. The number of customers that are buying repeatedly is all up and the indicators are strong. And they tend to buy more storage and they tend to buy more networking when they engage with us, a more complete solution. I hope that helped.

And we'll take a question from -- [Audio Gap]

That work has to be retained. Depending on what type of customer you are, there's compliance and regulatory requirements about how long that's got to be retained and what the protection policies are with that. So you have another new source of growth for storage. You see the same happening with KB Cash and how it's being used and driving more efficient inference. So we see multiple new lines or new paths for storage growth in our businesses. If we think about this across agentic workloads as we head towards physical AI and what's going to happen in manufacturing and IoT sensors and robotics, which drive tremendous amounts of multimodal unstructured data, Arthur likes to call it unstructured repositories. There's a lot of structured data and databases. The growth of that is immense, and we actually see it accelerate not slowing down.

And that bodes well for someone that's in the storage business, which we absolutely are and equally important in the data business and how do we start helping customers with forms of data management. We talk about this internally around creating a data semantic layer, a layer of intelligence that helps make something of all of that data to feed the AI engines to make sure that they can actually produce something even more worthwhile to help the agents be more efficient, et cetera, to help training.

That virtuous cycle, we believe, is just starting. And as we understand it today, we're very optimistic about the growth of storage going forward in the AI world. And we're positioned quite well across all of our storage assets. We protect data. We store all forms of data and as we build more of our data automation platform, we think about our data management work, I think we have a huge opportunity to grow and to be even more important to our customers as their data needs grow.

And our next question will come from Asiya Merchant with Citi.

Great results here. Can I just ask a little bit about supply? Jeff, I know you mentioned supply constraints. Maybe if you can just help us understand where the supply constraints have anything's changed from the last quarter? Clearly, some of the component makers are talking about supply agreements that have been signed. How do you think about your supply going ahead? And what we should think about where some of the incremental supply constraints are perhaps relative to last quarter in order to meet the demand durability you're talking about even going into next year?

You bet. How I think about supply, as I'm often reminded by our sales force, it's not enough. So we are doing everything we can to get more supply. In today's environment, that's a very difficult task. What we've been doing is, I think, optimizing the bits and bites that we have coming in, whether that be with configuration that being building match sets to maximize the output of the corporation out of the factories.

Our ability to increase guidance by the $25 billion is a direct reflection of our ability to optimize what's coming in, shaping demand, planning it accordingly and getting it out the door. One of the things that we did earlier this year as we saw the PC market showing signs of softening in the second half, we optimized the bits and bites we have towards the infrastructure business. There's a lead time associated with that. We're working through that lead time, which is part of why the second half looks a little better.

We've been able to realize greater shipments as a result of that. The constraints remain the same. DRAM, DRAM, DRAM, followed by NAND, NAND NAND. We have spotty CPU shortages. There are shortages with disk drives. If you go further down in the supply chain, just about every product going through a leading node is constrained, but churn nodes that are building MOSFETs, Power ICs, microcontrollers, drivers are constrained. There are shortages of ABF substrate, T-Glass, all of which we monitor. There are shortages in optical. The AI supply chain is working red line all out to build CDUs, power, racks. Welcome to the life of a supply chain person at Dell. This is what we do, chasing parts. We love it. trying to optimize the outcomes for the company. I think we've done largely a good job of that with the second half guide up, and we'll continue to focus on trying to get more supply and take the supply we have and optimize the output.

And the next question will come from Aaron Rakers with Wells Fargo.

This is Michael [ Spednoff ] on behalf of Aaron, thanks so much for letting me ask the question. I wanted to ask on the storage business, obviously, very strong. Within that, you mentioned several solidly performing DIP portfolio products, one of which I wanted to ask about is Lightning. How is that contributing at this point? And kind of what level of attach are you seeing to those cloud AI server deals?

Well, Lightning, our parallel file systems. Sorry, parallel file system designed for native AI use cases. We continue to have the product out in the field. We continue to see interest. It's still a relatively new product. It's in beta at several customers. We're in runoffs against other competitors with the product that will continue and as it builds momentum. I'm certain we'll give you an update in the future, but that's where Lightning is at the moment.

And our next question will come from Joseph Cardoso with JPMorgan.

Maybe can you guys -- and I know you guys have talked about the traditional business here in length, but maybe just curious if we break down the growth that you're seeing on the traditional server and storage side between volume and pricing. Relative to your earlier view, how much of the upside is coming from each of those vectors? And maybe more importantly, as you think about going forward and the momentum you're seeing in the business, how are you thinking about headroom for customers to keep on digesting these higher prices? And are you -- in your customer discussions, are you starting to see any pushback there?

Parsing servers and storage by revenue and growth. Let me try. So if I look at servers and what we're seeing in traditional servers, we're seeing, again, this notion of modernization that's driving higher core count, more DRAM and more storage. Those products are -- those configurations are part of this modernization or consolidation, and they continue to grow rapidly. They cost more than they did last quarter and the quarter before and the quarter before. So there's a notion of inflation inside our growth.

But the underlying demand for the technology is significant. I think about the new use cases, that's all new use cases, all new growth, which is being driven by agentic AI, essentially running the harness, if that makes sense. And we continue to be optimistic about the prospects. Again, demand outran supply last quarter, demand outran supply this quarter. The pipeline remains robust. David just gave an update on guidance of the server business, which is very healthy. And clearly, there's a component of that driven by the price increases as our input costs continue to go up.

Storage is a very similar story as I think I mentioned to one of the earlier questions, as we see a ramp down of our partner IP portfolio, it's being offset by more units in our Dell IP portfolio that come with higher revenue and higher margin rates. We're seeing a greater use of our storage products in AI applications, which is good to see most notably with our unstructured products, which had unprecedented growth again, but I rattled off a bunch of numbers earlier about every category of our traditional or core Dell IP portfolio growing multiple quarters now. The Dell IP stack has grown 6 consecutive quarters now ahead of the marketplace. We expect to take share again. And clearly, some of that is uplifted by the increased cost of the underlying material. Our software-defined products are doing well in the storage portfolio, which is incremental business for us. So I hope that gave some color.

And we'll take one more question before we go to a close.

And that question will come from David Vogt with UBS.

Maybe for Jeff and David, can you help us understand sort of the long-term margin differentials as customers modernize to next-gen servers off of older generation servers like 14 on the way to like 17G, 18G? And how much of the margin uplift that you're seeing in ISG comes from a like-for-like margin lift as we modernize data centers versus sort of your efficiency improvements and scale economics that you're seeing just from more volume going through the supply chain?

Yes. I mean, if you look at our Q2 results here, ISG up 15 points, obviously, tremendous performance. As Jeff outlined it earlier, the #1 driver here is a scale conversation given the growth that we're seeing in accelerated growth that we're seeing. That for the ISG business was a driver of just over 400 basis points. For the full year guide, it's worth over 650 basis points. So you see the leverage that we can adopt into the ecosystem.

Outside of that, you then -- as you do your storage growth and again, we drive that 13% guide, our 15% guide, excuse me, for the full year, that $2.5 billion of incremental storage is a huge drag in terms of revenue dollars that we pushed through. As you look at our guide for the second half of the year then for ISG, you'll see it's up over 1 point in the second half and you'll see it grow from Q3 to Q4 also even with the expansion of AI, which is over 3x growth year-on-year to $74 billion guide. So all of that's kind of contributing to a robust portfolio and then across the rest of the portfolio, it's really about mix, product mix, geo mix and just traditional execution, pricing discipline and operational rigor across our supply chain engineering and sales teams.

All right. We'll move it over to Jeff to close this out.

Sure. Thanks, Paul. Thanks, everyone, for joining us today. Our advantages are compounding, our addressable opportunity is expanding and our differentiated operating model is delivering significant leverage with our full year OpEx rate at a 42-year low. We raised our full year guide by $25 billion to $192 billion with $25.50 of EPS. We are optimistic about a stronger second half and the momentum we carry into next year. Thanks, everyone, for your time today.

Thank you. That does conclude today's conference. We do thank you for your participation. and have an excellent day.

Welcome to day 1 of Bank of America's Global Tech Conference. Delighted to see many familiar faces over here. I'm Wamsi Mohan, and I cover IT hardware and tech supply chain for the bank. Glad you could all make it. I'm delighted to welcome Dell here today. With us today, we have Arthur Lewis, who heads up the Infrastructure Group, which has been nothing short of amazing lately. So grateful to have you over here.

I do have a quick safe harbor statement to read from Dell. This presentation contains forward-looking statements based on Dell Technologies' current expectations. These statements involve risks and uncertainties that could cause actual results to differ materially. Factors that could cause results to differ are discussed in Dell Technologies' periodic reports on Forms 10-K or 10-Q filed with the SEC. Any forward-looking statements made today are based on assumptions as of today, and Dell Technologies undertakes no obligations to update them.

And with that, -- and I think Zach will give me some props from IR for reading that pretty fast. So -- but welcome, Arthur. Thank you for being here today.

Wamsi, it's great to be here. Thank you for having me.

Yes. No, absolutely. I know you're a busy guy, so we appreciate you being over here. And we've only got 30 minutes. So we'll make the most of it. So maybe to start, right, I would say like this, by any standards was maybe a really historical quarter for Dell, right? ISG was right in the center of all of that. How would you frame where ISG is today versus what you might have thought like maybe a year ago?

Well, clearly, the market is growing significantly faster than we thought 12 years ago, significantly faster than anybody would have predicted. And I think we've shown an incredible ability to navigate what is unprecedented sort of demand in the industry. When you take a look at Q4, overall revenue, $29 billion, up 181%. That's 9 consecutive quarters of double-digit growth. That's on top of growth of 73% in Q4 and growth of 12% in Q1 of last year.

And it was growth, Wamsi, across the portfolio. $8.5 billion on the core server side, that's up 92%; $4.5 billion on the storage side, that's up 8%; and $16 billion on the AI side, up 760%; and that's on top of $24 billion of orders and a record backlog of $53.1 billion and a backlog that still sits at multiples of our backlog. And operating margins, obviously, strong at 10.5%, even with the higher mix. That's largely due to scale and our growth in storage, right? That's been great. And overall operating dollars of $3.1 billion, up 206%, growing faster than revenue. And if you look over the last 2 years, revenue is up 3.1x and operating margins are up 4.2x.

So we like to see that the revenue is growing, but that the profitability underneath it is even stronger than the revenue growth. So you saw it in the absolute dollars, 206% versus 181%. And you see it over an elongated 2-year period where operating margins are up 4.2x versus revenue up 3.1x. So pretty pleased that we're executing in the face of unprecedented demand.

Yes. No, that's quite incredible, those stats that you just gave. And I remember, it was not that long ago that Dell was talking about $90 billion in full year revenue. And here, we are talking about double of that in a very short period of time. So most of that coming from the infrastructure group and particularly within AI servers. Maybe you can just give us some sense of like the confidence in this updated guide, right? Like I think going into the quarter, the perspective that people had was, "Yes, demand is strong," like you could do your channel checks and say, "Yes, like there was maybe pull forward of demand, like [indiscernible] was saying, a lot of demand is getting pulled forward."

So there was this notion of, "Look, it's going to be a great result, but we don't know about the guide." Well, we didn't know about the guide. The guide was order of magnitude different from what everyone thought. What's giving you confidence in projecting this guide?

All right. So the one -- first thing I want to communicate is the guidance went up $27 billion, $140 billion to $167 billion. That guide is only gated by supply. The demand that we're seeing far exceeds the supply that we have. So that's sort of point number one. But more importantly, what we are seeing -- and we now have deeper conversations with customers that are incredibly worried about their demand aspirations versus the supply that they're able to get. So we now have visibility into '26, into '27, into parts of '28.

And across the portfolio, whether it's GPU, whether it's core server versus storage or networking, we have a pretty good view as to kind of like where the demand is coming. And one of the inflection points -- I mean, artificial intelligence has been sort of this existential defining technology of our generation, maybe for all of humanity, quite frankly, something that we said was going to be bigger than mobile, bigger than the Internet, bigger than virtualization, bigger than all of those things combined. And that's exactly how it's playing out.

And I think in the fall of last year, when you think about how fast this technology is evolving, there was an inflection point with agentic technology, right? It was really cool when generative AI came out and you could train it on data, ask it a question, it gave you a really good answer. Then you get into these auto regressive long-thinking reasoning models, and it gave you a better answer. But now the technology can actually act, manage, orchestrate activities, which makes it even more useful than it was before. And you saw that as the enterprise accelerated AI adoption, we now have over 5,000 customers out there. More neoclouds are popping up. The demand for power around the world is skyrocketing. So yes, this is a pretty interesting demand cycle that we're in.

Yes. No kidding. I mean it is pretty unprecedented. So on your point on agentic, maybe just to talk about sort of how that is changing the business. You saw extremely strong performance in industry standard servers, 97% year-on-year growth. I mean, quite amazing. So you also said you're gated more by supply than anything else. As you think about, what is the evidence or what are you -- what are some of the signs that you're seeing in the market that's giving you that confidence that this enterprise adoption is starting to like finally pick up?

I almost got to 97%, 92%, but...

Yes, 92%. I'm sorry. Too many earnings.

Yes. So I would break it down into sort of 3 components as you think about it. Component number one is that the current server TAM sort of continues to grow in a lot of these legacy type workloads. We see things like these EDA farms continuing to grow, requiring more compute for all these really cool ASICs that are being created around the world across all these different semiconductor companies. We see modernization of current infrastructure. We see a modernization of the installed base. Majority of our installed base still sits on servers that are 7 years and older.

So like your core market is growing. At the same time, that core market is expanding with more AI-driven workloads, and agentic has accelerated that. And the reason for that is because prior to agentic technology, the majority of artificial intelligence was really more math matrices driven, which was really unique to the parallel processing that comes with a GPU. But as you get into agentic technology where these agents can act, manage and orchestrate activities, that is a serial sequential process that requires a CPU, right? And so we're starting to see those workloads come online.

And so we've seen sort of this kind of inversion of the GPU to CPU ratio. And honestly, I don't really like the CPU/GPU ratio because it's not like a scientific thing, and I know the finance community likes to model stuff. But just to give you conceptually how it works is if you have an agentic task where there's, say, 50 calls to the model, there could be 250, 300 calls to the tools, right? And so the model is a math exercise where parallel processing works great. But the tools are, "I got to go do this step, then the following step, then the following step after that, then the following step." That's all sequential processing more -- better run on a CPU.

So you have a market that's growing. You have a market that's expanding, and we're taking share. We grew 500 basis points of share in Q4. We expect to take meaningful share of Q1 when the data comes out in June. So you put all kind of 3 of those things together, and we think that there's durable demand and opportunity for us to continue to win.

What is the reason that you're gaining share fundamentally? Like, I mean, obviously, there's strong demand you guys are able to deliver. I guess that's the answer. But when you think about share gains, historically, people have not really associated. I know you've historically taken share in lots of markets, and you are standing at #1 share in most of your end markets. So -- but what is the ability? What is special at Dell that allows you to take share and you have confidence in that ongoing?

Well, look, there's the practical consideration that -- and obviously, you're talking to a biased individual, but there's the practical consideration that we have a great portfolio, right? We have great products. They're reliable. They're performant. They're secure. They're the best in the industry. You have a practical component that we have a good amount of supply, right? And so we're able to meet the needs of the customers.

But when I zoom out and kind of like what's changed for me dramatically over the last 4 years is the strategic nature of the relationship that we have with customers. As we entered into the world of artificial intelligence, customers brought us in to have more board level like discussions. They really wanted to know how do I think about ROI and use case selection? How do I think about model selection? How do I think about data preparation because obviously, data is the fuel that feeds AI. How does that inform my architecture? And then how do I think about infrastructure as a result, right?

So before AI, we would probably be brought in -- customer has an infrastructure need. They send us a request for a quote, we quote it, we have a great product, we win. But now the relationship has become more strategic and our role as a trusted adviser has elevated greatly over the last 4 years. And our ability to communicate a full stack system has also resonated well with customers. So the combination of a great portfolio, the availability of supply, but that trusted adviser where people have the confidence and that we have the credibility to help them work through what is a very complicated transition from legacy data center to agentic data center, I think that's been key.

I meet, Wamsi, with a lot of customers. I was joking in an earlier meeting like I love talking about product speeds and feeds, maybe 5% of my time is at. 95% of my time is talking more strategically around what does an agentic data center look like? What are your AI ambitions? How do I think about token economics? How do I think about my data? How does my data strategy need to change? Way more strategic conversations than individual product line speeds and feeds.

Any, like, insight you can share around those strategic conversations? And I ask this, and it's kind of ironic, like if you go back -- and we've covered Dell for a long time. If you go back some time ago when Michael decided that it would be good to spin off VMware, people were worried that Dell was going to be a pure hardware company with like not a strategic seat at the table. And today, we are talking about increasingly more strategic like seat at the table to determine the architecture on an agentic AI basis for enterprises. So any insights you can share around like token economics, time to first token? Like what are the discussions that are important to your customers?

Yes. So I think the way I would frame that is data centers over the past 10 or 15, 20 years have sort of been an amalgamation or accumulation of workloads, right? And these workloads were all very independent of each other, right? So you could have a data center with 15, 20, 25 different workloads and the infrastructure was sort of siloed based on that architecture.

As you move to more -- to an agentic data center, you're now in a world where all of these systems have to talk to each other, right? These systems have to be connected. Data silos have to be collapsed and the quality of the data needs to significantly improve in order to drive optimal results. So this transition from a legacy data center into an agentic data center is a big part of the conversation. Then token economics is obviously a big part of the conversation. There, customers want to talk about how do I effectively generate tokens and how do I effectively use tokens, right?

That's kind of like the equation we work with customers on because they want to know that, "Hey, not only can I generate them effectively, but I'm using them optimally." And again, the breadth of our portfolio from the data center to the edge and the PC gives us a lot of credibility in helping customers navigate where do I want to run a model, where do I want to run agents? You may want to run them in a data center, you may want to run them locally on a PC. We can have an end-to-end conversation with customers, which is also a pretty strong point of differentiation for us.

And then you kind of peel the onion on sort of those 2 big things. And then there's a bunch of strategic discussions. I think the third one I would probably call out is the data. This is a massive problem for companies. Majority of their data is sitting dark, sitting cold. We believe like in an agentic data center, there is no cold data. There is no dark data. This data is sitting in constant circulation, feeding engines, feeding artificial intelligence, driving optimal results. So the richness and strategic nature of the conversation has changed pretty dramatically over the last 3 to 4 years.

Yes. No, that's super interesting. You probably have a viewpoint on this as well. In general, like when you think about the adoption of AI, we've seen a lot of spending by hyperscalers predominantly that are kind of driving this infrastructure spend, right? And you saw the Google announcement overnight, hyperscalers spending a ton of money. So this notion of where AI is running today is largely in the cloud. But would love to get some sense from you whether just given that you mentioned data, how do -- how should enterprises think about bringing compute to data versus data to compute? So it's sort of like an on-prem like bring the compute to the data or more of a cloud where you're taking data to the cloud?

Yes. So we wrote down a couple of different hypotheses when we first started down this path of artificial intelligence in the beginning of '22. And one of them was that there's going to be strong gravity to data, right? And that's proven out to be the case. 83% of data sits on-prem today in the world across a majority of enterprises, and there's a strong proclivity to deploy on-prem for performance, cost, security reasons. And all 3 of those have very important characteristics for customers.

So we see more around, "Hey, over the last several years, I had a digital transformation strategy and a lot of it was cloud first." Now when we look at customers' digital transformations, it's data first, right? Because without the data, I don't have the fuel to feed the AI. And we have the portfolio to help customers turn that data into the premium grade fuel that they need to drive optimal results for artificial intelligence. And this is where we have new offerings like our AI data platform, which brings data management capabilities to bear where 5 years ago, you might have said, why is Dell getting into the data management space? Well, now it's a natural mandatory adjacency that we have to really solve agentic problems for customers.

Yes. The market has always been very competitive, broadly speaking, in infrastructure. You guys have done an amazing job in being able to manage supply chain, being able to like bring products to market quickly. When you think about the longer-term trajectory of the AI server business, how do you think about that? I mean we've just gone from like these -- we've kind of gone through these unprecedented growth rates, right? You're talking about $15 billion that went to $25 billion and $50 billion, now $60 billion. I mean these numbers are like growing at some very massive rates. So as you plan for growth in the future, how are you thinking about that?

As much as I liked earnings last week, I liked the Dell Technologies World the week before even more. The amount of innovation that we're driving across the portfolio is like nothing that we've ever seen before. We like to say our innovation engine is running incredibly hot. And it's not just across the GPU portfolio, which has been great: ARM, x86, air cooled, liquid cooled, but across the broader portfolio on core compute, where we launched 18G, including a new family of high-performance compute systems built for high density -- high core count density memory bandwidth, all kinds of innovation on the storage side with our private cloud platform, our AI data platform as well as our cyber resilience platform and bringing it all together within the AI factory to deliver a full stack architecture for customers.

And one of the things that we're seeing, Wamsi, is customers really appreciate the full stack solution because they don't want to be the integrator of systems and agentic requires connected systems that requires the compute, both the general purpose and the accelerated compute, the network and the attached storage. So our portfolio is broad. Our portfolio is deep, and our innovation engine is humming like never before, and it's incredibly exciting. From my perspective, the opportunity is too big not to be driving this very, very hard.

Yes. Yes. It's super interesting to watch sort of GPU transitions as they happen. It's no mean task to ship these things in high volume. What have you learned from past transitions? And how are you looking at like the next couple of transitions coming up?

Well, we've gotten pretty good about this, first to market with Hopper, first to market with GB200, first to market with GB300. And you may have seen a post from Michael over the weekend where we shipped our first QS-0 rack, full NVL72 working rack to CoreWeave, the first in the world.

So we've gotten pretty good at working very closely with NVIDIA, working closely with our customers to develop and bring up what is a very, very complicated technology. We have unique capabilities at the node level, but we have really unique capabilities at the L11 integration sort of capabilities. And we're now in a position where we can turn rack into production at a customer site in under 6.5 hours and maintain uptimes of 99.9%, which is kind of unheard of in the industry.

Yes. Those are some impressive stats. Arthur, maybe just talking about memory pricing, which has been obviously very inflationary. I think a lot of people felt that this would be a hurdle for companies like yours to navigate, which would create margin pressure. You guys have actually managed this incredibly well. So your ISG margins were very, very strong. And this is AI servers growing pretty strongly, industry standard servers growing pretty strongly at 92%. So as we think about that, like what have you been -- how -- why did people get it so wrong? What is it that you're doing that people don't have maybe full visibility into and rate and pace of that as you think about the future?

Yes. Well, Jeff made a call early in December, this -- some people call it a cycle, I think it's sort of the new normal. This was going to be -- regardless of how you thought about it, it was going to be a long period of time. And we needed to move quickly to communicate to customers what was going on in the industry. And on December 9, we moved the entirety of the business into the new cost structure that we were seeing. At the time, it caused a lot of consternation with customers, as you could imagine. But over time and over the next several months, that very transparent, clear message that we started to deliver on December 9 became appreciated and customers saying, "We see what you -- now we see what you saw -- we see now what you saw back then. And we appreciate you having the courage to have that conversation with us early, so we could think about this over the long term and plan for it accordingly."

And what you've seen since then is that we've driven pretty strong stability across AI servers. We've driven stability across server margins. And importantly, like we're in sort of historical ranges from a server perspective, and we're actually below where we were in COVID. So we've tried to drive stability there to make sure that we're not profiteering in any way. And then we've driven significant improvements in our storage portfolio where we're seeing storage growth, and we're seeing growth in Dell IP storage, and that Dell IP storage is more valuable than partner IP. And that's been one of the bigger lever in the overall ISG profitability framework and then scale would be sort of the last component.

Yes. No, that's really impressive. So maybe touching on storage for a bit, right? Like we investors are all trying to figure out like where the puck is going here. And it seems like we've seen this with compute and it was sort of interconnect, and we've seen it with CPUs and industry standard servers. Should we be rethinking how storage is going to look in the future?

Well, look, there's no question we've been talking about -- it was good to see the 8% growth that we saw in storage, but we've been talking about Dell IP growing on a demand basis well ahead of the market for 5 consecutive quarters now, and you're starting to see it in the P&L. And the innovation in that portion of the portfolio is probably some of the best innovation that we have anywhere within ISG.

If you think about the AI data platform, something that's incredibly relevant for customers, we've introduced a data management capability with the Dell orchestration engine and the Dell data engines that help to ingest, transform query data at great scale. Customers who are deploying this are seeing 12x faster vector indexing, 3x faster query, 19x fast times to first token. We've introduced new architecture with Lightning, which is not only the fastest parallel file system in the world, twice the throughput of our nearest competitors, but it also represents the context memory extension and data storage access that you need for disaggregated inference.

And when you think about creating a token generation engine, you really want to go down the path of disaggregated inference because you can split the prefill and decode stage of inference to create larger context windows, so you're generating less tokens when you do an AI type activity. And then we introduced Exascale, which is the world's only 4-in-1 unified rack architecture for storage that includes file, block, object as well as disaggregated inference. So a ton of innovation on the AI data platform.

We also introduced a lot of innovation into our Dell Private Cloud. At the core of that is PowerStore Elite. So PowerStore now, 8 consecutive quarters of double-digit growth. And the new PowerStore is absolutely amazing, 3x the level of performance with 1.5 million IOPS, 4x the throughput performance of the previous generation. We improved the world's leading data reduction guarantee from 5:1 to 6:1. We now enable up to 6 petabytes of storage in a single 3U form factor. And we've also built out the scale-out capabilities to include transactional file to allow customers to consolidate more workloads.

And importantly, PowerStore is a relatively new architecture and it's container-based. So it evolves with workloads and new technologies as they emerge. So customers know that this architecture is future-proofed, right? And then you have all the innovation on the data protection side with Cyber Resilience and PowerProtect One. From a table stakes perspective, customers get factory-to-site integration, Silicon Root-of-Trust, Zero Trust controls, end-to-end encryption, embedded immutability, all combined with PowerProtect Data Manager and PowerProtect Data Domain, which brings the data reduction of 75:1.

So like across the portfolio, whether it's private cloud, AI data platforms, cyber resiliency, the team's innovation is incredibly impressive. And we've seen it over the last 5 quarters with the growth in Dell IP and you're starting to see it in the P&L.

Yes. Those -- I mean, you're living and reading these stats as your customer conversations. I can see the pitch going right there on each of these stats, very impressive. So maybe this is sort of, maybe a different storage question, but you guys have like #1 position in enterprise storage, you have a broad portfolio. As memory pricing has gone through the roof here, so to speak, like are you seeing changes in the way customers are buying and changing their habits of all-flash versus hybrid versus disk? Like how is that changing from your vantage point?

It's not really changed. I mean all-flash grew extremely strong for us in Q1. I think what customers are appreciating is the efficiency and the density that they get with our storage portfolio, again, 5:1 -- or 6:1 now on PowerStore Elite, even more on PowerMax, improvements that we're seeing in PowerScale and ObjectScale, 75:1 on data protection. Look, storage requirements are only going to grow with agentic. Every single agentic conversation will be transcribed, recorded and kept. And so it might -- the growth there might be lagging a little bit on the compute side, but there's absolutely an explosion of data that's coming.

Yes. It's a shame, we've only got 30 minutes because there's just so much ground to cover over here, but maybe we're down to the last couple of minutes. So to close, like what are investors like maybe underappreciating about ISG post this quarter? It's hard to sort of say that the potential was not understood in the quarter, but what do you think -- what would you focus investors to, I guess, as a takeaway?

Well, look, I mean, I would say the stock price would indicate that investors are appreciating a lot.

Yes.

Which -- thank you, and we appreciate that. Look, I think for me, one of the most exciting things about the portfolio is the breadth and depth of full stack solutions that we can bring for customers. Customers are looking for technology that just works. And this is -- in an agentic world, all these systems are connected: the compute, the network, the storage, the data management, the ecosystem, the software, the deployment, the services capabilities. Our ability to deliver full stack validated solutions optimized to specific workloads, coupled with the credibility and knowledge that we've built over the last 4 years, deploying at the largest of the neoclouds around the world gives us a unique advantage that is incredibly hard to replicate.

Yes. No, that's amazing. Actually, I have 1 minute left. So I'm going to ask you something else before we close out, which is Tier 2 CSP demand has been very, very strong, and you guys obviously participate in that. Industry standard servers as you're seeing the rollout of agentic, are you seeing that across broadly both Tier 2 and enterprise. And within enterprise, like where do you -- what application set are people really leveraging these industry standard servers for? I'm sorry to close out on that, but...

Yes. No, no, that's a great question. Look, it's across everybody, right? It's across the Tier 2 CSPs. It's across the hyperscalers. It's across the enterprise. As we talked about a little bit earlier, when you get into the world of agentic, you're now into this -- by the way, it's another thing that Jeff called out 3 years ago that this AI is a heterogeneous mix of accelerated and general purpose compute because there will be parallel workloads that require parallel processing versus workloads that require serial sequential processing. And so as you move from technologies that can think and reason to technologies that can act, manage and orchestrate, it requires both the accelerated and general purpose compute. So we see that demand coming from both sides.

Yes. Well, we'll definitely have to get you a bigger room next time, Arthur, but we appreciate you taking the time. Thank you so much.

Thank you.

Good afternoon, and welcome to the Fiscal Year 2027 First Quarter Financial Results Conference Call for Dell Technologies Inc. I'd like to inform all participants this call is being recorded at the request of Dell Technologies. This broadcast is the copyrighted property of Dell Technologies Inc. Any rebroadcast of this information in whole or part without the prior written permission of Dell Technologies is prohibited. Following prepared remarks, we will conduct a question-and-answer session. [Operator Instructions] I'd like to turn the call over to Paul Frantz, Head of Investor Relations. Mr. Frantz, you may begin.

Thanks, everyone, for joining us. With me today are Jeff Clarke, David Kennedy and Tyler Johnson. Our earnings materials are available on our IR website, and I encourage you to review these materials. Also, please take some time to review the presentation, which includes additional content to complement our discussion this afternoon. During this call, unless otherwise indicated, all references to financial measures refer to non-GAAP financial measures, including non-GAAP gross margin, operating expenses, operating income, net income diluted earnings per share, free cash flow and adjusted free cash flow.

A reconciliation of these measures to their most directly comparable GAAP measures can be found in our web deck and our press release. growth percentages refer to year-over-year change unless otherwise specified. Statements made during this call that relate to future results and events are forward-looking statements based on current expectations. Actual results and events could differ materially from those projected due to a number of risks and uncertainties, which are discussed in our web deck and our SEC filings. We assume no obligation to update our forward-looking statements. Now I'll turn it over to Jeff.

Thanks, Paul, and thanks, everyone, for joining us. What a great start to FY '27. The first quarter underscored the strength and agility of our operating model and the advantage of our broad portfolio. Our team executed very well in a challenging environment, delivering record revenue and EPS. Revenue was $43.8 billion, up 88% and earnings per share was $4.86, up 214%. Demand was stronger than we anticipated across all lines of businesses and geographies with customers moving decisively to secure supply across a broad range of IT needs.

This drove meaningful scale, record cash generation and continued strong capital returns for shareholders. Our strong performance reflects not only demand in the quarter, but also the pace of innovation we continue to bring to market across the full stack of PCs, compute and storage. We have had a strong run of announcement since our last call. At GTC, we marked the 2-year anniversary of the Dell AI factory with NVIDIA and extended our leadership in accelerated computing. We introduced new infrastructure across NVIDIA's Vera Rubin rack-scale platform, the Rubin GPU architecture and RTX GPUs with form factors that scale the AI factory from the largest clusters in the world to the flexibility and efficiencies enterprises need. We also extended AI to the desktop with the new Dell Pro MAX systems supporting the GB10 and introducing the industry's OEM desktop with GB300.

At Dell Technologies World, we built on that momentum with new deskside server, storage and data management innovations. -- our deskside genic AI solutions help enterprises run production-ready AI locally, supporting use cases like coding, research and secure private assistance while keeping sensitive data and IP on-prem. Building on strong demand of our integrated rack scale systems, where Dell is the top rack scale infrastructure provider, we expanded the portfolio with the launch of Dell Power Rack, a turnkey factory integrated solution designed to accelerate deployment across compute, networking and storage.

In servers, our 18th generation of PowerEdge server portfolio expand support for AI, HPC and enterprise workloads with new air cooled systems that improve compute density and efficiency. On the data side, advancements in the Dell AI data platform help customers make enterprise data ready at scale with stronger orchestration, faster indexing of unstructured data and improved analytics performance. We further strengthened the storage foundation from modern and AI workloads. PowerStore Elite delivers up to 3x performance and density than prior generations with an industry-leading 6:1 data reduction guarantee. Object scale adds higher-density object storage and PowerFlex extends our exascale storage architecture with a unified approach across block, file and object workloads.

We continue to expand the Dell AI factory ecosystem with partners, including NVIDIA, Google Cloud, OpenAI, SpaceX AI, ServiceNow, Palantir, Mistral and CrowdStrike. For example, with Google Distributed Cloud, we are bringing Gemini models on-premises with confidential computes that customers can run AI closer to the data while meeting data residency, privacy and sovereignty requirements. The bottom line, Dell is expanding the AI factory from the data center to the desk side across compute, storage, networking, software and services. We're giving customers choice, helping them protect their data and enabling them to move from pilots to production faster. With that context, let me walk you through what we're seeing in the business. Our Q1 results and AI, the opportunity remains exceptionally strong, underscored by durable broad-based demand.

In Q1, we booked $24.4 billion in AI orders and recognized $16.1 billion of AI server revenue. We exited the quarter with a record $51.3 billion of AI backlog, and our pipeline continued to grow sequentially and remains multiples of our backlog, even after converting $24.4 billion into orders. Demand continues to exceed supply with memory as the primary constraint and we expect to exit the year with meaningful backlog. Our customer count surpassed 5,000 with growth across neocloud, sovereigns and enterprise customers. Our differentiated offering continues to resonate and our expanding platforms and capabilities are supporting continued share gain. We believe those share gains are rooted in things that have long differentiated Dell, strong engineering and design, the ability to deploy and install at scale, ongoing services and support and flexibility financing and consumption options.

