Western Digital Aktie

Good morning, everyone. Asiya Merchant here, day 2 of Citi's Global TMT Conference. Asiya Merchant here again. I lead the tech hardware and tech supply chain research here at Citi. Really delighted to host Kris Sennesael here. He is the CFO of Western Dig. We also have members of Western Dig's IR team here in the audience with us.

Before we begin, I'm going to hand it over to Kris to have a few opening remarks. This is a fireside. I'll leave a couple of minutes towards the end for questions. If you do have questions, please do raise your hand so we can bring the mic to you. Over to you, Kris.

Yes. And thanks, Asiya, for hosting us at this great event. And so today, I will be making some forward-looking statements based on management's current assumptions and expectations, including with respect to our product portfolio, business plans and performance, market trends.

These forward-looking statements are subject to risks and uncertainties. So please refer to our Form 10-K and other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations.

We also will be making some references to non-GAAP financials and a reconciliation between GAAP and non-GAAP can be found on the Investor Relations section.

Great. All right.

Back to you, Asiya.

All right. Thank you. Kris, clearly, it's been a great cycle, right, for storage, for HDDs in particular. I think investors who are, kind of, maybe still waking up to what's happened in the HDD space often ask this question about what makes this cycle different? You've lived through many other cycles before. You've been in the semiconductor space as well before. But what makes this current HDD cycle different than maybe prior cycles?

Yes. That's a great question to start. And I've said this many times before, you can basically forget everything about hard disk drives and WD until I joined in May 2025. And so when you look at 3 years or 5 years or 10 or 20 years ago, the business was mostly consumer and PC client where you have a lot of cyclicality. It depends on GDP, consumer confidence. It depends on the PC cycles and IT spending. And it's really a unit-based business.

Fast forward to today, 90% of our business is with the cloud, driven by massive amounts of data being stored in the cloud. You layer the AI cycle on top of that and the understanding by many of our customers that data becomes a lot more valuable. And so you're not really looking at cycles anymore as you've seen it before. It's more a secular growth business.

We have great visibility in the demand side. We talk to our customers. We have earned our seat at the table being a critical component to this AI data center build-out. And we have longer-term visibility, right? Customers are asking to sign LTAs all the way till 2030 and 2031. We haven't signed them yet, but that's the visibility that we have, again, being a strategic component and having earned that seat at the table. And so this is a totally different business.

Okay. And then, we've been asking pretty much everyone here who's at the conference, like the AI, the build is so huge, right? And then how do you think about a digestion period perhaps for this AI infrastructure spend? Are you, kind of, looking at that and saying, okay, a lot of demand here, but at some point, we, kind of, see the digestion of all this infrastructure, including the HDDs that are going into these data centers for data storage?

Yes. So currently, we don't see a digestion. It's actually the opposite. Every time we go and talk to our customers, myself or my CEO or our salespeople, they come back with a stronger demand signal, right? A stronger demand signal and better visibility for longer term, right? Today, and we have been very clear about that.

We do believe that the exabyte demand growth, compound annual growth rate for the next 5 years is greater than 25%. Actually, currently, I think it's a lot greater than 25%. We can't supply more than 25% right now year-over-year growth in exabyte. But the demand is there. The demand is strong. The visibility is good. And so we don't see any digestion right now.

I mean there is a lot of compute power that's being put in place and the CapEx spending of the hyperscalers keeps going up and will soon hit more than $1 trillion. The vast majority of that goes to the compute side. Now data and hard disk drive storage is slightly different, right? Because compute, you need a certain amount of compute to handle all the compute that needs to be done. But every time you use compute, the output of it is being stored.

And so the data that's being generated by AI keeps compounding even when you, so to speak, would stop adding more compute power to it. And so you will have to continue to add more HDD and storage.

By the way, as you know, 80% of all data that's being generated is being stored on HDD. And when you combine all of that, we continue to see strong demand for a very long period of time.

Okay. Great. And then maybe just coming back to your most recent quarter that you guys reported, I think investors were a little surprised that maybe exabyte growth just fell a little bit short of 25% year-on-year. Just there were some dynamics at play in the most reported quarter. Maybe you can just help investors understand what were those dynamics? And then given that you're seeing demand much stronger than that, obviously, you guys are supply constrained as well there. How do you think about the customer ramps and the product mix driving towards the 25%?

Yes. So first of all, again, I've been very clear about that. We do see exabyte demand greater than 25% compound annual growth rate over the next 5 years. By the way, this is a compound annual growth rate. That doesn't mean each and every year. But I do believe the demand will actually continue to grow more than 25% each and every year. But to your point, we are somewhat supply constrained today. And so my revenue is not gated by the demand, my revenue is gated by the amount I can supply.

Now the good news there is, I think, over the next 25 years (sic) [ 3 to 5 ] , we do believe that we will be able to grow the exabyte supply compound annual growth rate greater than 25%. Now again, that doesn't mean each and every year, right? If you look at last year, last fiscal '26, we grew exabyte on or about 25% year-over-year. Now that had 1 quarter where we had 34%, which we had an easy year-over-year compare. But we were on this, kind of, call it, low 20s, 22%, 23% year-over-year exabyte growth in supply right now.

We are obviously working really hard to accelerate that. And the way how we are going to accelerate our supply year-over-year growth is through technology and product transitions as we move to those higher capacity drives. Now that is not a linear progression, right? We are currently somewhat in between 2 product transitions.

Over the last 4 or 5 quarters, we flawlessly executed on a big product transition, our 32-terabyte ePMR solution that we progressively shipped more and more over 4 or 5 quarters. We are now somewhat in between the next product transitions, but we basically have 3 product transitions in front of us, right? We are going to start shipping.

We've actually shipped already a little bit in last quarter, our next-generation ePMR up to 40 terabytes, then followed with our first-generation HAMR up to 44 terabyte and then followed by 50-plus terabyte next generation as well. And so as we execute on those technology and product road maps, we will be able to accelerate the exabyte supply year-over-year growth as well.

Okay. And then we hear a lot about KV caching on the storage side. That's like a big buzzword, I guess, these days for AI-led storage, especially as we're talking about inferencing workloads and the context that you need to store in order for the inferencing workloads to meet certain requirements. So how does Western Dig think about that market? Like is that something that could be addressed with HDDs as well? Is that more for memory or storage that is maybe higher performance than HDDs? How do you think about KV caching as an incremental perhaps demand driver for HDDs?

Yes. So let's start at the highest level. As I said before, roughly 80% of all data today and in the future is and will be stored on hard disk drive. There is a small percentage that's being stored on tape and then you get 15% to 20% that's being stored on SSDs or flash. And so why is it not 100% hard disk drives? Well, flash has some performance advantages. You can get faster to the data, input, output, read and write. And so they have a performance advantage.

However, we have been investing not only on getting to higher capacity drives, but also have been investing in technology and product advancements to get to higher performance drives, getting the better throughput, faster read and write to the data as well. And as we continue to execute on those high-performance drives, we will be able to encroach into what's today the flash segment, right, as we get to higher performance. KV caching is one of the opportunities out there. Obviously, we need higher performance there. But as we execute, there is an opportunity there.

Okay. All right. And would that require a partnership with some flash makers or in that case, you'll just be maybe buying some components and -- or maybe it's just pure HDD solutions that you're talking about.

Yes. You need higher performance HDDs.

Okay. All right. Great. Maybe just on pricing. I mean, clearly, you've talked about demand significantly exceeding supply here. You've gone from an industry where you used to have like some price declines to rising pricing environment, clearly, on a per gigabyte basis when you looked at your recent results. You've talked about further pricing improvements. So maybe just talk about you have these build-to-order requirements and then you're putting in this upward pricing, potentially, there's some upward pricing. So how are you, kind of, balancing those 2 out?

Yes. So in terms of pricing, we look at it from a value base. We continue -- as we move to higher capacity drives, we continue to add more value to our customers. Higher capacity drives results in better rack density that results in lower real estate costs that results in better power consumption or lower power consumption. And so that all improves the total cost of ownership despite the fact that we are charging a little bit more on a price per terabyte.

The price per terabyte 2 quarters ago was up 9% year-over-year. Last quarter was up 18%, 19% year-over-year. And I think there is still a little bit more room there to grow that. Again, as we continue to add more value, moving to higher capacity drives, adding more features, higher performance to our drives, that will enable us to further improve the overall pricing environment.

Okay. And then you often talk about if there is demand that comes in intra quarter from your customers that's above sort of the commitments that you've already made to them, there is upward momentum in there, in pricing in those situations as well.

Yes, yes. So we do build to order, right? So we -- I mean, it takes 52 weeks to produce a hard disk drive. And so we ask most of our customers, but not all of it, to place purchase orders at least 52 weeks in advance. Now some of the business is being still quoted and priced in 3, 6 or 9 months, but the vast majority is within that 52 weeks.

Now even there, we have now entered into LTAs with many of our customers. I think the vast majority of calendar or fiscal year '27 is under LTA. Some of fiscal or calendar year '28 is under LTA and even a smaller part of fiscal or calendar year '29 is under LTA. I mean customers want more LTAs. Customers want to secure supply in 2030 or 2031. When we do LTAs, we typically have -- we commit to a certain base volume, which is typically less than what our customers want. And so we keep a little bit of volume flexibility to the upside at a certain base price.

But that also means that if there is upside, if we can supply more, that will be at a different price as well, right? And so there is both flexibility on a volume as well on the pricing side. And again, that's why you have seen some further ongoing improvements on a price per terabyte basis.

Okay. And you talked a little bit about earlier in your commentary about the value that you guys are driving the TCO for the customers that's beneficial as you guys move to higher capacity drives. Just -- and help us understand, like, the value, the alternative to HDDs is significantly priced higher, right? I mean that's part of the equation if I'm sitting there looking at what storage media do I want to purchase.

So just how is that -- when you think about that TCO and the price per gigabyte that you guys are probably charging, maybe just help us understand how that relationship's, kind of, evolved, like where it was as you're thinking about it now? And then obviously, when people are looking at the alternative, saying, well, in this case, it makes obviously a lot more sense to continue on HDD.

Yes. And so again, the name of the game here for us is to move to higher capacity drives, right? And we have been executing really well going from 24 terabytes to 26 terabytes all the way up to 32 terabytes. We are now at the cusp of introducing the 40s and the 44 terabyte drives. And then we do actually have a road map to get to 50 plus, eventually 100-plus terabytes per drive. And again, that adds a lot of value to our customers, right?