In AI, those advantages matter even more. Customers were not just buying components, they are looking for integrated solutions they can put into production quickly on infrastructure they control with the performance, security and data foundation their workloads require. Moving to traditional service. Revenue was up 92% as demand remained well ahead of supply in Q1 with strength across every region. The majority of demand was driven by large enterprise customers refreshing their compute environments and expanding capacities to support growing workloads. For many large customers, ensuring compute availability to modernize and grow remains their highest priority. Customers are also increasingly focused on infrastructure density as they optimize both spent and data center space which is driving demand and platforms that deliver more compute capacity, greater efficiency and better consolidation within existing footprints.

Additionally, we saw AI inference workloads driving incremental demand for traditional compute. The majority of the installed base remains on 14th generation or older servers reflecting the continued refresh opportunity going forward. The memory uncertainty is driving customers to proactively secure access to infrastructure across both traditional and AI workloads over longer periods of time. We also continued to execute the pricing and margin discipline we established in Q4. All in, we remain confident in the demand outlook for traditional servers and our portfolio is well positioned to capture that opportunity.

Turning to storage. Revenue was up 8% driven by continued outperformance in our Dell IP portfolio. Dell IP delivered a record demand growth quarter, making our fifth consecutive quarter of demand growth above market. In primary storage, we saw notable strength in Paramax and Power Store. We continue to see momentum in the mid-range ecosystem with PowerStore delivering its eighth consecutive quarter of double-digit demand growth and unstructured. We saw strong performance from Power Scale and object scale with 3 consecutive quarters of growth, including double-digit in each of the last 2 quarters. Dell IP storage continues to become a larger mix of Dell storage with its higher margins. And as a result, Storage delivered strong profitability was a key driver of overall ISG profitability in Q1.

Turning to CSG. Revenue grew 17%, and we gained share for the second consecutive quarter with broad-based demand led by large enterprise customers. Commercial revenue grew 18%, our seventh consecutive quarter of growth with demand up for the ninth quarter. Large enterprise customers continue to refresh with double-digit growth across all regions. We continue to see runway in the refresh cycle with roughly 1/3 of the installed base consisting of devices 4 years or older. Consumer revenue was up 9%. Our third consecutive quarter of demand growth, supported by continued strength in gaming. Overall, CSG profitability improved as better expected demand drove higher attach, greater scale along with improved consumer profitability.

In closing, Q1 was a strong start to FY '27 and another proof point in the power of our operating model. We delivered record revenue, EPS and cash flow and continued returning capital to shareholders while executing with discipline in a challenging demand and supply environment with notable commodity constraints, particularly in DRAM and NAND. Customers have come to rely on Dell during periods of significant disruption, and we expect that to continue over the course of the year. Our customers are investing in AI infrastructure, modernizing compute, expanding storage and refreshing PCs to support the next wave of workloads, we are well positioned with our portfolio.

I am proud of the team's execution and confident in our ability to create long-term value for customers and shareholders. With that, let me turn it over to David to walk through the financials and our outlook.

Thanks, Jeff. We delivered a record first quarter, which positions us very well for the year. Execution was strong across the business from supply chain to sales to pricing driving record revenue, EPS and cash flow, along with continued strong shareholder returns. Total revenue was up 88% to $43.8 billion. Gross margin dollars grew 57% to $7.9 billion. Gross margin rate was 18.1%, driven primarily by mix shift to AI servers with AI revenue up nearly 9x year-over-year. Excluding the impact of AI mix, gross margin rate was up. Operating expenses were up 9% to $3.7 billion. primarily from variable compensation tied to our outperformance. Importantly, we drove meaningful scale in the P&L, with OpEx down 610 basis points to 8.4% of revenue, the lowest level in over 20 years.

Operating income grew 154% to $4.2 billion or 9.7% of revenue driven by higher revenue and resilient margins across traditional servers, storage and CSG. Net income was up 194% to $3.2 billion, primarily driven by strong operating income. Diluted EPS increased 214% to $4.86, a record. Moving to ISG. ISG revenue was a record $29 billion, up 181%, marking 9 consecutive quarters of double-digit or better revenue growth. AI server momentum remained very strong. In the quarter, we generated $24.4 billion in orders, $16.1 billion in revenue and ending backlog of $51.3 billion. Traditional server and networking revenue was $8.5 billion, up 92% and demand continues to outpace supply.

Storage revenue was $4.3 billion, up 8% with strong demand across the Dell IP portfolio. Execution across the LIP portfolio was strong. was another quarter of growth above the market. Unstructured solutions were the fastest growing along with strong demand across primary storage. ISG operating income was a record $3.1 billion, up 206%, marking 8 consecutive quarters of double-digit or better growth, primarily driven by higher revenue across the business. Operating margin was 10.5%, and up 80 basis points even as AI service grew nearly 800% year-over-year.

Looking at the key drivers of margin performance. Storage profitability was up with a higher mix of del IP and rate expansion within the solutions. Traditional server margins remained stable despite a high inflationary environment. AI server profitability was in line with our mid-single-digit operating income rate target. Taken together, these factors along with stronger-than-expected revenue drove meaningful scale in the P&L. Turning to CSG. CSG revenue was up 17% to $14.6 billion. Commercial revenue grew for the seventh consecutive quarter, up 18% to $13 billion. while consumer revenue increased 9% to $1.6 billion. CSG operating income was $1.2 billion or 8% of revenue. This performance was driven by stronger commercial revenue and mix, which supported more higher-margin peripherals. And similar to ISG, this all drove meaningful scale in the P&L.

Looking ahead, we will continue to balance customer demand with supply while driving scale across the business. Moving to cash and the balance sheet. We delivered a Q1 record cash quarter with cash flow from operations of $4.1 billion. This was primarily driven by sequential revenue growth and higher profitability. We ended the quarter with $14.1 billion in cash and investments, up $0.8 billion sequentially. Our core leverage ratio is at 1.2x. We returned $2.1 billion to shareholders this quarter including repurchasing 11 million shares at an average price of $147 per share and paying a dividend of approximately $0.63 per share. Repurchase activity remains strong, and we remain committed to our shareholder return framework.

Turning to guidance. Customers continue to prioritize their IT infrastructure needs with an increased focus on securing supply. We expect that behavior to continue throughout the year. For Q2, our revenue outlook is similar to Q1's performance. At the same time, we have increased our expectations for the second half, while maintaining an appropriate level of prudence given that we're only 90 days into the fiscal year. From a profitability standpoint, the pricing discipline and margin stability we saw in Q4 and Q1 continue to hold. Excluding the impact of AI mix, our gross margin outlook is better than it was 90 days ago, and we continue to expect margin rate expansion through the balance of the year.

For Q2, we expect revenue of $44 billion to $45 billion, up roughly 50% at the midpoint of $44.5 billion. IHG is expected to grow roughly 75%, supported by $15.5 billion in AI server revenue and CSG is expected to be up roughly 20%. Operating expenses are expected to be down low-single digits sequentially. Operating income is expected to grow roughly 80%. We expect sequential improvement in ISG operating income rate while CSG operating income rate moderates to roughly 6% as we balance demand, share and profitability. We anticipate a diluted share count of roughly 652 million shares. Diluted non-GAAP earnings per share is expected to be $4.80, plus or minus $0.10, up over 100% at the midpoint.

For the full year, we expect revenue of $165 billion to $169 billion, up nearly 50% at the midpoint of $167 billion. IHG is expected to grow roughly 80%, driven by $60 billion of AI server revenue at the midpoint or approximately 2.4x year-over-year. Traditional servers expect to grow just over 60%, storage up mid-single digits and CSG to grow low teens. We continue to prioritize margin rate expansion. Excluding the impact of AI mix, our gross margin outlook is higher than it was 90 days ago and remains up year-over-year. We expect operating expense dollars to be up high-single digits, driven primarily by variable compensation. At the same time, our modernization efforts are paying off, simplifying, standardizing, automating and enhancing our operating model with AI, delivering significant operating leverage with OpEx as a percentage of revenue in the single digits.

Operating income is expected to grow over 55% with improvement both in dollars and as a percentage of revenue. I&O is expected to be between $1.4 billion and $1.5 billion. Diluted non-GAAP earnings per share is expected to be $17.90 plus or minus $0.25, up roughly 75% at the midpoint. In closing, we delivered an exceptional first quarter with record performance across revenue, EPS and cash flow. Revenue was $43.8 million, up 88%. EPS grew 214% to $4.86. We generated $4.1 billion in cash, and returned $2.1 billion to shareholders. The strength of the quarter reflects broad-based execution across the business, continued momentum in AI and solid performance across the rest of the business. Our portfolio and operating model continue to differentiate us in a dynamic supply environment, and we remain focused on supporting customers while driving shareholder value. We are well positioned for the year. Thank you to the team for their strong execution, and thank you all for your time. Now I'll turn it back to Paul to begin Q&A.

Let's get to Q&A. In order to ensure we get to as many of you as possible, please ask 1 concise question. Let's go to the first question.

We'll take our first question with Ben Reitzes with Melius Research. .

Well, let me say congratulations, guys. I don't think I've ever seen a Dell quarter like this, maybe Michael had 1 in the dorm room or something beating expectations. But congrats to you guys. The question that I have is with regard to the inherent level of real demand. when you see something like this, you think like there could be some pull forward, especially in the traditional servers and the PCs. But the way you guided for the year, obviously, by taking up the second half would imply that the pull forward doesn't have much of an impact stealing from the rest of the year. Can you just go through the puts and takes of the pull forwards in the major segments, please? And how you came up with still a higher second half?

Sure. And thank you, Ben. It was a good quarter. It's been here a long time. It's a good quarter. And when we look at the demand environment that we are operating in today, it's very different than historical -- we think of it as several factors that are driving demand. Clearly, what you said, there is a pulling component there's a buy ahead. Customers want to ensure they have access to supply. They're concerned about raising prices and they're acting. There's also a component of we have large installed bases, whether that be in PCs, where you have 1/3 of the units that are 4 years or older. We were lagging in a Windows 11 refresh and caught up through the quarter against historical refreshes, and we have a large number of 14G servers in the installed base that need to be upgraded as customers are moving to modernize. .

Customers are upgrading their edge, they're upgrading the infrastructure. They're looking for more capable PCs as Gentech workloads make their way to the edge. They're looking to consolidate space, power and cooling to drive efficiency. Our new 18G servers are a great vehicle to do that with its 13:1 consolidation. And we have -- we're seeing pockets of fundamental new demand. There's new demand driven by AI. There's an AI drag. There's inference. And the agentic AI is driving a new marketplace for traditional servers that we haven't seen before. And then I think the last 2 are what you would expect out of Dell are we're winning. We're taking share in all 3 segments. 4, if I count AI servers. So all 4 major businesses, PC, server, storage, AI servers, taking share, we're winning.

And then lastly, during these times of supply disruption and a lot of puts and takes in the marketplace. Customers tend to come to Dell to look for a calming hand looking for help, and we're certainly helping as many customers as we have. That's in PCs, that's in servers, that's in storage. So those are the demand levers of the demand dynamics that we see across all of the businesses. As we look at our forward-looking pipelines, the pipelines have never been healthier. They're actually growing at greater than historical rates, which gave us confidence to raise the guide by $27 billion of revenue for the year. I hope that helps.

And the next question will come from Mark Newman with Bernstein.

And congrats on a great quarter. we'll be glad to get a bit more clarity on the breakout of growth between units and pricing, particularly for the low outperformance we had on traditional servers and just like adding to some of the stuff you said in previous question, just trying to get a better sense for how much confidence you have that this is sustainable beyond just 1 or 2 quarters, but through this year and into next year.

Thanks, Mark. Very specifically, we grew units in PCs. The last reported quarter, we grew units in consumer PCs and commercial PCs and obviously, the aggregate market. We took share. So there's a baseline of growth there. We were already the industry leader in PC revenue. And clearly, the inflationary environment has driven up prices. And where we saw that primarily is on high price band products. and we were the market leader in high price band in PCs, and that expanded as prices moved up. I won't parse out the specifics, but we had a great growth quarter in units obviously, the inflationary environment we had the execution towards, for example, high-end and gaming and consumer. We had high-priced bands and commercial PCs. And don't forget the attached business. .

The attached business for us around peripherals and around services is very healthy when the base business grows, it drags more revenue with each and every unit. On servers, absolute server unit growth occurred. They'll probably kick me under the table here. We had significant unit growth in traditional servers. And then we had the content growth. We are continuing to see on a year-over-year basis more cores, more DRAM, more NAND placed in each and every server. So you have the uplift of more content. And then obviously, that content is growing as well in terms of the inflationary side. So absolute growth in units, absolute growth in the content driven by modernization and consolidation as customers are looking to upgrade and modernize their fleets and then we had the inflationary part.

The other part of servers that I think is important to call out is this notion of AI drag and seeing traditional servers move and take on AI workloads. -- those AI workloads we're seeing with very dense servers, making their way into the neo clouds into some of the more advanced enterprise users think semiconductor companies, big tech that are using it to actually drive some of the inference workloads and angentic workloads inside their environment. So that's how I capture it.

And the next question will come from Amit Daryanani with Evercore.

Congrats on a really nice set of numbers over here from my side as well. you folks talked about maintaining an appropriate prudence, I think there's the way you framed it, like maintaining appropriate prudence when it comes to your guide despite raising the back half outlook. If I sort of think about H2 versus H1 math for a second, right? I think the guide implies 48% of the revenues this year will come in the back half of the year. Historically, that number has been around 52%. Can you just touch on how much of that H2 drop you're expecting right now versus historically is from a pull-in versus you folks that perhaps just being a bit more conservative? And is that conservativeness coming more from lack a component availability? Or where could that level be?

You bet, Amit. I'm going to lead and David's going to punch the answer home, but I'm the problem. We have a supply issue. We are supply constrained in the second half. It is not a demand issue for us. .

Yes. I mean that's the story here. The demand continues to outpace the supply. That demand is broad-based, as Jeff said. So it's going beyond the GPU and there's more AI opportunities from a CPU perspective, traditional server in the PC. We continue -- obviously, these are complex designs. And as we go forward, we'll continue with operational execution to work with our supply chain teams, our go-to-market teams and our product teams fully execute and match up to the best execution wise, the supply that we have with the demand shaping that we see, the teams have obviously executed that tremendously in Q1. And we'll look to do the same as we head into the back half of the year. But the demand is there. As Jeff said, that's what we're looking at. We'll continue to look for more supply, we would like more supply, but the team will continue to go execute and go chase the pipeline that Jeff referenced earlier, which continues to be in very healthy shape.

And we'll take a question from Wamsi Mohan with Bank of America.

Very impressive set of results here. I guess if we think about the comment you made on the call, I think you said customers are prioritizing securing supply, and that is something that you expect will continue for the remainder of the year. in this kind of an environment where you just noted that demand is extremely strong. What's your take on the magnitude of the variation in IT budgets for this year? And do you think that some of this is coming out of the budget for next year as well on the enterprise side. And I'm also curious, on the traditional servers, I think, Jeff, you mentioned how agentic AI is maybe changing the usage. Is there a materiality of those to Tier 2 CSPs as well? You noted enterprise strength, but kind of curious if Tier 2 CSPs could be a potential offset to maybe any change in linearity of demand at enterprise.

I mean, Wamsi, I can't speak to next year and budgets for customers. I mean clearly, the longer-term conversations we're having with customers are multiyear in nature, of how they secure supply to provide their growth and upgrade their infrastructure and the discussions are multiyear in nature. Think 3, 4, 5 years. Those discussions are underway. And it really is about access to supply because quite honestly, I can't tell them what the price is going to be. But it's arrangements and agreements that we're working with large customers to ensure they have what they need to grow their businesses. And we saw that occur across some of the largest enterprises around the world.

And our pipelines indicate that's going to continue but the next set of customers and the next set of customers before that, which is breaking the historical norms of pipeline build within the quarter and the out quarters. Our pipelines are good 2 quarters out. And what was really interesting about traditional servers as an example, is we saw the pipeline grow in quarter greater than historical norms. And we saw the 2 quarters out pipelines grow and it's showing more customers looking to get access to the technology. We're seeing budgets grow. We're seeing budget shift. Obviously, we're 1 quarter into the year. We'll see how the second half plays out. I think that's part of the prudence that we're trying to convey of where is the demand signal. But the demands that we see continues to be robust. And we and the supply chain have to go find more parts for the businesses and for our customers to fulfill that demand. I hope that helps.

I'd add maybe a couple of things, Jeff. I think as we talk about demand outpacing supply in that environment, we would expect to exit the year with meaningful backlog as we enter next year. So I think that's an important point. The other piece would be from our financing and DFS element, which again is a competitive advantage for us. We're engaged with many customers who in normal course of business would not need to take advantage of financing offerings for the opportunity to partially use our facility to get as much gear in their environments in year to manage their potential budget issues is a great way to balance that, and we're seeing double-digit origination growth across our CSG business, our traditional server business, our storage business. as well as AI, which you'd expect. So we're seeing different ways we can help our customers get as much technology into their environments as quick as we can.

And the next question will come from Katherine Murphy with Goldman Sachs. .

I was wondering if you could further talk to the raised full year guidance for AI servers to $60 billion and where across your customer base, the 5,000 customers you mentioned you're seeing that $10 billion of incremental opportunity for the fiscal year? And as a follow-up, how much capacity can you support in the server space with your current manufacturing partners?

Yes. Look, a strong start to the year, Katherine, $16.1 billion of shipments, $24.4 billion of orders. Our backlog now sits at $51.3 million in just 90 days into the quarter, we're raising our full year guide by $10 billion. We'll continue to work through those deployments as we match up our supply. Obviously, these are complex designs that we're engaged on. We're actively involved in the technology transition as we get ready for Vera Rubin and obviously, working with all these customer bases in relation to data center readiness and making sure they can receive the product I guess, as you look at our portfolio, look, it's expanding and growing across all our verticals, whether that's neoclouds, our sovereign relationships, our enterprise customers, -- you would have heard Michael last week at Dell Tech World talk about our 5,000 customers, which is up over 50% in the last 6 months.

So you can see the traction that's coming. And I guess the other data point I'd add is as we look at our pipeline over the next 5 quarters, that's multiples of our backlog, and it's growing across each individual vertical there again, across Neo clouds, individually, sovereign individually and the enterprise space. So it's broad-based and it's present either geography-wise or vertical-wise.

On Capacity, yes. I mean we have the capacity but there's no capacity issue. It's parts, supply.

And the next question will come from Samik Chatterjee with JPMorgan.

Congrats from my side as well on the strong results here. Maybe just to focus on 1 specific comment that you made in your prepared remarks where the gross margin outlook for the year, I is better than you had 90 days ago. If you can just flesh that out a bit more in terms of is that a function of price increases that you've been able to take? Or is that a function of the kind of sort of mix of products that you're now selling or where customer demand is focused relative to what you envisioned 90 days ago. Just curious to hear what's driving that better outlook there?

Yes. Sure, Samik. I think it starts with our Dell IP storage portfolio. we've taken up our revenue guide not only for Q2 but also for the back half of the year. We're seeing our Dell IP portfolio resonate in the marketplace, whether that's the in structured product from a power store perspective in the midrange. All of that obviously drives from a Dell IP mix perspective, tailwinds from a margin perspective and a rate perspective. And the other elements of the business, both CSG and traditional server, we've made the commitment to make sure we sustain our margin rates. We see a path to that. We will continue to manage that as we see the growth. And you put that basket of goods together from a core business perspective, and you see the lift in the overall margin rate as a result. .

And we'll take a question from Saiya Merchant with Citigroup. .

Great. I'll add my congrats here too. If we could just talk a little bit about the attach rates that you're seeing for the AI servers, especially as you're talking about radar enterprise demand here. You talked about 5,000 customers now that's up from where it was a quarter ago. If you could just flesh out how you're seeing that attach rate for storage and services? And is that -- is there a recent uptick in Dell IP storage a function of attached to AI servers as well as on services, if you could just comment on that and how that fleshes out into the margin outlook for AI servers? I think I heard maintaining a mid-single-digit margin outlook for AI servers.

Sure. A lot is into the question. We'll start with the broader topic of we are increasing the amount of storage and services that we're providing AI customers. Period. Michael made reference at Dell Technology World last week with our unstructured data solutions, the portfolio of products and how we had won across several major customers, which I think are important bellwethers of what's changing in the marketplace and how our products are being perceived. We are making progress with our AI customers, neoclouds, sovereigns, the high-frequency traders and some of the biggest technology companies in the world, semiconductor companies in the world, where we are selling more storage, more Dell IP stores, in fact, only Dell IP storage. And you're seeing it in our unstructured portfolio of products had its best quarter in demand ever. .

Unstructured data is the dataset that feeds to be sort of speak in AI, and that's where we're seeing the greatest growth and making the most traction. If you think about the portfolio broadly, David mentioned 5 straight quarters of growth of Dell IP. It's 5 in PowerMax, it's 9 in power stores, 4 in power scale. It's 3 an object scale, and our data protection product is 2. We are seeing the entire portfolio gain momentum. A combination of more competitive products, products designed for the AI era. I would point to Lightning as being an example of an AI parallel file system, specifically designed for this class of devices and customers. We're seeing increased traction. We're certified across NVIDIA stack, which certainly is driving. We are engineering with them and how to make data ingest in data management and the whole data estate easier for enterprises to adopt to accelerate their AI needs.

And quite frankly, we're in an error where architecture matters more than it ever has. I think of our PowerStore Elite product would get really excited. We had fun with this last week at Dell Technology World. It's got 3x the performance of its predecessor, 1.5 million IOPS, 6:1 data reduction. Buy 1 petabyte of raw storage, store 6 petabytes of data. I think about its 70% faster in REITs has got 4x more throughput. I think about the exascale storage we built purposely for this class of customers. I think about the rack scale architecture that Arthur talked about last week on scale where we are on stage where we talk about the role of storage, networking compute coming together, driving more performance. I think about what's happening in the world of data protection architecture matters again in 75:1 compression rates.

And then ultimately, our fundamental architecture that drives fewer servers and fewer SSDs to store quiet amounts of information versus our competitors. All of that is being packaged up and presented to our entire customer set and then specifically targeted to our AI customers. Again, whether it's sovereign, whether it's neocloud or an enterprise, and we're seeing traction. Optimistic, not claiming victory here. We have a lot of work to do. We're committed to the space. If you look at the payload that we delivered at DTW last week, it was the biggest and broadest storage payload we've ever brought out at any given time. and there's more to come.

Using AI inside our R&D organizations, we are delivering larger payloads and shorter periods of time and storage is the primary vehicle to deliver that through. So I hope that gives a sense. And obviously, we're still seeing 1 of the differentiators we have in the marketplace is services our ability to deploy service product, keep up times greater than anyone else, continues to be a differentiator in the marketplace, and we'll continue to invest in that broadly across all customers.

And we'll take a question from Erik Woodring with Morgan Stanley.

A big congrats from me on the quarter, just an amazing result. I realize we're having a lot of conversations here about sustainability. But I'd love to maybe ask you, if we could go back to last October and knowing what you know now about the market in incremental agenetic and ways that traditional servers are being used perhaps in different ways and your ability to take share from peers. If we went back to the October Analyst Day. How would you change that 7% to 9% revenue guide and 15% plus EPS guide? And ultimately trying to get understanding the sustainability of what you're seeing across multiple years. We just I realize you might not have numbers, but would just love your thoughts on where maybe that 7% to 9% and 15% plus would go, knowing what you know today.

Yes. I don't think we'd work a 5-year program on the Q1 earnings call, obviously, as we kind of go through. What we'll do is obviously validate what we're seeing. We were very keen on the back of the Q1 momentum that we see where the growth is real, it's durable. It's accelerating. It's more broad-based. It's expanding beyond the GPU. All of those proof points as they evolve and emerge give us and gave us the confidence not only to take up our Q2 guide, which pretty much mirror images what we did in Q1. But also look at the second half and build out incremental guidance across every lab, whether it's PCs, servers, storage and AI as we do that. Jeff touched on earlier, we're always pretty confident as we look out over a 2, 2.5 quarter lens in our pipeline as we do that. There are obviously dialogue there as we talked about an AI pipeline over the next 5 quarters, building out multiples of our backlog.

So all of those, again, indicate that strong reference points of broader-based demand element and obviously, core to that, it will be the agility for our EPS over time. So that said, like I said earlier, we'll look to drive meaningful backlog as we exit this year and I think that's where we are, Erik, in terms of looking out any horizon for now.

Eric, I might add the following perspectives. I don't think applying historical models or historical views about the market and how it's going to act or appropriate today. we're finding new uses. I mean, the way that I get asked this old ask you this, is what's the value of adding intelligence into every workflow, every decision, every product every customer interaction. I would assert the value is pretty darn high. And that's what's been really, I think, the game changer since that October time is what's really happened in Agentic. And what you're seeing are new categories of TAM expanding, you had the 3 microprocessor leaders talk about an expansion of CPU TAMs. Why? It's driven by agentic. What's happening in agentic is really the movement of AI from an adviser to an operator. It's actually going to do something now. It's going to do something meaningful. -- to do something that agent needs support, just like a human need support. That agent needs support. In this case of a CPU, you have all of this wonderfulness that a GPU drives. .

But you have this work that has to be done around I/O, around branch, retries, managing state. They're very sequential. They're very serial in nature as a result of that. That's a workload that's for the CPU. So if you think about this notion, that's generally called the harness. So if you think about that harness, the CPU runs it. It's going to make those calls. It's going to manage memory. And it's in the loop and every decision that an agent makes. I didn't -- we didn't know this in October. This is a completely new marketplace. That's being driven by putting intelligence in every workflow and every part of knowledge work on the planet today, and we're just beginning.

Another way to describe this is -- the premium for computational capability, whether that be on the edge with a PC, smartphone, servers running this harness, GPUs doing magical, wonderful work creating all of this great value just continues to grow at a rate we've never seen, and it's pulling the rest of the ecosystem. That's what we see. And if I go from the trifecta here, all of that stuff has got to be stored. It needs high-performance storage to be able to ultimately have a receipt of what the agent is doing. So it can be corrected. You can understand what it did. That's where we're at. I don't know how we would have predicted that in October. And today, I can't sit here and tell you how big the TAM is other than I know it's bigger, it's growing, and we're in the early innings of it.

And the next question will come from David Vogt with UBS.

Jeff, I appreciate all the detail and David on servers and storage. I want to ask a question on CSG. Obviously, strong performance, taking market share. can you expand on sort of how you drove profitability dramatically both sequentially and year-over-year? Kind of going back and thinking through like the best margin I've seen in sort of the PC industry, it was probably not 8%. So just given sort of the drop-through, it looks like over 25% drop through, how do we think about what's driving that? How much is price versus maybe low-cost inventory? And how do we think the PC margins trends longer term? Do we go back to your normal historical long-term range that you talked about at the Investor Day? Or just how do we think about kind of where the market is in your competitive positioning from a pricing and margin perspective.

Look, Dave and I will tag team this. I think the way to look at this, clearly, we benefited from tremendous scale in the business. mean David made reference that the operating expense as a percent of revenue was down, I believe, 300 basis points on a year-over-year basis. .

Actually a sequential basis Yes, $600 on a year-over-year basis, on a sequential basis, $300.

Well, that's powerful in a business like the PC business. We also told you in our last earnings call, we were purposely late in making a price move because we wanted to build momentum with volume we did. We took share in Q4. In Q1, we purposely moved the price in earlier as we got our Q2 cost. I think about what we're doing today, we probably move a little too early in retrospect. We saw the temper a little bit of demand in the transactional business, I think a consumer, small and medium business. And we're looking to find the right optimum place for that, which is reflected in our go-forward guidance of operating margins for the PC business.

I mentioned we had TRU uplift, TRU uplift drives more profit. I mentioned that we had greater peripheral attach and service attached that drives profitability of the business. That's the package. We are not operating at COVID margins. far from it, in fact. And if you go back to the operating margins in that era, they were at this range or slightly better. We're benefiting from the tremendous scale of the Dell Company. -- a discipline in pricing that we're working to find the right optimum balance, particularly in that transactionally oriented side of the marketplace, consumer, small and medium business. Large deals are done priced deal by deal. And -- we like what we're doing. We think we don't have it perfect yet, still trying to find the right balance, but I'm optimistic.

And the next question comes from Tim Long with Barclays.

2-parter, if I could, hopefully, both quick on the more traditional enterprise business. First, you talked a lot about storage. Just curious with traditional server guided to 60% for the year and storage mid-single digits. Does that mean that we could see a longer pull-through or tail to storage as we look out a little bit further? And then secondly, you guys navigated the price increases very well. You've touched on that as well. Just curious, in your past history with this, if we do get another uptick that's kind of meaningful in the next several months or quarters, -- does it get harder to push pricing through another time? Or is it similar to the dynamic that you think you've seen over the last quarter or 2?

Yes, Tim, I guess a couple of things in there. First, if you look at our guide for the full year, again, I go back to the dynamic that we kind of referenced earlier. When you look at our second half growth in the plan, it's still, if you like, inhibited by the supply that we can get. So the demand is there. The demand is outpacing the supply that applies to across our ISG business as we look at that. The other element, and I think we've discussed this 90 days ago, as you look at our Dell IP mix in terms of our storage portfolio, we continue to do that crossover with the historical business. So we get more dell IP versus third party.

By the end of this year, that stops becoming any relevant element of our bridge in relation to that. So seeing growth in storage on a consistent basis and building that trend is something that excites us from a P&L perspective as we move forward and as we kind of go execute that piece of it.

On the pricing side, Tim, we're repricing, it feels like every day. And I'm sure our customers feel that pain. Unfortunately, I don't see that changing given the world that we're living in today where you have an inflationary environment, whether it's fuel, whether it's raw materials, whether that's DRAM, whether that's NAND, CPUs, we are living in an inflationary environment that is changing at a rate that, obviously, we've never seen before. And everything that we see suggests that continues. There will be a point where some customers, it's enough and they'll wait it out. And we're seeing that in some cases. .

In other cases, we're seeing an acceleration, the notion of that was called out earlier, where folks are trying to secure that supply now and over multiple years because it's going to be more constrained.

And the next question is from Simon Leopold with Raymond James.

Great. Appreciate it. I wanted to come back to the risk and the supply constraints in that I think everybody understands memory at this point but I'd like to get a sense from you as to what other elements or factors are limiting any upside beyond the memory constraint? And I'm thinking about things like printed circuit boards, et cetera. Just help us understand sort of the rank orders. Appreciate it.

Sure, Simon. Yes. I called out the 3 that we're spending a tremendous amount of time on. Obviously, NAND and DRAM microprocessors if you went down the list, next, it's likely hard drives. If you go down the list beyond that, there's lots of things. If you look at what's happening in the semiconductor network, you're seeing utilization of the trailing nodes beginning to fill at greater rates. Leading-edge node stuff is full as fully allocated. Lead times are a year. So all of those are pressured, but the most pressure comes across the 4 that I described and the first 3 primarily DRAM, NAND, CPUs. And then hard drives and then ultimately, the basket of goods that sit around that. Our supply chain has clearly worked through this.

This is what we do, never run out of parts, got a sales force that's out selling lots with the demand and pipeline that looks very encouraging that we tried to convey through the call. We have our work cut out for us to work with our partners to drive more supply and every bit and bite matters. Every microprocessor matters, and that's what we all try to do every day.

We'll take 1 more question. We'll take our final question from Krish Sankar with TD Cowen.

Jeff, again, congrats on amazing results. I just wanted to find out on your servers, -- is there a way to think about what is the mix of x86 versus on? And does it matter to you? Or is there any margin differential between those 2 architectures from a Dell standpoint?

Traditional servers are x86 today or excited about the opportunity with Vera in the future, particularly as we talk about these advanced workloads that drive increasingly more apputational need running this harness the CPU managing the scaffolding around every GPU call is going to be more performant and there'll be more choice here more opportunity. We need the relief of microprocessors. So we're excited about that. On the GPU side, if my memory serves me right, it's biased towards ARM. So when you think about the big GB 200, 300, obviously heading towards Vera, you think about direct liquid cooling the large deployments bias towards ARM.

If you think about enterprise and air, so I think B200, B300, RTX-6000 Pro, you think those x86.

Jeff, we'll turn it over to you for the close.

Thanks, Paul, and thanks, everyone, for joining us today. Q1 was an exceptional start to FY '27, highlighted by strong execution across ISG and CSG and continued momentum in AI. As we look to Q2 and into the second half, our pipeline indicates demand is not slowing but accelerating and meaningfully outpacing supply as customers prioritize securing the infrastructure they need across AI, traditional compute, storage and PCs. Reflecting that strength, we raised our FY '27 revenue and EPS guidance by approximately $27 billion and $5, respectively. We are operating with discipline in a challenging supply environment, scaling the business and continuing to return capital to shareholders. We feel very good about our position, our momentum and our ability to create long-term value. Thanks again for joining us today. .

And this concludes today's conference call. We appreciate your participation. You may disconnect at this time.

Please welcome Michael Dell.

Good morning, everyone, and welcome to Dell Technologies World. Thank you for joining this extraordinary community of innovators, dreamers, builders and leaders. Together, we are driving one of the most important transformations in history. Abundant intelligence is here. It's not coming. It's here right now. Intelligence is becoming infrastructure. And just as electricity transformed the world when it left the power plant, AI will transform the world when it leaves the screen.

We are making intelligence real. local, secure and useful on the oil rig, in the ambulance and on the factory floor. Dell Technologies is building the distributed infrastructure that turns isolated insights into intelligence in action. And we are opening the door to a new era of progress across every field of human endeavor. The models are smarter than all of us, and they're improving exponentially.

Since ChatGPT was launched in 2022, we've gone from a chatbot that could write a decent essay to self-improving AI agents that write code. They run workflows and they operate around the clock. Coding is now AI native. You describe it, verify it and ship it. And the barrier between imagination and execution is collapsing, and this is unlocking human creativity on a scale we have never seen before.

For organizations, AI is no longer a feature. It's becoming the operating model for the modern enterprise. And that is where we are going to spend our time together this morning on the journey through the modern enterprise. All right. So here we are in Austin, Texas in our boardroom, where modern enterprises begin with a conversation and a commitment and a vision for what comes next. It's where we learn about each of you about your journey, your opportunities from health care to manufacturing, to finance, to agriculture, from small businesses to large enterprises and governments around the world.

This is buying all-time favorite visual. I use it every year, and I'm just wondering, can we make it just a little bit bigger? There we go. This is what I am most proud of. It's the trust that you place in all of us that we take so seriously and the relationships that we've been able to build and the work that we're able to do together. It's incredibly meaningful. This is where trust begins and where our partnerships start to reinvent industries. AI becomes real when it shortens the distance between discovery and impact. And when you're at the frontier of modern pharma, the stakes could not be higher.

So now let's look at Eli Lilly, where science becomes life-saving medicine. Please welcome Diogo Rau.

Diogo, great to have you here. Thanks for being here. And look, I mean, Lilly has an incredible history and story of remarkable scale and impact in the world in medicine over many generations. And you've been at the center of sort of bringing this super important company into the modern age with the latest technology.

Tell us a little bit about how technology is shaping Eli Lilly today and what that's look like.

Well, fantastic. And what a journey it has been because Lilly actually just celebrated a birthday party on Thursday last week. We turned -- thank you. Really don't feel like a day over 140 years old. But Lilly is not a tech company at our core. We were started by a kernel from the civil war in the United States who saw people needlessly dying on the battlefield, not getting medicines that they need. And we're actually getting sham medicines, and he had this crazy idea that why don't we actually go use the scientific method figure out what medicines actually work and then figure out how to make them at large volumes.

So from the beginning, really was always a company about scale. So in 1923, when insulin was discovered, we had a breakthrough medicine that could save so many lives, but nobody could figure out how to make it at volume. That's where Lilly came in or actually penicillin in 1928, as when it was discovered. We needed it for World War II when people were dying on the battlefield. U.S. had ramped up supply by 1942, 14 years after penicillin was discovered, guess how many patients the United States could treat with its supply?

I don't know. 10.

Not 10,000.

That doesn't sound good.

No, it doesn't. That's definitely not good. So the United States government called in Lilly and said, "Hey, can you do that thing you did with insulin? Can you do that again with penicillin?" And that same thing happened again with the Salk polio vaccine. It happened again with the first COVID antibody. It's happening again today with GLP-1s because medicine isn't just about discovering something new. It only works if you get it out in mass numbers to help the people that they need.

The only way you ever get scale in medicine is to actually have technology underneath it all. And so that's why we have always been huge technology adopters at Lilly. Early in bid, we were one of the first to adopt the IBM mainframe back when it took up a space probably like this size to run. And we were one of the first adopters -- actually, we're the first commercial adopter of the Cray-2 supercomputer. And this year, in February, we introduced when online with a pharmaceutical industry's largest supercomputer. We've always gone early, and we've always gone big because technology is how you get that scale for medicines.

And give us a sense for what the impact is. How AI is making a difference today in Lilly?

Sure. Well, it makes a difference across the board, of course, because I think if we took it away from people, they would quit right now, right? Nobody wants to go back to the pre-AI days. But one of the exciting places, I think, is in manufacturing we probably don't talk enough about what happens in the world of manufacturing.

To be able to scale up all these medicines.

Exactly to get that scale of medicines again. And one of the things that's really cool that I like is when the medicines come down the filling line, they're like a social media influencer getting glamor shops the whole way from all kinds of different angles. It's not like 70 different photographs with budgets of milliseconds to detect. It's something that's super cool or actually...

The optical inspection is mature quality.

Yes, exactly. Because we have humans do that and that's what humans mess up. Like these machines, they don't. And so it's amazing what you can do. Digital twins, I mean, how long have we been talking about digital twins, that's been like forever.

Now we are actually really doing it. And we've taken processes that humans have said are completely fully optimized. And what we've done is we've replicated the machines that are in there as digital twins move them around. And guess what the AI the humans. It's not completely optimized. So manufacturing is just a fascinating world. But the thing about it is, we're so dependent on technology now that it just can't go down. That's the difference about this technology revolution from anything that we've ever seen before.

And that's why we've been very happy about our partnership with Dell because -- when we open a new site, we don't have to go out and figure out what we need to do. We've already got a plan together. We already have the architecture. We already have the stack, everything has been validated, and it just works.

It's copy exact.

Exactly.

So that's the operations side. What about drug discovery and research and accelerating really the plot Eli Lilly at its core?

I love the challenges in research and discovery because that's another kind of scale challenge just like we have never seen before.

I mean I think -- when we think of labs and if you're not in this world, you picture people running around in white lab coats with notebooks, taking notes and maybe looking for a microscope. Now today, a modern lab, it's all technology in there. And every single one of those things is connected. Scientists aren't looking at some fuzzy image and trying to make something out. They've got 4K resolution, 120 frames per second of whatever they're taking. Everything in that lab is just showing us with data, petabytes of data as it goes. And it's fascinating, it's overwhelming, but at the same time, that's the data that we need to use to power AI. That's the whole reason we have such high-performance compute needs.

And what happens in a couple of years when you get orders of magnitude, more computing power to the research and drug discovery at EI Lilly? I mean how do you think about the effect of that on cancer and other sort of challenges that exist in the world?

I love that because we're -- today, we're already there. We're solving problems that like -- that we weren't thinking about 10 years ago. The problem was of 10 years ago are just frankly boring today. So everybody knows about AlphaFold, right? Super cool. You figure out genes and you figure out what protein would come out and what shape it -- that makes Well, you know what, that's boring, Nobody really cares about that now. It's like, who cares what the protein looks like, what is it going to do for me? That's what you need to know, like, okay, so you've got a protein now. How is it going to interact with another protein.

They're all moving rapidly in time and space, orbiting each other, bumping into other molecules. That's how we're actually modeling things that are much more advanced or we're -- and by the way, that's actually factor cancer and things like that. That's actually how you build something that's so selective. It can only attach to one class of cell. And if you can do that, you can drop off a payload, chemical or nuclear actually or other things like that to kill a very specific cell. That's the kind of stuff we never could do before. It's just an absolute breakthrough of what we have in technology now.

So another example where we live in the most incredible times and RNA vaccines, gene editing, all these things are incredible. So tell us a little bit about how Dell is supporting all these efforts and sort of what it means to be able to have the technology to power this.

Well, it all the scale, everything I've been talking about only happens with technology and all that scale is actually what we are getting from Dell. We are still proud of our AI supercomputer that we have or...

Is it 1,000 GPUs?

1016. The newest generation. By the way, if you haven't seen 1 of these things, it's crazy there. Like it's 105, 110 decibles. I clocked the wind speed coming off the fans at 30 miles an hour. It's like you're...

Who's trying trying to make them quieter?

They're amazing. But the Dell infrastructure underneath is actually what's powering all of that. But I want to make it -- I want to make it clear because it's not actually about the hardware. What we did first is we tried this out in a couple of smaller versions. And Dell was there with us as we were learning. We weren't just learning about how to set up the hardware.

Dell was actually helping us with the whole management process. And what we really -- what we really benefited from was not the models that we trained but actually the training we got from you. And so I really would love to say thank you to you and to all of the Dell team members because that's actually what helped us get the skills that we need to be able to build the supercomputer that we have today. And that's how we're going to make the kinds of breakthrough discoveries that we've been talking about. Because you know what, in 150 years from now, we're going to have new things that we can't -- we're going to have breakthroughs that we can't possibly imagine.

I think we're on the verge of maybe being able to end disease as we know it. I mean something like that was completely unimaginable 20 years ago. But today, we can imagine it. And I think it's what you said at the beginning, -- now the challenge is how do we shorten that distance between the discovery of the imagination and the execution.

Diogo, amazing. Thank you for a life-changing partnership and capabilities that we're really privileged to be a part of. We're inspired by everything that you and Eli Lilly are doing with technology. Thank you so much for being here.

Wow, more from health care soon. So thank you, Diogo. It's -- what you just saw is not a chatbot. It is intelligence in the physical world. It's a faster path from science to medicine. And now every organization faces the same challenge. How do you turn intelligence into impact at speed. And I hear it from the CIOs, the CTOs, engineers, operators, CEOs and boards. It's coming from the top down. I can tell you it certainly is a Dell. I see Jeff Clarke. That's because we understand how AI native companies work, how they operate, what they look like and what they're truly capable of.

This is an existential moment. The companies that redesigned their work around AI are going to compound advantages faster than at any time in history. And every leader in this room feels that pressure right now. It's exciting. It's exhilarating. It's sometimes exhausting, but it's also unavoidable. And I want you to know that you're not alone. We are by your side with the expertise, the global supply chain and the execution to deliver when it matters.

The supply chain is modernizing too. And some of our most important partners and suppliers also are our most important customers. For our next modern enterprise, let's look at the intelligence behind something we all rely on pretty much every single day, modern semiconductor manufacturing.

[Presentation]

In companies like Samsung and across every industry, AI is accelerating from proof-of-concept into production. And it's flipping the traditional buy versus build equation. For years, the trend was off-the-shelf software and public cloud. But now AI is collapsing the cost and time to express your competitive advantage through software and so guess what, there's going to be way more software everywhere.

A recent Dell technology survey on AI adoption shows that 67% of AI workloads already run outside the cloud either on-prem, on device at the edge or in a colo and 88% of respondents are running at least one AI workload on-prem. The CIOs are aggressively pivoting to hybrid AI.

The risk is not the cloud. The risk is losing control of your data, your cost, your security, your intellectual property and your speed. In the agent era, walk-in does more than slow innovation, it actually limits what your company can become. As soon, every company will deploy fleets of agents, composing workflows on infrastructure that they control. It's estimated that the worldwide AI infrastructure spending could hit $3 trillion to $4 trillion by 2030.

AI is fueling a renaissance in enterprise hardware, a shift from bits back to Adams. The question is, how do you deploy the world's best models, where you need them with security and governance built in. We give you model choice without infrastructure chaos, you get the open models, the frontier models, the specialized models running where your data lives, available starting today in the Dell Enterprise hub on hugging face representing major architectural leaps across the AI ecosystem, new models are available, including MiniMax M2.7, DeepSeek-Pro, DeepkSeek-V4, GLM 5.1 and Kimi K2.6. These joined the Gemma 4 family of models, the NVIDIA Nemotron, Super 3 series, the RC Trinity large thinking and mistrial small.

In addition, ServiceNow customers will be able to leverage the Dell AI factory to operationalize, their AI-focused business outcomes. And we have some very big news, you'll be able to bring the latest frontier models to your enterprise data on the Dell AI factory. Google and Dell are bringing the Google distributed cloud with the Gemini 3.0 models on-prem with the PowerEdge XE9780 servers.

You can run advanced AI workloads in a confidential computing environment using the new Gemini models. And open AI and Dell Technologies are building a solution based on open AIs codex to bring their latest agentic harness on-prem with the context provided by our Dell AI data platform using a set of connectors and skills. You can securely connect to the latest GPT 5.5 and GPT Codex models. And SpaceXAI and Dell Technologies are delivering rock advanced reasoning and multimodal capabilities at secure, enterprise-grade AI systems, deployable on-prem or in a hybrid approach.

Palantir's foundry AI platform are also coming on-prem with ontology deployed on our ObjectScale and PowerFlex platforms, you can connect all your data sources and optimize operations with the full weight of AI connected within the boundaries of your own environment. And reflections open source frontier AI models are now coming to the Dell AI factory.

Regulated industries, including governments and sovereign entities to deploy AI in a fully controlled environment. Look, every nation is waking up to a new reality. AI capacity is a strategic asset like energy, communications and secure supply chains. And we're delivering the secure air gap nationscale AI infrastructure without lock-in and without worsening your data into somebody else's black box.

AI can become the most concentrating technology in history or it could become the most democratizing technology in history. We are choosing democratizing. Powerful AI wherever it's needed, open, secure and yours. This is hybrid AI. It's not a compromise. It's a competitive advantage running at your Dell AI factory.

Now how big is the demand for intelligence, it may well be the largest market that's ever created. By 2030, estimates are that token consumption will explode, increasing by 3,400%. And as organizations scale their AI workloads, there's a new economic reality, cost curves, compute demands and data movement patterns are all being rewritten. AI is not just changing technology, it's changing the economics of technology in favor of enterprise infrastructure.

And now is the time to decide how you can most cost effectively generate the tokens that you're going to need for the long term. We now have 5,000 customers running their Dell AI -- their workloads, AI workloads on Dell AI factory, turning their AI ambition into scale production from data to tokens to outcomes. Customers turn to Dell Technologies for the design and innovation, the engineering, the supply chain, the services, the support and the financing, all working together.

Time to first token is incredibly important with investments of the scale. And we can deliver hundreds of AI racks a week to a given customer up and running in hours, generating outcomes immediately. Now manufacturing, particularly advanced manufacturing is where AI stops being theoretical. And it has to work with super complex and precise machines, sensors, supply chains, safety systems and make real-time decisions. Join me in a visit to modern industrial innovation. Please join me in welcoming Suresh Venkatarayalu.

Suresh Venkatarayalu of Honeywell. Suresh, it's so great to have you here. When some people think of Honeywell, they may think of the Honeywell for in the past, but we know Honeywell is a different company today. modernizing and far beyond the roots in industrial controls, Tell us a little bit about what the modern Honeywell is today.

Sure, Michael. Before that, true to be here. This is my first ever then being with you right on stage, fantastic to be here. Honeywell under the 140-year-old company and one of the most important transformation we are going through at this point in time. We are positioning 3 independent companies, advanced materials. We have already spun them off, which is called Solstice. We have coming up with aerospace and then new Honeywell company, which is going to be a pure-play automation company.

And then what do we do? It's a controlled system company, which runs industrial oil and refineries, commercial building, utilities, data center, we do the best, connecting to their installed base, assets, controls platform. That's what we do, optimizing and building application all along 140 years of automation. Few things that's happening right now and why I'm so excited about the future Honeywell is the customers are looking for improving their throughput much bigger and better. Not only that, their systems are fragmented, siloed systems all over the place. What -- the current technology and our exploration and innovation is going to -- there's an opportunity. So we have come back and build this whole Forge as a platform to integrate their entire backhaul to optimize their operation. So the future about automation autonomy is going to be so real, and I see the possibilities for us to really serve them better.

We've been working together for decades, and we've been honored to be a part of your systems. And now tell us about Forge and what capabilities that's going to bring to customers, what is it going to enable? How are they actually going to see it in their environment, be meaningful and impact them?

Michael, I'm going to be breaking down into 2 chapters. Chapter 1, 2016 to 2024, I would say. We will forge like any IoT platform, connecting to our installed base improving the visibility of our operation assets and we built enterprise SaaS-based software application. I think we brought in a lot of improvisation in terms of service upgrades, software upgrades and we brought in what they call visibility and transforming their operation. Chapter 2 for the last 18 months, we said it's time that we actually not only have a connectivity tissue to the operation, but you have to contextualize and understand assets, processes, systems much better.

I think with an advent of all the foundational models and AI at scale at this point in time, we are venturing into what we call a Chapter 2, which is all about you connect your operation, you have to contextualize process data and system. And then if you are able to simulate and emulate you have a better prospect in optimizing the operations. So we believe that time has come in to drive something at scale.

Let me give one example that will give you a context. We work with a number of customers across the board. Let me take an oil and gas customer in Middle East. We have done for the last 10 years. Having our Forge platform, connecting to their assets, we would predict their uptime, downtime, any downtime for an oil and gas refinery remains impact to their throughput in year, we've done a lot more work right there. But we believe that there's a second phase, which is all about contextualizing all types of assets that are close to 100-plus asset types your ability to contextualize asset characteristics and names in terms of oncology and then taking their process units and contextualizing them, can you predict the yield. Can you predict the throughput, assume that if you have a feedstock availability, can you predict what those outputs would be. That's pretty much what we are venturing at this point in time, Michael.

It's in a way analogous to the entire sort of AI macro theme that's going on in the sense that in the past, all these things were individual silos. But now you're bringing them all together with data and intelligence and helping drive yield and safety and better outcomes. And how is Dell engaged in this and give us a sense for how we can help in making all this real?

There are two things. I think we partnered with you and NVIDIA in the last year or so. We said there are two things that's required. One, you've got to extend your intelligence beyond cloud and SmartEdge. You got to extend it across your enterprise network. So we actually believe that introducing this concept called a physical server at each site level, you have an ability to really balance out your intelligence workload across the board because the decision-making for any autonomous future, latency matters, data transaction matters, so we started creating this concept call, Forge cognition with Dell and NVIDIA.

I think we're happy to say we are piloting at this juncture, getting great results at this point in time. For me, partnering with Dell and NVIDIA is not just about getting an infrastructure, high-compute services, but this whole AI stack which is a combination for me is right from an ontology simulation, getting this entire model, as you talked about the there's a proliferation of models and then getting an agentic AI platform that you could run at scale, secure and ability where our customers can trust.

I actually believe that we are going to be bringing this whole industrial AI at scale that is so essential to transform whatever we're talking about, which is automation to autonomy.

Awesome. So customers are going to feel something very different. Sort of give us a vision of 5, 10 years from now, what does this likely look like for 1 of your customers?

The first thing that I would actually say is autonomy for us means it's not about taking human beings off. We still believe there are untapped potential in every operation we touch. I'm sure there's a pressure from our customer side. improved throughput, help us to expand the margin. So we believe that the new age of AI industrial, AI is going to help the workers perform their task much better.

The assets to extend their life cycle to their peak performances, the operation processes, we expect that to actually be expanded. So I think we have brought in a category at this juncture where we believe it's no more connected intelligence systems. Our operation is going to be a self-learning and evolving operations of the future, which we believe that we have an opportunity to co-create along with Dell and rest of the partners moving forward in the world of industrial side, utility side and the commercial building side, and we are extremely happy and thankful for everything that you all have done.

I'm sure I said this in the past that Dell had helped us to really deploy our control systems for the last 20 to 25 years, Experion EBI in Niagara, which we call a Wintel strategy. And then now you bring in a GPU accelerated and extend this across the value chain. I don't integrate something interesting for the world.

Fantastic. Thank you so much, Suresh. -- it's a great -- it's really a great honor a privilege for us to be part of the solution that you're delivering to so many important customers and activities around the world. Thank you.

Happy to partner.

So not long ago, AI meant assistance that could write faster and summarize better and answer questions. But that was sort of the age of 20% to 30% productivity gains, it was valuable and kind of amazing, but really only the beginning. Now we're deploying agentic AI, autonomous agents that plan and reason and execute and adapt and close the loop. And agents are not something you just simply bolt on to your legacy systems. That's not going to get the optimum results.

They are digital workers. They have memory and credentials and access and the ability to take action. And this requires a new architecture for work itself, trusted data and governed action and infrastructure close enough to be able to make real-time decisions. The old systems and processes and the patient decision-making from the past are no longer enough. And it is time to completely rethink and reimagine your company's workflows for agenetic automation and for recursive self-improvement.

And that is going to lead us to gains of 20 and 30x in terms of productivity improvement. So who gets there first will rapidly distance themselves from all the rest. And the companies that do not become agentic AI-driven businesses, I think they'll struggle to survive.

There is a massive AI investment boom that's already underway. And a productivity boom is beginning. And in some companies, including ours, I see you, David, it already has. The rate of change has gone parabolic, and it's not slowing down. This creates monumental opportunities, and it also creates new responsibilities. The agents have credentials. They have access and autonomy.

Security must expand beyond the human users to also govern the nonhuman workers operating at machines feed. If a bad actor can influence an agent or if an agent is trained on bad data, the blast radius is no longer contained to just a single system. It can propagate across all your workloads and your infrastructure and, in fact, the business itself.

This is already changing the boardroom conversation. It's not just are you secure, but do you understand what your systems are doing on your behalf. You can't protect what you can't see and you can't manage. So we give you visibility and control, no matter how fast these threats are evolving.

Zero Trust principles keep you adaptive and resilient and are built in from the endpoint to the core. Speaking of the endpoint, you can see this into the deepest layers of the PC with our Dell trusted device platform. You can store the key credentials. In fact, the keys to the kingdom, if you will, in control vault so that only authorized identities, whether they're human or machine have access to your systems.

Agentic AI is only as good as the data that it can trust access and act on. And if your data is siloed, your agents are blind. They can deliver the value that you expect. Everyone has access to the same models. The differentiator is your data, the unique proprietary hard one knowledge inside your business.

The agents are hungry for your data. Now imagine tens of thousands of them going to millions of them, running 24/7 thousands of times faster than any person. The Dell AI data platform with NVIDIA has been engineered to address the data bottleneck. Within the platform, the data orchestration engine is the intelligent control center that terms that raw fragmented data into production-ready AI. You can transform and clear your data with NVIDIA's close to now 1,000 models and Q2X libraries integrated natively and you get 12x faster vector indexing, 6x faster data querying at 19x faster time to first token. And we keep pushing the limits of what's possible.

Today, we are super excited to announce that we are working with Starburst to enhance the analytics engine with NVIDIA with their vera CPUs for 3x faster SQL and with Blackwell GPUs for 6x faster SQL, leading institutions like Bank of America, who already have a partnership with Starburst, NVIDIA and Dell Technologies are going to be able to use these new capabilities to support AI-driven analytics while meeting they're important governance, regulatory and resiliency requirements.

And for extreme scale AI and high-performance computing environments, we're introducing Dell Exascale Storage a unified RAC architecture supporting power scale, object scale, PowerFlex and the Lightning File System. Exascale is a 4-in-1 storage platform built to deliver up to 6 terabytes of throughput per second per rack. And we all know that AI devours data.

And our family of unstructured data solutions including power scale, object scale and the Lighting File System, the fastest in the world, are being deployed extensively by thousands of customers around the world, including some of the AI native companies like CoreWeave and Nscale, Iron, Fluidstack, Boost Run and many others and many other leading firms that demand incredible performance like McClaren Racing, Hudson River Trading, Quadrature, Two Sigma, Optiver and many, many more.

Now of course, all that data needs to travel at blinding speeds. And so AI factories need AI grade networks. The new power switch Spectrum 6 gives you the massive scale, speed and efficiency and reliability required for AI. And we've added the NVIDIA Quantum-X800 liquid cooled with co-packaged optics also to our networking solution.

Now all that storage and networking feeds the beast of accelerated computing. The PowerEdge XE9812 built on NVIDIA's Vera Rubin NVL72, delivers 10x lower cost per token than Blackwell for massive scale genic AI inferencing. And the new lineup of PowerEdge XE servers built with NVIDIA HGX Rubin NVL8 are an industry leader, supporting up to 144 GPUs per rack with 100% direct liquid cooled compute nodes, 5.5x more powerful than the previous B200 GPUs.

And to bring it all together, we're excited to announce Dell PowerRack, a turnkey rack solution where the AI compute, the networking and the storage are all engineered together designed, tested, validated and delivered as 1 system. With PowerRack, you can deploy a fully integrated system where the thermal design, the power management and all the software have been engineered and tested to work perfectly as one.

Now one reality we're all facing is that AI is energy intensive, even though the amount of energy used per token is decreasing quite rapidly. A single rack of NVIDIA Rubin GPUs can draw over 130 kilowatts. And as the industry builds hundreds of thousands of these racks, the pressure on the power grid is real. And we are committed to being a part of the ongoing solution. Our PowerCool Heat Exchanger delivers up to 60% reduction, including energy consumption and cuts the annual energy cost by 1/3. And our PowerCool CDU is capable of cooling Vera Rubin-NVL72. It delivers more than kilowatts of cooling capacity.

We designed for performance per watt, so you could deploy more AI within your existing footprint. Now of course, the most efficient token is the one that is generating closest to where your data is. So we're expanding the Dell AI factory with NVIDIA to accelerate agenetic AI in the enterprise. We're announcing Dell Deskside Agentic AI, which brings together the highest performance Dell Pro Precision workstations, NVIDIA NemoClaw and Dell Services.

You can test, build and fine-tune agents locally while running the latest open weight models in the 70 billion to 250 billion parameter range, up to a trillion parameters. And you can do it without unpredictable cloud costs, bandwidth costs or risk of IP leak. This is unmetered intelligence, and you can break even versus public cloud APIs in as little as 3 months.

NVIDIA Open shell is also now supported across the entire AI Delhi factory, giving developers a sandbox to secure and fine-tune their AI agents from workstations to servers. And Dell technology support for NVIDIA AI-Q 2.0 blueprints gives you a tested foundation to deploy multi-agent workflows. We are providing a consistent path to accelerate the move from deployment to pilot to production at scale.

From the PC to a full rack of 144 GPUs, to the biggest data center you could imagine in the world, all of those racks, we are engineering the accelerated architectures to work together as one system agenetic and autonomous. The Dell AI Factory with NVIDIA, the backbone and the brain of the modern enterprise.

All right. Now the whole family's here. Now no company does this alone. And the next era of infrastructure is going to be built by deep partnerships between companies that are advancing accelerated computing and the companies that know how to deploy it across the real world. Please join me in welcoming a great partner friend, a true leader and visionary of the AI age, NVIDIA's Jensen Huang.

Thank you, Jensen. Jensen, it's great to have you back here at Del Tech World. We've been talking about agents.

I'm here every year selling Dell.

We appreciate it, man. We appreciate it. So give us your perspective on where we are in this agentic recursive self-improvement world. It looks like every time we wake up, there's been some leap in the model and the capabilities and sort of it doesn't seem to be plateauing. Give us your perspective from the front lines.

Well, 2 years ago, when I was here, we had just started with the agentic -- excuse me, generative AI journey, right. Generative AI can, of course, generate content but remember, it can also generate content to think with, generate thoughts, which led us to reasoning, which led us to planning, which led us to agentic systems.

So now we have -- we now have, for the very first time, useful AI, which is the reason why your demand, my demand is going parabolic, utterly parabolic because -- right, it's going parabolic because agentic systems -- the AI has to understand, has the reason, has to think it has to plan, use tools, look at the results of the tools, think about it some more, come up with maybe an improved plan. And so it iterates until it can get the job done using a whole bunch of different tools.

Well, the amount of computation necessary because it's running autonomously for so long, instead of just responding to a query, the amount of computation has grown 100x, 1,000x and depending on the work that you're doing, sometimes we'll kick off a software programming job, it doesn't finish for a week. Of course, it did a week what would have taken a whole team a month to do. And so big deal in productivity but gigantic leap and computation requirements. So that -- let's just say that computing went up by 100x or 1,000x. Meanwhile, because it's so useful -- the number of people using agents now all over the place. Every enterprise, our company, your company, everybody is using agents all over the place to do software development, devops, SRE, all of our CI/CD work, QA testing, the amount of software work that we do in the company now supported by agents is incredible.

One engineer, a really good engineer today is working with an agent, but a really great engineer in the future is going to be orchestrating a whole bunch of agents who are going to be orchestrating a whole bunch of sub agents to do work. Well, between the amount of computation and the amount of demand use going up, the product of that, that's our demand. And so we've now arrived at the era of useful AI, which is really just really exciting for all of us because until now, it's been novel, interesting, incredibly exciting.

But really at the enterprise level, many of you have said before last couple of years, the impact of AI is -- the potential is incredible, but the actual use was minimal. Now it's taken off. And we're starting to see now -- we didn't -- a couple of years ago, we started -- we didn't have 5,000 enterprise customers. Now we have the biggest companies in the world like you saw with Lilly at Samsung and Honeywell, they're piling into this in a big way. But really, that's just starting, right?

And to reimagine their workflows to be able to understand the trajectory of how this is improving and how it can affect ultimately what the companies could become, that really is just an idea. I mean we haven't seen that in any scale. I mean there are some companies that are doing it, but it's a very small number today.

No doubt. Well, you're experiencing, and I'm experiencing, our company has always gone fast. But it's gone -- it's really going fast now. And just the amount of content that's being created internally, the progress that's been made, people said that AI is going to make us more productive. There's no question about that. What took months now takes weeks, what took weeks, now takes days and what takes days, now it takes hours. And things that would take an hour, you and I pretty much expect it instantly now.

And so what has really changed is that our ambition has changed. There's no question my ambition has changed. I want it to be somebody to do something, make a contribution. But that was the old Jensen. The new Jensen, I got big ambitions now.

We all have to ask ourselves the question, how high is up. Well, it's pretty high, right?

Yes, yes.

So we've got all these great products. We're embedding AI in everything and enabling this distributed inference and intelligence. This is the best room I've ever been.

What is your favorite?

Well, it's hard to -- I got -- I love all my kids.

How about the one on the end?

You've got to love that one, most others, yes.

Well, this is incredible. And so we -- Michael and I have been working on a whole new line of computers to run agents. The way to think about agents is this. There's a large language model. It's gigantic. It's the most computationally intensive piece of software in the world's ever known. And then we've run in that system towards the end. That NVLink72, the world's largest scale-up single domain computer, okay? It's just 1 giant system operating as 1 computer. And that system has a large language model in it, 1 terabyte, 10 terabyte of parameters, no problem. And so that's one giant system. That's the brain.

However, an agent starts with the harness. That harness has to sit in a secure and governed container, we call it a sandbox. And so the NVIDIA OpenShell open source sandbox is the security system that just about everybody in the industry is using, inside, we have a reference harness we call NemoClaw. NemoClaw runs on a CPU. And that CPU could be a CPU and there, a CPU here a CPU in there. It can also run Nemotron or any of the models, open source models that you would create your own specialized agents for your own companies trained for your own special domain of data or skills. And that would run locally, if you like, and then the large language models could run in the cloud. And so you have this hybrid AI.

The thing that's really cool is that NVIDIA's architecture is the only architecture in the world that runs every Frontier AI model and lately -- recently, in the last year or so, Anthropic has been really leaning into the NVIDIA architecture. So now we support every single Frontier model. We support every open source model and they -- we support them in the cloud or locally. As you can see, our computer is the first 1 in the world that runs in every cloud, but it also runs locally.

And if you have your models that you're quite sensitive about, our systems are built with confidential computing. That way, you don't have to trust the operator that's operating the data center with your secure data. And so all of these architectures now run at every cloud runs every model, runs hybrid AI and runs agentic systems. And so you have your harness that runs on a CPU.

The CPU also uses tools and the CPU we created, I think you mentioned it earlier, called Vera Yes. The CPUs of the past were built for hyperscale clouds. And so you're renting the CPU cores and so you're optimizing for us many CPU cores as possible.

Well, agents in this new world, you're generating tokens. You're not running course anymore. You're generating tokens and marketing that. That's the economics of this AI era. And so the AI wants to generate runs work and generate as many tokens as possible as quickly as possible because that's the output of the intelligence. Vera CPU has the highest single traded performance of any CPU in the world. It has 3x the memory downwidth of the fastest CPU in the world.

And as a result, Starburst, DuckDB, all these databases run incredibly fast because the agents are pounding on the databases, so the CPUs better be super fast. The agents want to get through its work, so the CPU has got to be super fast.

Otherwise, the big machine down there is waiting for the agent to get its work done. And so now you have the harness running on the systems here. Local AI models running on the systems here and giant models running in the cloud or in your own data center on the big machine down there.

Well, let's go check it out. But as we do that, what's exciting is to think about in the past as humans, we would do work, right, and we would pass it on to the next human. But now we have all these agents that we're managing, right? And an individual can manage, I don't know, 1,000, 100 agents. And so the possibilities that, that unlocks in terms of human creativity and what humans are going to be able to do. And I love this idea of the unmetered intelligence, right? Where you've got the power in your own PC in your own data center and you can use it with your own data, that's sort of super cool.

You don't have to struggle with token anxiety. You tell your employees you ran out of tokens or get that [indiscernible].

But Jensen since you're here, would you do us the honor of autographing this latest.

All right. We've got to climb here, I'm not as tall as you are. This is -- what's the date today?

18th, May 18.

We're not going to sell this one. This one is -- you got to sign it, too.

I already signed it.

You did? Well, you got it right here. Dell, Dell, Dell. Here you go. Michael has already signed up right here. And so this is -- Michael, this is 100x larger than this one, the station, right, exactly the same architecture.

The GB 300.

GB 300 in here. That's 100x larger than this one. This is the only computer deskside computer in the world that can run a terabyte a $1 trillion parameter AI model, okay? So this is notable.

It would have been unimaginable even a year or 2 ago.

I know it. One trillion -- I mean, this would basically be a cloud just 2 days ago. And so -- so this is the station, and this is 100x bigger. So yes, this one, and that's 5x, 6x larger than this one. And this is one of my favorites. This is the smallest GB10. Same architecture is that one is that one. It's incredible, right? One architecture, Incredible.

Well Jensen, really appreciate you being here. We treasure the incredible partnership with NVIDIA, all we've been able to do for customers around the world. And look forward to...

And we grew up together, practically.

Yes, we did. 31 years, we've been doing this.

Thank you. Guys, keep up the great work. Thank you, everybody.

Thank you, Jensen. Well, thank you, Jensen. His enthusiasm is infectious. And his passion for innovation and its belief in accelerated computing and what that can bring to the power of AI around the world. Now let's look at that power in action. In modern health care, intelligence is not measured in tokens. It's measured in time, in accuracy, and compassion and in live safe.

[Presentation]

Ladies and gentlemen, Dr. Chuck Fraser is here with us today. An incredible pediatric, cardiac surgeon. Thank you, Dr. Fraser.

This is why we do what we do, why we are so passionate about solving your toughest challenges. For too long, the AI conversation has been trapped inside the screen. The real story begins now. AI is moving into hospitals, into factories, into schools, energy grids, laboratories, cities, homes and yes, even into orbit, solving problems at the scale of humanity. Over the next few days and in tomorrow's keynote with Jeff and Arthur and in our technical sessions, our trailblazers and the solution showcases, we want to inspire you to be bold and to move fast. The road ahead is bright, and it is beautiful, and we will be with you on it every step of the way. Thank you and enjoy the show.