Lower real estate cost through better rack density, better power consumption and the ability to manage vast amounts of data in a very predictable, scalable, reliable way. And at the same time, as I said before, we're also adding more performance to the hard disk drives. And the combination of all of that creates a vast amount of value creation for our customers that they need, right?

Because let's be clear, there is no AI without HDDs, right? HDDs might only be 4% or 5% of the total CapEx spending of the big hyperscalers, but it is a critical component, right? And without HDDs, there is no data centers, there is no AI. Again -- and that's why 80% of the data is being stored on HDDs. It's the most economical, scalable, reliable way of storing data.

Okay. As we talk about HAMR, you guys have laid out you're doing quals, you expect to ramp here. What are the most important customer qualification milestones that investors should look out for as you're continuing -- you're going into HAMR as well?

Yes. We're doing really well with our HAMR technology and product road map. We've been working on for almost 10 years by now. And we're getting to the point where soon we will start shipping our first-generation HAMR. We're currently in qualification with 4 large hyperscale customers, and the qualifications are going really well. We are getting great feedback from our customers. They like the product. They like how it behaves inside their production environments. They like the areal density, the reliability and the quality of the product. And so we are ready to go.

We've said before, we intend to start shipping in the first half of calendar year '27, probably a little bit in Q1, a little bit more in Q2 and then continue to ramp that in Q3 and Q4 of calendar year '27 and beyond that as well. And so things are well on track. And so I can't wait to start shipping the first HAMR drives.

And just again, from those who, kind of, follow the HDD qualifications, like could you like double-click a little bit what exactly needs to happen right now? Just because one of your competitors is already up there on HAMR, but the fact that these hyperscalers have already tested HAMR before, would that speed it up a little bit? There's obviously demand for more capacity. And could that speed it up? Like what are some things that have to happen for those qualifications to be met before you do volume production?

Yes. It definitely helps a little bit that we are not the first one introducing HAMR. So our customers are familiar with HAMR drives as well. Having said that, our customers are, I think, rightfully so, very conservative as it comes to qualifying new products. They don't want to lose my data or your data or anybody's data here in the room, right?

And so they have a very rigorous process. They're not cutting corners or anything like that. They just follow the process. And so we are going through it right now. Again, we're executing well, great feedback and all of that. In the meantime, we are at the cusp of ramping the 40-terabyte ePMR, right? And so customers love that product. They have more than 10 years of experience with that product. But there as well, we went through the same lengthy qualification cycles as well.

It's just a little bit earlier in terms of the time line. And as I said, we shipped a little bit last quarter. It will start ramping this quarter and then progressively more over the next couple of quarters as well.

Okay. All right. And then just from a competitive opportunities, you guys did talk about high bandwidth drives at your Innovation Day. You talked about dual pivot architecture. You talked about opportunity within AI data lakes. Just help investors understand where -- has there been adoption on some of those innovations that you've talked about? Are they part of some qualifications? And when do you see that ramp for some of the additional technologies?

Yes. So our high bandwidth drives was very well received by most of our customers. We are shipping, I believe, 5 of our customers. We're shipping engineering samples as well. And again, so far, the feedback is really great. They like the characteristics and the features of the product. We will continue to further enhance, right?

We are now targeting a 2x improvement on the performance, but we have a 4 and an 8x that's in the works. And as I said, like so far, customers really like the product. The introduction of the high bandwidth drives will be in the plus 50 terabyte range. We're not introducing that in the 40s or the 44 terabyte. But once we move to the higher terabytes, there it becomes even more important to have those features of high performance.

Okay. All right. So yes, so you're saying it will be timed with when you do the 50 terabyte, right? Okay. That's great. Maybe I'll just turn it around to the questions. If there is any in the audience, please do raise your hand. We have one here in the front.

You talked a lot about maybe some of the differences today versus maybe prior cycles in HDDs. It's clearly more structural demand for exabytes, higher capacity drives. It seems like the industry is a little bit more focused on no new supply. And so it would be interesting to hear from your perspective, 3, 4, 5 years down the road past this huge surge in capital spending, what does the industry look like? Is there sort of more of a predictable algorithm of volume and pricing growth over time? Or do you expect there to continue to be cycles as there have been in the past?

Yes. Again, the cycles in the past were totally different, right? It was consumer PC cycles and other things. Here, this is a lot more structural, right? This is a planned economy. Our customers, they plan their data center builds more than 5 years in advance because they need to go and secure the land, they need to go and secure the permits. They need to go and secure all the XPUs. They need to secure the memory and they need to secure, of course, the storage side on the SSD and the HDD side. And so they really need to plan many, many years ahead. And they are doing that.

And they -- and again, we have -- with our deep customer engagements, we have earned our seat at the table. We do get insights and we are able to challenge them and ask them to explain. And that's why they want to secure the supply for up to 5 years out, right? Because they have this all planned out.

As I said before, data is still somewhat different than compute because data keeps compounding, right? When you have compute power installed, every time you use your compute for 3 or 5 or 10 seconds, you can recycle it and reuse it. But every time you use it, you spit out data that needs to be stored and that typically gets stored forever.

Maybe you can switch a little bit to talking about margins, Kris. I mean, incremental margins, 70% plus, I think you guys have talked about. And on top of that, you have operating leverage as well. So just help investors understand, like, margin expansion, sort of what you're thinking about over the next several years? Is there some theoretical limit there that investors should keep in mind, clearly, as supply continues to undership demand here?

Yes. No, we've done a great job at improving the profitability. Again, the name of the game is move to higher capacity drives. As you move to higher capacity drives, you add more value to your customers that's reflected in a better price per terabyte. At the same time, higher capacity drives also lower your cost per terabyte. And so the result of that, the gross margins have continued to increase, and we are now in the mid-50s with operating margins now in the mid-40s. And so this is a very highly profitable business.

And is there -- I get the question a lot, is there a ceiling to the gross margin? Is it like a magical number where gross margins could not get over? I don't see it, right? Again, we need to continue to execute and continue to drive the technology and the product, move to higher capacity drives and continue to add more value, which will also continue to lower the cost. So I think we're very well positioned.

We are obviously very thoughtful in terms of supply. We want to make sure that we fulfill the needs of our customers, right, with the strong exabyte demand growth. But again, as we execute on our technology and product road maps and introduce the 40s and the 44s and the 50 and the 50-plus and eventually down the road, the 70s and eventually the 100-plus terabyte drives, we will be able to continue to grow the exabyte supply somewhat in line with the exabyte demand growth that we see as well.

Okay. And then, as you're talking about transitions towards these higher capacity drives, 40, 44, 15 (sic) [ 50 ] , how should we think about these cost per terabyte reductions? I think it historically was around 10% or so roughly in terms of cost per terabyte. As you're transitioning towards these higher capacity drives, you have SMR, UltraSMR as well on that. How should we think about the cost per terabyte reductions?

Yes. So in the long term, we do believe that we will be able to reduce cost per terabyte on or about 10% year-over-year, right? That's very clear. And we've demonstrated that in the past as well. Now you need to be careful, right? That doesn't mean each and every year or each and every quarter, it will be down 10% year-over-year, right? Again, the biggest driver for cost per terabyte to be down is move to higher capacity drives.

As we explained before, currently, we are somewhat in between 2 product transitions. And so when you're in between 2 product transitions, that slows down temporarily a little bit the cost per terabyte reductions that you see.

In addition to that, I mean, we do see input cost increases, right? As you probably know, inside the hard disk drive, we use a little bit of DRAM and we use a little bit of NAND, and that has become a lot more expensive. And across the board, the supply chain, you do see a little bit of input cost increases. And so that's, again, not each and every quarter, you will see the 10% cost down per terabyte. And currently, we are facing a little bit of headwinds. But then again, in the long term, mid- to long term, as we move to those higher capacity drives, as we continue to execute on our value engineering initiatives, taking out cost on the drive as well.

I have strong conviction that over time, cost will be down on or about 10% year-over-year on a cost per terabyte. And even in the near term, given where the pricing dynamics are and where the cost dynamics are, gross margin will continue to increase at a pretty good pace.

Okay. All right. Great. And then, just you do have -- even though your CapEx is very small as a percentage of your revenues, you are still investing. What are some of the areas that you need to invest in, especially as you're going towards the 40 -- the higher capacity drives, maybe they have more heads or platters in there as well? Like just help us understand the investments that you're making right now. What are they -- how are they in line with the technology road maps that you have?

Yes. So our long-term CapEx forecast is in the range of 4% to 6% to revenue, right? Now again, that doesn't mean each and every year, it's going to be within that range. Actually, the last couple of years, we've been below that range. I believe the next couple of years, we might be actually slightly above that range. But when you look at it over 5-year periods, it should be in or about that 4% to 6% CapEx to revenue.

And so where do we spend the CapEx? Well, most of the CapEx spending goes to head and media as well as automation, right? And why do we have to continue to invest in head and media? Well, as we execute on our technology and product road map, right, we -- and as we move to those higher capacity drives, we -- how do we get to higher capacity drives, there's 2 factors. The most important one is areal density, right? Get to 3 terabyte or 4 terabyte or 5 terabyte, eventually way down the road, 10 terabyte per platter, right? In order to do that, you need better head and media, right? And so we need to continue to invest in the head and media.

Now secondly, we also -- the way we get to higher capacity drives is to add more platters in the box without changing the form factor of the box. And so as you add more platters in the box, you also have to make sure you have more capacity for media and heads as well. Media, by the way, for us is 2 sides. We do our own substrate and then turning substrate into media, into the magnetic platters as well, right? And so there's no hesitation there. We will make the necessary investments in there.

And then, of course, we also focus on automation, right? We are big believers in AI. We use a lot of AI inside our factories as well, but that also requires a lot of automation. All of that translates into better yield, better productivity, more uptime of the equipment that actually enables us to produce a little bit more units each and every quarter without actually having to spend major CapEx in expanding the unit capacity, we are able to produce a little bit more units just by investing in automation and productivity inside our manufacturing sites, which, again, all helps us to further drive down the cost per terabyte over time as well.

Okay. And then, just all of that comes back down to great free cash flow generation, and you guys have been great at that. There's a net cash balance sheet right now. You've been talking about an optimal -- you've been generating a lot of free cash flow, returning that to shareholders. Just help us understand what is that optimal capital structure for Western Dig?

Yes. First of all, I mean, the free cash flow is very strong. Our target free cash flow margin is plus 30%. We've already a couple of quarters operating at that level. Now there again, each and every quarter might be a little bit up and down depending on how much CapEx and what's going on in the business as well. But we have a business model that throws off a lot of cash. That's one.