Crazy train. I know the team doesn't know what I'm about to say, but they have no idea that brings back such fine memories. This song hit the United States in February of 1981, but ready to date myself. I graduated high school a few months later. On August 12 of that year, right before I entered engineering school, something very important happened that's changed my life forever. IBM launched the PC. And I've had the privilege of spending most of my career at Dell, nearly 40 years now with a front row seat wave after wave of innovation from PCs to servers, to storage, the Internet, mobile. And here I am today in front of all of you getting to talk about artificial intelligence and Agentic workforces. What an incredible ride. Welcome to day 2 of Dell Technologies World.

Michael set the stage yesterday and as usual, set the bar high, so I have a bit of work to do. He talked about what the modern enterprise looks like. Today, we're going to talk about how to build one. Last year, you might remember, I've said token use was likely undercalled. And I ended my keynote by saying we are on the forefront of the Agentic era. It turns out I was a bit too conservative, which seldom happens, but I was. AI crossed a real inflection point. The models, the software, the hardware, all improved faster and more capable than anyone predicted. What was thought to be a 3-year journey happened in less than 12 months, happened in less than 12 months, utterly amazing. And you all know I love a good top 10 list. I'd like to start these conversations with them. The team has continued to try to break me of this habit. Yet again, they did not succeed.

I have 10 things that I want to talk about that changed our world since the last time we were together, and they're profound. The first one, number one, AI went from an adviser, Michael talked about that yesterday to an operator. AI executes the answer, manages the exceptions and escalates what it can't resolve. Number two, model prices have dropped roughly 40% -- excuse me, 80% but token consumption is up 10x, surpassed 100 trillion tokens in a year in 2025. Equally important, the context windows crossed millions of tokens. You can now enter an entire code base, a year of contracts, a full operations history, all done in a single pass. Number five, training built the models, inference runs the business. Inference workloads account for nearly 2/3 of all AI compute this year. The compute curve bent and quite honestly, it's never going back.

Number six, generative AI software and enterprises, the spend tripled in 2025 to $37 billion. Companies that are spending aren't experimenting anymore, they are building. One of my favorites, physical AI left the lab; robotics, autonomous systems, and embodied agents are now showing up in factories, warehouses, hospitals and farms. One of my personal favorites, the PC is part of the AI stack. Power at the desk side matters because workloads like software development, media workflows have all moved to the endpoint. And quite frankly, the talent conversation has flipped. It's no longer where do I find AI engineers. It's how do I get everyone to think and work like an AI engineer.

And then lastly, the conversation changed. Every CIO I meet has stopped asking me, "Should we?" and I started asking, "How fast can we?" 10 things, 12 months, any one of them in any other era would have been the headline of the year. This time, they are the context of what comes next. Think about that. So what I wanted to do now is share with what I'm seeing inside Dell, inside our customers, inside the 1,000 environments that we're helping customers implement today, 3 patterns, 3 distinct patterns that I think needs to sit with each and every one of you as you leave this morning.

The first observation kind of buried in the top 10. Token costs are falling fast. Token use is exploding. Token costs are down roughly 80% year-over-year. Token use alone for reasoning is up 320x. Think about that. Model cost, token cost down 80%, explosion of use 320x, so much that you see the unit prices collapsing, leading to this volume is exploding. And if this pattern sounds familiar, it should. We've seen this before in bandwidth, storage, compute for decades now. Cheaper units unlock so much new usage that the total spend goes up, not down. And we've never seen it move this fast at this rate. And we see here at Dell.

We've had groups of engineers who have consumed a month's worth of allocated tokens. Yes, we were allocating them in the past. I'll get to that in a bit, and they consumed it all within a few hours. And not because anything was broken because it's working. This is what success looks like with agents and you better be ready for it because it's coming. Observation number two, AI productivity is extremely nonlinear. Inside almost every company deploying AI today, 5% of the people are driving 95% of the value. 5%, 95% of the value. Why? Because when someone learns how to use agents effectively, they use them for everything, research, coding, modeling, strategy analysis. Their day looks completely different from the individual right next to them. We call them super users of intelligence, a small slice of the workforce driving most of the gains, compounding their advantage week after week.

So if you're measuring AI ROI, mouthful, by averaging across your entire organization, quite frankly, you're doing it wrong. That's like measuring system performance by averaging the idle cores. The right questions are, who are my super users? What are they doing? And how do I get everyone to become one of them? Observation number three, if it already hasn't, tokens are about to be a line item in your P&L. Agents are doing in minutes and hours, what took multiple sprint teams of multiple people days and weeks, writing code, testing code, deploying it, planning the next iteration. That's not a 30% productivity gain. That's a 10x gain, that's a 100x gain and oftentimes even greater than that. That's structurally a different company. As the agents take on more of this cognitive work, cost migrates from headcount to tokens. I'm going to say that again, it's very, very important.

As agents take on more of the cognitive work, cost migrates from headcount to tokens. Tokens are going to be much greater than a rounding error in the future. And the companies who are planning for this shift now are going to have a big advantage over everyone that waits. If I take those 3 observations together, infrastructure demand is exploding as model prices collapse. The gains are concentrated in a handful of super users who are pulling away from everybody else week after week and the cost structure of cognitive work is inverting in front of us. Cognitive work, the cost is inverting in front of us. Every enterprise in the world was built on one assumption, cognitive work scales with human hours. You want more analysis, hire more analysts, more code, more developers. Agentic AI has broken that ratio forever. And here's the part that nobody wants to talk about out loud.

Most of our operating models that are built for the world that we lived in yesterday are now broken. You can feel it. Your teams can feel it. Nobody has completely figured this out yet. I know we certainly haven't. But what's clear to us is the companies that win in the next decade will be AI native. What does AI native mean? I want to be very clear with this. You don't have to be born in the AI era to be an AI-native company. It's not a birthright. It's an operating model. It's built on the premise that intelligence is a utility that it flows through every workflow, every decision, every product, every interaction. In my home state of Texas, we often hear the transplants say, "too darn many of them", but the transplants do say, "I wasn't born in Texas, but I got here as quick as I could." That's my challenge to you.

Just because you didn't start with AI doesn't mean you can't get there. And the question you're going to have to ask yourself, are you really ready and willing to disrupt what you've built? Are you ready to tear down to the ground the old ways and rebuild in a new way? Or as I like to say, break some s*** and start over. That is ultimately what we're talking about. Things have to fundamentally change. This takes me to the next part. So as we think about those observations and the impact they're having, it's going to have a profound impact on how we build an AI native enterprise. So I have 5 imperatives, things that you need to do right now and you got to get it right or you're not going to be prepared for what's coming.

The first thing, Arthur will expand on this in a bit later, build an AI-ready data foundation. Michael said this yesterday, AI runs on data. However, in most enterprises, the data is scattered across dozens of systems. 80% to 90% of it is unstructured. None of it's connected in a way that an agent can use effectively. To power agents at scale, you need a real-time connected knowledge layer. And here's the structural and architectural decision you're going to have to make as a result. Don't move the data to the AI, move AI to the data. That's fundamentally a different approach and, quite frankly, a decision that you need to make now.

Number two, build a distributed AI infrastructure. Training and inference are very different workloads with very different physics. You heard Jensen and Michael speak briefly about this yesterday. Training needs massive GPU clusters, centralized by necessity. If I think about inference, where the growth is, you have reasoning models executing multistep change. You have agents calling models over and over planning, evaluating, iterating. These workloads are 10x to 100x more compute-intensive. Jensen said yesterday 1,000 times. 10x, 100x, 1,000x more compute-intensive than what you were running 18 months ago. An AI native enterprise has to be built for both.

Number three, secure autonomous systems. Here's the thing about agents. They just don't call the model. They call your CRM, they call your ERP, they call your financial systems, they call your customer databases. Every one of those touch points has to be secured, logged and must be reversible. Because when an agent takes action on your behalf, changes a price, updates a customer record, initiates a procurement workflow, you need to know what it did, why it did it and how to undo it if it got it wrong. In an AI workforce, every action needs a receipt. And that's not a compliance check box. That's how you build trust into the system that will act on its own.

Imperative number four, integrate the enterprise stack. The agent becomes the coordinator. The stack has to let it. Agents need to plan task, call tools, execute work, handle exceptions across your entire stack. That means an API-first architecture, workflow orchestration and agent frameworks that can do multistep execution. If your stack can't do that, your agents are going to be siloed and expensive. And that's simply not a conversation you want to have with your Board.

The last imperative, restructure for Agentic AI and tokenomics. The question isn't whether consumption grows as it absolutely will. The question is, are you running the right tokens on the right infrastructure? A routine summary doesn't need a frontier model. A sensitive financial analysis shouldn't leave your building. And a complex reasoning chain might need the most capable model wherever it lives. You need to think right workload, the right model on the right tier, tier being edge, data center cloud. And if you do that correctly, you're going to get the optimum performance, privacy and cost efficiency. Run everything in one place because it's easy, be ready for a surprise, and that surprise is a large bill that's only going to get larger.

And Michael said this yesterday, it's fundamentally something we believe the notion of token routing, where to put that token is going to be one of the most important infrastructure decisions you will make. 5 imperatives, one direction, the AI native enterprise isn't a vision anymore. It's a blueprint. And the reason I'm confident you can build it today is because the ecosystem is here to support you. So now let me show you what that looks like with one of our most important partners.

[Presentation]

Gemini, on-premises, air gapped if you need it, Dell's customer zero available today. Pretty exciting. So I've been talking this morning about the AI native enterprise and what it has to look like, those 5 imperatives, the data foundation, the distributed infrastructure, the security, the stack integration and the tokenomics. And Thomas just shared what happens when that comes together with a world-class partner, extraordinary and more to come. Now we're going to shift the conversation a bit because building the modern enterprise isn't just an infrastructure story. It's an agent story, too. And I want to bring someone up here who's been living this from a completely different vantage point. Our next guest is the Co-Founder of Offline Ventures, Co-Founder and Board member of OpenClaw Foundation, one of the most important open source projects in enterprise AI right now, built on the simple principle. The enterprise should own its AI, not rent it. Everyone, please welcome Dave Morin. Good morning, Dave, how are you doing?

Great. Good to be here.

Glad to have you here. So we spent some time backstage this morning talking about what you're doing. It's pretty exciting. You've had a front row seat on some of the changes in mobile, social in your career. You had this epiphany with OpenClaw. Tell us about that "aha" moment and why you realized it was going to change everything.

When I was 7 years old, my grandfather put a personal computer on my desk. And the first feeling that I had with it was I could create anything that it was mine and I could hack it, I could turn it into whatever I wanted. And throughout the rest of my career, I've been chasing that feeling. I went to college and I took a Dell computer with me, and I built my first business in my dorm room using that Dell computer. After that, I went on to work at Apple and Facebook. And in each of those places, we were trying to give the power of computers to people so they could create new experiences for themselves.

Over the holidays, I was at home, visiting family and a friend of mine pinged me and said, "Dave, you got to check out this OpenClaw thing." And I said, "Oh, it's the family time, it's the holidays. I don't have time to check this out." I decided to install it. And within 24 hours, I had that feeling again. I felt like this is the first AI experience that I've had since ChatGPT launched or maybe even since the iPhone that I felt like I had a computer of the future in front of me on my desk. And OpenClaw was the first time that I felt that way. And ever since then, I called up the founder, Peter Steinberger, and we started working together every day to bring that feeling to everybody in the world.

That's pretty cool. I've been using OpenClaw in my dashboard at work. It's made things much easier. The team is a little surprised where I'm using that. But if you think about the capability of an agent and its applicability to enterprises, we were talking backstage that you've done some research on the use cases of OpenClaw in the enterprise. How do the 2 merge from what may have started as a personal assistant now into an enterprise class capability.

Yes. I mean I think when you put an agent into the enterprise, the first thing that comes to mind is security and observability. You need to understand what does this agent know. We all know that when AI wakes up, it doesn't know anything about you. And so you have to fill it with context. You have to give it access to critical data to help people do better, work better in their workflows. In order to do that, you have to know what it has access to and you have to also know what it's doing. And so what we've learned in building OpenClaw over the last several months is figuring that out, giving the agent read-only access at the start and then also building observability layers in order to understand every single action and every single thing that the agent does helps you figure out where it can go inside of your enterprise and how it can help people either in the cloud or on the edge.

We talked backstage a little bit about one of the biggest challenges is connecting data.

Yes.

Share your experiences maybe with our audience today of how connecting the data and the challenge with that, but also the benefit once you get it right, the efficiency and productivity you get.

Yes. We've seen 2 different things. We have people building massive multi-agent architectures in the cloud, trying to figure out how do I build adapters for each different type of data. It can be a really intense task and people are going about it a lot of different ways. One of the cool things we're doing at OpenClaw and our own workflows is we actually build crawlers that go and grab data from the cloud and pull it on to the edge device and store it in SQLite or other lightweight database and simple data structures so that the agent can rip through it really quickly and answer questions locally on device and make it really fast, really easy to use. And you can also know that it's happening on that device, the data, the workflow right there, right next to the person, you know it's secure. And so we're doing this internally. It's how all of us at the central core of OpenClaw, the maintainers, our key engineers. We use workflows like this every single day.

Pretty cool. Maybe we should wonder over to the devices.

Yes, I saw this thing over here.

We sent a GB10 to you. We're going to get to this one. I know what you...

You want to stay next to that one.

How about here? We sent you one to our GB10 device. Tell us your experience with it, particularly as we think about moving more of the capability to the edge.

Yes. I mean these things are amazing, 128 gigs of RAM. You've got a Blackwell in there. I've got one set up with OpenClaw on it. It will run any OpenClaw workflow. You can use 30 billion parameter models, run data right on the device. You can also set it up as a satellite of your core claw. So I have a really advanced claw that I've been working on for 3 months. It runs all of my key workflows. It has access to all of the foundation models, but it now also has access to my GB10. And so when I want to run a really sensitive workload, something that's important, secure, I don't want it leaving my own machines. I use this as the local inference layer and have my main claw talk directly to this and run load.

But from our conversation, maybe you want a little more and why would you want more accessibility?

I used to think that I wanted a bigger monitor, but now I know I want a bigger computer.

And why?

Well, what's interesting is that this thing puts you about 1 year in the future. This thing it puts you 5 years in the future as an individual. And this thing, you could run an entire company on. It depends on the size, of course. But a 3,800 under your desk can run a 1 trillion parameter model. You're talking frontier level performance under your desk and the ability to run it 24/7 multi-agent workflows, absolute frontier engineering. I mean, I want to put one of these underneath every single one of our core engineers' desks so that we can be running in the future today.

Available on www.dell.com. Maybe a last parting question. If you think about a thought to leave the audience today, what would it be? And maybe tie it to our conversation about the notion of needing more capability at the edge and the role the edge plays in the future with agents and AI.

Yes. I mean I think give people the tools to try this. If it needs to be at the edge, do it, start here, start here. Models need context in order to provide the best, most powerful possible answer workflow. We're seeing that people are coordinating their work in vastly more efficient ways, put an agent into your workflows, give it read-only access to the data to start and just get it in the game, see if it can organize a simple workflow, help people do things faster more locally and just get started, don't be afraid.

Get started, don't be afraid. Everybody, Dave Morin. Thanks for joining us today. My pleasure. Thank you. So Dave just confirm something I've been saying all morning. The enterprise has to own its AI, not rent it. But it raises a very, very practical question. Where do I run this stuff? Yesterday, Michael announced the Dell Deskside Agentic AI. The principle of it is very straightforward. The most efficient AI token is the one produced closest to the data. And most of your data is not in the cloud. It's on-prem, it's at the edge. That's where the AI needs to run. For every developer and IT leader in this room, that means a secure sandbox where your teams can build, test and fine-tune agents locally. No unpredictable cloud cost, no bandwidth issues, no bandwidth penalties, no IP leaving the building, and I'd like to show you what that looks like.

[Presentation]

It's pretty cool. That's right. That 45-year-old device, the PC, a free unmetered token generator. That's the future. One developer, 10 agents, 1 billion tokens in 24 hours in the cloud, $3,400 on a workstation, 0. That's not a demo. That's a deployment model we're helping customers build right now. We kind of stole the thunder a little bit earlier with Dave, but I want to show 3 products that actually build upon one another to do exactly what we just described. The GB10, 1 Petaflop of performance. It's optimal for models in the 30 billion to 200 billion range. It generates 40 tokens a second. This is where most team should begin. If I go to our Dell Pro Precision Towers, this is a T4, 7 Petaflops of performance, scales to a 500 billion parameter model, generates 400 tokens a second. This is where serious developers should start or move next.

And then the big boy, my personal favorite, the GB300, [ 20 Petabytes ] of performance. Jensen said this yesterday, up to 1 trillion parameter model, generates 700 tokens a second. This Grace Blackwell Ultra Superchip draws over 1,500 watts of power. You can't cool that by air. So we engineered a closed-loop liquid cooling system that doesn't exist anywhere else. This one makes me smile. This is special. The big boy is fascinating. That's our new portfolio of Dell Pro and Dell Pro Precision shipping available today. And as I mentioned today, www.dell.com, you can get the GB300, and we've already shipped the first last week.

So I've been talking about what it takes to build the AI native enterprise. The desktop is where your developers start. But to scale it from the desk to the data center to the cloud, you need infrastructure underneath it. That's Arthur's world. I'm going to let him tell you what he's built because, quite frankly, some of the stuff they've done has even surprised me. Everybody, Arthur Lewis. Ask him about his socks.

Don't ask me about the socks. Good morning, everybody. It's great to be back here at Dell Technologies World. Jeff just laid out a clear and compelling demand signal and 5 very important imperatives. There is little question that the Agentic workforce is emerging and that the token explosion is real. And we have a lot to cover today, and I want to jump into something that I think is incredibly important, which is building a data foundation. And let's start with a very important fact. The vast majority of enterprise data resides on-prem. This data is your strategic advantage, yet no frontier model has seen it. It is sitting cold, it is sitting dark, invisible to AI, generating little to no value. And leveraging ever-evolving techniques like active learning, fine-tuning, reinforcement learning across the entirety of your data is not a nice to have in the world of Agentic. It is an absolute must-have. Not only will it optimize the results, it will drive significant token efficiency.

For example, a task that's run against a frontier model that consumes 4,000 tokens and chain of thought and 3 tool calls can be done on a fine-tuned model with 400 tokens and a single tool call. And of course, this calculus multiplies daily across millions of inference. So how do you build this muscle? How do you build this engine? And how do you do so at scale? The answer is simple. The Dell AI data platform. This is an engine that is fueled with your data, shaped by your intelligence and tuned to your outcomes. And it comes in 3 simple layers: Layer #1, data preparation. In order for your data to be AI-ready, it must be discovered, cleansed and labeled and transformed into something that the AI can actually interact with. And this is exactly what the Dell orchestration engine was designed to do, built from our data loop acquisition and in close partnership with NVIDIA, it can handle structured, unstructured multimodal data, turning raw enterprise content into rich, curated and governed data sets ready for training and inferencing.

This orchestration engine comes with a built-in marketplace with over 200 applications, models and data pipelines, including NVIDIA NIMs as well as no-code pipelines for RAG, active learning and fine-tuning. You can go from an NVIDIA Blueprint to production deployment without writing a single line of code. And powering all of this are the Dell data engines designed to transform and query data and to do so at great scale. With NVIDIA CUDA-X libraries natively integrated, customers are seeing 12x faster vector indexing, 6x faster queries and 19x faster time to first token. Let's take a look at the solution in action.

[Presentation]

How cool is that? So layer #1, data preparation. Layer #2, disaggregated inference. Once your data is AI ready and available to agents, speed now comes down to the model's ability to access and query data to generate new tokens. And this is exactly what Lightning with its container-based and high availability architecture was designed to do. Lightning is the fastest parallel file system in the world, delivering 150 gigabytes per second of throughput per rack, more than twice that of our nearest competitor. Lightning also provides the context memory extension and storage access needed for disaggregated inference, delivering 20x the performance of flash-only scale-out alternatives. Coupled tightly with NVIDIA Dynamo, Lightning is capable of bypassing the prefill stage of inference to intelligently route KV cache data between GPU memory and storage leveraging RDMA in real time.

The infinite storage capacity that Lightning brings to bear enables agents to retain long-term context over extended durations, something that is impossible to do with limited GPU memory. And the third layer is the underlying storage platforms, PowerScale, ObjectScale and something new. You may have heard about this yesterday. Over 1,500 customers today deploy GPUs on PowerScale. It is NVIDIA certified, SuperPOD validated and supports GPU direct and RDMA. And when compared against other NVIDIA reference architectures for saturating up to 16,000 GPUs, PowerScale is incredibly differentiated. 80% less rack space, 8x fewer switches, 72% less energy, saving customers $1.5 million along the way.

Next, ObjectScale is the world's most cybersecure object storage, delivering 40 gigabytes of throughput per node, more than twice that of its nearest competitor. And our unique implementation of S3 over RDMA and GPU Direct shows significant gains versus traditional S3-only implementations, 70% less CPU utilization, 230% more bandwidth and 80% less latency. And when coupled with advanced features like S3 tables and S3 vectors, ObjectScale is primed for a broad array of data sets for AI training and inferencing. And this new thing was announced yesterday, Exascale, a 4-in-1 -- a unified rack infrastructure built to support PowerScale, ObjectScale, Lightning and PowerFlex.

Exascale is the only 4-in-1 storage built platform specifically designed for extreme scale AI, high-performance compute and enterprise workloads, delivering 6 terabytes of throughput per second per rack unit. So whether you choose a software-defined architecture or an appliance architecture, whether you choose file, block, object or disaggregated inference, Dell has the broadest portfolio of data storage in the industry. Your data is prepared, it's moving at speed and it's stored at scale, a holistic workflow designed to drive down your token cost.

Now that we've built a strong storage foundation, let's talk about security. Agents don't just analyze data, they access systems, they execute, they modify data. Thus, security has to be a prerequisite. It cannot be a secondary concern. And this high level of protection must exist where your data lives at the edge, in the data center or in the cloud. Our research shows 94% of ransomware attacks seek to compromise backup workloads. But more alarming is the success rate, 57%. Why such a high success rate you ask? The reason is that the vast majority of backup workloads today land on generic infrastructure. This is a huge problem.

Number one, it's a horrible TCO. But more importantly, the software and hardware are not jointly integrated and hardened. Rather, security controls are layered in after the fact instead of being engineered in from the beginning. Security must start very early from the supply chain and extend into jointly integrated and hardened software. Backup workloads must land on purpose-built systems like PowerProtect Data Domain, systems that are specifically engineered for security, for performance, for recovery under attack. This means customers get factory-to-site verification, silicon root of trust, Zero Trust controls, end-to-end encryption and built-in immutability. These are table stakes in a purpose-built appliance.

Equally important is the software stack and the ability to securely and effectively manage a complex environment across an entire estate. And this brings me to another announcement, PowerProtect One, Cyber Resilience simplified. PowerProtect One is a unified integrated architecture that brings PowerProtect Data Manager and PowerProtect Data Domain together. PowerProtect Data domain, of course, the world's leading target appliance, rated #1 in customer satisfaction and supporting 650 exabytes of data across more than 15,000 customers around the world. PowerProtect One will enable you to scale components independently to improve security and drive down operational cost.

With PowerProtect One, customers should expect 0 to first backup in 1/4 of the time that it takes with a nonintegrated appliance, 50% less management -- less daily management overhead, 75:1 data reduction, enabling 150 petabytes of logical capacity per node, 2x faster restore time and 2x faster replication. This is cyber resilience for the AI era. Now that we've built a strong data foundation and we've secured it, let's talk about modernizing your infrastructure and freeing up critical resources with the Dell Private Cloud.

As customers think about building a modern private cloud, they typically think about 3 things. Number one, they think about flexibility because we're living in a multi-hypervisor world, spanning virtual machines, containers and bare metal. Number two, they think about simplifying IT operations and significant advancements in automation and application orchestration. And number three, they simply want to know that their architecture is future-proof and cost effective. And with the Dell Private Cloud, they get all 3. And the savings are significant, upwards of 65% TCO when compared against legacy hyper-converged infrastructure.

And today, we are expanding our offers across broad ecosystem of partners, including the latest software from VMware and Red Hat as well as PowerStore support for Nutanix and Microsoft Azure local. We are giving customers real choice and automation-driven simplicity. And of course, at the center of any private cloud is storage. And today, we're incredibly excited to announce PowerStore Elite, a new class of modern data platform. This is -- I haven't gotten through the specs yet. This is a new class of modern data platform, a full third-generation refresh of hardware and software. We've added more powerful CPUs. We've added faster memory and significant enhancements to our software stack and the results, amazing.

PowerStore Elite delivers 1.5 million IOPS, which is 3x that of the previous generation. It delivers 80 gigabytes of throughput per node, more than 4x the previous generation. We've improved the world's leading data reduction guarantee from 5:1 to 6:1, tripling the density generation over generation, enabling 6 petabytes in a compact 3U form factor. And we've extended PowerStore scale-out capabilities to include transactional file, enabling customers to consolidate even more workloads. Let me be clear on this. PowerStore Elite is the gold standard. No competitor in our industry can deliver this level of density, performance and functionality in a single product.

And now -- and it's not just about the capabilities that we deliver today. PowerStore's container-based architecture means that everything is modular and upgradable, which means that it will evolve as workloads evolve and adapt as new technologies emerge. PowerStore Elite is the definition of future-proof. Now let me -- and every single Dell private cloud is managed through the Dell Automation Platform. And today, we are extending it with AIOps-driven intelligence and an Agentic layer, enabling customers to proactively observe, manage and optimize their infrastructure. Let's take a look at the solution.

[Presentation]

So freaking cool. The London Stock Exchange, now known as LSEG, chose Dell to build a new secure modern private cloud, integrating our servers, our storage and our automation platform and LSEG using this solution can now deliver capabilities related to data services and trading operations faster than ever before. Now let me bring it all home with the Dell AI factory. We launched the Dell AI factory 2 years ago with a very simple premise, a full stack, modular, validated architecture, the compute, the network, the storage, the data management, the software, the ecosystem, the services, designed, delivered and serviced as a single product, built to support workloads at the edge in your data center and in the cloud. And the adoption speaks for itself. 5,000 -- over 5,000 customers have deployed a Dell AI factory.

In the last year alone, you've seen 160-plus product releases. And perhaps most importantly, customers are seeing first year ROI of 269%. The model is working. It's working exceptionally well. And today, I want to introduce you to the next generation of infrastructure that will power the Dell AI factory, starting with the compute. And let me begin with the workhorse of the portfolio, the recently authored -- signed XE9812 built on NVIDIA's Vera Rubin NVL72 platform. This absolute beast of a platform delivers 10x cost per token reduction gen over generation and 260 terabytes of GPU to GPU throughput, which is more bandwidth than the entire Internet.

The new lineup also includes 3 new liquid-cooled PowerEdge servers built on NVIDIA's HGX NVL8, supporting up to 144 GPUs in a single rack and delivering 5.5x the performance of B200 GPUs. And as we've said all along, AI requires a heterogeneous mix of accelerated and general-purpose compute. And this week, we are announcing the broadest refresh in PowerEdge history with our 18th generation portfolio, including a new series of ultra-performance servers supporting CPUs from NVIDIA, AMD and Intel built for high core count density and memory bandwidth to support everything from databases and virtualization to AI workloads. The entire portfolio one common platform of management, the entire portfolio post-quantum cryptography compliant day 1.

Now compute alone will not solve. You need a high-speed network. The most powerful GPUs in the world require a communication super highway and Agentic technology will tax the network like never before, agents exchanging context, scheduling, making decisions in real time. Your network must be lawless, adaptive and fast or GPUs will stall and a stalled GPU is an incredibly expensive proposition. To build this communication super highway, we are expanding the power switch portfolio to include NVIDIA Spectrum -6 and Tomahawk -- and Broadcom Tomahawk 6 Ethernet technology, enabling 409.6 Terabit per second of switch capacity with co-packaged optics, which will enable 5x more power efficiency, 10x more reliability and 5x more uptime in AI workloads.

We are also introducing NVIDIA Quantum-X800 InfiniBand solution, liquid-cooled co-packaged optics, enabling 10,800 gigabit per second host connections and a 9x boost in performance. But network hardware alone is not enough. Your network operating system must be purpose-built for AI. It must be high bandwidth and low latency. 20 years ago, Linux became the standard operating system for servers. It was open, multi-vendor and a global community continued to improve it. The same thing is happening today. SONiC is to networking as Linux was to servers. And we have enabled the entire power switch portfolio. We have enabled SONiC across the entire power switch portfolio, giving customers access to the same high-performance fabrics that hyperscalers use with enterprise class support and life cycle management.

And now that we've talked about a lot of the individual components, it's important to not lose sight of the fact that bringing -- that the complexity in bringing all of this together is immense, and we make it simple for customers. We design and validate the full stack, the compute, the network and the storage as one system. The rack is now the product. And this engineering and design philosophy, coupled with the expertise and capabilities of our services organizations means that we can deploy rack scale infrastructure in under 6.5 hours and maintain uptime of 99.9%. And the market has taken notice. IDC ranked Dell #1 in rack shipments last year, shipping more than 2.2x the number of racks of our nearest competitors.

The Dell AI factory is the proven path from AI investment to business outcome, and the model is working. Hudson River Trading trades millions of shares a day across 200-plus financial markets, deployed their AI research platform on a Dell AI factory with NVIDIA. PowerScale as the underlying data foundation, coupled with liquid cooled PowerEdge GPUs. The results HRT's AI research now keeps pace with the markets in which they serve. SanDisk is another great example. They went all in building an advanced AI and generative AI platform on a Dell AI factory with NVIDIA, and the results were simply amazing. Factory costs were down 32%, energy costs were down 46%, CO2 emissions were down 45%, defect rates were down from 800 parts per million to 100 parts per million and factory operation -- lighthouse factory operations was up to 95%.

Ladies and gentlemen, this is what good looks like. In closing, Jeff gave 5 imperatives. Number one, build a strong data foundation. We've built it. Number two, distributed AI training and inferencing at the edge, the core of the cloud. We've built it. Number three, secure autonomous systems, we've built it. Number four, a full stack solution. We've built it. Number five, an -- a full stack solution ready for an Agentic world. We've built that too. The Agentic enterprise is no longer a vision. It is being built right now, and Dell is leading the way. Thank you for your time this morning. And Boss, back to you.

Arthur Lewis, everybody. So a home stretch. I'm going to get this home quickly here. 3 years ago, we found ourselves with thousands of shadow AI projects running inside Dell. People across the company experimenting on their own with whatever tools they could find. It wasn't a failure of governance. It was actually a signal of demand. So we got disciplined. We identified 5 use cases with real outcomes. We created an AI office to kind of gather it all together. We went to 90-day sprints that turned into 60-day sprints that turned into 30-day sprints. My challenge to the team is now 3 days heading towards 3 hours. Today, our service assistant is alive across our global services organization. It's actively closing cases. It's reducing dispatch rates. It's lifting customer satisfaction, all deployed on-prem, all on Dell infrastructure, all in an existing data center. No additional power, no additional cooling. The ROI was in less than 3 months. And our current infrastructure deployed at scale agents ROI in less than 3 months.

And every observation I shared with you today showed up on our own data. We blew past every forecast we had for token consumption. We pivoted and are now running our agent workloads on our own Dell AI factory because the economics demanded it. We can name our super users, a small group of engineers and operators who became force multipliers and reshaped how we're thinking about AI adoption entirely in the company. And we restructured how the company runs in real time. We are mapping workflows. And even more important, we are tracing the data path and the logic that powers our operations from the supply chain to services, to sales, to software development, all of it. This is not a pilot. This is how our company is running today.

So let me leave you with 3 go-dos. 3 things I really want you to think about as you exit the building, so to speak. One, go look at your AI cost model. Ask honestly, whether you're budgeting for both compute and token consumption. If you can't answer that, you're going to have a very uncomfortable conversation with your CFO in about 6 months. Number two, go find your super users. Don't count them, go find out what they do, study them, build alongside them. The gap between your super users and everybody else is the gap between your future and your past.

And then lastly, decide on whether you're going to lead the operating model change, the disruption or be organized by it. Both options are in front of you, but only one of them is yours to choose, I believe. So that's the modern enterprise. It's not a destination. It's a permanent operating model. We're doing the work ourselves. We're working with thousands of customers right now. The opportunity is massive. Let's get to work. Thanks, everybody, for joining us today.

Okay. All right. Good. Sorry about that, everyone. Well, hello, and good morning to you all. Welcome to our View From the Top CEO Call Series. I'm honored to welcome back Founder, Chairman and CEO of Dell Technologies, Michael Dell, on this View From the Top Series Call. Michael is joining us again for the third time. So we're especially privileged to have him on this call. Before we get started, I do need to mention that conflict disclosures as related to the individual companies or securities discussed on the call today can be found on the call invitation.

Additionally, this presentation contains forward-looking statements based on Dell Technologies' current expectations. These statements involve risks and uncertainties that could cause actual results to differ materially. Factors that could cause results to differ are discussed in Dell Technologies' periodic reports on Forms 10-K or 10-Q filed with the SEC. Any forward-looking statements made today are based on assumptions as of today, and Dell Technologies undertakes no obligation to update them.

We have a lot of ground to cover over here today. We're going to talk through, hopefully, a broad breadth of topics, including AI, memory, PCs, cap allocation, lots of exciting stuff to cover. We're here with Michael. So as you all know, and Michael really needs no introduction, but Michael really started and founded Dell Technologies with $1,000 in 1984 at the age of 19. He became the youngest CEO ever to earn a ranking on Fortune 500. He has navigated many cycles, lots of portfolio changes, including the acquisition of EMC, the go-private transaction back in 2013, spinning out VMware in 2021, Perot Systems buying and selling somewhere in there and now re-architecting the company for the Age of AI.

And beyond the success in business, Michael, alongside his wife, Susan, founded the Michael & Susan Dell Foundation in '99 based on really expanding opportunity through initiatives spanning education, health, and family economic stability. And most recently, I'm sure you've all seen the foundation announced as $6.25 billion, that's with a $1 billion philanthropic commitment to help seed-investment accounts for 25 million U.S. children, and that was incredibly generous, Michael, and we hope many others will follow your lead here to lead this next generation to success. That's just really incredible. So we're really fortunate to have you here.

Welcome, Michael. We really appreciate you taking the time to be with us here today.

Well, thank you, Wamsi. Great to be with you, and thank you for the kind introduction. Happy to discuss all this.

Thank you so much, Michael.

So to get started, you have reinvented Dell from PCs and peripherals to server, storage, networking and now AI. And you seem pretty all in on AI now with revenues growing to 1/3 of the company in just 3 years. So how do you see AI changing the tech landscape? And what do you see as Dell's opportunity here?

Yes. I think you have to step back and understand that we're shifting from calculating and computing to thinking and intelligence, which is just a fundamental change. And the value of that and the opportunity for that is enormous. And certainly, if you look at our AI server business went from $2 billion to $10 billion, to $25 billion. We're expecting $50 billion this year. We're still in the steep part of the S-curve adoption of the technology. We have 4,000-plus customers using these AI factories. And look, I think if you think about this thinking intelligence platform, you've got $114 trillion economy. If you get a little bit of productivity improvement with AI tools, it's worth an enormous amount.

And so while there's a ton of investing going on here, it's at least possible that the world is underinvesting in this given its potential. And it turns out that you need a lot of what Dell Technologies has built over 42 years in terms of infrastructure capability, server, storage, networking, and the support and services, and supply chain, et cetera, to be able to deliver that. And I think maybe 10% of customers maybe 15% understand what this can do and the rest of them are still figuring it out. It takes time for all this to kind of happen in the real world. I also think there are distinct phases of it. There's kind of the tools phase, which is what most people are doing. They're adding a tool here and there, and it helps people do their job and be more productive. That's great. There's a whole another phase of this, which is really reimagining the workflows in companies.

And this is very, very different than tools, right? It involves really rethinking how you get to a given outcome. And in many cases, this is just a big change inside companies. Again, it doesn't happen quickly, not easily, but you go from, let's say, a 10% or 20% or 30% improvement with tools to a 10 or 20 or 30x improvement. And there's certainly sightings of hundreds-of-X improvement in various processes and outcomes where you're just doing it very differently with Agents and Recursive Self-Improvement and all of that. So super exciting time. And certainly, there's a lot of demand for what we do.

Yes. No, that's incredible and so many things to touch on over there. Maybe let's start with the comment you made about potentially underinvesting. I mean a lot of investors are worried about hyperscaler CapEx budgets that probably exceeded anyone's forecast from 2 or 3 years ago. And these companies have gone from a capital-light to a capital-heavy model, which is pressuring their own cash flows at the moment. So how do you see the sustainability of this CapEx cycle? And how do you position Dell if the spending on AI was to either accelerate or decelerate?

Well, it's certainly not decelerating. I can tell you that. We took $64 billion in orders in the past year, and the opportunity and the pipeline keeps growing. And to be clear, we're not putting it on our balance sheet. We're still operating with a capital-light model. But when I talk to the hyperscalers, each of them view this as an existential issue for their business, and so they're all investing super aggressively. And again, I go back to the size of the services economy and the demand. What I can tell you is that when we deploy this infrastructure, Time to First Token is incredibly important because they're putting the infrastructure to work immediately and it gets consumed.

And so there's just a ton of demand here. I still think we're in the early stages of the adoption of all this. And we haven't even gotten to Agents and the Physical AI and Recursive Self-Improvement. And so certainly, there are going to be ups and downs here, but we feel we're well positioned. I mean we don't actually make any commitments until we have actual orders. And I think we've been pretty careful in managing our capital commitments and demand supply to make sure that we're ultimately converting this into free cash flow, which is what we focus on every day.

In some ways, this unprecedented CapEx and demand for AI data center build-out, it's just creating a lot of tightness across the supply chain, whether it be from labor and cooling equipment to memory, just to mention a few. So where do you see the biggest bottlenecks? And what are some of the opportunities and risks associated with these from Dell standpoint?

So we kind of love it when there's a supply chain challenge because that's kind of -- it's kind of like the Super Bowl for us. We're ready for that. Yes. I mean you have to understand that while you can easily identify, let's say, 10 or 20 suppliers that work with Dell, there's actually thousands and thousands of them. And so it's a complex thing. We have built a supply-chain machine, and it's built on relationships. Obviously, we took our guidance up a lot. We have the supply for our guidance. We're out looking for more supply.

And if you just get back to what's going on here, there is extraordinary demand growth across the industry. And so you've got all sorts of constraints. one way I would describe the memory issue is, I would say memory and advanced silicon, it's kind of the 25 x 25. And what do I mean by that? So when NVIDIA came out with the H100, it had 80 gigabytes of high-bandwidth memory. The current part has 288 gigabytes.

Next year, you'll hear about parts with a terabyte, and the year after that, you'll hear about parts with 2 terabytes. So from 80 gig to 2 terabytes is 25x more memory per accelerator, okay? And in that time frame, you'll have about 25x more accelerators. So 25 x 25 is 625x. And it also takes about 4 years to build a new memory plant, assuming you don't have shells already built. And if you wanted to have memory capacity in 2027, you would have made the investments in 2023.

Now if we dial back the clock to 2023, you might recall, it was a horrible year for semiconductors, particularly memory. Micron, in particular, had negative gross margin percentage, their sales went in half from 2022 to 2023, and the industry collectively lost $40 billion. A lot of them still have PTSD from that. And they're like scared little puppies. Some of them almost went bankrupt. And so they're not -- they're very careful about investing. And then if you go to the logic side, TSMC didn't increase its CapEx in '23 from '22, didn't increase it in '24. In '25, they started to increase it, but not enough, they're pretty conservative, and they're sold out.

And so yes, there's big supply constraints. The good news is that we are not a monoline company. We can move wafers around across multiple product families. And again, we've had relationships. I mean, Jeff Clarke's relationships and my relationships with these companies go back literally to the 1980s. And it turns out a lot of these companies are quite relationship-oriented, particularly some of the Asian ones. So we feel that the environment advantages us. And of course, scale.

I mean, if you look at our scale, our server and storage business is 4x larger than any single competitor generally, and we're larger than #2, #3, #4, all combined together. And we've also gotten really good in terms of our reaction time in dealing with the changes in cost. And customers also know that we have a supply chain that works. Now nobody likes it when the price goes up, but even worse is if you can't get supply. And so we feel we're well positioned for this kind of environment.

No, that's great. And yes, you guys have demonstrated time and time again when there is supply chain disruption, you guys just somehow managed to out-navigate everyone else, and that's just a testament to a very strong supply chain legacy that you've built at the company. I want to come back to memory, Michael, but staying on AI, your AI revenues have just grown extremely fast, right, like from almost nothing to $50 billion in just a few years. Most of this is Tier 2 CSPs. So when you talk to boards and CEOs, CFOs, how are they framing the ROI for AI? And when do you see enterprises start to maybe more broadly leverage AI and start to hit an inflection point? And really like maybe worked into that, how does this change the margin and ROIC profile for Dell as you think about those moving pieces?

Well, I would say we're seeing it, Wamsi. I mean there's substantial uptake in enterprise, and it's continuing to grow. Certainly, the margin profiles will be more attractive than the CSPs because they're going to -- they're smaller deals. It tends to be more services and storage and networking. And the lowest cost token is the one that's generated closest to where the data is. And enterprises have figured out that it's a hybrid world and these tokens are pretty expensive. And again, we haven't even gotten to the whole agent activity. But inference is definitely taking off.

And a lot of the use cases are not super complicated and a lot of the smaller models or open models work super effectively. And you probably know that the open models are not that far behind, maybe 6 months behind, depending on who you ask. And so we see robust growth in the enterprise. And look, I think companies are going through an understanding here where I would say there's different kinds of companies, right? There's companies that have said, wow, this stuff is total game changer, and we better do this or we're going to have a big problem if our competitors do it and we don't.

And then there's other companies that -- or organizations that say, well, okay, this is our budget, and it doesn't really matter what's happening in the outside world. We're sticking with our budget, okay? I mean that will kind of work until it doesn't work. But I think over time, there's going to be fewer of those. If you really believe that this is a game changer, and put me down for that. We see it in our own organization. We see it in our own productivity. If you look at our own ROI, if you want to measure free cash flow per person or revenue per person or gross margin per person or however you want to measure it, there's a ton of ROI here.

And it's -- we're, I believe, still at the beginning of this. Also, you need to think about in the past, right? Each of us processed data and tasks, and we had projects and tasks and e-mails and things, and we do those at human speed and we pass them off to the next person and organizations and technology that are in organizations are a function of what was available at the time they were established, okay? And they get updated from time to time. So now fast forward to, let's say, 2027, we have these agents that are able to process all these tasks and they work way faster than humans. They're way more accurate. They never sleep.

And so you can supervise tens or hundreds of those Agents and fundamentally change the way work is done. So there's just a whole rethinking and reimagining going on inside companies, but that will happen at very different speeds, right? Not every organization is going to be AI-pilled and just throw everything and redo their company immediately kind of like we are. But I think you'll see more and more of that over time. And you'll see a stark contrast in the performance of companies based on the rate and pace that they do this. We're kind of already seeing that.

Yes. No, it seems like the deluge of usage of Agents is just starting now, and we're starting to explore what these Agents could do and especially after OpenClaw and other recent developments, there's obviously a lot more focus on it.

Yes, that's right. We've gone from LLMs to Reasoning to Agents, and now we have this Recursive Self-Improvement. And if you think it's going to end there, you would be sadly mistaken. It's going to keep going. And a way to think about it is we have this platform for thinking and intelligence. And you're going to see a significant number of innovations on top of that. And can somebody predict exactly what those are going to be? -- not a chance. But hold on, it's going to be exciting.

Yes. No, that sounds very much like I think when we first entered either the Cloud era or the Internet or Cloud or Mobile, like each of these, we did not know this whole ecosystems that would form on top of that, which kicked off like many multimillion-dollar businesses. So it seems like there is a lot of innovation yet in store, and you guys are going to be obviously part of that. Maybe, Michael, to switch gears a little bit. Fortunately, world is a little bit in a tumultuous place and with the Middle East turmoil, do you see any sovereign build-outs that are potentially slowing? Or do you think that now with maybe more of each country looking to secure its own infrastructure, there's more of an urgency to become self-reliant and actually the demand is becoming even stronger from a sovereign perspective.

Well, geopolitical tension is the root cause for sovereign AI demand. And so let's suppose there's a scenario where there's tension in the transatlantic partnership. Who knows? Maybe that would be a thing. Well, all these countries in Europe, they don't want to be reliant on U.S. AI or U.S. anything actually. And so now we remind them that, well, actually, we can't make these things without ASML and -- there's all these companies in Japan that make all sorts of gases and chemicals in South Korea and Taiwan, and it's not just the U.S. And so dear customers, we're dependent on the world to make these things. It's not just the U.S. involved, but we have extensive factories in Europe.

And yes, I think Sovereign demand continues to grow, some really good partnerships like with Palantir and customers are looking for the ability to run AI inside their country with their own private data for all sorts of applications. And we've had some great wins there. And think of it as any country in the top 25 of GDP, there's some Sovereign thing going on there, whether it's a -- which whether it's the government themselves or a telco, or a separate company that is somehow affiliated or connected back to the government, there's definitely demand for Sovereign AI.

No, that makes a lot of sense. Maybe, Michael, like just looking at your AI server business, historically, Dell has had fairly significant negative cash-conversion cycle, and you kind of engineered that the whole concept across the whole PC industry to start with. But as you think about AI becoming such a sizable part, almost as large as the PC business, probably going to exceed it very soon here, how do you think about whether or not it's important to have this negative cash-conversion cycle across your company? Or does it matter in your long-term calculus on how you manage this business as you think about maybe the different capital requirements associated -- working cap requirements associated with the business?

It absolutely matters, and we're super focused on cash flow. And last quarter, our cash conversion cycle was flat sequentially. And we're taking all the best practices that we have and applying that to the AI business. And we're, as I said, pretty careful. We don't buy material until we have a PO from customers. And there's some CapEx, but it's on our part, but it's pretty limited. And we feel very good about our ability to generate strong cash in this environment.

Okay. Amazing. Maybe one of the questions we get often, right? Asked is just around the competitive differentiation in the AI market. You guys have just really taken this to a new level. I mean you've sort of beyond anyone what anyone could have expected, the amount of revenue acceleration that you've shown, your order book, you just had very strong performance. So what is it that's driving that differentiation? I know Jeff, sometimes and you talk about Level 11, 12 and deployment in data centers, you just mentioned time to first token. It just seems as though you're doing something different.

And there is something that stands you apart from a lot of your competitors who are going through several issues of their own, not just execution issues, but other issues beyond that. So as you think about, a, where your differentiation is; and b, what are some of the opportunities given maybe some of the missteps and other issues that are there with other companies? How do you think about that?

Yes. I would say we keep getting repeat orders from the same customers, and we keep winning over the customers that, let's say, were earlier reluctant or had decided to go with somebody else. And there can be many reasons for that. I don't think it's like just one thing. Certainly, we would start with our engineering. We actually do it and lots of it. And it's not just in the compute space. Obviously, storage is a big element of what we do. And we now have our Lightning file system, our PowerScale and ObjectScale are doing super well.

And building the whole system -- building all these systems to be reliable, the networking, then you get into the deployment, installation. We learned a long time ago that if you just ship these things to the customer, bad things happen. So we show up with an army of people and deploy them and install them. And by the way, we get paid for that. And then you have service and support, incredibly important, and these things are deployed in all sorts of places around the world wherever there's power. We have broad ecosystem of partners, obviously, our supply chain, DFS comes into play.

And look, we've been first-to-market now with the GB200, with the GB300. And I think the proof is in our consistent execution and customers coming back for more. Jeff Clark and I at GTC, we had a dinner with a bunch of the Neo-clouds and CSPs, maybe all of them actually were there. They said it was the first time they were all in the same place at one time. And the demand from those customers is very strong, and they're happy. They're happy with what they're getting. I spoke with one of the largest ones this morning and orders are -- their business is very strong. And look, they know they can rely on us. I mentioned Time to First Token. I mean we build these things ahead of time. These things are unbelievably complicated.

And we've perfected the precision logistics and supply-chain engineering where we can deliver hundreds of these racks in a given week like clockwork and have them show up and within 24 to 36 hours, they're up and running and they're generating money for the customer. So our competitors don't seem to be able to do that reliably. And I would also tell you that Jensen at his various performances, he does a great job on stage and things come up and it's like, here come all the servers and everything looks fantastic, but it actually doesn't quite work that easily. They have these things called Reference Designs. And the Reference Designs, I'll let you know a little secret that they don't actually work -- and there's a lot of bugs in them.

And we find all the bugs. And some of the bugs only they can fix and we tell them about those because only they can fix them. And the rest of the bugs that we find that we can fix ourselves without them, we don't tell them about those. And I think that's why we keep winning because we're building a more reliable product at the end of the day, and that comes down to engineering and discipline inside our whole Engineering Organization. And I think it's very different from other -- and look, we do what we say we're going to do. We don't over-promise. We say we're going to deliver it, we deliver it, and it works and it's reliable and you can count on it. And apparently, the other guys, not so much.

Yes. No, that's a super interesting point, Michael. So when you think about maybe an analogy to industry-standard servers where I think maybe the level of complexity is definitely lower and the level of differentiation might be lower. So the ODMs actually have meaningful share in the -- at least at the hyperscalers for industry standard servers. It sounds like the differentiation and the engineering could be so different here with AI servers that maybe the OEMs don't have the same kind of foothold. So the Hon Hais and Quantas of the world perhaps don't actually take as much share of the AI-server market potentially. Would that be something that you foresee as likely in the future?

Well, the way I would describe it is slightly differently, Wamsi. What I would say is that if you're a hyperscaler, you can afford to build a massive engineering organization to go and do some version of the kind of work that we're doing. And those companies do work directly with the ODMs, but I don't think there's a whole lot of companies that can do that. And certainly, it would not include the Neo-clouds. You've got also hundreds of these -- now cloud-native businesses that are consuming enormous numbers of tokens. No enterprise customer is ever going to try to do that. I mean you're talking about an enormous engineering effort to do that. And we still have some business with the hyperscalers, although that's not our main priority.

Yes. Yes. No, that makes sense. From a share perspective, right? You clearly have been growing very, very fast in AI servers. How do you think about a natural state of share for Dell in AI servers? And there's been some disruption definitely at some of the other larger peers, so to speak, who kind of struggled with various issues beyond execution. Do you -- in times past, has Dell been able to capture incremental share because of those worries and not just on a temporary basis, but more structurally over time?

Yes. We've gotten some phone calls recently, pained phone calls from some customers. I think in the enterprise, we're certainly advantaged there and just don't see the other competitors as much there. And yes, I don't really know on the share. I mean we're often asking the question how high is up in terms of the size of the opportunity because we just keep seeing incredible waves of demand and the lead times keep getting longer and the orders keep coming in. So, we're basically seeing some super robust forecast for future demand and responding accordingly. But in core servers and storage, as I said earlier, we are quite a bit larger than others, and we seem to be growing. So...

Yes. Yes. No, for sure. Any going back...

I think we're still in the early stages of Agents and Multi-Agent Systems, and Recursive Self-Improvement that will drive demand even further.

Yes. Yes. No, that makes a lot of sense. Going back maybe to Michael, your point earlier about the near-death experience these memory companies have and had like back a few years ago, 3 years ago. And now looking at where we are with memory price inflation, it sounds like not much new capacity is going to come online. I think overnight, Samsung reported some preliminary numbers, extremely strong memory price increases, both in DRAM and NAND for Q1, Q2 expected.

So can you maybe -- like I think people were very surprised to see your fiscal year guidance because people felt like with this memory price escalation, I mean, there is no way that anyone can navigate that to not have a down year-on-year for earnings. You had committed to a long-term framework of 15%-plus EPS growth. And frankly, you've driven higher than that now. We're talking about 25% earnings growth in this fiscal year, which is just astounding.

So is it like maybe you can help explain to investors that the secret sauce behind your ability here to overcome this intense margin pressure that the rest of the industry is seeing? And I'm sure you're seeing increased costs as well, but you're somehow able to navigate it much better than others. So what's actually happening behind the covers? What are some of the levers that you are using in, maybe across your business lines that is helping more than offset the challenges of memory-price increase?

Well, so -- it shouldn't be super surprising that we're raising prices to protect gross margins. And we're very agile in doing that. So, we're able to execute that very quickly. And we kind of saw what was happening, and we made appropriate changes. The other thing is going back to what I said earlier, we're in a good spot in terms of supply. Everyone is subject to the price increases. But again, let's say you're at Bank of America or whatever bank you want to pick or whatever customer, it would be bad if the price of memory went up, but it would be worse if you couldn't get any, right? And so we've translated that situation into a successful outcome. I'll say there's probably some tradecraft in there that I'm not going to explain, but we're doing very well in this environment.

Michael, you mentioned price increases. And obviously, some of this is pretty visible externally. This industry-standard server pricing has gone up 30% to 100% in some SKUs. There's been significant price increases across the board. So when you think about the impact of these price increases on customers, what sort of buying activity are you thinking? How does that change the buying activity of these customers? Do we anticipate a pull-forward of demand? Or is the demand so strong and the value proposition of replacing a server is so great that we're going to see continued demand?

Or is it like given your secret sauce and tradecraft, like you're able to take share. So even though there might be a pull-forward in the first half, your share gains in the second half will let you do better than the rest of the industry. Like can you frame how to think about how the customer demand might be changing and the seasonality of that could be changing because of what's happening with memory prices and your price increases?

Yes. I'm sure there's some pull-forward, but we don't know how much. We also know that there are some customers who will say, oh, the price has gone up too fast, I think I'm going to wait for it to come down, right? And who knows? Maybe it will come down. I mean I'm sure at some point, it will, but probably not. It doesn't look like it's happening anytime soon. So, the problem, of course, is you can delay the purchases for a while, and there is some demand destruction, certainly in low-end phones and low-end PCs. But there are about 1/3, about 500 million of the 1.5 billion PCs in the world are 4 years old or older. Okay? And so if you work at a company, let's say, with knowledge workers and now you've got a 4-year-old PC, this is a bad situation, right?

And so you're paying somebody $100,000 or whatever to do their job, and they've got this decrepit tool that doesn't work very well, and they know it. And so yes, maybe you don't like that the price went up, but eventually, you're just going to pay the price. You can delay, you can defer. But -- so yes, it's almost a question of when are they going to buy, not if they're going to buy. And then with servers, we still have the majority of our installed base of servers are 14G or older. We're on 17G now. So that's a 7:1 consolidation benefit.

And so I think customers are starting to figure out you're going to -- your budget for tokens has to go up. And so this just costs money. Obviously, we've put all this into our guidance and with respect to the second half. But yes, I do think these are the kind of environments where we tend to gain share. And I also think different from the kind of 2020-2021 cycle, you've also seen us scaled our cost structure into a pretty extraordinary way. So we're kind of fit to fight here for the future.

Yes. Yes. No, great points. Michael, you mentioned, obviously, some of your relationships with these memory companies go back 4 decades. When you think about what the hyperscalers are doing, they're starting to try to have long-term agreements with a lot of these parts of the supply chain, whether it be components, whether it's memory, whether it be optical. And they're investing aggressively into these companies in some ways.

And so you've obviously done a phenomenal job. You're just saying like, hey, you've got enough supply to take care of your guidance and you're trying to get more supply. So is there -- as you think about your discussions with these suppliers, is that changing in any way because of what these hyperscalers are trying to do, which is just really go in and try to lock up as much supply as possible?

What I'd say is we have had long-term agreements with a lot of these companies going back decades, and we continue to. Obviously, the nature of those adjust given the environment and the demand from the overall industry. Also, I think you have to understand this whole intelligence thinking platform, it just uses memory in a very different way. And so that's a big change here. It's a lot less expensive to take a token that's already been generated and store it and pull that out of memory than to regenerate it.

And so you end up with a much faster-growing, pyramidal system of memory from SRAM inside the logic to high-bandwidth memory to DRAM to SSDs, to ultimately rotating storage still around. And so yes, we have agreements with them. The demand is extraordinary. We have the supply for the demand that -- for the guidance that we've provided for the year, and we're working on getting more.

And generally, from your perspective, the flavor of these long-term agreements that you've had is generally where there is a supply guarantee for Dell and then pricing could be some variable in nature. Is that the right way to understand most of these longer-term agreements? Or would you characterize that as not right?

Yes, that's generally right. I think also while some of them can be very opportunistic in periods like this, they're also quite concerned with the long-term health of their business and who the offtakers are over time. So what do I mean by that? These guys are investing tens of billions of dollars in these plants, and they want to know how they're going to sell their parts in the next quarter or 2, sure. But they really want to know how they're going to sell them in the next 3, 5, 7, 10 years.

And we've always been a predictable, reliable, let's say, low-volatility customer, right? We don't show up when there's a disaster and say, "Oh, we've got to have a whole bunch of memory. We're there all the time. And we've been there all the time. And so that, I think, is super helpful and also just the breadth of our demand signal and the quality of our demand signal is appreciated.

And these guys have been whipsawed in the past by the hyperscalers where they overbuild, then the hyperscalers turn everything off and stop buying. That does not happen with our demand. Our demand is a lot more stable, and that's appreciated, and that's part of the relationship. And also, we treated them like long-term partners, and they've generally reciprocated.

Yes. Yes. No, that's a great point. Time is flying. We only have 5 more minutes here, and there's so much more to talk about. So maybe quickly here, I do want to touch upon what Dell is doing internally with AI. You mentioned Agentic AI. You mentioned productivity, different metrics, whether you look at it on a per employee, profit metric, revenue metric. What are you doing with Agentic AI within the organization? And where are you in that deployment like cycle?

So, I would say we're rapidly hurling toward it. And if you look at our modernization, we've been on this theme for several years. You've been hearing us talk about it. Our OpEx has come down for 4 years in a row, which is pretty unusual. And this kind of started going in 2023, and we think there's still more to go and significant scaling benefits. And we've guided OpEx to a single-digit percent of revenue. That hasn't happened in 20 years.

And of course, we have an even more R&D-intensive business than we did 20 years ago. It's 100% structural. There's tons of opportunity and -- we're doing this while we're investing in our R&D, in our sales capacity, in our support, and in our supply chain. So I would say I'm super bullish on our ability to generate productivity. And ultimately, that makes us a much more competitive company, opens up paths to profitable growth, as you've been seeing. So...

That's quite amazing. I mean, yes, no, you're adding $30 billion more on top line with very little incremental OpEx, so quite amazing. Maybe Henrik, in the few minutes that we have left, on a prior call, I think with me, you said you are the ultimate long-term investor. How do you define success for Dell in the next 10 years?

I would say the last 5 years, we outperformed the Mag 7 with the exception of NVIDIA. But in aggregate, we outperformed the Mag 7, including NVIDIA. But we outperformed every component, except NVIDIA. Certainly, we look at our relative performance to the S&P IT index and we see tremendous opportunities to continue to grow, generate strong cash flow. You've seen our capital return policy has been super shareholder-friendly.

Buying back stock, we bought back 54 million shares of our stock. We increased the dividend by 20% -- and I think you'll see us continue to generate strong revenue, earnings growth, free cash flow, and the business should be a whole lot more valuable. And certainly, if we -- certainly, we would endeavor to outperform all the relevant benchmarks by a lot.

Yes. You've been -- you are continue to be a technologist at heart, and you have taken this company from basically you starting it to this amazing scale of now $140 billion a year in revenue with very fast earnings growth. So I can see you're very excited about the future of Dell. Any final thoughts for investors why they should be just as excited as you are?

Yes. So, in the last 5 years, right? We doubled EPS, 15% EPS growth. And we talked about that as our long-term framework for the next 5 years. ISG is super well positioned. Obviously, you know about our guidance for this year. Our portfolio has never been in better shape. Our operating model is super strong. We've got relationships with customers and supply chain. And we have a ton of levers in this business to drive EPS and cash flow, shareholder return. So I feel very good about the future opportunities for our company.

Amazing. Well, Michael, we really appreciate your time. This has indeed been a great discussion. Always love your insights, love the work you're doing for our country, for the future. Just incredible always to speak with you, and it's been an honor and a pleasure. I look forward to catching up with you again. And thank you so much for all your time and your insights.

Thank you, Wamsi. Thanks, everyone, for joining us today. Take care. Bye-bye.

Thank you, everyone.

Well, let's get started here. Welcome to day 3, again, the afternoon of day 3. My name is Erik Woodring. I lead the hardware coverage here at Morgan Stanley. I'm delighted to be joined by Dell Technologies' CFO, David Kennedy.

Before we get started, a few things. So from my end, please see Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative.

And from the Dell side, statements in this presentation that relate to future results and events are forward-looking statements and based on Dell Technologies' current expectations. In some cases, you can identify these statements by such forward-looking words as anticipate, believe, could, estimate, expect, intend, confidence, may, plan, potential, should, will and would or similar expressions. Actual results and events in future periods may differ materially from those expressed and are implied by these forward-looking statements because of a number of risks, uncertainties and other factors, including those discussed in Dell Technologies' periodic reports filed with the SEC. Dell Technologies assumes no obligation to update these forward-looking statements.

So David, first time at the TMT conference. Earlier in your career here as CFO at Dell Technologies, a lot to talk about, dynamic world. Thank you for joining us.

Yes, I appreciate it. Thank you.

So I think the best and clearest way place to start, you reported earnings less than a week ago, last Thursday night. There's a lot to unpack, but I want to start kind of high level, which is you guided to over 20% revenue growth, 25% EPS growth in fiscal '27. You have tremendous momentum in AI servers that we'll kind of dig more into. You're also extremely bullish on kind of core PC server and storage. So high level, just give us a lay of the land as you see it today and how that influences as you look out over the next 12 months?

Yes, sure. And I think, look, I guess, from a top-down perspective for guidance, let's start with revenue, right? So as we did our results, we talked about our $43 billion of backlog from an AI perspective. That's a great anchor tenant from a revenue perspective as we get into the year and look at the growth that's potential that's out there. You look at our Q4 results, record in terms of revenue and EPS, both on an accelerating curve as we went through the year and back ended through that. So you come into the year ahead and you think you're pretty bullish then from a revenue and an AI perspective. When we look at the core business, as we've guided Q1 is more near term. As we look at Q2 to Q4 and the rest of that year, look, we've tried to stay prudent from a revenue guide perspective in the core. It's pretty dynamic out there, as everybody knows.

So from a supply chain perspective, making sure from a cost perspective, pricing, I'm sure we're all going to learn in terms of elasticity and how that looks for the year ahead. But given the trajectory, the customer conversations we're having, the problems we're trying to solve for our customers, the revenue piece is in that range then for guidance. EPS, last year, 27% EPS growth, guiding to 25% growth for EPS this year. And look, we feel we have a robust set of tools in the toolbox. So 4 key levers really. One, the revenue lever, which is a big component, and we've just touched on that. Two, from a margin rate perspective, as we look through there, we'll talk about, I'm sure, by line of business, but really enthused with our Dell IP storage portfolio and the size of that and the capability of that to mix margin rate enhancement for us.

Number 3, we've been a consistent execution around OpEx management, particularly as we scale the business and show this decoupling between the revenue growth and controlling OpEx dollars, which obviously gives you the scale that you get in the P&L. And then look, fourth, we've been very, very consistent in our execution from a working capital perspective in relation to giving back the cash that we generate back to our shareholders. So 54 million shares repurchased last year, 14.9 million shares in Q4. We're going to keep a consistent posture on returning our cash flow to our shareholders in the year ahead. So you put those 4 things together, and we think we've good agility from an EPS perspective as we head into the year.

Okay. All right. That's a perfect place to start, and I'll kind of dig into each one of those. So again, you started kind of what you were talking about with AI servers. I'm going to do the same here, $34 billion of orders in the quarter, backlog of $43 billion, a pipeline that is growing sequentially despite the fact that you're able to convert a significant amount of that pipeline into orders. And again, your AI server business, as you guided, is effectively doubling again, something that's gone from nothing 3 years ago to expectations for $50 billion in the coming year. That doesn't make the audience bullish on AI CapEx after Jensen earlier today, I don't really know what does. But outline for us what you're seeing in terms of trends, demand at the market level, but more importantly, what you, Dell Technologies are doing to be that clear market-leading OEM?

Yes, sure. You've given all the stats there, which kind of bear fruit as we've executed. I think the key one for me in that is we went from a net new $34 billion of demand sequentially in the quarter, but yet our dollar -- pipeline dollar value has never been higher now than it has been in history, right? So you think of converting the pipeline that we saw in Q4, and we've replenished that and see the opportunity that comes. That has come in a fairly broad base. So the statement holds true whether you talk neoclouds, whether you talk sovereign opportunities or whether you talk the enterprise expansion.

From an enterprise perspective, we talked 90 days ago when we did Q3 earnings about having 3,300 enterprise AI customers in our installed base. Just last Thursday, we updated that 90 days later to be 4,000 plus. That sequential jump, that was the strongest enterprise quarter we've had. And now you look at our pipeline, all pipeline verticals grew. The percentage increase that was the fastest was that enterprise cohort as well. So we kind of see it across the various customer stacks that are within the AI portfolio. All of that points to an accelerating appetite for compute pretty much.

And can you maybe touch on the secret sauce that Dell has been able to use to differentiate itself from others in the market? Again, I look across my coverage, nobody is doing what you're doing in terms of AI servers. So where does the secret sauce come from for you guys?

Yes. Look, from us, from an execution perspective, number one, from an engineering standards expectation, you go into power cooling, extend out into the installation, the deployment, the service. You put all those suite of execution together to build, deploy and make sure that when we turn the lights on, it turns on, it stays on and it stays up over 99% of the time. All of that's executing. And I think we've plenty of reference big rollouts in our favor now to showcase that's a differentiated capability that we've seen. So we want to build on that as we kind of go forward. And look, the exciting piece is we work with these customers well upfront. We're part sometimes of the data center design process. We work the way through the architecture and the deployment schedules, complexity we expect to see. We're taking the learnings in and replicating that as we kind of go forward. So it's serving us really well right now.

Okay. Maybe a follow-up to that. A few years ago, Arthur was on stage here talking about kind of the robust storage and services attach that you get with AI servers. Talk to us about that opportunity now 2 years later. Does that opportunity exist? Compute very much seems to be the priority right now. But maybe talk to us about how -- whether that opportunity exists, contextualize it for us a little bit.

Yes, for sure. Look, again, the majority of our AI server deployments that we're doing, we have service attach tied to it. So it's all part of the pricing, the value that we extract as we go execute. A huge opportunity from a storage attach perspective as we go forward. We talked about Project Lightning. That was one from a Dell perspective. We're on track to be in production and deployment of that here in the first half, and the parallel file system that -- it's part of the ecosystem. And as an extension of that from a use case from an enterprise perspective as we work with those customers, a unique opportunity there also. So yes, it's very much an opportunity in front of us.

Okay. Okay. Maybe last question specifically on AI servers is you had your Analyst Day 5 months ago. You outlined a path to 25% annual growth for your AI server business over the next 4 years. You're doing that this year. You've pulled it for 3 years. Has the market accelerated just that quickly in the span of time? Is there something you're doing differently or been able to capture share that maybe you didn't necessarily expect to. Just talk about how you've been able to kind of pull that forward...

I think it's all of the above, quite frankly. Look, if you go back to our Analyst Day, I think the joke at the time there was when we called the AI number 2 years previous to that, we were wrong and we were wrong by a lot. I think we predicted when we did last October the chances that we'd be wrong again on the upside were realistic, if you like, right, as part of it. But obviously, just from then to now, an accelerating appetite for investment for sure. We would also dovetail on that in terms of the deployments that we've done and the execution that we've done as part of that. We think we're starting to differentiate ourselves in terms of taking on some of these big large projects that are in place, doing them at scale, showcasing the value proposition that we can bring along those things I mentioned earlier.

In particular, once we turn it on, it stays on and it works and it deploys at scale. And I think from there, as a lot of our customers go into purchase #2 or 3 or 4, we're building a strong reputation with them and the building blocks keep going then. So yes, a lot of opportunity.