Second, the balance sheet is in really good shape right now. We've completed the monetization of the SanDisk shares. We've cleaned up most of the debt on the balance sheet. As you've seen from some of the filings that we've done, we're still working on the convertible debt and cleaning that further up. But we are already in a positive net cash position right now. And so there is no hesitation.

All the remaining accessible free cash flow is being used to return back to the shareholder. And we do that through a combination of our dividend program and share buyback program. The vast majority of the cash goes to the share buyback program, but we are, of course, fully committed to our dividend program. We've already increased the dividend a couple of times, and we'll continue to do so.

And so when you put that all together, I think we have a very shareholder-friendly capital return policy in place that on top of the strong revenue growth for many years based on long-term demand visibility, the strong gross and operating margins improvements that all translate to very strong free cash flow as well as strong earnings per share growth that's being further accelerated with our shareholder-friendly capital return policy. I think that's a great setup.

That's a great reason to own Western Dig. Thank you very much. That wraps up the call.

Good afternoon, and welcome to the Western Digital's Fourth Quarter Fiscal 2026 Conference Call. Please note, this event is being recorded. I would now like to turn the conference over to Mr. Ambrish Srivastava, Vice President of Investor Relations. Please go ahead.

Thank you, and good afternoon, everyone. Joining me today are Irving Tan, WD's Chief Executive Officer; and Kris Sennesael, WD's Chief Financial Officer. Before we begin, please note that today's discussion will contain forward-looking statements based on management's current assumptions and expectations, which are subject to various risks and uncertainties.

These forward-looking statements include expectations for our product portfolio, our business plans and performance, ongoing market trends and our future financial results. We assume no obligation to update these statements. Please refer to our most recent annual report on Form 10-K and our other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations. In our prepared remarks, our comments will be related to non-GAAP results on a continuing operations basis, unless stated otherwise. Reconciliations between the non-GAAP and comparable GAAP financial measures are included in the press release and other materials that are being posted in the Investor Relations section of our website at investor.wdc.com.

Lastly, I want to note that when we refer to we, us, our or similar terms, we are referring only to Western Digital as a company and not speaking on behalf of the industry. With that, I will now turn the call over to Irving for introductory remarks. Irving?

Thanks, Ambrish. Good afternoon, everyone, and thank you for joining us today. Let me begin by reflecting on our first full fiscal year of WD as a focused pure-play HDD company. We drove strong year-on-year revenue growth of 36%, while expanding our gross and operating margins significantly. We doubled EPS and generated $3.5 billion in free cash flow and strengthened our balance sheet to a net positive cash position while continuing to make significant capital returns to shareholders.

These strong financial results an outcome of meeting our customers' growing storage demand by focusing on innovation and driving operational excellence across the global WD organization. We enter fiscal year 2027, and with robust customer demand, increased visibility and continued confidence in the durability of demand as well as our ability to service this demand with our industry-leading products and technology road map.

Demand for storage is being driven both by AI as well as our core cloud services. Within the AI infrastructure market, there is a key dynamic taking shape. While compute cycles can be reused, data compiled, training and inference workloads share and reuse compute resources over time, yet the data generated by these workloads, including model inputs, outputs, locks and retained contacts continues to accumulate. As AI usage scales, this creates a growing need for storage infrastructure capable of economically storing and managing this massive data sets.

The inflection we have discussed in the past from AI training to inference to Agentic AI has only become more pronounced. Training models create significant initial data requirements, but inference generates and retains data continuously. Today, the largest AI platforms process tens of billions of tokens per minute, and billions of prompts per day, creating a rapidly expanding body of data that must be stored, managed and accessed over time. Recent disclosures show token volumes growing several full year-over-year underscoring the pace at which inference is scaling.

Meanwhile, AI is moving rapidly from merely answering questions to agent AI that does the work coordinating task accessing data and operating continuously across multistep workflows. This transition creates a fundamentally more data-intensive workload and one that is increasingly persistent rather than transient. The storage implications are significant. Agents generate data at every step of the workflow, increasing both the volume of data created and the amount that must be stored over time.

This is why we continue to view Agentic AI as a structural step function driver of capacity orientated storage demand. Beyond inference and Agentic AI, we are also seeing the emergence of physical AI, autonomous vehicles, robotics and industrial automation systems and humanoids, where the volume of real world data needed to train these systems is insufficient. Thereby requiring the generation and storage of synthetic data sets, creating another driver of storage demand.

Infrastructure investment is important, but it's only the beginning. Training creates the initial data foundation. Inference generates data continuously. Agentic systems multiply the volume and frequency of that data and physical AI accelerates the cycle further. Together, these trends create a more durable demand environment for data storage, driven not just by building AI infrastructure, but by the continuous creation and retention of data once that infrastructure is deployed.

As AI workloads move from deployment to sustained use, storage demand becomes less about a onetime infrastructure bill cycle and more about the compounding of data, that is the underlying secular demand growth driver for our business. Today, roughly 80% of data stored in hyperscaler data center resides on hard disk drives. That is likely to continue. That reflects what HDDs do exceptionally well, delivering the scale, economics and power efficiency required a long retention large-scale data storage.

Let me now turn to why these secular growth drivers play to WD's strengths. WD's robust technology road map is based on the industry-leading are density per platter with a focus on innovation to meet our customers' capacity needs at scale, along with our growing ability to meet their requirements by providing cost-effective storage solutions in additional layers of the AI storage stack. We are on track to ship our 44 terabyte HAMR product in the first half of cal year 2027. Customer feedback on the qualification process continues to be very positive with the capacity, performance and reliability of our drives exceeding customer expectations.

For our next-generation 40 terabyte ePMR drives, we commenced shipments in our June quarter, and are now entering volume production with two customers. We are currently ramping our out Ultra SMR technology with the third major customer. We expect that Ultra SMR will make up around 60% of our near line Exabyte shipments as we exit fiscal 2027. Since our last update, we have continued to broaden our customer engagement and qualification pipeline with additional hyperscaler and cloud customers advancing through qualification and deployment planning.

Beyond capacity, we are also extending innovation into new layers of the AI storage stack. We are making progress on improving drive performance before high bandwidth drives and are now sampling with five customers, now getting up to 8x the throughput of today's drives without the corresponding increase in power drop, exactly the kind of performance the workloads require.

In closing, the opportunity in front of us is real. And at WD, we are well positioned to capture it. Data creation isn't slowing. It's accelerating. And as the belly of data grows, does the infrastructure storage requirements to store, manage and protect it. That's a durable long-term tailwind for our business, and we intend to fully capitalize on it. Our technology road map is strong, our customer relationships are deep, and we have the operational discipline to translate this opportunity into sustained earnings and free cash flow growth and long-term shareholder value.

With that, let me hand it over to Kris to walk you through the financials and our outlook for Q1.

Thank you, Irving, and good afternoon, everyone. Fiscal 2026 was an outstanding year for WD, driven by broadening demand, deepening customer engagements and disciplined execution. We grew revenue 36% to EUR 12.9 billion while expanding gross margins 970 basis points to 49.1% and increasing operating margins by 1,290 basis points to 37.3%. We more than doubled earnings per share to $10.22, and we generated $3.5 billion of free cash flow, delivering a robust 27% free cash flow margin. We returned $3.1 billion to shareholders, reflecting our confidence in the durability of the business and our commitment to long-term value creation.

Let me now turn to our fourth quarter of fiscal '26. Revenue remained at $3.75 billion, up 44% year-over-year on the back of strong Exabyte growth and favorable pricing dynamics. Earnings per share grew 109% year-over-year to $3.56. Revenue, gross margin and EPS all came in at or above the high end of the guidance range. We delivered Exabytes to our customers, up 22% year-over-year. Nearline continued to drive our Exabyte growth, complemented by solid non-nearline Exabyte growth in the quarter. We began shipping the next-generation EPM drives with capacities up to 40 terabytes in our fiscal fourth quarter and expect a strong ramp over the next few quarters.

Cloud represented 89% of total revenue at $3.3 billion, up 43% year-over-year as demand for our high cast nearline products was strong with a favorable pricing environment. Client represented 6% of total revenue at $225 million, up 61% year-over-year. Consumer represented 5% of revenue at EUR 187 million, up 38% year-over-year. Both segments benefited from improved pricing. Gross margin expanded 1,310 basis points year-over-year to 54.4%, resulting in strong year-over-year incremental gross margin.

This was driven by a mix shift towards higher capacity drives, favorable pricing across our portfolio and disciplined execution in our manufacturing operations. During the quarter, the blended average year-over-year price increase per terabyte improved from high single digits last quarter to high teens this quarter, reflecting the impact of our predictable and sustainable pricing strategy as we deliver greater value to our customers.

Operating expenses were $382 million or approximately 10% of revenue, a 170 basis point sequential improvement, demonstrating further operating leverage in the model. strong top line growth, expanding gross margins and leverage in the model drove operating income to EUR 1.66 billion, up 126% year-over-year, translating into a durable operating margin of 44.2%, up 1,610 basis points year-over-year. Interest and other expenses were $10 million, and our effective tax rate was $0.16.

Taking into account the diluted share count of 388 million, earnings per share was $3.56, an increase of 109% year-over-year. Operating cash flow was $1.4 billion and CapEx was $108 million. This resulted in free cash flow generation of $1.3 billion for the quarter and a strong free cash flow margin of 34%. During the quarter, we completed the monetization of the remaining 1.7 million shares of SanDisk exchanging them for 4.8 million WD shares. Additionally, we repurchased 2.3 million shares of our common stock for a total of $1 billion.

Our full year and fourth quarter repurchase numbers that we are describing on this call include $328 million to settle the conversion premium for some of our converts in cash rather than in stock, avoiding the issuance of roughly 773,000 new shares. We also did $54 million of dividend payments and we ended the quarter with $1.1 billion of debt and $1.6 billion of cash, resulting in a net positive cash position of $500 million.

Lastly, the Board has declared a cash dividend of $0.15 per share of our common stock, which will be paid on September 17, 2026, to our shareholders of record as of September 8, 2026. I will now turn to the outlook for the first quarter of fiscal 2027. We continue to operate in a strong demand environment with improving longer-term visibility. And favorable pricing dynamics across our cloud, consumer and client end markets.

We expect revenue to be $4.1 billion, plus/minus $100 million. At midpoint, this reflects a growth of 45% year-over-year. Gross margin is expected to be in the range of 55% to 56%. We anticipate operating expenses in the range of $390 million to $400 million. Interest and other expenses are expected to be $15 million. The tax rate is expected to be 17%. As a result, we expect diluted earnings per share to be $4, plus/minus $0.15 based on a non-GAAP diluted share count of 388 million shares.

As we look ahead, we have high conviction in the drivers of demand, AI, cloud and the proliferation of data-intensive workloads continue to accelerate. On the supply side, our industry-leading technology and product road maps, combined with strong operational execution enabled us to deliver substantially more Exabytes to our customers. This does not require spending CapEx to add unit capacity, but we are making the necessary investments in our heads and media operations as well as an automation to increase our productivity.

In summary, durable demand, disciplined supply and expanding margins position us to deliver sustained earnings growth, strong free cash flow and long-term shareholder value. With that, let's now begin the Q&A. Ambrish?

Thank you, Kris. Operator, you can now open the line to questions, please, and to ensure that we hear from as many analysts as possible, please ask one question at a time. After which, we will give you an opportunity to ask one follow-up question.

[Operator Instructions] And our first question comes from CJ Muse from Cantor Fitzgerald.

It's hard not to compare your results with your main competitor, where they're seeing better sequential top line growth and targeting gross margin is nearly 200 bps higher than your September guide. So curious what you make of this? Is that due to their early ramp of HAMR, is it a difference in timing of contracts, perhaps you're selling fewer bits into the open market? Would love to hear your thoughts there and how you think about kind of gross margin acceleration from here? Do we need to kind of wait for HAMR in calendar '27? Or are there other drivers that can allow you to push higher?

Thanks for the question, C.J. there are always quarter-to-quarter variations in terms of gross margin because there's a whole range of factors that go into driving it. Obviously, timing of pricing in relation to the long-term agreement contracts is one. So there will always be timing differences on when existing LTAs expire and new LTAs kick in with different pricing regime. So that's one factor.

Second will be the introduction of next-generation platforms. So we're very excited that we started shipping in Q4 up to 40 Terabyte ePMR drives. Right now, we are very much on track to hit 50% of nearline bids on that form by third quarter of year '27. So that will, again, will drive our ability to ship more higher capacity drives into the marketplace and provide some additional upside opportunity from pricing as well, and obviously, we continue to work on operational efficiencies like we normally do.

So in short, there will always be quarter-to-quarter variations, but we feel very confident with the things that I mentioned in terms of the pricing structures that we have put in place with the introduction of our higher cap drives coming out in the second half of the year and our ongoing operational efficiencies that will lead to stronger Exabyte growth continuous top line growth and obviously ongoing margin expansion.

Do you have a quick follow-up, CJ, which wouldn't be a multipart question, please.

Yes, of course. Marvel hosted this session at Flash Memory Summit talking about the economically indispensable nature of HDDs in the AI data center. So curious with that as a backdrop, how is your visibility kind of improved? And how do you think it could improve further as you start signing contracts that renew.

Yes. Thanks for the question,. Look, I think we have strong visibility and increasingly strong visibility it's being customer driven to a large degree as well. The last time we reported on LTAs, we talked about having one LTA of a large customer all the way up to calendar year '29. But we're very much in the coals of discussions with customers to establish LTAs for calendar year '29, '30, and '31 as well. So visibility remains very strong. customer-driven demand for LTAs extending all the way out to '31 remains very strong. So we continue to work on them.

I think we have very good line of sight in terms of the exabyte demand. What we're working through with them is more in terms of the pricing commercial construct of what those LTAs would look like going forward.

Our next question comes from Amit Daryanani from Evercore. .

I guess maybe the first one is exabyte shipments were up about 21%, I think, year-over-year to 21%, that's below the 30% exabyte growth you folks have had the last several quarters. Can you just talk about -- is this that can be pause related to timing or supply? Or does this sort of print a more normalized growth rate? And how do you think about exabyte growth through fiscal '27? .

Yes. I mean I think what we've guided to is a exabyte. We see demand growing at above 25%, right, 25% plus going forward. So we clearly have a road map that's designed to support that. That will, again, as I mentioned, similar to margins, we quarter-to-quarter fluctuations in terms of exabyte growth rates depending on the mix of products that we ship. And as we move into the second half of the year as we really ramp up the 40-terabyte shipments. As I mentioned, that will be over 50% of the nearline exabyte ship, we expect exabyte growth rate to accelerate. And then obviously, we have HAMR coming on in the second half of the year and in calendar '27 as well at a 44-terabyte level.

Yes. And just to add, I think in Q4, exabyte shipments were up 22% year-over-year and when you look at it on a full fiscal year '26, exabytes were up 25% year-over-year.

Do you have a follow-up, Amit? .

Just on the pricing side, right, price per provide went up pretty dramatically high teens year-over-year, Kris, I think you talked about this. Can you just talk about how do you think pricing is going to keep evolving from here? And maybe specifically the new LTA that you folks are signing, are they enabling this kind of high-teens pricing? Or is there more upside as we go forward? .

So I mean, I think we were beneficiaries of across-the-board pricing improvements, both in terms of our nearline products and obviously in the client consumer space as well. In terms of the near line, especially with our large customers, we've been very consistent in talking about predictable icing. And that's the strategy we continue to adopt. This will give them visibility to ensure that they can make long-term architectural decisions with the right economics underpin that. So -- but even then as we increase higher capacity drives, we provide better TCO value to them, that gives us an opportunity to increase pricing, but you'll see that on a very predictable basis going forward. .

In terms of the non-nearline space, particularly in our client and consumer space, we saw a higher opportunity to increase pricing, predominantly driven by the pricing structures on alternative products that are flash based as well. So that resulted in a slightly higher pricing environment. Kris, do you want to -- anything you want to add to that?

No, I think you summarized it well. .

The next question comes from Aaron Rakers from Wells Fargo.

I guess in the building on the prior questions, it seems to me like you've got a product transition dynamic going on with the 40 terabyte ePMR drive. If I look at your cost structure, it looks like you saw a plus per terabyte that was more maybe flattish, down a little bit sequential. So I'm curious, as we move through this product transition, how should we think about cost per terabyte or cost of goods sold per terabyte progressing through '27? .

Yes. So Erin, the cost per terabyte in Q4 of fiscal '26, the quarter that we reported was down approximately 8% year-over-year. We've indicated before that cost per terabyte in the long-term will come down on or about 10% year-over-year. And the main driver, as Irving already explained, right, is the move to higher capacity drives better aerial density, which gives us a substantial reduction on a cost per terabyte. And as we execute in the near, mid and long-term, execute on our technology road map, execute on our product road map introduced the next-generation ePMR and introduce our HAMR products into the market, we will, over time, continue to see cost per terabyte coming down. follow-up.

Do you have a followup, Aaron.

I do. Maybe Kris, back to -- you've done a lot on the balance sheet, right? You've got a $1.1 billion left on kind of the debt. You've got $1.6 billion in cash. I'm just curious, how are you thinking about the capital structure from here? Is it pretty much 100% free cash flow return? Do you want to put more debt on the balance sheet? I'm just curious if any updated thoughts on that.

Yes. No change there to our strategy and our commitment to return the free cash flow back to the shareholders consistently with what we have been doing over the last 4 or 5 quarters through a combination of our dividend payments as well as our share repurchases. And so we're fully committed to those 2 programs.

Our next question comes from Wamsi Mohan from Bank of America.

I was wondering, Irving, just if you could maybe drill down a little bit on the mix comment that you made in the answer to one of the prior questions on the 22% exabyte growth because previously expected that those room deliver more exabytes in the quarter and the deceleration seems pretty strong. So I was hoping maybe you can double-click a little bit on your mix comment. At the same time, you're expecting obviously an acceleration here in your comments, both on a macro level as well as from your product cadence perspective. So anything you could give us in terms of what is changing in that mix very specifically, which caused maybe the slower growth in the quarter? I mean, 22% is pretty strong, but like definitely a deceleration and then further acceleration, like is there a near-term cap on that before you maybe start to ramp HARM?

Yes. So thanks for the question, Wamsi. A lot of it, as you know well, we have the large customers. And so different customers take different types of technology. So if there's a particular customer in a particular quarter, that takes a bit more CMR products, Obviously, that will mean we ship fewer bits into the marketplace for the number of units that we have, right? So in a different quarter where we have customers that are more ultra-smart heavy and for the same number of units, we're able to deliver much more exabytes into the marketplace.

So there will always be variations quarter-to-quarter, and they don't buy on a very linear basis. It's a bit lumpy. So you will see sort of quarter-to-quarter variations within that. But from a medium- to long-term trajectory, we feel very confident about the ongoing exabyte growth, the ability to hit percentage growth trajectory that we shared, both in terms of the 40 Terabyte ePMR that we're ramping up very aggressively. Then obviously, we have the 44-Terabyte HAMR coming out in the first half of calendar '27, which we're ramping and obviously, we've indicated in our road map, we will have 50-Terabyte products coming out towards the second half of calendar '27.

And then if I could just follow up. When you look at your quarter-on-quarter gross margins from your guidance, not to beat a dead horse here, but year-on-year, you still are suggesting pretty strong incrementals. I get that. But is there anything specific within, again, mix and sort of pricing dynamic that is changing on a sequential basis? Or is the pricing dynamic something that should persist and it's more about just sort of the mix that's creating maybe not further upside to the gross margins? .

So definitely, again, quarter-to-quarter, there's going to be valuations. That's why I'm not really fixated on my sequential incremental gross margins. I look more on my incremental gross margins on a year-over-year basis. And just to put some numbers around that, right, in fiscal '25, on a full year basis, our incremental gross margin was 60% and in fiscal '26 -- the fiscal year that we just ended, incremental gross margins were 75% year-over-year, ending in Q4 with incremental gross margin year-over-year, 84%, 85%. And implied in the guide for Q1 is a year-over-year incremental gross margin of 80%, 81%.

So I think we're executing really well on driving incremental gross margins, further improving the gross margins. I have high conviction that we will continue to improved gross margins for many quarters going forward based on everything what we've explained. There is very strong demand. We're moving to higher capacity drives that provides more value to our customers, and that enables us to increase our price per terabyte while at the same time driving down the cost per terabyte. And I think that's a great recipe for further gross margin improvements.

Our next question comes from Krish Sankar from TD Colin. .