And maybe just a follow-up to that, the longer-term follow-up is that -- so is there a way that you can kind of help us think about now that trajectory from here? Obviously, there's a lot of dynamicism when it comes to kind of the compute market. Again, you heard Jensen earlier today talking trillions and trillions. So what's the longer-term outlook? And then second to that is, how do you make sure that you protect yourselves against some of the lower-priced Taiwanese ODMs, making sure that you can defend what you've done and stay with the customers that you have and grow with them?

Yes. So in terms of the future, I guess let's talk next 5-quarter pipeline, which is generally kind of the metric we'll be staring at in the meantime. The dollar value of that pipeline, like I mentioned, has never been higher. So as much as we did $64 billion of orders last year, $25 billion of shipments, $43 billion of backlog. We've committed to $13 billion of Q1, which is effectively $1 billion a week of shipments going out the door. But if you look at the size of the pipeline, a big mix of Vera Rubin opportunities within that now, whereas obviously, our backlog is predominantly Grace Blackwell backlog that's there from that. So that's a good technology changing point that we're observing.

So we're going to be aggressive to get out there and look to partner and win that business in addition to try and secure supply to kind of go drive that level of scale. So we'll do that. And back to your second part of the question, look, again, our value as we look at the production cycle here from the L11 scale and beyond, those engineering standards, those activities, I'm sure we might touch on OpEx in our P&L later. But within our OpEx framework, we are making sure we're investing, investing in our go-to-market teams. So as AI opportunities expand in the enterprise, we're building out the right pod structures and the right capacity to go execute that. But also two, giving Arthur Lewis and his team, the engineering capability, the labs and the investment. So they're not working just on, let's say, Vera Rubin, but the next gen beyond that and beyond that. We're ahead of the game right now. Our job is to stay ahead. So we maintain the value that we can find in our P&L.

Okay. That's a perfect way to end kind of the AI server discussion there. Let's move to the traditional side of the business. So you guys outlined a scenario in your fiscal '27, calendar '26, so to speak, where there is an unprecedented memory cycle, but you are extremely confident about both being able to grow and expand margins across kind of that core business -- just talk about how you're able to do that.

Yes. I think this is not business as usual. So all playbooks don't necessarily work. We're probably taking a lot of what we learned from the last cycle, albeit not as acute from a COVID timing and accelerating new things as part of that. Traditionally, in this environment, you might have heard a lot of narrative around, hey, if you can recover 2/3 of the cost increase in 90 days, that's good execution. That would be failure in the environment we're in. This is event management. It has to be totally different than that. So the speed and decisiveness of decision-making just has to be different. We think we're really good at execution, and I think we can differentiate ourselves.

And to give you one tangible example. So back on December 10, we would have looked at thousands of our traditional server quotes ready to go out, pulled every one of them, every single one of them back out of the ecosystem, went dark for about 36 hours, replenished them all back out into the field and ultimately saw minimal demand destruction in that initial period, but saw the instant immediate margin protection and stabilization within that period, if you like, right? So we've executed that. That muscle and that speed, we're going to rely on as we go through the rest of the year given we expect costs will continue to go up, hopefully at a slower rate, but continue to go up. But we're going to learn. We're going to observe different lines of business, react differently to elasticity and various things like that. And we're going to back ourselves to be really decisive in our decision-making and make good calls.

Okay. And I don't want to kind of ask the high-level question across elasticity because each business is going to respond differently. So let's maybe touch on PCs first. You're guiding to low single-digit year-over-year revenue growth, 1% year-over-year revenue growth. We're coming off of what was -- maybe we had the strongest PC growth rate in quite some time, although it was a bit out of a way to get there, right? You had tariffs and you had Win 11 and you have old PC refreshes. Talk about the drivers of demand as you see them this year. And especially, you mentioned prices are going to move higher. How do you make sure that, that elasticity kind of stays below 1, so to speak. So no matter how much you're raising prices, even with the unit offset, you're still growing.

And that's revenue. Yes, correct. I mean -- so what we guided to was revenue growth. Purposely, we didn't guide to unit growth because, again, there's a vector from an elasticity perspective there. Look, if you step back and look at the PC market, we have a refresh to Win 11. That's probably the slowest refresh we've seen, about 10 points behind historical refreshes. There's still hundreds of millions of Win 10 PCs in operation out in the field. So based off that, there's still a robust refresh to take place. But now in the event management of the supply environment, we expect units to start the year, we talked last week on our call, minus 11%, minus 12% market, which kind of is at odds with a refresh cycle that should be happening.

So there's still a lot of units to refresh, and it looks like it will take longer to do that as part of the conversation. The next question then is as and when prices and costs change and go up, what's the reaction to the unit velocity as part of that -- we're going to learn as part of that. But we do expect the back half of the year to be deeper, probably high teen double-digit negative growth. But we're going to stress test elasticity. I mean, to your point, we have priced. If we have to price more and it takes some units out of the ecosystem, well, then we still feel our revenue guide is pretty solid. And obviously, in our modeling, we've kept a few points of prudency in that to give ourselves a little wiggle room.

And Jeff has taken the helm at CSG, obviously, you hear him on the call. He is fired up about what he's doing with that business. Is there any change in approach that you guys are taking with CSG because you'll hear him be critical of the team on a backwards look and say, "Hey, we are an execution machine, we should be doing better for ourselves." What are you guys doing that's different to drive kind of that excellence that he speaks about?

Yes. So I guess 2 things. First on the products. So we announced probably 90, 120 days ago, we needed to expand our TAM. So the low end of commercial, education and in consumer, we launched new products and set out a broad base to touch more of the TAM from a share perspective as we do that. So that's launched, that's up and running. And then from a perspective, it's getting back to what we talked in terms of execution, but trying to do it at scale that the teams haven't seen, if you like, right? So I mentioned that 2/3 of cost in 90 days, we're showcasing, look, that doesn't work. It's not good enough to execute at those levels.

But if you think of even when we executed really well in a supply-constrained environment like COVID, we've modernized our company a lot more since then. So we have way more systematic tools now to raise discounting, move discounting up and down to change list pricing, sort of the speed of change that we can get through the system, we measure it in hours, everything from like weeks or months. So it allows us to get to market much, much quicker. And then it's nothing like event management to focus the mind and if you like, war room it and be ready to go in terms of execution.

Let's put kind of the memory dynamic to the side. And if we just take a kind of bigger, broader look at the CSG business, you've guided at your Analyst Day to kind of low single-digit growth, high end of low single-digit growth there. How does Dell -- or maybe the question is, is what's going on in the marketplace, for example, a Win refresh cycle that's just taking a bit longer new products leaning into expanding your TAM. Does that change at all how you think about CSG on kind of a multiyear period?

No. I mean, look, CSG is part of our DNA. It is the most capitally efficient line of business we have in our P&L. So very much part of the complementary piece that we have in terms of our broad stack. Now more than ever, you see that opportunity in a supply-constrained environment, where we sell a broad-based structure all the way from a consumer PC up to the biggest data center evolution. That's giving us leverage to a degree in relation to getting supply and having those strong relationships that allows us to balance some of those things off. So for lots and lots of different reasons, notwithstanding the big refresh of hundreds of millions of PCs, it's going to be a vital part of our P&L for a long, long time.

Okay. Let's shift to storage. Your outlook implies, call it, around low single-digit year-over-year growth -- revenue growth this year, kind of a multipart question, which is how much of that is kind of pure market momentum behind enterprise AI? How much of that is just Dell outperformance or expected outperformance and share gains? Why don't we start there?

Yes, you're right. So we've guided to low single digit for the year as part of the opportunity. Look, I think we're at a really cool inflection point for our storage business. We have a Dell IP portfolio. So it's not new. It started with PowerStore. We are now on -- from a mid-range perspective, we are now in 8 quarters of growth, 7 quarters of double-digit growth. As we've expanded out with PowerMax, PowerScale, PowerFlex, -- that portfolio as an entity grew 12% on a demand basis last year, almost 2x the market and exited the year at the fastest acceleration.

So we've been working hard with Arthur and his team to up our R&D and kind of get those products rolled out. We're seeing that now, and that's resonating. So that opportunity is probably the biggest piece from a scale perspective. You see the outcome of that in terms of profit already in our P&L. If you look at the Q4 results, our Q1 guide, the mix of Dell IP storage is the biggest lever as to why we say our core margins, excluding AI, can show a bit of expansion. It's that Dell IP storage factor that's there and building on what is now multi-quarter growth trends as we look out into the year ahead.

And the follow-up to that is just you touched on it earlier, but Project Lightning was something that Michael teased at Dell Technologies World. I think it was 2 years ago -- a year or 2 years ago. It's going GA in the first half of this year. What does that do for the storage portfolio?

Yes. So yes, it's here. First half, we're on track, and we'll be putting it in customers' hands and shipping here in the first half of the year. Obviously, it's very complementary from an AI workload perspective. So actually, it's really great timing as we see this accelerating demand from an AI GPU perspective. We can talk about a solution of technology that goes beyond just the compute piece and actually build it up through the stack that we can offer. So again, we're looking to showcase to be the #1 technology partner as customers build out their enterprise suite, not just the GPU piece itself, but we can expand now into storage and having the products there like Project Lightning to enable that is exciting for us.

Okay. Okay. And finally, on the traditional server business, maybe the question is, does kind of a once-in-a-generation kind of [ price hike ] hinder the momentum behind Dell, what you're seeing in kind of the CPU-based server market? Just talk us through how you see the market -- the CPU-based server market evolve in 2026 at both kind of the macro level, but then Dell-specific level.

Yes. I mean if you look at the macro level of the market, there is still -- the vast majority of units are still 14G and haven't evolved into the 16-17G opportunity that's there. So the installed base and the desire to continue to modernize the data center is still prevalent, which means pretty strong growth, if you like, in terms of that market. It also then benefits just from the tail of general AI workloads that's there as it comes off the edge in relation to a standard enterprise customer. So that showcases a robust environment. We're seeing that. So if you take the price increases got into the marketplace since December to today, it's had very little impact on demand. The demand is still outstripping supply in the marketplace in relation to that.

So that's how we've kind of guided our year ahead for traditional server. We've guided to kind of high single-digit growth. But I would say I would kind of give you 2 pieces to that. One, strong double-digit growth in Q1, very reminiscent to Q4 and just holding a slightly more prudent view for the rest of the year. We have opportunities to expand beyond that. But again, we're going to learn a lot in terms of elasticity as we go.

[ Fair enough ]. Something that I think is underappreciated is how much you guys have been able to scale OpEx. So if we look point-to-point on the last 5 years, your revenue base has grown $40 billion. Your OpEx base is down $1 billion. That's kind of unprecedented as I've seen it in my coverage. And you are still investing in future growth opportunities. So just talk to us how you've been able to do that, but then also how sustainable that is as you look forward and you think about these emerging growth opportunities that are still on the come?

Yes. The scale is obviously pleasing and it gives so much flexibility to the P&L. Obviously, you get the scale from the revenue piece. We'll put that to aside, you guys see that. Look, from an OpEx dollar perspective, we started and almost like a customer 0, we started our modernization, standardization, optimization over 2, 3 years ago. And in the early streams of that, a lot of it is doing reprocessing work, like literally just pulling apart different ways we do work and being aggressive with that footprint to look at that. And that gives us a lot of outcomes, if you like, right, in all functions. And I can give you some examples as we kind of chat through that.

So that enabled us to take OpEx in dollar terms down the last 2 years in a row. So we've guided for this year, OpEx, it's slightly up, like flat to plus 1%, 1% is the guide that we've given. But there's really no change in the posture of what we're trying to do. So as we go forward, our modernization efforts are going to continue. The next wave is to layer in more and more Agentic scale into the company. We have lots of use cases already. As an example, even in my own finance world, using Agentic capability to book journals, invoice, collect cash, do all those various things. So very much alive and well. All my peer group within the company are aggressively modernizing at the same time.

So we're going to continue to do that and find scale from an OpEx dollar perspective. But we've guided to make sure we have flexibility to invest to, like I said earlier, the sales side and more importantly, in Arthur's R&D side from an AI perspective to keep that value proposition ahead. So yes, we still think there's plenty of road to travel here in terms of opportunity. And what better way to showcase the power of AI to an enterprise and actually showed in our own company's performance. And we're giving lots of those use cases as we talk to many reference accounts now across the enterprise.

Okay. Before I touch on capital allocation and kind of a final question to end on, I just wanted -- the speed to market that you guys have been able to have in terms of this canceling purchase orders overnight, reprice -- going down for 36 hours, repricing, it is kind of unprecedented, so to speak. You talked about how the old model was offsetting 2/3. Now it's -- we're stabilizing the night of. How does Dell just protect against demand pull forward? I know you've kind of said there might be a little bit in there, but how do you protect against making sure that, that doesn't become a headwind?

Yes. Look, I think one of the gating factors is obviously the supply in the marketplace. So I guess I should have mentioned at the start, one of the anchor tenants of our guidance for the year is we have supply to achieve our guidance, not just for Q1, but for the full year. If we were to look at some of the demand asks that we would see from customers today and replicate that for a full 12 months, there isn't enough supply to fulfill that. Our job and what we want to do is try and help as many of our customers as we can. So within that, there's a relationship thing that matters there where, hey, if -- because it's pretty obvious we can see if there's a demand ask pull in. We know the historical buying patterns and AMO that's in those accounts that you can see.

It's like we're going to ask you to trust us to say, Hey, show us the real trend of linear requirement you have from your perspective. And if we can get into a supply access conversation then lock you in over multiple periods, and you can lock on that and then we can go to the next customer and work our way down. So we're just not in a position to -- and we don't want to put all the supply into a more finite set of customers who don't need the compute power yet and leave some really reference accounts and customers that we have in our installed base stranded, quite frankly. We're trying to maximize the number of customers we can help with the supply that we have. We're going to keep doing that all year as we can do.

So turn to capital allocation, a really exciting story at Dell Technologies. So you've raised your dividend more than 10% annually for the last 3 consecutive years. The Board just authorized another $10 billion buyback authorization. I think your goal is 80% plus of free cash flow return to shareholders. You always leave that plus at the end to make sure it's kind of open-ended on the upside. Just talk about the philosophy behind that, the opportunity to perhaps deliver on top of that, which you've been able to do opportunistically. Where could we see that 80% actually trend into that plus part of the 80% plus?

Yes. I mean, I guess inception to date since we launched this program about 3 years ago, we're at 84% inception to date, so slightly ahead of that 80%. It can go up or down a bit in any given quarter given cash performance or if we can see opportunistic areas from a stock repurchase perspective to do. If you look back at the year we've just completed, there was -- our solid framework of repurchase, which is programmatically set. That's our baseline. And then we did 2 opportunistic buying patterns in FY '26. One was last Q1 when the market was pretty bearish. We -- I think, repurchased as many shares in Q1 last year as we did the whole previous year.

And then the quarter we just finished, we actually increased sequentially our share buyback by $700 million to $1.9 billion, took 14.9 million shares out of the market. Again, that's a reflection of us seeing a strong cash position coming in that we could see, looking at what is a kind of a bearish or volatile market that's out there and being opportunistic then to give back and find that value for shareholders. So we don't guide to cash, but we expect a pretty solid cash here again as we go forward, nothing from normal levels you would see from us. And then our commitment is to make sure we put it to good use and continue to reward the shareholders.

Okay. We're budding up against time here. So we covered a lot of ground today, servers, traditional business memory costs, all of the good stuff associated with Dell Technologies. I appreciate the candid insights and just kind of want to leave you with a final word here for everyone in the room, which is either what most excites David Kennedy as we look forward, what you think might be underappreciated as we look forward, and we can end there.

Yes. Look, we're coming off what was a record year for our company, highest revenue, highest EPS that we've had. And we want to make that all news, and we're guiding to make FY '27 the best ever, right? So as we look at that, we've never been more confident in our agility in terms of the ways we can get there and how we can do it to go execute. So we mentioned at the start, our revenue opportunities is an accelerating appetite in the market. Our execution around margin and Dell IP storage from a margin rate accretion perspective, continuing our OpEx scale, which we have a very tight rein on the dollars being spent and then putting cash to good use. You put those things together, we think we have a pretty robust framework to kind of go execute and gives us flexibility to lead into different things and try different things as opportunities emerge.

So we couldn't be more excited about -- we see the environment right now as an opportunity. We want to solve as many customer problems as we can. The phone is ringing and they're looking and coming to us in times of uncertainty, and we want to be there for them again this year. So yes, a great year ahead already.

That's a perfect way to end, David. Thank you very much.

Appreciate it. Thank you.

Good afternoon, and welcome to the Fiscal Year 2026 Fourth Quarter Financial Results Conference Call for Dell Technologies Inc. I'd like to inform all participants this call is recorded at the request of Dell Technologies. This broadcast is the copyrighted property of Dell Technologies, Inc. Any rebroadcast of this information in whole or part without the prior written permission of Dell Technologies is prohibited. [Operator Instructions]

I'd like to turn the call over to Paul Frantz, Head of Investor Relations. Mr. Frantz, you may begin.

Thanks, everyone, for joining us. With me today are Jeff Clarke, David Kennedy and Tyler Johnson. Our earnings materials are available on our IR website, and I encourage you to review these materials. Also, please take some time to view the presentation, which includes additional content to complement our discussion this afternoon. Guidance will be covered on today's call.

During this call, unless otherwise indicated, all references to financial measures refer to non-GAAP financial measures, including non-GAAP gross margin, operating expenses, operating income, net income, diluted earnings per share, free cash flow and adjusted free cash flow. A reconciliation of these measures to their most directly comparable GAAP measures can be found in our web deck and our press release. Growth percentages refer to year-over-year change unless otherwise specified.

Statements made during this call that relate to future results and events are future-looking statements based on current expectations. Actual results and events could differ materially from those projected due to a number of risks and uncertainties, which are discussed in our web deck and our SEC filings. We assume no obligation to update our forward-looking statements.

Now I'll turn it over to Jeff.

Thanks, Paul, and thanks, everyone, for joining us. FY '26 was a defining year in our company's history. We delivered record full year revenue and EPS. Revenue reached $113.5 billion, up 19% and EPS grew 27% to $10.30. We converted that performance into record annual cash flow over $11 billion and returned $7.5 billion to shareholders, including 54 million shares repurchased more than doubled last year.

The AI opportunity is meaningfully growing and transforming the company. In FY '26, we closed $64.1 billion in AI orders, shipped $25.2 billion and exited with a record $43 billion in AI backlog, powerful proof points that our engineering leadership and differentiated solutions are winning. We are executing with discipline and speed. We are gaining share in our PC business and strengthening ISG with strong margins in traditional servers and storage, all while positioning the company for the AI era. It was a monumental year. We exit with strong momentum, and I couldn't be more proud of this team.

With that, let me turn to the key highlights for the quarter. We delivered a record quarter. Q4 revenue was $33.4 billion, up 39% and earnings per share was $3.89, up 45%, driven by disciplined execution and demand for our AI solutions. While operating in a dynamic environment, we saw record cash flow generation and above-trend capital returns for shareholders.

Now let's move to AI. We capped an already strong year with an exceptional quarter for AI, record orders and broad-based demand. In Q4, we booked $34.1 billion in AI orders, evidence that demand is accelerating as customers deploy AI at scale. We shipped $9.5 billion in AI servers in the quarter. We exited Q4 with a record $43 billion in AI backlog, and our pipeline continued to grow sequentially even after converting $34.1 billion of orders, a clear sign of sustained momentum. For the full year, AI orders reached $64.1 billion.

Our customer base surpassed 4,000 with growth across neoclouds, sovereigns and enterprise customers, evidence that demand is broadening across all customer types. We're winning for the reasons we've outlined all year, engineering for performance and time to market while optimizing TCO for AI workloads, deployment and installation at speed and scale, ongoing life cycle support that keeps clusters up and running and DFS financing. We're doing this with discipline. Profitability is in line with our mid-single-digit operating margin target. We like our position, the line of sight we have with our backlog and pipeline and the advantages our scale and supply chain bring.

Moving to traditional servers. Demand significantly outpaced supply in Q4 with strong double-digit demand growth across every region and momentum accelerated through the quarter as customers prioritized access to compute for critical workloads. We saw broad-based strength with units up, a larger active buyer base and a richer mix of our 16th and 17th generation platforms as customers shifted to dense high-performance configurations. The ROI to refresh is compelling. Even at higher ASPs, customers see a 7:1 consolidation when upgrading from the 14th generation to our latest platforms.

The runway is substantial. A majority of the installed base remains on 14th generation or older servers, creating a significant opportunity to modernize, improve performance and lower the cost of ownership. Traditional x86 is benefiting from AI infrastructure build-outs. While many AI workloads rely on specialized GPUs, traditional compute remains essential for orchestration, data processing and inference support. As customers deploy AI, they are modernizing broader AI estates, refreshing and expanding general-purpose environments.

Turning to storage. Revenue was up 2% with continued outperformance from our Dell IP portfolio. We saw double-digit demand growth in Dell IP with momentum across PowerMax, PowerStore, PowerScale, ObjectScale and data protection. All-flash arrays delivered their third consecutive quarter of double-digit growth. PowerStore, our primary midrange platform, posted its seventh consecutive quarter of double-digit growth. Profitability improved, supported by our higher Dell IP mix. Lightning, our parallel file solution remains on track for general availability in the first half of the year with early customer deployments already underway.

Turning to CSG. Revenue grew 14%, and we gained share as the October momentum carried through November and December. We leaned into share and expanded our buyer base, broadening the portfolio to reach more of the market, including the lower end of the commercial market, emerging markets, consumer and education and by targeting strategic accounts. These actions expand our installed base and position us for future refresh cycles.

As we leaned into growth, we saw a higher mix of competitive large bids and customer expansion than planned, a higher-than-normal industry channel inventory levels, which delayed price increases. We have already taken actions to address each of these. We implemented pricing moves effective January 6 to reflect our higher input costs. Orders margins improved and are the basis for all new orders.

We remain confident we can operate CSG within our long-term value creation profitability framework. Commercial revenue grew 16%, our sixth consecutive quarter of growth with demand up for the eighth quarter. We are seeing growth across geographies, strong large enterprise demand and traction in the lower end of commercial where we set to expand. The refresh cycle remains a meaningful opportunity given the large installed base of devices that are over 4 and 5 years old. Consumer revenue was roughly flat with demand up for the second consecutive quarter, supported by strength in gaming.

Before I wrap up, a quick update on the supply chain and component costs. Across the industry, the environment remains highly dynamic with unprecedented AI demand creating sustained supply tightness and frequent pricing resets. In Q4, we did what we said we would do. Shorter quote validity periods, more dynamic pricing and a tighter alignment between our supply chain sales and pricers. We saw the benefit of this in ISG and expect it to extend to CSG. Given CSG's higher transactional volume and deal velocity, repricing to reflect multiple cost changes takes longer to flow through.

We're executing our operating model with urgency, securing supply as the first priority. Our scale, direct model, world-class supply chain and long-standing supplier relationships are a real advantage and they become even more visible in periods of disruption. We are managing this environment in real time, applying lessons learned from prior cycles to improve resilience and to strengthen our position.

In closing, FY '26 was a pivotal year for Dell. We delivered record performance, converted it into record cash generation and returned significant capital to shareholders, all while building a stronger company. We outexecuted across the portfolio. ISG is at record levels with accelerating AI demand. Traditional servers is growing sharply with demand outpacing supply. Dell IP storage continues to outperform the market, and CSG is gaining momentum with share gains in Q4. We are operating with discipline, lower OpEx alongside meaningful double-digit revenue growth. We've made the company more agile, which is on display in this commodity environment.

Bottom line, we enter FY '27 with momentum, a strong backlog and pipeline and a proven operating model. I'm excited about the road ahead and proud of our team's execution.

With that, let me turn it over to David to walk through the financials and our outlook.

Thanks, Jeff. It was a record year for Dell. And as Jeff mentioned, we're very excited about what's ahead. The team executed extremely well this quarter, delivering record revenue, EPS and cash flow with strong returns to shareholders. Total revenue was up 39% to $33.4 billion. Gross margin dollars increased 18% to $6.8 billion. Gross margin rate was slightly better than anticipated at 20.5% and reflected a mix shift to AI servers with AI revenue up more than 4x year-over-year and improved profitability in storage.

Operating expenses were up 5% to $3.3 billion, primarily from variable compensation tied to outperformance. We continue to drive meaningful scale within the P&L with OpEx down 320 bps to 9.9% of revenue. Operating income grew 32% to $3.5 billion or 10.6% of revenue, primarily driven by higher revenue. Net income was up 36% to $2.6 billion, primarily driven by stronger operating income. And our diluted EPS increased 45% to $3.89, a record.

Moving to ISG. ISG revenue was a record $19.6 billion, up 73%, marking 8 consecutive quarters of double-digit revenue growth. Before I get into the categories, we are now breaking out AI server revenue from overall server and networking line, reflecting the scale of the business and growth we expect to see going forward. AI server demand remained exceptional with records across the board. We had $34.1 billion in orders, $9.5 billion in shipments, $9 billion in revenue and an ending backlog of $43 billion.

In traditional server, demand improved throughout the quarter, outpacing revenue with stable profitability. Traditional server networking revenue was $5.9 billion, up 27%. Storage revenue was $4.8 billion, up 2%, with strong demand across the Dell IP portfolio. Dell IP demand outpaced market growth and PowerStore remained a bright spot with 8 consecutive quarters of growth, 7 of which were double digit. ISG operating income was a record $2.9 billion, up 41%, marking 7 consecutive quarters of double-digit growth. This was driven primarily by higher revenue. Operating margin was 14.8%, up 240 basis points sequentially. The sequential improvement was driven by scaling and strong storage profitability due to higher mix of Dell IP.

Turning to CSG. CSG revenue was up 14% to $13.5 billion. Commercial revenue grew for the sixth consecutive quarter, up 16% to $11.6 billion, while consumer revenue was roughly flat at $1.9 billion. CSG operating income was $0.6 billion or 4.7% of revenue. As Jeff mentioned, profitability reflects strategic share capture in a highly competitive market. This is building our installed base and expanding our services and attach opportunities, positioning us well for the refresh cycle ahead.

Moving to cash and the balance sheet. We delivered a record cash quarter with cash flow from operations of $4.7 billion. This was primarily driven by higher profitability and sequential revenue growth. We ended the quarter with $13.3 billion in cash and investments, up $1.9 billion sequentially. Our core leverage ratio is at 1.4x, in line with our target. We returned $2.2 billion to shareholders this quarter, including repurchasing 14.9 million shares at an average price of $125 per share and paying a dividend of approximately $0.53 per share. For the year, we returned $7.5 billion and repurchased roughly 54 million shares, more than double the amount of shares we repurchased in FY '25.

Looking ahead to FY '27, we are raising our annual dividend by 20% to $2.52 per share, well above our long-term value creation framework. Additionally, the Board of Directors approved a $10 billion increase in our share repurchase authorization. These actions reflect our confidence in the business and our ability to generate strong cash flow in any environment.

Next to guidance. Looking ahead to FY '27, we expect to build on a record FY '26 and deliver another exceptional year. AI demand continues to accelerate, and our value proposition is resonating with customers and driving continued wins and success. This is demonstrated by $10 billion in shipments in FY '25 and 150% year-over-year growth to $25 billion in FY '26 with an exiting backlog of $43 billion. For FY '27, we expect $50 billion in AI revenue, about 100% growth year-over-year. This outlook reflects the composition of our existing backlog, customer readiness and delivery schedules.

Across the rest of the business, customers are assessing their needs and priorities in an environment where component demand is outpacing supply, which is elevating input costs and extending lead times. We have priced to offset these pressures, and our guidance incorporates a prudent view of second half demand while navigating this dynamic environment.

For full year, we expect revenue of $138 billion to $142 billion, up 23% at the midpoint of $140 billion. ISG is expected to grow in the mid-40s, driven by roughly 100% growth in AI revenue. Traditional servers and storage are expected to be up mid-single digits with growth concentrated in traditional servers and more weighted towards the first half. CSG is expected to grow roughly 1%. Margin rate expansion remains a priority.

We're maintaining pricing discipline and our transition to Dell IP storage is accretive to margins. Excluding the impact of AI mix, our gross margin rates are up year-over-year. We expect operating expense dollars up low single digits, delivering significant operating leverage as we continue to invest and modernize, simplifying, standardizing, automating and enhancing our operating model with AI. We expect ISG and CSG operating income rates to be at the lower end of our long-term framework, reflecting the rapid mix shift to AI and the near-term CSG margin dynamics we discussed earlier.

Operating income is expected to grow approximately 18%. I&O is expected to be between $1.4 billion and $1.5 billion. Diluted non-GAAP earnings per share is expected to be $12.90, plus or minus $0.25, up 25% at the midpoint. You're seeing the operating model at work with strong EPS growth driven by significant expansion of our AI business, growth and improving profitability across the rest of the portfolio, meaningful OpEx scaling and EPS leverage from our share repurchase program. We're leveraging our strength in a dynamic environment.

For Q1, we expect revenue of $34.7 billion to $35.7 billion, up 51% at the midpoint of $35.2 billion. ISG is expected to grow over 100%, supported by $13 billion of AI server revenue, and CSG is expected to be up roughly 2%. Operating expenses are expected to be down low single digits. We expect operating income to be up roughly 60% with sequential improvements in CSG operating income rate. We anticipate a diluted share count of roughly 664 million shares. Diluted non-GAAP earnings per share is expected to be $2.90, plus or minus $0.10, up 87% at the midpoint.

In closing, we delivered an extraordinary year with record revenue, EPS, cash flow and capital returns. Revenue reached $113.5 billion, up 19%. EPS grew 27% to $10.30. We generated over $11 billion in cash and returned $7.5 billion to shareholders. Our focus is clear: drive durable shareholder value through consistent execution, profitable growth and robust cash generation through any cycle or environment. I'm excited about the year ahead. We have the portfolio, operating model, discipline and multiple levers to deliver growth that exceeds our long-term value creation framework.

Thank you to the team for their outstanding work, and thank you all for your time. And now I'll turn back to Paul to begin Q&A.

Thanks, David. Let's get to Q&A. [Operator Instructions] Operator, let's go to the first question.

We'll take our first question from Tim Long with Barclays.

Yes, I'll try to be concise. I did want to just follow up on the AI servers. Obviously, pretty huge order number there, really good performance. Just curious if you could kind of provide us a little color. Clearly, given the demand backdrop, it seems like you're not really seeing any memory price impacts on that business. But I am curious, as the business scales continually, it seems like you're maintaining that mid-single-digit operating margin, which is excellent. But as we see scale and all the diversification you're talking about, maybe you could just provide a little color on other opportunities either on the margin front or pull-through of other business as that business reaches a much broader, more diversified customer base?

Yes, Tim, I'll take the question. I mean let's start with -- we had an extraordinary quarter, taking $34 billion worth of orders. And I think equally important, and we tried to call this out, the 5-quarter pipeline grew as well. We didn't drain it. We actually grew it, and we grew it across all customer types, CSPs, sovereign, neoclouds as we call CSPs and enterprise. Enterprise, in particular, I'd call out in Q4 demand is very strong. We now have over 4,000 customers. We saw our enterprise AI business grow significantly quarter-over-quarter and very encouraging. And we're seeing AI deployed across many use cases in enterprise. So the backdrop of demand continues to be strong. And you called it out, we tried to call it out as well. We operated throughout the quarter and over the course of the year in that mid-single-digit operating income.

With what we see in front of us, there's no reason to change that. That is our guidance of where we can operate this business, and we're going to continue to grow it. It's exciting to say that we can grow it twice what it was this year. We grew orders 6x over the previous year, and we're excited about the prospects of growth. And probably a backdrop from a technology perspective, as we see inference ramp, inference is driving more tokens, tokens drive more compute capacity and intensity. And ultimately, that is good for the revenue stream of the company.

And we'll take our next question from Mark Newman with Bernstein.

Great numbers, very impressive numbers on AI servers. I wonder if you could talk about the profitability of AI servers given the huge numbers you're posting. Any change directionally versus previous quarter and going forward that we may anticipate given the huge upside on the top line? And then similarly, I guess we'll get to this later in the call, but a little bit more explanation on how rising memory prices are impacting profitability in CSG and traditional servers would be appreciated.

So two questions. The first one, I'll reemphasize what we just talked about. We maintained throughout the quarter in the guidance that David just talked about that we can operate AI at the velocity we're seeing in the mid-single digits. We have a significant technology transition in front of us. We see our ability to work our way through that and maintain mid-single digits. We clearly have a lot of backlog to clear with $43 billion in backlog. Backlog will ship at mid-single digits. So hopefully, those are 3 data points of backlog, future demand and technology transition that -- and in our guide that we can maintain the profitability that we've talked about.

When we look at the other two businesses, traditional servers and CSG that you called out, I think we are operating at a high level of proficiency of changing our price as our input costs are rapidly changing. We began to change price, most notably in servers in mid-December, December 10 to be specific, and we saw our margins stabilize with higher input costs coming our way. CSG, we purposely -- we were deliberate in delaying moving price because we leaned into a share position.

We began that in October. We gained momentum in October, November and December. You saw that in the share results. Our business grew 18% in a market that grew 10%. We took 100 basis points of share. We chose not to take our foot off the accelerator. We saw a change in the composition of the business in December, in particular, with a higher mix of large deals as well as acquisition pricing of new customers. We grew our customer base in CSG. And then we made a price change on January 6. When we made a price change on January 6, our business normalized, margins stabilized and they're where they need to be, so we can operate within the profitability framework that we've committed to. I hope that was clear enough.

And our next question comes from Louis Miscioscia from Daiwa.

Great numbers, guys. Congratulations. If you look at traditional servers, obviously, you're growing exceptionally well in the AI ones. Some of our checks suggested faster growth in that area. So I guess I'm just trying to understand why number is not higher for that more into the solid double-digit range. And are you only seeing inference deployed on AI servers and not yet as much on traditional servers?