My first one, Irving, clearly, your pricing is improving pretty well. You kind of mentioned customers asking for LTAs into 2029, '30, '31. I'm just wondering, with 90% of your bids going into cloud and most of them unto LTA extending longer, the velocity of price increases that you are seeing going from high single digits to high teens last quarter and probably 20% in September quarter, would that slow down because most are under LTA? Or do you think there's still flexibility where the velocity of price increases could still increase? .

Yes. Thanks for the question. As I pointed out, I think if you talk about the cloud business, as you pointed out, quite a significant amount of it is under LTA. Obviously, not every LTA starts sense at the same time. So we will see price adjustments from LTAs even as we go through the fiscal year depending on when the contracts kick in with the new pricing, and it's obviously dictated by some of the new platforms we introduced into the market. And to reiterate our whole pricing strategy there to make sure it's predictable, and we give a lot of visibility to our customer.

Where we've seen a bit more ability to drive more rapid price increases has been in the non-nearline space. If we fast forward that to what we're looking at into what we're discussing with customers on '29, '30, '31, we've -- we're working through the specific commercial constructs of the LTAs, but we have a good line of sight and visibility to the volume requirements of our customers that sort of further strengthened our conviction on that 25% plus exabyte growth. What we're working through is the pricing regime of how we would effect that for those years going forward. In short, I think there's still a lot of opportunity for us to deliver more value to our customers, deliver better TCO and be able to benefit that through pricing as we've always stated we would do.

Do you have a follow-up, Krish? .

A quick follow-up for Kris. You kind of mentioned how the cost per terabyte downs are like probably like high single digits right now. and it can get to like down 10% carve downs. Is that in a post hammer qualification world? Or do you think that's something you can get to by in the next few quarters?

So I have said in the past and repeated today, right, that the mid- to long-term cost per terabyte decline is on or about 10% year-over-year. If you look at it the last couple of quarters, we've been executing to that. And I have high conviction again over the longer term that we will be able to execute to that. Again, there are always going to be some variations quarter-to-quarter, but mid- to longer term, -- and it's all driven by our technology road map and our product transitions to higher aerial density and higher cost drives that will bring down the cost per terabyte. .

Our next question comes from Asiya Merchant from Citi.

This is Mike Cadiz for Asia merchant at Citi. So my first question is that given your LTA conversations, and of course, the visibility that they provide. Can you perhaps talk about the types of workloads your customers are planning to accommodate, be it agentic or physical, et cetera, and how, if at all, the workloads differ through time. from LTAs encompassing more near term 2027 to LTAs and conversations covering '28, '29 and even '30. Just wanted to see the progression of workloads from your perspective.

Sure. I think the bulk of the workloads will be the same. And there are a couple of growth drivers that we see, we are definitely getting visibility from customers. One, and first of all, something that's underappreciated. -- core cloud services continue to grow. As well, especially video-driven applications. So an example is the collaboration tools that many of us use each and every day. A lot of these video meetings that we're on are getting sort very increasing that's driving a lot more storage demands and on core cloud services. .

And obviously, in AI, the primary drivers over the last 24, 36 months have been model training and development. And what we are seeing now and going forward is growth driven by inference, Agentic AI. And actually, we're starting to see the early innings of growth being driven by physical AI as well. And we are engaging with many large enterprise customers who are in the physical AI space, and we have good visibility in terms of the growth trajectories in those as well.

So definitely growth coming from those vectors that I just shared. And on top of that, obviously, with the introduction of our innovations like the high bandwidth drives that we have now sampling with five customers that opens up the ability to also able to deliver hard drive capabilities with the superior economics that it has in higher bandwidth foot loads as well. So we see that also as a future driver of growth that we are very excited about that customers are engaging very closely with Aon.

Can we go to the next question, please.

The next question comes from Erik Woodring from Morgan Stanley.

Last quarter, you mentioned you could see some interest from sovereign and neocloud data centers for nearline drives in your systems. Can you maybe just tease that out a bit more now that we are three months beyond that comment. Just what are they buying? How price sensitive are they? How big are they? How big can they get, et cetera? And then a quick follow-up.

Sure, Erik. So it's a great question. I think we looked at that as an opportunity, and the opportunity is materializing today. And again, I would highlight 3 areas. We definitely see increasing demand from neo clouds, actually even frontier AI labs are coming to us for supply. And as I just touched on to the earlier question, physical AI company. So we've seen very strong demand from an autonomous vehicle company for more storage as they ramp up the autonomous vehicle capabilities, we are seeing more demand for sovereigns from new clouds from AI frontier labs as well as the particularly see the economic benefits hard drives and the increasing storage demands that their business models are creating as well. .

In terms of pricing, I would say they are in the very tax supply environment that we're in, that provides us an opportunity for increased pricing leverage as well.

And Erik, you said you had a follow-up?

Yes. Just a quick one. Just a clarification, Irving, on your kind of pricing commentary. Is it -- am I correct in that you have sequential pricing escalators in some of your contracted pricing? And just given that demand has strengthened and you are adding more value to your customers now with these higher-capacity drives, is it the right way to think that these pricing escalators are accelerating? Or is that maybe over extrapolating too much? .

Yes, you're probably over extrapolating a bit too much, Erik. I would say that the LTAs obviously have a base price associated to a base volume. And then as we were able to deliver upside to our customers, that's subject to different pricing constructs. So that's generally what we've consistently said as a general construct of how the LTAs are commercially organized.

And the next question comes from Ben Reitzes from Melius.

I wanted to ask you, given the product ramp and improving demand environment, should we expect the first quarter fiscal to represent the low point for both sequential revenue growth and incremental margins with acceleration as we move throughout the year with regard to FY '27.

Yes, Ben, as you know, we only guide one quarter at a time. And again, I think this quarter, we are guiding to some sequential growth as well as year-over-year growth and some good incremental -- strong incremental gross margins on a year-over-year basis. Again, we have -- based on the strong demand environment that we see based on our great execution on technology and product road map based on the fact that we are at the beginning of a ramp of the introduction of our and new generation of ePMR up to 40 terabytes followed then by our HAMR introduction, we have high conviction that we can continue to ship a lot more exabytes at better pricing and driving down the cost over time, which will lead to a continuous strong revenue growth as well as further gross margin improvement. I'll leave it at that.

Do you have a follow-up, Ben.

Yes. I just was wondering if you could tease out physical AI a little more. I mean I think that what might be useful for investors is there's a ton of training video but also synthetic video. And you're mentioning it now more prominently than you did, I believe, in the prior quarters. And I was just wondering, did something change? And how do you see that your tier being impacted by what's probably going to be a surge in both synthetic and video data used to train these things?

Thanks for the question, Ben. Yes, we have highlighted it a bit more this quarter, specifically because we're getting a lot better visibility from this space as well. I highlighted an example with autonomous vehicle player where I would say that the increase in exabyte demand from them for this calendar years -- or sorry, calendar year '27 has increased multiple for, right? So we're definitely seeing that. And if you extrapolate that by the number of players in this space, you extend that to human or you extend that to industrial automation systems as well.

And as I highlighted in the prepared remarks, in many of these cases, they are generating new data from the tools and vehicles that they are collecting data from, but it's still insufficient to train their models as well. So they're also using the existing data feeding it into AI tools to generate synthetic data to further train the models and drive reinforcement learning, and that's requiring a lot more storage data. as well.

So the cycle that we saw in how model training development inference was being created will extend into what is happening into physical , but it's becoming very real now and we see that being a driver of growth of '27 and beyond as well.

And our next question comes from Karl Ackerman from BNP Paribas.

I have 2 questions, if I may. First, how should we think about the gross margin delta today between nearline versus non-year line? I ask because given the greater ability to raise prices in the non nearline HDD market, I would imagine that gap closes. And then what share of these LTAs reprice for nearline over the next 12 months?

So the gross margin between nearline, non-nearline, which includes consumer and client and all of that, it's all in the same ballpark right now in the 5 ZIP code. So there's not a lot of differentiation there anymore.

Irving, obviously, hyperscale demand is robust. How large is on-prem enterprise now, and are you able to sign LTAs with them, too?

Yes. Thanks for the question, Karl. I would say again that the vast majority of the new line bitter shipping are going to hyperscale customers. Increasingly, there's actually increased demand from enterprise OEM players in the storage space, especially as they are pivoting to more hybrid-based storage solutions where most recently, they were looking at all potentially 100% flash array systems. There's a shift back towards hybrid systems that's driving more demand from us. And so we are also ontheforward-looking requirements into the projections that we're working on in terms of supply that's giving us better visibility as well.

And obviously, our engineering teams are also working with them to be able to adopt our Ultra products. and next-generation ePMR HAMR products that will give them high capacity drives and therefore, more exabytes faster as well.

And operator, we'll go to the last caller, please. .

Question is from Ananda Baruah from Loop Capital.

Irving, Kris, just was wondering what's a useful way to think about sort of the China hyperscale and the China AI lab opportunity going forward with all the activity that's been going on there?

Yes. Thanks is the question, Ananda. I think the Asian market represents obviously a very exciting opportunity for us. We've clearly been hearing and seeing a lot about the proliferation of AI frontier lab models there. that are open source. We think that's actually going to be good for the overall industry because that's actually going to drive greater access to different models that will help to proliferate AI a lot more with superior economics.

So depending on the use case, you can use different economic models to be able to drive growth and value in the different applications. But at the heart of it, whether it's the traditional frontier models that you've seen in the U.S. that are premium or some of the new open source more economical models. The underlying requirement that they're going to be able to require of data to support the training of these models. These models are going to generate a lot more data that requires storage for us.

So we view it as very positive because even as compute maybe drives greater efficiency in terms of compute and memory resources. The requirement for storage is just going to compound and grow. So we view it very positively for demand going forward.

A quick follow-up on hand?

Yes, quick one. And just dovetailing from there, the interplay between mixing up meaningfully higher aerial density points, counterbalance against we're at the front end of inference, which is the real kind of killer use case for hard drives. And so how do you see those two things sort of impacting the supply-demand gap in the coming years? .

Well, we think the growth in inferencing will be positive in terms of the demand for higher capacity drives, course influencing is actually going to generate even more data. that's going to be required to be stored both to support reinforcement learning, and the fact that storing the context of data that's generated is going to be more economical than rerunning it through compute and memory resources. So we think actually is inferencing in the genetic growth is going to drive even more demand for storage and the best way to that storage with superior TCO and the faster it is to deliver high capacity drives, which is definitely something we're working towards delivering to our customers. .

This concludes our question-and-answer session. I would like to turn the conference back over to management for any closing remarks.