Louis, let me try to break that down, and I'll ask for David's help on sort of the forward guide. In the quarter, I think we were very outspoken about demand outpaced supply and traditional servers grew double digits. And we saw that across all use cases in small business, medium business, large multinationals across all geographies. We saw the continuation of the consolidation that we've talked about and modernization that we've talked about in traditional data centers where you get incredible efficiency upgrading old technology. For example, our 14G is converted to a 17G converts at 6:1 to 7:1. So you're getting incredible efficiencies in power, space and cooling and that continued throughout the quarter, and we're projecting that to continue into fiscal '27.

Additionally, we are seeing AI workloads on x86 pick up. I think that is significant. And we saw that in enterprises. A phrase that we use inside the company is AI forward. So the most advanced enterprises that are really adapting AI are moving quickly. And we saw them use that across a large number of use cases, whether that be software development, scientific computing, whether that be some of the financial traders that are using very sophisticated algorithms. We're seeing that deployed broadly as well as in some inference use cases. I'll let David take the guide.

Yes. So to build on that, Jeff, like you said, demand far outweighs supply in relation to the Q4 results. Q1 guide, a strong reflection. We see that momentum continuing at strong double-digit growth in Q1. We have moderated that slightly as we go into the back half of the year. There's a lot of dynamics out there as we watch the supply-demand dynamics. Part of that will be us out there looking for more supply. But if this demand continues, obviously, there is the potential for growth. But you're going to see in Q1, we're guiding to a particularly strong quarter similar to Q4.

And our next question comes from Amit Daryanani with Evercore.

Clearly, you folks did not get the memo that you're supposed to miss numbers, by the way, when memory prices go up. But it's good to see these results. And I'm hoping you folks can talk about. As we shift to Rubin, could you just contrast how you think the Rubin cycle plays out in '27 versus the Blackwell cycle? And really hoping you can touch on, a, do you think operating margins will be much more smoother through Rubin? Or will it have the same cadence that we saw with Blackwell? And then b, do you think the revenue lumpiness could be less severe as you have a much more diversified customer base perhaps? So I'd love to just contrast those things. And then David, perhaps I missed this, free cash flow expectations for fiscal '27 would be appreciated as well.

Sure, Amit. On Vera Rubin, look, we're excited about the technology that's in front of us. We have a significant technology transition to go through. There's been a lot of lessons learned from the Grace Blackwell implementations. We're expecting a smoother transition. We're expecting all of the manufacturing lessons and lessons around test that are implemented into the next architecture to allow us to ramp with more velocity and speed. So that's -- we believe that's the case. We have our early engineering samples. We're working with customers now on designs. Very encouraged with what we're seeing thus far.

When I look in terms of operating margins, the question, I guess, will be consistent here. We believe we can operate in the mid-single digits with the backlog at hand with the new orders that we will take on over the course of the year, which will include Vera Rubin. And that's our outlook and guidance and what we've implied in the numbers that David has given. And I think it is a smoother transition. We've learned a lot. We're more enabled, more skilled. We've invested in more engineering capability. Our forward deployed engineers, again, are working with customers right now on advanced designs. And I like what we're doing. And I expect it to ship in the second half of the year because we know Vera Rubin is in production.

And to touch on cash. I mean obviously, we're coming off the back of a very strong FY '26, $4.7 billion of cash flow from operations. That totaled up to $11.2 billion for the full year, and that allowed us to maintain our commitments to our shareholders, right, to return over 80% of our adjusted free cash flow back, repurchasing over 54 million shares in the year. And while we don't guide to cash, but what I can say is we expect another really strong cash or cash year ahead our net income to adjusted free cash flow will be at or slightly ahead of our commitment in our long-term value creation framework. And that consistency and ability to reward the shareholders will be strong again.

You can see that with our dividend commitment today, 20% growth in our dividend per share. That's the fourth year in a row of double-digit increase in our dividend per share. That with our execution on the share buyback program means we should have a very strong and stable working capital year ahead.

Our next question comes from Ben Reitzes with Melius Research.

Nice execution here, guys. I'll echo that. My question is on storage. It sounds like it's turning a little bit. You beat the Street by a little bit, 2%. And then you said, I believe that it will grow in the mid-singles for the year, and it looks like it may outgrow servers in the back half of the year. So can you just talk about what's really going on with storage, your highest margin business? Is it really turning? Is it going to be a contributor to mix in the upcoming year that allows you to keep gross margins pretty flat for the year in a tough component environment?

Sure, Ben. Look, we're excited about our storage business. Again, we're reporting that on an orders basis, our Dell IP portfolio grew double digits. That's the entire portfolio. PowerMax, PowerStore, PowerScale, ObjectScale and our data domain platforms all grew double-digit demand. Our all-flash grew double-digit demand. It grew in all regions, and we acquired new customers. PowerStore grew its eighth consecutive quarter, 7 of those -- the last 7 double digit.

Half of those new customers that we are winning are new to PowerStore and nearly 30% are new to Dell buying storage. We saw tremendous demand for our unstructured products as AI inference and AI continues to grow, grow, grow. Our Dell IP portfolio is now a greater percentage of the mix year-over-year. We expect it to grow FY '27 over '26, it will be a greater percentage of mix next year than this year. That's part of the profit contribution that David has outlined in our guidance.

And we're entering an era in my humble opinion, where architecture matters. So where we have the leading data rate reduction 5:1 with our PowerStore product, we're going to increase customers' effective storage capacity. We're the leader there. You take our data protection product where we have up to 75:1 compression in dedupe, we are going to help customers through this memory crisis shortage, whatever you want to call it, with advanced architectures that require fewer servers and fewer drives to back up customers' information. That architecture difference, we believe, is fuel that will continue to serve our Dell IP portfolio well in FY '27.

I would also add there, Ben, I think you asked about the contribution and the margin that we see going forward. Well, we see it today, right, the results from Q4 illustrate that in relation to the [indiscernible] performance of the ISG P&L, which was tremendous. As we look into the guidance, particularly in the back half of the year, we talk about mid-single digit between traditional server and storage, maybe a smidgen more on the traditional server side, but growth nonetheless. So I hope that helps.

And our next question comes from Erik Woodring with Morgan Stanley.

Really impressive guide. And Jeff or David, I guess the question is, can you maybe just help us understand what kind of memory price inflation you're assuming in your 2027 outlook and -- or fiscal '27 outlook, excuse me. And since you're talking about margin rates improving year-over-year, maybe just help us also understand what you're doing different this cycle versus past cycles to protect your profitability?

Yes, a few questions in there. I think if you go line by line in terms of the business, so we talk CSG in relation to guidance, our current assessment of the market is double digit, probably in around that minus 11%, minus 12% range. As we look at the year ahead, probably more acute in the second half. So obviously, we believe our guide, which is 1% growth in revenue for the year ahead is a prudent guide to start with, aligned with our commitment to manage and maintain our margin rates within that portfolio. So you can see the difference between the units and the revenue there. We do and will take share just like we did in Q4. But we'll obviously learn as we go to the price increases have been there. Costs will continue to go up to a certain degree. We'll make quick and decisive decisions and execute.

Jeff outlined earlier the ability and the execution of the team, particularly on the server side in relation to Q4. We're already seeing the signs of that for Q1. So all the actions we discussed around time validity of quotes moving at speed, increasing list prices, moving to discount off list pricing, limiting to little or no promotions. All of those actions are strong operational execution levers we've got to work the business as we kind of move forward into the guidance.

I mean, Erik, you asked about memory assumptions. Obviously, our price is proprietary. I'm not going to share our specific percentages. But look at the spot market. I mean the spot market for a gigabit of DRAM over the last 6 months is up nearly 5.5x at $2.39 a gigabit. If you look at NAND, the cost is $0.20 a gigabyte, that's up nearly 4x over the last 6 months. The industry analysts have Q2 up over Q1 in a range of 20% to 50% at Q3, 5% to 15%. Q4, 5% to 10%. Those are estimates. Those are probably in ballparks where things are. So we haven't changed anything. We continue to work with our long-term partners. We've had LTAs in place. We've had capacity agreements in place. We know how to budgetary price.

I mean, David's team has our best estimate of cost for Q2, Q3 and Q4 and how we're pricing deals today. We're working with our memory partners to be as flexible and as agile as possible. We are working through things of how do we minimize our complexity, how do we improve our mix, how do we sell what's coming, how do we improve our designs to take whatever parts that are available.

And then to answer really to maybe put the bow around this, the pricing actions, look, we became very proficient during COVID. And all of the best practices that we learned during COVID, as I mentioned in the last call, we put in place and we put in place faster. We changed the entire pricing of our server business on December 10 in a couple of days. We had tens of thousands of open quotes in the PC business and changed them all on January 6. This notion of we recover our cost in 2/3 of it in 90 days, we moved that quickly. That's what we learned in COVID. That's what we put in place here.

So we made list price changes across the board. We've changed in our vernacular, our internal mechanisms around smart price and margin floors all changed instantaneously. We're moving to discount off list price. We're compressing discounting. Our quotes are valid for the shortest period of time they've ever been. And we're reducing promotions and all sorts of special pricing going forward. That's what we've done. It's been in place.

As I mentioned with one of the previous questions, we saw our server business stabilize with the higher input cost. You saw that in the performance of the business in ISG, which was extraordinary. In PCs, we purposely delayed implementing that price move to stay in the hunt to take share and to drive growth, which will serve us for the long run. And then when we made the change on January 6, it wasn't 90 days later, it was that day we stabilized margins. I hope that helped.

And we'll take our next question from Krish Sankar with TD Cowen.

Congrats on the amazing results. Jeff, I had a question on enterprise AI adoption. You kind of said that it's very strong in your server business. Clearly, agentic AI, long horizon agents, all of them has implications across your server, Dell AI factory and your storage business. I'm just wondering, is there a way to break down your AI server orders between enterprise, neoclouds and sovereign? And when do you expect enterprise AI adoption to spill over to your storage segment and probably start a new storage cycle?

Well, of course, there's a way to parse it. I won't do this on the call, but we absolutely keep track of the health and growth of those 3 parts, our neoclouds, sovereigns and enterprise. We try to give you a sense of the enterprise adoption with the number of customers, which is now over 4,000. We gave, I think, some breadcrumbs, if you will, around the buyer base grew. It was a record quarter in enterprise revenue in Q4. We're seeing usage models expand.

I take a look at our own company, an example that I would give, 2 years ago, we deployed coding assistance and used some GPU capacity. Mid last year, we started deploying agents to write the actual software with basically specifications from our software developers and architects. And what we saw was an incredible need for more compute power. The amount of tokens that is required to do that well is significant. And that's just one use case in one company. And what we believe is we're seeing that broadly in the leading companies who have deployed AI and seeing the tremendous benefit and potential of this technology that we're going to see AI and enterprise continue to ramp. We'll continue to give you signals of our customer expansion, the revenue going forward.

I'd also just to make sure that I'm very clear, the enterprise portion of our 5-quarter pipeline grew and actually was the fastest-growing portion of the 5-quarter pipeline. I hope that helped put some context around it.

And we'll take our next question with Wamsi Mohan with Bank of America.

Jeff, maybe you can talk a little bit about what you're seeing from a purchasing behavior standpoint as you're implementing these price increases. It sounds like you deliberately delayed your CSG price increases to take share, obviously, indicates that, that is elastic in that sense. From a server perspective, you did implement and you've had some time now to look at sort of what the reaction from customers has been. I'm kind of curious to see if you've either seen material elasticity, both on CSG and ISG side. If you could like maybe put some bookmarks around that?

And also, did you see any pull forward behavior given that the expectation is that you shorten sort of these windows of [indiscernible] validity, you're talking about continued sort of price escalation from a component standpoint. Are customers worried about supply? And is that creating any change in terms of pull forward across your portfolio? And quickly, if I could, for David, the inventory stepped up a fair amount. I'm just kind of wondering if you could break that down for us a little bit on the composition of those things.

I'll let David, you take the last one first because it sounds easier. I have a multifaceted question to answer.

Yes, sure. I think, look, if you look at our cash conversion cycle, minus 32 days, that's actually flat quarter-on-quarter and actually improvement of a day year-on-year. So if you think building on the expansion of our AI business and shipments that we've done to maintain our cash conversion cycle in that position shows the diligence that we have from a working capital perspective. We have guided to $13 billion in Q1 of AI shipments. That results in reality in February and March, we're shipping billions of dollars of gear. And obviously, we're positioning inventory to do that. So it's purely a function of the size and scale and growth we're seeing in the business that we've got.

Wamsi, I'm going to take a run at all the parts of that question. So how are customers doing reacting? I mean, clearly, early on, there was a wide range of emotions. As it wasn't completely understood. And there's a dynamic that I think it's important for me to communicate a different reaction in PCs versus infrastructure. So I'm going to talk initially about infrastructure, then I'll pivot to PCs.

In infrastructure after the sticker shock and our customers began to understand the gravity of the situation, the conversations quickly turned to access to supply. It was not literally a light switch, but it was a pretty quick order after the emotions of price increases, it was, oh, this is real. And as you know, over the course of the quarter, the understanding of this situation became better understood. Customers began to see that. And the largest customers in the world, the most sophisticated customers in the world began to move aggressively to protect their infrastructure build-outs. And we saw that over the course of the quarter in AI and in traditional servers and in storage.

PCs was a little different because you had inflated inventory positions in the channel. So the cost did not hit that inventory, which is another reason why we stayed in price position for growth. And when we began to see it in PCs was in large bids, where they would be fulfilled over the course of the year, course over the first half and what have you, customers began to see the reality that this was going to be tight costs were going to go up depending on when you wanted product and delivery, there was an associated cost with it. And then when you started having conversations, what's this going to cost me in the first half versus today and you give an answer, I don't know. It certainly heightens a buyer's awareness and understanding the cost today is likely better than the price that will be tomorrow, the next day and so on.

So that clearly has driven some amount of pull ahead. I don't know how to quantify that. What we do know is IT budgets are generally fixed at the beginning of the year. So this pull-in is going to obviously drain those IT budgets to some degree. That's sort of what we put into our guide, our best understanding of that and why you saw some of the numbers around our PC business and traditional server business. But clearly, technology has to be replaced. If budgets aren't sufficient this year, that just means replacement cycles will be elongated and extended. And I think the result over the next couple of years is we'll see product bought early, but we'll also see the replacement of some technology extend over time. I hope that answered the multiparted question.

And our next question comes from Samik Chatterjee with JPMorgan.

Congrats on the outlook as well. Jeff, maybe I just want to get sort of any more color that you can share on the AI order backlog of $43 billion. How does it break down between Blackwell and Vera Rubin? And what are the implications of when some of that backlog ships based on the backlog mix? And you're now guiding to $50 billion of revenue on that front. How should we think about capacity? You're doubling sort of revenue, but to the extent that demand is higher, how should we think about your ability to add capacity over time?

So on the $43 billion backlog, Samik, it is predominantly overwhelmingly Grace Blackwell. There is no Vera Rubin in the backlog. The largest percentage of our 5-quarter pipeline is a combination of Grace Blackwell and Blackwell where we're seeing a rise in x86 Blackwell in the 5-quarter pipeline, driven primarily by enterprise deployment, air being the #1 consideration. The second consideration that's driving that demand in the 5-quarter pipeline tends to be around some of the scientific work. And some of the -- as I mentioned in one of the earlier questions, some of the financial trading models and algorithms, the high-frequency traders and using some of the x86 air-cooled solutions to meet their AI needs.

So I think that's sort of the composition of both the backlog and the ability to convert that 5-quarter pipeline. That is clearly our job is to take that 5-quarter pipeline and convert the potential into POs. The corresponding next part of that is we'll go find parts to match the POs. The $50 billion guidance that we gave is the alignment of what we believe at this point in time, 4 weeks into the fiscal year of our best understanding of our customers' deployments and build-out of buildings and power and infrastructure, the availability of DRAM and E1 drives and E3 drives, our ability to deliver that yields the $50 billion number that David gave. But leading the operational part of the organization, we're out looking for more parts. Our job is, again, those orders get converted to be able to fulfill our customers' needs and to do that in a timely fashion. That's what we're working on.

And we'll go to our next question from Aaron Rakers with Wells Fargo.

Also my congrats on the quarter. I guess one just housekeeping thing. When I look at the slide deck and you talk about $9.5 billion of AI shipments, you're now disclosing an AI revenue number that's a little bit different, right, $8.95 billion versus $9.5 billion. Can you help me understand what that difference is, number one? And then number two, on the traditional server side, I'm just curious, Jeff, with all of the pricing stuff going on, it sounds like a very healthy demand backdrop. How do I take the context of mid-single-digit growth and maybe separate that between what that underpins in terms of unit growth versus what I would assume to be a pretty healthy ASP uplift environment through the course of this year?

Yes. So look, the first piece I'd add in relation to shipments versus revenue, normally, in any given quarter, they'll be close to the same number. It's simply in transit. So as we ship out $9.5 billion, some of that obviously would have been happening time-wise right at the end of January. So literally just in transit will show up days later from a P&L perspective. So just normal run the business there in relation to that piece.

On traditional servers outlook reconciling unit growth of the industry TRU expansion, we clearly saw TRU expansion in Q4. Customers are continuing to migrate towards our 16 and 17G server and buying them with a lot of DRAM and a lot of storage. That's part of the consolidation play, replacing 5:1 if it's a 16G, 7:1 if it's a 17G, and we'll continue to see that behavior of buying servers with more CPU capability. More memory, more storage.

Our best estimate in demand for units next year -- or I guess, the year we're in, sorry, excuse me, in FY '27, calendar '26 is units are clearly down while TRUs are up. To reconcile the difference of that spread of prices increasing, I think what we tried to outlay in our guidance, and David can chime in here is the uncertainty. We're being prudent in our planning in our guidance to you because of the uncertainty associated with the second half, we try to put that in our best reflection in our guidance.

And that's how you can reconcile between what we talked about in Q4 that demand was out ahead of supply, double-digit growth, TRU expansion, clearly seeing that into Q1. But as we see going into the second half of the year, we're trying to describe that uncertainty.

And again, just to dovetail on that, again, it's linked to the articulation is if we enter the year, we have sufficient supply to support and meet the guide that we've laid out. Obviously, demand is far outstripping supply right now. If that continues, like Jeff operationally said, we'll -- and he'll be out hunting for more parts to try and find that. But right now, I think it's a good guide. We see this trend, double-digit growth for Q1, and then we'll take care of the rest of the year as we go.

And we'll take our next question from Asiya Merchant from Citi.

If I could, just -- if you could provide any incremental color on the attach rate that you're seeing as you're shipping out these AI servers. Are you seeing a better attach rate perhaps than what you have? And any other further color on what else we can attach to those AI servers that you're shipping?

Well, clearly, the attach items for us around an AI server, I think, lie in 3 distinct groups: storage. And given the enterprise momentum, we're seeing more storage with enterprise customers. Networking, our networking business continues to grow. And the third bucket would be around all of the types of services. Installation deployment services, break services, which are proven to be a huge source of differentiation for us in the marketplace. Our ability to deploy and install these very complex customers is unmatched in the marketplace today. Our uptimes are the best in the industry. And then our ability to maintain them with Dell badged employees on site, taking care of any challenge, any miscue, again, is a differentiated capability that we have in the marketplace. So those are the 3 areas that we focus on and attach. We're seeing continued growth and acceptance of that, and we're optimistic that will continue into FY '27.

Thanks, Asiya. And we'll do one more question before we have Jeff close the call.

We'll now take our final question from David Vogt with UBS.

Jeff, just a quick question on the structural share gains in PCs. Is it sort of the premise here that your ability to dynamically adjust price and steer demand based on your supply chain availability and your expertise relative to maybe some of the smaller competitors put you in a position to structurally have a better sort of CSG backdrop in fiscal '27 calendar '26 and more likely than not in fiscal '28 as we think about where you're positioned? Is that really where you're going to see share gains? I'm just trying to get a sense for how you're going to outgrow a market that's going to be down double digits. I get the pricing umbrella that's going to happen, but I just want to get a better sense for where that's the gains are going to come from.

David, I couldn't have said it better myself. If you look at the last industry-wide shortage, Dell excelled and took share across the board, most notably in its PC business. Our long-term relationships, supply agreements with our partners, we believe, position us to take share in all of our businesses and in particularly PCs. Again, which is a reinforcing point why we didn't back off on the pricing position and the posture that we had during the quarter.

We believe we are entering and changed the trajectory of the business. We had lost share for 3 years. We exited the year with tremendous momentum. You saw it in the Q4 results, 14% revenue growth. IDC number was 18% unit growth for Q4, calendar Q4 to be specific. That momentum is important to us. We grew customers, and we're going to continue to focus on driving and winning in that business. And we think there is a structural share gain opportunity for us certainly over the next couple of years as our supply chain team has positioned us quite well. I believe we can do that in servers, and I believe we can do that in storage as well.

Jeff, go right ahead.

Sure. We closed FY '26 a defining year for the company with record results and strong execution, and we are entering FY '27 with clear momentum. We have tremendous AI traction entering the year with $43 billion in AI backlog. The supply environment is tightest we've ever seen and input costs are moving higher. Our priorities are straightforward. First, secure supply; next, price to protect our margin rates. You've seen this in ISG. CSG will follow with improvements beginning in Q1 and continuing through the year.

In Q4, our gross margin rate came in slightly better than anticipated. Excluding AI mix, we are guiding FY '27 gross margin rate up on a year-over-year basis. The operating model we've executed for the past 4 decades allows us to move fast and adjust as demand evolves. With a broad portfolio and several levers at our disposal. Our FY '27 guide reflects that 23% revenue growth at the midpoint and 25% EPS growth, driven by the expansion of our AI business, growth and improving profitability across the rest of the portfolio, meaningful OpEx scaling and EPS leverage from our share repurchase program. We are positioned for another record year.

Thanks for your time today.

This concludes today's conference call. We appreciate your participation. You may now disconnect at this time.

All right. I think we're live. So thanks for everyone joining. Welcome back to the UBS Tech Conference. I'm David Vogt. I'm one of the tech analysts here at UBS, and we're excited to have with us today, Dell Technologies, Jeff Clarke, Vice Chairman and Chief Operating Officer.

Before we get into the hard-hitting questions, I need to read the Dell Technologies safe harbor statement.

So statements in this presentation that relate to future results and events are forward-looking statements and are based on Dell Technologies' current expectations. In some cases, you can identify these statements by such forward-looking words as anticipate, believe, could, estimate, expect, intend, confidence, may, plan, potential, should, will and other similar expressions.

Actual results and events in future periods may differ materially from those expressed or implied by these forward-looking statements because of a number of risks, uncertainties and other factors, including those discussed in Dell's periodic reports filed with the SEC. Dell Technologies assumes no obligation to update its forward-looking statements.

Jeff, thank you for joining.

Thanks for having me.

So I think I just took Paul's job. So all right, here we go. We just reported earnings last week. So I'm sure you've gotten these questions today already. But can we start maybe just as an overview of how you saw the quarter play out from a demand perspective? Maybe we can touch on AI server to start, and then we can touch on the other ISG products. And how you're thinking about kind of the macro conditions and the demand drivers as we step forward into your next fiscal quarter?

Sure. If I think about the quarter we just reported, we're pretty excited. We posted $27 billion in revenue, up 11%, $2.59 of EPS, up 17%, still showing the ability to grow as we're managing our expenses down, modernizing the company. We returned $1.6 billion to shareholders. ISG grew 24%. Service and networking grew 37%. But more importantly, we improved operating margins quarter-over-quarter by 360 basis points from [ 8.8% to 12.4% ].

Then we had some AI good news, record orders at $12.3 billion, $30 billion year-to-date, shipped $5.6 billion, $15.6 billion of shipments year-to-date. A growing backlog, we added almost $7 billion to our backlog to $18.4 billion. 5-quarter pipeline continued to grow across Neocloud, sovereign and enterprise customers. And then we upped the annual guide for the remainder of the year, up to $111-plus billion and increased AI shipments to $25 billion.

And then over the course of the quarter, we saw strengthening demand across our business. Month 3 was better than month 1. And we saw that across commercial PCs. We saw that across storage servers. We saw server demand improve in North America quarter-over-quarter. It was up double digits on a demand basis in our international markets. So all in all, a solid quarter. We're pretty excited about it.

Great. So we get often this question, Dell has been putting up incredibly strong AI orders. You just incredibly strong quarter, most recent quarter. That doesn't even include the [ IREN ] deal that was announced fairly recently. That will spill into your next fiscal -- post quarter close.

So maybe can you just help us level set what does Dell bring to the equation that creates that winning sort of dynamic relative to the peers in the marketplace, whether it's on the enterprise side, sovereign or Neoclouds?

And along those lines, being able to bundle it together, we were talking a little bit about this earlier, like margins are a key debate. So again, how do you think about Dell's competitive advantage in terms of maintaining profitable growth, profitable orders in the context of what's going on in the marketplace where maybe some of your competitors are a bit struggling?

Well, we've consistently said, our differentiation starts with the engineering side. We have this concept called forward-deployed engineers. We do that in a series of pods where our engineers are working upfront with the largest-scale users in the industry to get the jump on the design to understand their data center design and how rack-scale clusters will get deployed.

And it's that upfront engineering that is really differentiating us, the ability to optimize for power, optimize for performance. We look at performance per watt, performance per dollar driving density, which then gets you into the components of the design around the rack scale architecture itself into the power, the cooling, density, its serviceability, the ability to manage that. We build management software around that. And then we have the services that deploy, install and support it and then financing for those customers that need it.

That combination of capability, we believe, has been differentiating us. We believe it continues to differentiate us today. And we believe as we look at the technologies going forward, we'll continue to do so. And we think those are all competitive advantages that can't be replicated easily.

And I like our hand. And it's enabled us. We've not been a price leader here. We've been -- we're adding value. We're helping customers deploy this gear. Our stuff is rolled off a truck. It's put in line in 24 to 36 hours up and operational at the highest quality levels, and our uptime is 99% or better. Those are all differentiated markers that we believe we're getting a premium to, which is why we communicate we can operate this business in the mid-single digits.

And you mentioned you're not a price discounter, you're a price leader effectively. Is that...

Value leader, I hope.

Value leader, maybe it's a better way to say it. So that -- is that a reflection of the holistic service and solution that you just referenced from hardware to software, to service, to maintenance, to reliability that allows you to command a premium relative to maybe some of the other participants in the market who are maybe just more hardware-centric effectively?

We believe so. I mean to deploy thousands and thousands of GPUs at scale in a short period of time, there's the upfront engineering work that I described to be able to compress that cycle time, then turn it over to our world-class supply chain to be able to build it and turn it into systems and then a service organization that literally takes it and installs it and then is there to handhold it and support it over time; again, it's differentiated, it's why we believe we are winning in the marketplace, and we'll continue to work and win in the marketplace.

So when you think about the $12 billion plus of orders, what gives you confidence? I know you've talked about revenue being lumpy. It's hard to predict quarter-to-quarter. But the pipeline, you've talked about many, many multiples of your orders. Now that's, I would assume, still kind of an unweighted probability...

We define it as an opportunity pipeline.

Right, an opportunity. Is the optimism underpinned by how strong hyperscaler CapEx then -- I know that's not your target market, but then what you're seeing with the Neoclouds and then enterprise customers and sovereigns, is it just an arms race that you think from a multiyear period is we're in the early stages at this point? Like how do you frame the longer term?

We get questions on how does '26, '27, '28 play out. I know you don't have a crystal ball, but would just love to get your perspective, given that you guys are closer in touch with many of these customers than anyone in the room.

Yes. Maybe a way to describe it is we entered this marketplace a little 2.5 years ago with the [ 9680 ], you may recall. We shipped $1.5 billion in our fiscal '24 calendar '23. Last year, we shipped $10 billion. We now communicated we'll ship $25 billion this year. So tremendous momentum.

Every one of those quarters, I believe I've communicated that the 5-quarter pipeline has continued to grow. So our view out in time continues to show, and we went to a multiple of our backlog to multiples of our backlog is how we've communicated that, and it continues to grow.

And within that, we talk about Neocloud customer base is growing dollar value, sovereign customer base is growing dollar value and enterprise. We've deployed now over 3,000 Dell AI factories to enterprises, as an example. So our view in that 5 quarters continues to expand rapidly.

And then you have the fact of token demand generation is not slowing. In fact, I'd argue we're in the infancy of this curve that looks like this that the token demand generation or token need is going up exponentially. Then you're beginning to see real-time value be extracted out of deploying AI in the enterprise.

To me, all of that bodes well. And then we've barely scratched the surface of [ agentic ]. So we even have the next evolution, which will drive more token demand, which will drive more need for computational intensity to be able to provide the capability. So that's why we remain optimistic. And our 5-quarter pipeline continues to look, I think, very encouraging to us.

And just from a supply chain perspective, it doesn't feel like you were hit by any sort of supply chain constraints recently. Obviously, there's a lot of discussion about higher memory costs, other component maybe shortages. Dell has relatively strong supply chain expertise over the last 20 to 25 years.

So how are you thinking about your kind of product order growth, your ability to deliver revenue rec in the context of what seems to be a very tight supply for a lot of key critical components going forward?

Well, we run the supply chain by one fundamental rule, never run out of parts. So first and foremost, our supply chain experts, who I think are very good, are out looking to get the supply and mix that's required to operate our holistic business. And we've been obviously doing that for some time now.

Clearly, this is a -- I've been at this for almost 4 decades. This is the most unprecedented mismatch in demand and supply that I've ever seen in the memory industry, which you're seeing it correspondingly reflected in spot price. But there's an equal component is the material going to be available. We're out with our long-term agreements, our partnerships that we've had for many, many years. I've grown up in this industry, know most of those CEOs for a long period of time, and those relationships matter in these times.

Once we get access to the material, then it's our ability to go generate demand. If you have the material, that's an advantage. And then you have to be able to extract an increasingly large cost basis through our customer base. The products either have value to our customers that they don't. They either need them or they don't. And if they need them, as our input cost grows, we'll have to be able to pass that through and extract that in the form of price to our end users.

We believe we've learned a series of tools over the past several cycles that prepare us for this. We've taken all of those lessons that we've learned and we put in place weeks ago. We're making dynamic pricing changes. We're making changes to how we quote and how we provide information to our customers. And I'm encouraged that we can be able to get the value out of having the material that our customers need.

And there will be some part of our sector that I do think will get somewhat -- probably the TAM compresses a little bit is the best way to describe it. They tend to be in the lower price bands, I think consumers, consumer PCs. But if you're building a large-scale deployment of storage and/or servers and if you need it, you're going to buy the material, the infrastructure to solve the workload or just solve the task at hand.

Does that make sense?

Yes, that makes sense. So maybe just sticking to Neoclouds for a second and enterprise on AI server, obviously, this is a new phenomenon over the last several years. We don't have a historical analog, I guess, to think about how demand changes with changes in pricing.

Is it your view or is it Dell's view that you will be able to work with them? And even if you do pass on pricing, there's probably de minimis demand destruction because of the critical nature of these products that you're selling them at a critical point and juncture in their business cycle, is that a fair way to think about it? Like Neoclouds need this today, right? There's no reason why to think that they don't.

Well, I think, first of all, it's unknown, right? To your point, I think rightly called out this is a new sector that has not been through one of these cycles. History would tell us, and as I was trying to describe, perhaps not clear enough, the lower price band parts of our infrastructure business, which is consumer PCs, that's impacted the most, as you move up the infrastructure stack, less price sensitivity if it's project-based, workload-based trying to solve a project.

When you look at the way that AI has unfolded to date, most that are making these investments, large capital investments, it's viewed as an existential threat. If they don't complete the task, the workload, train the model, whatever it might be, they fall behind. I think that is going to be a different lever in how this infrastructure will be deployed and thought about even as the cost basis goes up.

Got it. And I would imagine this is a wonky nuance. She'd be able to offset, obviously, the higher component costs. My guess is that would be -- it doesn't change your dollar profit, but it does have maybe a potential impact on your margin rate per se. Is that maybe fair?

Well, we will continue to try to drive the value associated with having the material, which inherent value...

And the price.

We will try to stay in that -- if we're talking about specifically AI, we will try to stay in that mid-single-digit margin that we've communicated broadly to our shareholders and value holders. And I don't see that changing.

Got it. Along those lines, we get questions from investors about the long-term relationship of GPUs, custom ASICs and Dell's position. Obviously, you've been very successful deploying GPUs at scale.

How important is the relationship with potentially alternative accelerators, custom ASICs down the road? I mean, is that strictly going to be a hyperscaler solution in many regards, so it's a little bit less relevant? Or do you see Neoclouds potentially going down this road and maybe diversifying away from sort of the NVIDIA stack, if you will, and you respond accordingly?

Well, we're going to be customer-driven. And if our customers ultimately have access, which is key, to the ASICs that have largely been proprietary and kept in specific...

Specific use case, right?

Exactly, use cases by specific customers and that becomes more readily available and our current customers have access to that, we absolutely want to be their infrastructure provider. We've learned a great deal of how to scale infrastructure. How to build it at speed, high quality, incredible uptime with all of those other attributes that I just described, absolutely, we want to be there, and we'll help our customers do so.

And same with other merchant GPUs like AMD, et cetera?

Absolutely. Again, we're a customer-driven organization. If there's customer demand and those customers obviously have that demand because they have access, whether that's AMD, and it works better in their workload or how they're doing inference or what have you or for that matter, some form of a TPU; I want to be their infrastructure provider.

Yes, the reason I was asked -- I mean, TPU is in the news last week, this week, it's all over the place. So -- but the way I think about TPU, it's obviously -- at least historically, it's been customer-specific.

How do you -- I know this is not your bread and butter today, but obviously, would that require significant engineering investment on your part if TPU becomes more broadly available to a set of customers that you can address? Like is it different in terms of integration, software stack?

Obviously, it's been customized for what Google has done historically. How do we think about that longer-term opportunity? And does that require like an investment?

I think of it as another form of an accelerator. What we've been witnessing for the past 2.5 years is a bifurcation of the traditional infrastructure to, for lack of a better way to describe it, a traditional data center built in the way we've built them for the past 20 years. And then we have this notion of accelerated computing.

And we've been putting accelerators in our computers for as long as I've been doing this. We had a math coprocessor back to the [ 80 88 ], if you go back in this business as long as I've been in it, we've been putting all forms of acceleration to it for 4 decades. And what's happening now is these dedicated accelerators are driving a new architecture. That architecture now is a balance between the compute node, the networking stack and the storage stack.