Thank you. So at WD, as we look ahead to our new fiscal year and beyond, we do so with a very clear sense of purpose. Our customers are counting on us to deliver storage infrastructure at scale and speed that matches the pace and adoption of AI, and we are very much release that challenge. We're confident that our technology leads our execution is disciplined and the entire WD team remains focused on delivering. So I want to really take this opportunity to thank the entire WD team for an outstanding year, to our investors and analysts for joining us today, and thank you for your time and your continued confidence in Western Digital. .

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

All right. Good morning, everyone. Really delighted to have with us Kris Sennesael, CFO of Western Dig. I guess before I get into questions, Kris, you have fair disclosures you've got to read.

Yes. So today, I will be making some forward-looking statements based on my current assumptions and expectations, about our product portfolio, the business plans and performance and the future financial results. These forward-looking statements are subject to risks and uncertainties. So please refer to our Form 10-K and other SEC filings, so -- because risks and uncertainties could cause actual results to differ materially from expectations. We will also be talking about some non-GAAP financials, and the reconciliation is available on our website in the Investor Relations section.

Perfect. Thank you for that. All right. So we have about 20 minutes on this. It's a fairly short fireside chat. I'm going to kick it off with some questions, but anyone in the group here has questions, feel free to raise your hand. Happy to integrate you into that as we go forward.

Kris, thanks a lot for being here. I appreciate your time. There's a lot going on. The [ drive ] names have done extremely well. Western Digital has been extremely well in the last couple of years. Maybe just talk about at a very high level, right, what are the messages from you folks have been the underlying data need for data storage is getting -- growing faster, and it used to be low 20%, maybe it's over 25% CAGR right now. Just talk about what are you seeing from a data storage requirement perspective? And to the extent you've sounded a little bit more positive on the data growth in the last couple of quarters, what is driving that upside to data growth?

Yes. Yes, it's a great time to be at Western Digital and be part of this AI data-driven economy. And so let's try to unpack that. So currently, based on all the discussions that we have and frequent discussions that we have with all our customers, we are getting more and more comfortable that the exabyte growth over the next 3 to 5 years will be greater than 25%. So very strong growth in exabyte. And so what is driving that? Well, first of all, there is still -- sometimes we forget about that, there is still the cloud, right? And there is 8 billion people on the planet that continue to take pictures and video and upload that to the cloud through their phones and other devices and then, in many cases, multiply that through their social media, right? In addition to that, I think every company on the planet by now has understood what the value of data is and is storing all the data that's being generated inside the company.

We at WD do that. And so there is definitely still a lot of growth just by data being pushed into the cloud. In addition to that, yes, there is a lot of growth as a result of AI. And AI, there are several factors in it. First of all, and that's what it started was big data lakes that were created to support the training of the multi-model large language models, right? It's a multi-model because it was not just on the tax base, but also pictures and more and more video based. And that is still ongoing, but there's still a lot of relearning, retraining of the models, and they are training the next generation models. Some of them are very broad. Some of them are more specific. And that all still requires a lot of data, the vast majority of that being stored on hard disk drives.

In addition to that, we've now moved into inferencing, right? And inferencing can have multiple forms from a simple chat box inferencing all the way to agentic AI, right, where one agent or multiple agents go through some very complex inferencing and all of that. And so what we have seen is that the output of all that inferencing is being stored, right? And it's being permanently stored. They want to remain the history of every question you've ever asked to a chat box, but they also want to use the output and some of the logic, how they came to the output to retrain the language models as well.

And then last but not least, and that's just beginning, you have physical AI, right? Physical AI from autonomous cars, robotics in the future, humanoids, right? A lot of those devices, they have multiple cameras. They shoot video footage 24 hours and that video footage is then being used to retrain and learn the algorithms. There is actually not enough video footage available today. So that's now all the hyperscalers and people involved in physical AI. They use AI to generate synthetic data to complement the real data to further train their models. And so when you combine all of that, we have, again, high conviction that exabyte growth is greater than 25% for the next 3 to 5 years.

Perfect. Data will grow north of 25% to your point for the next several years. Talk about where do we end up storing all this data? Because I think one of the dynamics has been, you have this number better than I do, but 80% of the data gets stored on hard disk drives, I think, traditionally, how much capacity does the industry have? How much capacity is Western Dig have? And how much of this data growth can you really support from a supply basis?

Yes. Yes, if you look at hyperscalers, roughly 80% of all the data is being stored on HDD and roughly 20% or so is stored on SSD. By the way, both segments are growing very strong, right? It's not that one segment is growing faster or slower than the other. Both segments are growing very fast. And to the other part of your question, right, how do we support greater than 25% exabyte growth over the next 3 to 5 years? What is very simple. We don't have to add unit capacity to support that, but we can support this very strong growth of greater than 25% through technology and product transitions, right, as we move to higher capacity drives, right? And we have a great track record there.

Just to put it in perspective, right? The average terabyte per drive that we shipped last quarter was on or about 23 terabytes, right? Despite the fact that we have a 32-terabyte available, and we're actually shipping it in really high volume. And in addition to that, we are working on qualification of our next-generation ePMR solution that will go up to 40 terabytes. And in parallel, we're working on the qualification of our first-generation HAMR drives that will go up to 44-terabyte. In addition to that, we have a road map that goes out multiple years, where we have a path to 50 or 60 or 70 or 100 or plus 100-terabyte drives. And when you take that into account, there is sufficient exabytes there to support the strong demand growth that we see from our customers.

Got it. One of the things that's really changed. I think this industry is historically price per terabyte revenue sold to your customers would be down 10%, maybe more in a bad down cycle. That's fundamentally shifted. And a little bit of this is what you said, supply is good -- demand is good, you don't want to add that much supply. How do we think about price per terabyte as we go forward? And maybe if you can contrast that a bit with what we're seeing on the NAND side, for example, where the price increases are very dramatic.

Yes. Well, first of all, our industry and the hard disk drive industry has changed drastically compared to 3 or 5 or 10 years ago, where a lot of the business was consumer or client, some of that was cloud, but it was done through OEMs. Fast forward to today, 90% of my business is cloud, still 5% consumer and 5% client. And we -- within the cloud, we're now dealing directly with the large hyperscalers, all of them, right, on -- in the U.S. and outside of the U.S. And so that has changed drastically. And our relationship with our customers is really -- and our pricing strategy is really value based as we provide more value to our customers, we want to get paid for the value that we deliver to them.

And moving to higher capacity drives is adding more value to our customers because that is good for them in terms of rack density, lower real estate cost, better power efficiency and less power consumption. And so as we move to higher capacity drives, we provide more value, and you will see the price per terabyte goes up last quarter. Our average ASP per terabyte was up 9% on a year-over-year basis. We do it slightly different than on the memory side. The memory side is more of a supply-demand driven pricing. It goes up in some cases, very drastically, but in some cases, it comes down as well as the demand supply dynamics change over time.

Got it. Maybe just on the pricing side, if I stick for it right for a minute. One of the enablers has been these long-term agreements, LTAs and build-to-order processes that drive -- Western Dig and drive industry has implemented. Just talk about a little bit of what does the LTA provide to you folks? And as you look at that over the next several years, what sort of pricing visibility do you end up having?

Yes. For me, the LTAs -- first of all, we're not asking for it. It's our customers who are asking for it because they see the very strong demand and need for storage multiple years out, and they want to secure the supply from the suppliers and from Western Digital. And so some of our customers, they want to have LTAs all the way till 2032, 5 years LTAs. For us, the LTAs are important to create more visibility in both ways. We want to better understand what the long-term demand trends are from our customers multiple years out. We also want to provide visibility to our customers on how much supply we have, and how we grow the supply over time, again, as we move to those higher capacity drives. And so in addition to that, I mean keep in mind that it takes on or about 52 weeks to produce a hard disk drive, 9 months to produce the wafers that goes into the heads and then 3 months to produce the hot disk drive itself.

And so as a result of that, we've educated our customers and most of our customers, they place purchase orders now 52 weeks in advance. And so we have pretty good visibility, at least one year out. But even beyond that, right, in part because of those LTA discussions. Those LTA discussions, they have a volume component and a price component in there with some flexibility, right? And so it's not all really set hard, but there is some flexibility around that. But again, most important thing is better visibility both ways.

Got it. Perfect. You touched a little bit on like the ePMR road map and then you obviously have a HAMR road map as well. Maybe just -- and you touched on areal density improvement you folks can go after. Talk a little bit on cost per bit. What does that mean for the company as you go forward? And what's the right way to think about cost per bit decline maybe on ePMR and then also as you go to HAMR eventually?

Yes. So again, if you look at moving to higher capacity drives. Forget what the recording technology is that's not the most important thing. The most important thing is moving to higher capacity drives that does a lot of good for everybody. I call it win-win situation, right, moving to higher capacity drives. It provides more exabytes, right, for storage at a higher value for our customers. Of course, you get better rack density and lower real estate costs and better power efficiency. Also for us, it has a benefit because we can -- as we provide more value, get a better price per terabyte. But at the same time, moving to higher capacity drives also results in a cost per terabyte that comes down because it doesn't cost that much more to produce a 40-terabyte drive versus a 32-terabyte drive, right?

And so it's a win-win situation all over the place. Our cost, we think, over the longer term, costs will come down on or about 10% year-over-year on a cost per terabyte basis, right? It's not necessarily going to be like that each and every quarter. But in the mid- to longer term, right, we see a cost down of on or about 10%.

Got it. If I take a couple of these things together, right, you said win-win, I imagine this. There's a win on your revenue side because pricing goes up, there's a win on your COGS side because cost per bid comes down as well. What does that mean for incremental margins and gross margins for the company as we think about this on a multiyear basis?

Yes. So at the Innovation Day in February 2026, I indicated that gross margins in the next 3 to 5 years will be greater than 50%. We actually -- in the March quarter, we entered the new ZIP code, and we got into the 50% already. So we are now operating in this greater than 50% environment. Incremental gross margins, depending on how you look at it quarter-over-quarter or year-over-year, I tend to focus more on year-over-year because it's more meaningful. It's in the 70%, 75% range. So very strong, and it's driven by price per terabyte that goes up and cost per terabyte that comes down. And so I think -- yes, we're in a, I think, in a very strong position to continue to see further gross margin improvements for many, many quarters to go.

Got it. And clearly, you have good top line growth. You have 70%, 75% incremental gross margins. OpEx, I assume it's going to be fairly limited as well. What does that mean from an EPS and free cash flow basis eventually for investors?

Yes. I mean, if you look at the story, right, we have very, very strong top line growth. The gross margins and operating margins continue to increase. We have a really strong capital return policy as well, leveraging our very strong free cash flow. I mean, the free cash flow margin is approaching 30%, and so that generates just last quarter, close to $1 billion in free cash flow. What do we do with the free cash flow? Well, we return it all back to the shareholder, right, through a combination of our dividend program and the share buyback program. On the dividend program -- we fully committed to that dividend program.

We've already increased it twice. And there is a lot more opportunity there in the future. But the vast majority of the free cash flow is being returned through the share buyback program in -- for me, there is no hesitation, right? I do intrinsic value calculation. I look at my forecast, my strong revenue growth, my gross operating margin improvements, my free cash flow. There's no hesitation. We're buying back almost every day.

All right. Perfect. You folks bought back some of the convert today, money, the 8-K leasing, [indiscernible] a bit of redemption. Maybe just touch on that and kind of how does that play into your capital allocation process?

Yes. So in addition to the free cash flow, we also have a very strong balance sheet. At the end of last quarter, the only thing what was left was $1.6 billion of debt. We wiped out all the other debt. And we had $2 billion of cash. So we were in a positive net cash position at the end of March. At the end of March, we also still had 1.7 million of SanDisk shares. Subsequent to March, we already monetized on or about 600,000 SanDisk shares in an equity-for-equity exchange. And we have the intention to further monetize the remaining slightly more than 1 million SanDisk shares that we have. But we also have the $1.6 billion convert out there, that has a maturity in November '28, but we can call it in November '26, but the convert is so far in the money, the converters which can actually put it to us as well. And so we are having discussions with some of the note holders. And as we've announced this morning, we did a private exchange with a good chunk of that $1.6 billion of the note holders and will pay the principal in cash and pay the premium in shares.

Perfect. One of the questions I get a fair amount when we talk about Western Dig investors has been, what makes them add more capacity, what makes them add more units to the ecosystem over time? Just touching your CapEx philosophy and maybe ask the question like what would it take for Western Dig to actually add more units to the ecosystem?

Yes. So again, we -- let's start with the demand, right? So we see strong demand greater than 25%. And so I'll leave it up to you to pick what that means, what -- but it's greater than 25%. And so -- but we believe that as the demand is even greater than 25%, we can fulfill that demand without having to spend CapEx for adding unit capacity, right? We will have to spend CapEx on or about in the long term, 4% to 6%, right? In some years, it might be less than that, some years it might be more than that. But in the long term, 4% to 6% of CapEx to revenue, we will have to spend that in head and media, right? Because we need better heads, and in some cases, we need more heads, and we need better media, and in some cases, we need more media, right? And we are -- there's no hesitation there. We are making the necessary investments to make sure we can, again, support the growth of more than 25% exabytes.

Perfect. One of the challenges, I think, for folks always have the drive space has been like you're dealing with 5 or 6, maybe a bigger number, but a constant number of very large buyers, and how much leverage do you really have in these negotiations with them over time versus not? I think, Kris, at your prior job you dealt with Apple as your big buyer, if I'm not mistaken, which probably is the toughest one on supplier sometimes. Just talk about how does that engagement work? And is it just they come into your pricing, and you have to agree to it? Or is it more of a collaborative thing at Western Dig?

It is -- I mean, we have changed the engagement model as well, right? We have a very strong collaborative engagement model with our large hyperscalers. They share their multiyear data center road maps and technology and product programs, we do the same from our side. And I think we have earned our seat at the table. They fully realize to build out this AI data center infrastructure. They need a lot of elements. They need a lot of GPUs and TPUs and CPUs. They need a lot of HBM and memory, but they also fully realize they need a lot of HDDs, and that is a very critical component in the overall buildout of the AI data center. And that's why we have this very strong mutual respectful relationships. We're working hard to provide as much value to our customers. We obviously want to get paid for that. And I think that's a win-win.

The thing is going right on me, but I'll ask you a quick question, if you don't mind. How do you know all the stuff you're shipping to these hyperscalers actually getting deployed in the data center versus sitting in shelf somewhere?

Oh, yes. No, no, no. We -- I mean, we have, again, a strong relationship with our customers. Customers are pounding the table. They want hard disk drives as fast as they can. And through the collaboration we have with them, we have great visibility that all the hard disk drives we are shipped are being deployed as fast as they practically can do it.

Perfect. We're up on our time. So maybe I'll stop there. Kris, turning back to you. Any closing comments, anything we did not touch on that you want to flag to investors tell us how great greater than 20 really means?

No, again, I leave that up to the investors to figure that out. But it's still improving. It's still improving. Again, like I've been with the company for 12 months now. I talk to the customers as well. Many of our employees talk to the customers every time they come back. It's with a forecast that continues to go up. And so again, we are -- I think we are well positioned. We're a technology leader. We have a great technology that our customers love and appreciate, and we're well positioned to be one of the winners in this AI data-driven economy.

Perfect. We'll stop with that. Thank you very much for your time, Kris. Thank you.

Thank you.

Good morning, everyone. Welcome to the first day of the BofA Global Tech Conference. Delighted you could all make it. See a lot of familiar faces here. Welcome, again, I appreciate you all joining us today. I'm Wamsi Mohan, I cover IT hardware and supply chain for the bank.

Today, I'm delighted to welcome WD CFO, Kris Sennesael. We also have the WD IR team. We have Ambrish and Amitesh sitting back here as well if you have follow-up questions after the session is done.

Kris, glad you could join us. I know you have a little bit of a quick spiel on your disclaimer to go through first, and then we'll kick it off.

Yes. And good morning, Wamsi, and thanks for hosting us here at your conference. And before we start, I just want to remind everybody that today, I will be making some forward-looking statements based on my current assumptions and expectations, including related to our product portfolio, our business plans, performance and future financial results. These forward-looking statements are subject to risks and uncertainties. So please go look to our SEC filings, our Form 10-K and that provides more information on the risks and uncertainties that could cause actual results to differ materially from expectations.

We will also be making some reference to non-GAAP financial measures. So please go to our website where you can find a reconciliation between GAAP and non-GAAP. So with that. Let's dive into it.

All right. Great. That was read quickly, so I appreciate that. Well, Kris, it's been a phenomenal year for you guys. And I would say that the market is starting to look at HDDs somewhat differently. You're seeing that in your results. You're seeing that in the numbers. The question that we generally get is, sustainability of this. And so maybe to kick it off, right, I think for baseline HDD exabyte growth of 25%, it can be a little bit higher, lower, somewhere in that range. But when you think about that growth rate at these kind of levels where now we're starting to ship materially more exabytes. I mean we're talking about 1.5 zettabytes going to closer to 2 zettabytes, which is just an incredible amount of storage. What are some of the underpinnings that you're looking at that give you confidence around the fundamental premise of this growth in data?

Yes. No, it's a great time to be at WD and be part of the storage environment, that's part of the bigger AI data center build-out, right? And let's just go back a little bit in February of 2025, just on or about the time when the company separated their Flash business and became a strategically focused hard disk drive company. The company did an Analyst Day. And at that time, we were expecting the exabyte growth to be mid-teens, and that was mostly driven by the cloud, right? The cloud, which is 8 billion people taking pictures, taking video, uploading that into the cloud, multiplying that through social media as well as every company on the planet that store all their data more and more into the cloud. We also indicated at that time that there was a possibility for stronger growth if and when AI kicks in, and we didn't have really good visibility into that.

But a lot of things has happened since then. First of all, we've created a lot better visibility by deeper customer engagements, right, having deep ongoing conversations with all our customers, mostly the hyperscalers that provided a lot better visibility to us also longer term, multi, multiyears out. And as a result of that, in February of '26 at our Innovation Day, we've indicated that we think exabyte growth is going to be in the mid-20s, call it, on or about 25% CAGR for the next 3 to 5 years. Although since then, we had further discussions with our customers and at our last earnings call, we've now indicated that we believe exabyte growth could be well above 25% CAGR for the next 3 to 5 years.

And so where is that coming? Well, first of all, the cloud continues to be a strong growth driver, just the traditional cloud. But in addition, yes, we have seen AI kicking in, right? And AI kicking in, in multiple ways. First of all, it started by training and vast amounts of data lakes were created to support the training of the multimodel large language models, right? I say multimodel, because it was not just only text-based, but also pictures, still pictures and video. And so massive amounts of data was being stored to create those large language models.

By the way, that is not over, right? The hyperscalers continue to train, retrain, relearn, create new data sets to create the next generation of large language models. But we've also started moving into inferencing. And for us, at the beginning, it wasn't clear what is inferencing going to do for us. But now as we see more and more inferencing, probably 2/3 of the compute power that's being installed is being used today for inferencing. We see a huge demand for data storage because all the data that's being generated while doing inferencing and the output of it with much of the logic, how they got to the output is all being stored. It's being stored.

And that's the case for simple chatbot activity, but also more and more agentic AI, where you have a lot more complex inferencing interactions, a lot more rich output and all of that is getting stored. That is not being stored is being fed back into the models for retraining and relearning as well.

And then, last but not least, and that's just the beginning, it's physical AI, right? Physical AI, think about autonomous cars, robotics, all the way to humanoids, right? What does those devices have in common? They have multiple cameras on it, right? And they constantly shoot video. All that video footage is being stored for training and upgrading and learning of the algorithms. There is actually not enough data that's being -- or video footage that's being captured so that now those companies use AI to create synthetic data to feed into their learning algorithms. And so the amount of data just keeps growing exponentially. And again, as we talk to our customers, we get more and more conviction about the plus 25% exabyte growth for the next 3 to 5 years.

Yes. Yes. No, a lot of incredible drivers of data growth over here. So maybe just to -- you mentioned sort of the visibility that you're getting from your customers. Can you talk a little bit about how is this different from what you've seen in the past in terms of visibility? How do we ensure that we're not getting ahead of ourselves in some way that, indeed we are shipping to demand and not sort of shipping in excess of demand anywhere because that's generally how these cycles tend to get broken in some ways and not that there are any signs of that at the moment, but can you just talk about visibility, the type of contract structures, the things that you're looking at to make sure that we're shipping sort of maybe even below demand levels probably, but like if you can maybe elaborate on that.