That engineering, I can replace the different accelerator and put a new one in. My software stack probably has to change a little bit, which we have the wherewithal to do it to drive the different management, the different way to optimize and maximize performance. But our ability to work across this accelerated computing new architecture that's been derived, I feel very confident we can do that without a tremendous incremental investment.

Got it. One final question on AI. So as we sit here today, obviously, you mentioned full rack scale solutions, liquid cooling. How important is the ability to offer cooling to -- outside of Neoclouds to maybe that long 3,000 customer list from an enterprise perspective. Is that critical to that part of your vertical? Or is it really dedicated towards your Neocloud customer base today that's looking for a more advanced robust solution that requires like a liquid cooling environment?

While there are exceptions, the vast majority of the liquid cool infrastructure is rack scale towards these massive deployments. Most, not all, enterprise deployment is air and PCIe.

And what's interesting and why I think this continues to perpetuate in a good way for us is the engineering challenge only grows generation over generation. We started this at 80 kilowatts of rack. We're at 200 kilowatts of rack today. Next design is 500 kilowatts of rack. Now we're talking about 1 megawatt and 1 megawatt behind. Now we're talking about a 400-volt rail becoming an 800-volt rail.

the engineering hurdle for each successive generation is coming, is significant, and it's coming faster, which is why I think it gives us an opportunity to continue to differentiate long term.

How about traditional CPU server? So that -- we've talked about this. Customers are looking for more cost-effective, power consumptive, performative solutions. There seems to be a bit of an upgrade cycle from a value perspective, maybe not a unit perspective.

In the context of what we're seeing on the back end of these networks, how do you see that market evolving on the front end for traditional CPU-based servers, given those kind of parameters that we just kind of laid out?

I think you've nailed that we are in the middle of a modernization consolidation cycle. It's started 2 years ago. Those traditional workloads running traditional architectures, you're seeing an opportunity to save floor space, power, cooling by upgrading to a modern server.

An example, our 17G server today displaces 6, 7, 8 to one of our 14G server. It's a single server. It's got more cores, it's got more DRAM, it's got more NAND in the combination of those, and it's far more power efficient and it's actually more performance. That's what we believe has been driving this cycle that we've seen, where we are in the middle of a consumption cycle of traditional servers, and we see that continuing.

Is that -- is there a view that the power constraints on AI side lead to customers looking for power savings and performance across the entire hardware stack, if you will, and that's part of the driver? Or is it the fact that the traditional CPU that was deployed 5, 6, 7, 8 years ago is just highly inefficient and too much of a power draw, where it's a natural upgrade cycle? Or has it been accelerated or maybe amplified by what we're seeing from the AI perspective?

I think it's both, but I think this notion that you described, you can make a case that it's been accelerated because of AI, because they need the floor space, they need the power, they need the cooling. Many modern data centers and enterprises today don't have large empty spaces waiting for stuff to be plugged in. They have to go make space, they have to consolidate, they have to modernize, get efficiencies to be able to deploy gear in their data center.

So are those 3,000 AI customers a prime feeding ground for upgrading traditional CPU-based servers? Is there an opportunity to use that as a lever, if you will, to drive continued demand in '26 and '27?

You might say there is a relationship between them. That's correct.

Got it. And then on storage, we get this all the time. Obviously, the storage market has been a little bit challenging. I know historically, you've been incredibly strong in the high end of the market, high performance, and that's been a little bit of a challenging market. But your Flash business has done exceptionally well. You've called out, I think PowerStore did exceptionally well last quarter, PowerMax.

How should we think about the storage dynamic in your overall portfolio, given what we just described in server, right? So we're seeing more power efficient, more performative CPU upgrades. Historically, I think people view storage as kind of a 1- to 2-quarter lag from a demand perspective. It seems like we've kind of maybe -- not broken that relationship a little bit, but how do we think about that tie-in from a storage perspective vis-a-vis server?

I'd like to communicate. I think we are seeing it, but it's masked. It's masked because we had built a very significant part of our storage business based on third-party IP, our VxRail business. And that business continues to decline. It's offsetting the progress we're making in our Dell IP portfolio.

I think we communicated last quarter that PowerStore had grown its sixth consecutive quarter, 5 of those double digit. That's taking share with that product. If I look at our all-flash array portfolio, I think we communicated it grew double digits for the second consecutive quarter. That's outpacing the marketplace.

It's not offsetting the secular decline of our PowerMax, the high end and the decline of our VxRail business, which is why overall, our storage business performed at minus 1% revenue growth. But if I look at the underlying characteristics of the Dell IP portfolio, where we have invested, it actually grew. It's continued to grow, and I'm very encouraged by that portfolio.

I know you're not going to give this, but I'll ask anyway. So if I strip out third-party IP, VxRail, for example, would it be fair to say that your core storage business is growing more akin to what historical growth rates would look like at this part of the cycle for storage, like it would grow, right, like 2%, 3%, 4% potentially without that headwind?

I'll answer that. Yes, I think that's fair to say. He can -- he'll probably give me a lecture. That's okay.

So then as we inflect next year, I know you don't have guidance for next year, but just based on simple arithmetic, I would imagine that storage is a better business next year than it has been in the last 4, 5, 6 quarters. Is that a reasonable?

We -- look, we -- the most valuable dollar for our company when you look across the portfolio is in our storage business, our Dell IP storage. It's in our best interest to grow and to grow up market. We believe the market will grow next year. We need to grow at or better than the market rate. So it is -- every bit a reasonable expectation that we should grow at the market rate next year.

And that underpins what -- I think you laid out the Investor Day that ISG margin of 11% to 13%, if I'm quoting it correctly?

10%.

10%, sorry, I don't want to misquote you. And that takes into consideration the AI server margin that we talked about earlier.

Correct. That's why it's -- one, it's an important category. It's part of the portfolio mix as we see AI continue to grow. Our traditional server business and storage business need to deliver the profit profile that keeps us in that ZIP code.

Got it. In the few minutes that we have left, I'd be remiss if we didn't talk about PCs. I know that's been an area where you're refocusing your effort to help re, sort of, I think, position the portfolio from a product perspective and end market perspective. But now we have this component issue that's potentially leading to either higher prices, maybe some demand destruction, maybe the inability to get product effectively components.

How are you thinking about where we are in the cycle from a Win 11 components? We'll just merge everything together. Like how do you think about the puts and the takes in the PC business? Because it hasn't played out like we thought it would play out this year. I think people had higher expectations that Win 11 would be a bigger short-term tailwind, and that hasn't really played out. Just how you're thinking about where we are in the cycle?

I think we've all seen that the Windows 10 end of life is taking longer than about any other operating system transition that I've been associated with, which is most of them.

We have an installed base of just under 1.5 billion units. You got 500 million of them, give or take, that are 4 years or older. You got 500 million, give or take, that aren't capable of running Windows 11 that need to be upgraded. So there's a still pile of product that needs to go through an upgrade cycle.

Then if you put -- in addition to that, I believe OSs and more importantly, there'll be an application suite coming over the lifetime of the PCs being bought today that will have AI. So the AI computer with an NPU, the ability to do new and exciting things that we've not thought of is coming, that there is encouragement that there's still more replacement runway or upgrade in front of us.

Now our headwind is the one that I described earlier with an increased cost base is what impact does that have on the TAM? Our view today is the TAM is roughly flat year-over-year. Now that will be give and takes. We'll see what happens as this dynamic cost situation and supply situation, as you called out, impacts us.

Look, I know what we're doing. We're planning to win. I plan to outgrow the marketplace, take share. I communicated during our Analyst Day that we were going to more broadly cover the market because we'd shrink our coverage. I was going to expand coverage. That was going to include consumer.

This is a scale business. It's not lost on us. We need to scale and grow. It is our most capital-efficient business. It allows us to sell a computer and then being able to sell a dock, a display, a keyboard, a mouse, a microphone, a speaker, a headset, et cetera.

So that estate is very important to us. And it drives the completion that we need for our customers from the edge of the infrastructure to the deep backbone of how companies are run and everything in between uniquely positions us. So it's important, we're committed to it. And our goal is to grow and overcome the challenges thrown at us.

So 30 seconds. Since you reported earnings, what has surprised you with the feedback, the reaction or maybe the questions or comments that you've gotten since you reported earnings last week? It's only been a week now, not much time, but what was misunderstood or not appropriately appreciated?

Maybe a couple of things. One, I think this -- there was -- perhaps we didn't communicate clearly enough the performance of our ISG business improving 360 basis points in its operating margins quarter-over-quarter while growing 24%, service and networking growing 37%, ISG growing 24% with a storage business that was minus 1% that underlying the key components of our Dell IP grew. I think we got to do a better job of communicating that plot.

There is a lot of concern about can we -- I keep hearing, "Oh my gosh, NVIDIA is going to standardize the design and there's no room for innovation in the next generation. Oh my gosh." We see the engineering hurdle so significant that we think we're in business for a while with the opportunities we have to continue to innovate and drive differentiation.

And we're optimistic about that and continue to operate in that mid-single-digit margins, are probably the two things that kind of stick out the most. And if there was the third since I like things in 3s, the lack of confidence that we're going to be able to get parts and be able to take increased cost basis and not win when we won in every other cycle seems odd to me, but we have to go prove it. I like a challenge. I'm up for it. We're going to go do it.

Great. Well, I think, Jeff, we covered a lot. We're out of time. So Jeff, thank you very much. Thank you, everyone.

Thanks for having me. A pleasure.

Yes. Look forward to it.

Good afternoon, and welcome to the Fiscal Year 2026 Third Quarter Financial Results Conference Call for Dell Technologies Inc. I'd like to inform all participants, this call is being recorded at the request of Dell Technologies. This broadcast is a copyright property of Dell Technologies Inc. Any rebroadcast of this information in whole or part without the prior written permission of Dell Technologies is prohibited. [Operator Instructions]

I'd now like to turn the call over to Paul Frantz, Head of Investor Relations. Mr. Frantz, you may begin.

Thanks, everyone, for joining us. With me today are Jeff Clarke, David Kennedy and Tyler Johnson. Our earnings materials are available on our IR website, and I encourage you to review these materials. Also, please take some time to review the presentation, which includes additional content to complement our discussion this afternoon. Guidance will be covered on today's call.

During this call, unless otherwise indicated, all references to financial measures refer to non-GAAP financial measures, including non-GAAP gross margin, operating expenses, operating income, net income, diluted earnings per share, free cash flow and adjusted free cash flow. A reconciliation of these measures to their most directly comparable GAAP measures can be found in our web deck and our press release. Growth percentages refer to year-over-year change unless otherwise specified.

Statements made during this call that relate to future results and events are forward-looking statements based on current expectations. Actual results and events could differ materially from those projected due to a number of risks and uncertainties, which are discussed in our web deck and our SEC filings. We assume no obligation to update our forward-looking statements.

Now I'll turn it over to Jeff.

Thanks, Paul, and thanks, everyone, for joining us.

Before we get started, I'd like to congratulate David on his appointment to CFO. We've worked together closely for the past couple of decades, and I look forward to what's ahead. Now moving to our results.

We delivered a strong third quarter, with a record for both revenue and earnings per share and an all-time high in AI server orders.

Total revenue reached $27 billion, up 11%. CSG and ISG combined were up 13%. Year-to-date, total revenue was up 12%, with ISG revenue up 28%.

EPS was up 17% to $2.59, driven by improved profitability in AI and storage and continued operational scaling. Our strong performance and operational discipline led to continued robust cash flow and significant capital returns for shareholders.

Now let's move to AI, where momentum has accelerated meaningfully in the second half of the year building on an already strong first half. AI server demand remained exceptionally strong. We booked $12.3 billion in orders in the quarter, bringing year-to-date orders to $30 billion, both record figures.

The large-scale customer base continues to broaden, with expansion across Neoclouds or Tier 2 CSPs and Sovereigns. Our strong orders and customer base expansion clearly shows customers value our unique ability to design, deploy and maintain large, at-scale AI factories, especially our engineering and rapid deployment capabilities.

We have AI racks operational within 24 to 36 hours of delivery, with uptimes exceeding 99%. We shipped $5.6 billion in AI servers during the quarter for a total of $15.6 billion year-to-date. We ended the quarter with a record backlog of $18.4 billion.

Our 5-quarter pipeline continue to grow sequentially across Neoclouds, Sovereigns and Enterprises, and remains multiples of our backlog, even when accounting for the robust demand we've seen. As expected, AI server profitability improved sequentially.

Moving to traditional servers. Overall, demand grew double digits, with growth accelerating sequentially in both EMEA and North America. We saw growth across units, TRUs, our buyer base and the mix of the 16th and 17th generation platforms, reflecting customers' preference for dense, high-performing compute configurations. Traditional x86 compute demand continues to benefit from workload expansion and AI, driving broader IT modernization and consolidation.

Moving to storage. While revenue declined 1% year-over-year, demand for our Dell-IP portfolio remained strong. For 2 consecutive quarters, our all-flash array portfolio has delivered double-digit demand growth, supported by strong double-digit growth from PowerStore, PowerMAx, ObjectScale and PowerFlex.

PowerStore demand has now grown for 7 consecutive quarters, with 6 quarters of double-digit growth. Profitability improved as we increased both the mix and margin of Dell-IP offerings, underscoring the differentiated value of our platforms.

In CSG, we saw momentum continue. CSG revenue increased 3%, with commercial up 5%. International growth accelerated sequentially, up double digits year-over-year.

North America also showed improvement. Demand for small and medium business remained strong, and we now have 5 consecutive quarters of P&L growth and 7 consecutive quarters of commercial demand growth.

Consumer revenue declined 7%, although the demand environment turned to growth as we refocused on expanding where we play in the market.

Commercial profitability was stable, while consumer and education were competitive. The PC refresh cycle remains durable, supported by an aging installed base and a significant portion of systems not yet upgraded to Windows 11.

And before I wrap up, I'd like to briefly touch on the commodity supply environment. We are well positioned across our commodity basket. Q3 was deflationary, and our outlook for Q4 is largely unchanged from last quarter.

Looking ahead to next year, there will be dynamics that we will have to navigate, but we are confident in our ability to secure supply and adjust pricing as needed. As always, we'll leverage our world-class supply chain to deliver the best outcomes for our customers and shareholders.

In closing, we delivered a record third quarter, with strong performance across all segments and continued operational discipline.

Revenue and EPS reached Q3 highs, supported by growth in ISG, CSG and improved profitability in AI and in storage. AI momentum remains exceptional, with record orders backlog and a growing diverse customer base. Our competitive edge in AI is our ability to engineer bespoke, high-performance solutions, deploy large-scale clusters rapidly and support them globally, all backed by an unmatched ecosystem and flexible financing offerings. This end-to-end capability is why Dell continues to win in AI. We are well positioned to capitalize on AI infrastructure build-outs, expanding traditional infrastructure demand and the ongoing PC refresh cycle.

Now let me turn it over to David to talk more about Q3 in detail.

Thanks, Jeff. I'm pleased with the team's strong execution this quarter, delivering Q3 records for both revenue and EPS, along with strong cash generation and above-trend capital return.

Total revenue was up 11% to $27 billion. ISG and CSG combined grew 13%. Gross margin was up 4% to $5.7 billion or 21.1% of revenue. Gross margin rate was driven primarily by a mix shift to AI servers with shipments doubling year-over-year, partially offset by improved profitability in storage.

Operating expense was down 2% to $3.2 billion or 11.8% of revenue, as we continue to drive scale within the P&L. Operating income grew 11% to $2.5 billion or 9.3% of revenue. The increase in operating income was driven by higher revenue and lower operating expenses, partially offset by a decline in our gross margin rate. Q3 net income was up 11% to $1.8 billion, primarily driven by stronger operating income. And our diluted EPS increased 17% to $2.59, a Q3 record.

Moving to ISG. ISG revenue was a Q3 record $14.1 billion, up 24%, marking 7 consecutive quarters of double-digit revenue growth. Servers and networking revenue reached a Q3 record $10.1 billion, up 37% and is up 43% year-to-date. AI server demand accelerated, with a record $12.3 billion in orders, $5.6 billion in AI server shipments and a record ending backlog of $18.4 billion.

In traditional servers, we saw demand improve throughout the quarter and stability within the P&L. Storage revenue was $4 billion, down 1%, with strong demand across parts of our Dell-IP portfolio. PowerStore continued its double-digit growth trajectory, with 7 consecutive quarters of growth.

ISG operating income was a Q3 record $1.7 billion, up 16%, marking 6 consecutive quarters of double-digit growth. This was driven primarily by higher revenue. Our ISG operating income rate was up 360 basis points sequentially to 12.4% of revenue. This improvement was driven by mix of AI servers, sequential improvement in AI server margins and stronger profitability from storage.

Turning to CSG. CSG revenue was up 3% to $12.5 billion. Commercial revenue grew for the fifth consecutive quarter, up 5% to $10.6 billion, while consumer revenue declined 7% to $1.9 billion. CSG operating income was $0.7 billion or 6% of revenue. Commercial profitability was stable, driven by steady pricing sequentially as customers prioritize rich configured, AI-ready devices. In consumer, profitability improved year-over-year and demand returned to growth.

Moving to cash and the balance sheet. We delivered another strong cash quarter, with cash flow from operations of $1.2 billion. This was primarily driven by profitability and working capital improvements.

We ended the quarter with $11.3 billion in cash and investments, up $1.6 billion sequentially. Our core leverage ratio is 1.6x. We returned $1.6 billion of capital to shareholders, including 8.9 million shares of stock repurchased at an average price of $140 per share, and paid a dividend of approximately $0.53 per share. Through 3 quarters, we have returned $5.3 billion and repurchased over 39 million shares.

With record Q3 results in hand, I'll now walk you through our outlook for Q4. In ISG, we expect to ship roughly $9.4 billion of AI servers in Q4, a record, bringing full year shipments to roughly $25 billion or over 150% year-over-year.

Our Q4 outlook for traditional server and storage remains unchanged from last quarter, supported by continued data center modernization and consolidation and above-market growth in Dell-IP storage.

In CSG, with the ongoing PC refresh cycle, we are improving our execution to drive revenue growth and gain market share. Given that backdrop, we expect Q4 revenue between $31 billion and $32 billion, up 32% at the midpoint of $31.5 billion. ISG and CSG combined are expected to grow 34% at the midpoint, with ISG growing mid-60s and CSG up low to mid-single digits. Operating expenses will be flat sequentially. We expect operating income to be up roughly 21% with continued sequential improvement in ISG operating income rate. We anticipate a diluted share count of roughly 672 million shares and an 18% non-GAAP tax rate.

Our diluted non-GAAP EPS is expected to be $3.50, plus or minus $0.10, up 31% at the midpoint. Our Q4 guidance implies a strong FY '26, with revenue of $111.7 billion, up 17% and non-GAAP EPS of $9.92, up 22% at the midpoint, both well above our long-term framework.

And briefly on FY '27. It's still very early in our planning process, we wanted to give you some context on how we are thinking about next year. We have strong conviction in our AI business, supported by what we see in our backlog, the pipeline and ongoing customer discussions. We've proven we can execute and deliver for our customers in this space.

For the rest of the business, the long-term framework we outlined at our Securities Analyst Meeting remains a solid starting point as you think about next year. We are highly confident in our ability to drive EPS growth, supported by multiple levers, including leveraging our go-to-market engine, improving gross profit, scaling operating expenses and ongoing share repurchases.

In closing, we delivered a record Q3, with revenue of $27 billion and EPS of $2.59, both quarterly highs, driven by strong execution across ISG, CSG and disciplined cost management. ISG continues to see sustained double-digit growth and accelerating AI demand, evidenced by $30 billion in AI server orders over the past 3 quarters. We are focused on capitalizing on the ongoing PC refresh and expect continued growth from CSG. We remain focused on driving shareholder value through strong cash generation and capital returns.

Thank you all for your time. Now I'll turn it back to Paul to begin our Q&A.

Thanks, David. Let's get to Q&A. [Operator Instructions]. Operator, let's go to the first question.

We'll take our first question from Samik Chatterjee with JPMorgan.

Jeff and David, I mean, maybe since this is sort of the topic of investor conversation mostly at this point, if you can flesh out your thoughts on the kind of reaction you expect from customers in relation to the pricing discussions by sort of the product categories, where do you think it's more sort of easier to take some of those pricing actions versus not relative to your overall portfolio?

And David, if I heard you correct, you're saying to sort of use your Investor Day targets for about mid-teens EPS growth as still a starting point for next year despite those sort of dynamics sort of headwinds on the memory side? Can I just clarify that as well?

Sure, Samik. Let me wade my way through that. I suspect it will be the first question this afternoon or the only question I should say. Look, we're in a very unique time. It's unprecedented. We have not seen costs move at the rate that we've seen. And by the way, it's not unique to DRAM. It's NAND. It is hard drives, leading-edge nodes across the semiconductor network. There is a -- if you will, I'd categorize it as demand is way ahead of supply. And as we wade our way through that, we're going to lean on the things that we've always done. We have a lot of experience at this. This isn't our first DRAM cycle. There have been 7, I think, in the last 40 years.

Michael and I have been here navigating the organization in various ways through that time. Our senior leadership team and our supply chain has been through everyone this decade. First rule of our supply chain is to get the parts, supply matters, mix matters. And as we get to supply and mix, our job is to minimize the impact of that to our customers.

But clearly, we're in a situation that is not typical. We've learned a great deal since COVID, since previous cycle of this -- last super cycle of this magnitude was 2016 through 2017. And we're going to do everything we can to minimize the impact. But the fact is the cost basis is going up across all products. No one more unique than others. Everything uses a CPU, has DRAM, has storage in it.

So with that said, we're going to do things we've always done. We're going to work on configurations. We're going to work on availability, adjust mix. Our direct model allows us to move demand where supply is. Our direct model allows us to act to the market signals it gives us quicker than anybody else, allows us to price accordingly, reprice when needed. And we will make our way through that across consumer PCs, commercial PCs, into server storage and through our AI servers. No product category is not going to be impacted in terms of the aggregate cost basis moving.

Again, our number one rule is get parts, secure supply, secure the mix we need to meet the customer demand. The world needs more computational intensity, needs more compute. If you're in the world of AI token growth is going, we see a consolidation in servers. Consolidation in servers is driving denser servers with more DRAM, more storage. And we're in the middle of a PC refresh that's not complete. So I don't see how we will not -- what's the best way to describe it? I don't see how this will certainly not make its way into the customer base. We'll do everything we can to mitigate that. As we mentioned earlier, our cost outlook for Q4 is largely unchanged.

And David just gave you what our guidance is that we believe that you'll see sequential profitability improvement in our company across the broad portfolio while managing an increase in our cost basis. That's what we're going to do. That's what we know how to do, and we have all of the tools in our company to be able to do that effectively and fast.

And David, I think...

Yes. Samik. Yes, like we said, it's very, very early in our planning process, obviously, but the framework from our Security Analyst Meeting is a good reference point to start with. So I think EPS included in that is the ZIP code will be in. We'll be looking obviously to leverage our go-to-market engine, which is differentiated. All the things Jeff has just outlined there in relation to our supply chain. We'll continue to drive significant scale in our OpEx and then obviously stay committed to our capital return KPIs, right, whether it's share repurchase or stay committed to our dividend. So look, we feel we have many tools in that toolbox to allow us to stay agile and deliver on our EPS numbers. But like I said, it's still very, very early in the planning process here.

And we'll take our next question from Mark Newman with Bernstein.

Congrats on a great quarter, particularly impressive on the AI server orders. I wondered on AI servers, if you could talk about some of the recent comments that have been coming from NVIDIA around the potential vertical integration that they're doing, getting a little bit more involved in the supply chain and how that may impact or how Dell is navigating around that? And also on AI servers, any color on the mix of AI servers? Any change on the mix, for example, enterprise as a portion of AI server orders would be useful?

Mark, let me make my way through that. I mean, first of all, as we look forward to the new technologies that are in front of us, we remain excited. We think there's ample opportunity for us to continue to differentiate. These large-scale deployments are very complex. Our value add is at the rack level, is at the solution level, L11 and beyond. That differentiation, we believe, remains for the next several cycles easily. In fact, our ability to engage with customers early, which we can on the next-generation technology to work through their needs to bring these very complex offers to the marketplace fast and at scale with a significantly better uptime and outcome, we believe is a differentiation.

We focus on optimizing performance per watt, performance per dollar at the data center level. We focus on our services, our value-add and deployment, our financing side, the ecosystem that we bring to our customer base, none of that changes in the next generation of technology. And to be honest, I think the opportunity for us gets greater in the future as we head towards 500 kilowatts of rack of power density moving to a megawatt and beyond.

The engineering skill required to do that at rack scale is significant. We've invested in that ahead of the curve, and we believe that gives us the opportunity to differentiate, remain the leader in time to market, drive broad installation and deployment capabilities ahead of our competition at a higher level, uptime of 99% or better. That's why we win, and I don't see that changing.

When I look at the mix, 2 forms of the mix that I'll address is we saw a change in the quarter towards GB300. So in our backlog of $18.4 billion, there's been a significant shift towards GB300 as expected. And then lastly, we continue to see great build on our 5-quarter pipeline around Sovereigns and around Enterprise and remain very encouraged about the opportunities in both.

And the next question will come from Ben Reitzes with Melius Research.

Great. Good execution with the commodity environment, guys, and I'll try to be concise for Paul. The question is around your AI server margins. You mentioned it was up sequentially. I was wondering if you guys could talk about order of magnitude there. And is that going to continue into the 4Q? And are you starting to see more product attached, more high-margin attach to that end?

I'll take a run at it, Ben, and then David can certainly add to this. Clearly, we made reference in Q2 that we had some onetime cost elements that hit us. If you recall, we talked about expedites and supply chain reconfiguration. Those went away in Q3 as expected. We also talked about shipping a lot of the early aggressive GB200 deals in the quarter. Those went through the system. And we continue to now see the ability to add differentiation, as I just mentioned in the previous question, that we see in the GB200 and 300 designs and our margins move to stay right in that range that we've talked about, mid-single digits. We see that continuing as part of our long-term value creation framework that we laid out 8 weeks ago. It's what we'll continue to talk about here, and we believe that we can operate going forward in that range. In fact, we're very confident of that.

And then we also had a mix -- if you will, a change in customer mix to the good. When you look at the broad portfolio and diverse customer set that we have within the AI portfolio, shipping to a broader set of customers across a greater range of solutions helps margin. Hope that answered your question about AI margins.

And our next question will come from Erik Woodring with Morgan Stanley.

I wanted to touch on PCs. Jeff, you sound very bullish on the PC opportunity into next year. Some of the channel partners earlier in earnings were talking about maybe the seventh inning of a PC refresh. And I'd love to just get your comments because you sound more bullish. So where do you think we are on the PC refresh? And is that still Windows end-of-life upgrades that still need to get done? Or are there new factors that you think could elongate the PC cycle well into 2026?

Erik, a couple of things. One, we have not completed the Windows 11 transition. In fact, if you were to look at it relative to the previous OS end of service, we are 10, 12 points behind at that point with Windows 11 than we were the previous generation. So we still have ample opportunity to convert. If memory serves me right, the installed base is roughly $1.5 billion -- 1.5 billion units. We have about 500 million of them capable of running Windows 11 that haven't been upgraded. And we have another 500 million that are 4 years old that can't run Windows 11. Those are all rich opportunities to upgrade towards Windows 11 and modern technology.

Equally important, AI PCs, small language models, more capable applications, improvements in operating systems and their capabilities and the embedded AI there, the use of an MPU, the capability of an MPU and future PCs gives me the view that the PC market will continue to flourish going forward.

Now let's define flourish. We have the PC market in our outlook roughly flat year-over-year. That's after a year that we grew mid- to high single digits. I think it's flat as we look into next year's planning horizon, and we're building plans accordingly that would take share against that outlook.

And the next question will come from Wamsi Mohan with Bank of America.

I was wondering if you could just maybe give some color around this AI business. You noted very strong conviction going into fiscal '27. Obviously, you just raised your guide here from $20 billion to $25 billion. Can you just put that in context of some of the financing issues at Neoclouds? And how much of sort of your conviction and growth is predicated on some of these Neoclouds being able to procure financing versus maybe other customers that you might have visibility into?

And Jeff, if you could just clarify, you mentioned the cost base moving up across the product portfolio. And I was wondering if you could maybe just share at the highest level, how much of that conceptually could you recover from pricing versus how much of OpEx reductions are possible to offset some of these pressures?

Maybe I'll start, Wamsi, and Jeff can add some color. Look, I think if you start answer first, you look at our Q4 guidance, $9.4 billion. That represents $25 billion, obviously, for a full FY '25. So you look at that appetite for AI demand and it's across Neoclouds, Sovereign opportunities and obviously, within the Enterprise, shipments of $5.6 billion in Q3, orders of $12.3 billion, that's year-to-date at $30 billion, backlog at $80.4 billion. And as Jeff referenced in his opening remarks, the next 5-quarter pipeline is multiples of that.

So every conversation we're in, which is also us being very aware of all the opportunities that are out there, is about demand. It's about opportunity and eagerness to work and see the opportunities in front of us. So we actually see huge scale to come. Every opportunity, reality is we're not going to win them all, but we love our momentum that's there, and we think we're well positioned to meet the expected needs of the customer base.

Yes. I would add to that maybe some color, $25 billion this year, 150% increase over last year. On the guidance that David called out, we will ship nearly as much in Q4 as we did all of last year. I think that gives a reflection on the need for compute, the need for and what we see as token generation increasing at an incredible rate and the corresponding compute that has to be behind that to generate those tokens. It's reflected in that 5-quarter pipeline that David said that is up across all 3 customer types, Neoclouds, Sovereigns as well as Enterprises, and we're seeing progress in all 3. So I think that's very important for us to make sure that we communicate that the momentum as we head into Q4 continues. You saw that in the orders in Q3, the backlog building and significant shipments in Q4.

If I flip to your other question about the cost basis and our ability to recover, it's an interesting question. We've set over the years in normal times, when our input costs go up, we can recover roughly 2/3 of that cost in a 90-day period. I would tell you, this is not normal times. This is extraordinary times, and we put extraordinary actions in place weeks ago as we saw this to be able to mitigate the impact upon our company, our customers and our shareholders.

And again, it goes back to our business model, direct signals, we understand the demand, our long-term partnerships and agreements with our partners that make DRAM and make NAND, the agreements we have in place around capacity, those relationships are meaningful and impactful as we navigate these types of situations again that are unprecedented. And then our model gives us tremendous flexibility, whether that is to reprice, whether how we set out quotes, whether that's to reconfigure, redirect to different products, the ability to determine how long price will be in effect, the ability to understand where we're going to drive demand to and change our demand generation vehicles to drive that. It's important that our engine and the way we run, I think, is very different than others in our ability to respond.

And those of you that took note, you saw that in COVID in a very similar situation, the experience that we had there where there was material shortages and increased cost, our ability to navigate that I think was unmatched in the marketplace. Our supply chain is very good at this, and we're going to lean on them. Those lessons learned from the COVID time and most recently what happened with tariffs, I think, show that we can operate with the right sense of urgency. We're managing this real time, actively managing it. I was on 3 pricing calls today alone, and we're driving to get a better outcome. So our belief is we will do better than our normal 2/3 in a 90-day period, given the actions that we've put in place and our understanding of demand and our understanding of supply.

And our next question will come from Amit Daryanani with Evercore.

I guess maybe you could just spend a little bit of time on ISG margins improved rather well by about 350 basis points sequentially. Can you just touch on like what drove the strength in ISG margin in Q3 versus Q2?

And then your guide, I think, reflects the largest AI server revenue number you guys are going to put up in Q4 at $9.4 billion plus. How should we think about that impacting your P&L? And do you think gross margins should remain in the Q2 levels? Or is there kind of further movement from there as we think about the P&L impact from the AI numbers in Q4?

Yes. Thanks, Amit. Yes, look, really pleased with the team's execution in Q3 around ISG up at 12.4%, like you said, up 350 basis points quarter-on-quarter. So a lot to like here. I guess a couple of things to call out. First, on the storage side, look, Q3 was no different than what we've seen year-to-date, where we've seen demand growth at a premium to market for our Dell-IP storage portfolio. Probably a strong call out there will be PowerStore also, 6 consecutive quarters with double-digit growth. So obviously, that Dell-IP portfolio gives us better operating margins as you'd expect. So there's a natural mix effect that creates a tailwind there.

Secondly, in storage, our pricing discipline was something I was very pleased with also. And then thirdly, look, the focus of the teams looking to find improvements at a -- by product level within the portfolio also. So again, like I said, a lot to like on the storage side.

Also within that, on the AI margins, like Jeff said earlier, Q3 on track to what we've consistently committed to mid-single-digit in here in relation to that, and we obviously didn't have those Q2 one-timers that were there, and we'll keep that consistency as we go into Q4.

And then your reference, I think your question was Q3 into Q4, then from a guidance perspective. Look, we expect to continue to make progress. Our Q4 profit guidance is anchored again through the storage P&L. With the Dell-IP storage growth, we expect to make it 4 for 4 in terms of quarterly growth in that portfolio. That should allow us to grow at or likely slightly ahead of normal sequentials, which will allow us to see an uptick in our op rate sequentially also into Q4.

Yes. Amit, just again to emphasize that, AI shipments $5.6 billion to $9.4 billion quarter-over-quarter. Our strategy of focusing on Dell-IP storage, the mix is up, the rate is up and increased velocity of our traditional server business is the recipe for the performance that we expect to have in Q4 while increasing AI shipments significantly, as we mentioned.

And the next question will come from Aaron Rakers with Wells Fargo.

I want to shift gears a little bit away from the AI to the more traditional server business. Jeff, I think in your prepared comments, you mentioned double-digit demand growth. I think if my math is correct, I don't think revenue grew necessarily at that clip. So I'm curious if you could talk a little bit about what you're seeing as far as the aged installed base, where we're at in the upgrade cycle for traditional servers. And do you think double-digit growth is a good baseline that we could think about going into fiscal '27 as that demand follows through to revenue?