Yes. So first of all, I mean, today, our hard disk drive business is totally different than compared to what it was 3 or 5 or 10 years ago and the multiple cycles that we have gone through. I mean it's not that long ago, 3, 5 years ago, more than 50% of our business was still consumer and client PC, right? And fast forward to today, 90% of our business is with the cloud, with the hyperscalers, right? And so that's a different business. That's a business where it's almost like a planned economy, right? Those guys, they think 3, 5, 10 years out and how much data centers, gigawatt compute power, but also exabytes or zettabytes of storage they have to build over the next 3, 5, 10 years.

So it's really a lot longer visibility, a lot more planned in advance, which takes out somewhat of the cyclicality, right? And so that is a big thing, again, like 90% of our revenue tied to hyperscalers. With those hyperscalers, they are -- again, they think ahead a long time. And it's actually the customers that came to us who wanted to secure the supply. We were not the suppliers that were pushing long-term agreements on to our customers. It was a collaboration, customers, again, who see that the demand is growing extremely strong. The supply is there, but it's tight. It's a very tight environment. And so customers wanted to secure supply, but in a collaborative way, right? So for them, it's important that they provide the visibility about the data centers they build, about their technology and product needs that they have. And we did the same thing.

We shared our capacity expansion plans, which we can talk more about that later. It's based on technology and product road map, right? And so it's a very collaborative environment where you create a lot more visibility multiple years out. I'm not nervous about overbuilding or inventory builds. I mean everything what we ship today is getting deployed right away, right? And I think there's multiple industry reports out there that indicate that demand is at least today is bigger than supply. And even in the next couple of years, many industry analysts expect that demand supply will continue to be very tight, again, part because the demand growth is so strong.

So maybe to touch on that a little bit, right? You mentioned sort of this multiyear visibility. These are customers who are planning out well in advance. Can you talk a little bit more about the specifics around that? Like, for instance, if you have a hyperscale customer, do they talk about in the next 1 year, we want 100 exabytes. In the next following year, we want 100 exabytes. As you think about the planning for that, what is the variation or variability around these demand levels that you and your customers kind of talk about on average? And then what about pricing? So those are obviously the 2 big variables that as you're managing the business, you're looking at how many exabytes and what you can charge for that. These long-term agreements that you have, how does that play out across your entire set of customers?

Yes. It's -- I mean, today, there is -- in terms of variability, every time we talk to them, they seem to be asking for more. So that's kind of the variability we see today. But let's -- I mean, we -- again, we changed our business model drastically. Again, if you look at it 3, 5 years ago, I mean we were getting orders for shipment within current quarter, right? That is no longer the case.

We've educated our customer base that it takes 52 weeks to end-to-end to manufacture a hard disk drive. It takes 9 months to produce the wafers that go into the heads and then 3 months to put everything into the box and create a hard disk drive. So it's end-to-end, it's 12 months. And so our customers realize that. And now most of our larger hyperscalers and larger customers, they place their orders 52 weeks in advance, right? And so we have pretty good visibility for the next 12 months or so.

Now, again, some of those customers, they wanted to secure supply longer out, '27, '28, '29. Some of them would love to sign LTAs all the way until 2032. Now for us, it's not important, necessarily the structure of the agreement. It's all about the visibility, right? We, and our customers, we want visibility about the exabytes multiple years out. We also want to have some visibility about the pricing and the pricing environment. Now, again, there is -- it's slightly different customer to customer, and there is some variability there. Again, the most important thing for me for the LTAs is the mutual increased visibility, predictability of the business.

Yes. So as you think about -- I mean, we get this question often around pricing, right? So if really demand is so much in excess of supply, then why is it that pricing is -- and pricing has been very strong. I mean relative to history, we've seen dollars per terabyte decline 15%, 10%. Now for the rest of the year, it's probably going to be high single to like maybe potentially higher than that. So we're entering a new phase of pricing.

But when you look at other areas within the industry that have kind of become some type of bottleneck for whether it's NAND or DRAM, pricing is going up 100% quarter-on-quarter. Optical, like pricing is going up a lot. So it feels like pricing potentially has the potential to go up a lot. So how do you manage that balance of visibility versus pricing and being able to capture some of the upside, which you are, but is there room to do more?

Yes. And so we definitely, as WD, like a strategically-focused hard disk drive company, we play it differently than what you see in the memory where you have multiple participants that sell into multiple end markets and where you actually do have some spot pricing, right? And demand and supply is really dictating what the price are and prices go up and prices go down all the time.

That's not how we play it as WD and the hard disk drive business. We really want to create long-term value for our customers, for ourselves and our shareholders, right? And that means being very thoughtful in the short, medium and long term. And our pricing, I would really say, is value-based, right? The more value we provide to our customers, the more we can charge to our customers because the more value it creates for our customers. And we do that, again, based on a technology and product transition, right.

As we move to higher capacity drives and/or higher performance drives, that creates more value for our customers. And customers are willing to pay for more value, right? We actually can charge a little bit more on a price per terabyte and still increase better total cost of ownership for our customers, right? And so that's what we have been doing.

If you look at in February of 2026, I indicated that I expect price on a price per terabyte to be mid- to high single digits increase year-over-year for all 4 quarters. We announced our March results and ASP per terabyte was up 9% on a year-over-year basis on a price per terabyte basis. So mid to high, we're definitely at the high end. And so that really indicates that, yes, the environment is very strong. And we see it continue to be strong for multiple years, again, based on the fact that we have a technology and product road map that will continue to deliver more value over time as we move to higher capacity and higher performance drives.

Okay. One more question on pricing, which is when you think about the 90% of the bits that are going into the data center, there's probably 70% of that, that's hyperscaler, and then 30% of that, that's enterprise-centric. These long-term agreements that you're striking are generally with the hyperscaler capacities. The enterprise, you have more discretion in pricing where the demand could be up or down a little bit more cyclical relative to the hyperscalers, but then pricing could also have more upside. Is that the right way to think of it?

Yes. Yes, absolutely. I mean we have been raising prices as well in our consumer and client business, and we have been raising prices across the board. Again, as we provide more value to all those customers. And -- but in certain areas, if you think about consumer and client, the ASP increases actually were above the company average. So yes, there's definitely opportunities there. And we're constantly testing the market, right, and see what value can we extract for the value that we provide.

Yes. I mean you're in a completely different regime when you look at gross margins, right? Like in the sense of putting a historical lens on it seems absurd at this point because historically, the targets were in the low 30s and you're at the 50% gross margin levels. Incremental margins are even more impressive. Incremental gross margins can be 70% to 100% range. So what is it about your road map that gives you confidence and comfort around very strong continued both gross margins and incremental gross margins?

Yes. Yes. No, gross margins, I mean, last quarter, we entered the new zip code, right, starting with the 5.

Yes. Congratulations.

Which is great. But again, it is -- to me, it's all based -- it starts with our technology and product road map, right? As we move to higher capacity drives and work on performance improvements, but especially move to higher capacity drives, that provides more value for our customers. And as we provide more value for our customers that translate in a better price per terabyte, right? And that, again, despite a higher price per terabyte, we still, for our customers, creates more value.

At the same time, moving to higher capacity drives lowers the cost per terabyte, because it doesn't cost that much more to produce a 40-terabyte drive versus a 32-terabyte drive, right. It's all by increasing the areal density per platter that gets you to the higher capacity drives. And so that results in a lower cost per terabyte. And we have a road map in front of us for many years.

I mean, as you know, today, the highest capacity we ship is 32 terabytes, but we are qualifying our next-generation ePMR at 40 terabyte, as well, in parallel, we're qualifying our first-generation HAMR at 44 terabyte. But we have a road map to drive that to 50, to 60, to 70, 200, 200-plus terabyte in not too far distance, right. And just a combination of that, again, will provide more value to our customers, allow us to increase price per terabyte while at the same time, reduce cost per terabyte.

And that results in those, yes, a very nice incremental gross margins. The last couple of quarters, to your point, incremental gross margins was in the 70%, 75% range, right? That's how we have moved the overall gross margins now in the low 50s. But I think for many, many years, there is further improvement possible there.

Yes. That's definitely exciting. Maybe just to talk a little bit about the technology end of things, right? You mentioned areal density increases effectively are what are driving capacity increases. So as you think about your road map here, you kind of have a dual road map with having both continued PMR, but also HAMR, which you just spoke about. How do you think about as you go over the next 2, 3 years, the mix shifting between these? Would you be standardizing on one over time? What does that time frame look like?

Yes. And so I mean, customers don't really -- let's start with that, right? Customers don't really care what the recording technology is. They don't care if it's ePMR or HAMR or CMR or UltraSMR, the 2 flavors of ePMR and the 2 flavors of HAMR. They want reliable, scalable, high-performance exabytes of storage, right? And so that's the most important thing for our customers. But we, again, we have an industry-leading technology road map, right? And we're -- again, today, we're shipping 32 terabytes. We are in qualification with ePMR with 3 customers. That will ramp in the second half of calendar year '26, in a pretty steep ramp because, again, ePMR is a known technology. Our customers have 10 years of experience with that technology, comes in a CMR and an UltraSMR version. And so we can -- we and our customers can actually scale that very quickly. And again, that comes all the way up to 40 terabytes.

In parallel, we have been working for many years on our HAMR technology. Great progress being made. We're in qualification with 4 customers right now, and we intend to ramp in the first half of calendar year '27, initially coming at up to 44 terabyte HAMR drives. We still believe that we can continue to drive further capacity improvements on ePMR 50, maybe 60, but probably that's where ePMR recording technology will -- it's going to be hard to do more than that. I mean, you never know engineers. We have a lot of smart engineers. They always try to find more, but that's kind of like the current understanding.

And that's why we need HAMR, right? HAMR for us is the path not only to get to 50, 60, but to get to 70, 100, 100 plus. And it's all, again, driven by areal density. That's the most important part, right? We want the industry-leading areal density. Today, we're at 4 terabyte per platter, but we have a path to get that to 5, 6, 7, 10, 10 plus.

In addition to that, and that's a little bit unique to WD, we always look at the opportunity as well to add more platters in per hard disk drive, right? Because there is economics. If you can get from 10 to 11 platters, which today we are at 11 platters, that gives you an additional cost advantage, right? If you can get it to 12 or 14 or more than 14, there is an additional benefit. But for sure, the most important one is driving that areal density from the 4 terabytes we have now to 10 and more than 10 terabytes in the future.

So I know a lot of investors in the room are probably thinking about sustainability from a different standpoint too, right? The CapEx levels that we're seeing are from hyperscalers are advanced. I mean, notwithstanding the Google announcement from yesterday, right? Like so this commitment to go build more seems to be extremely strong at the moment.

As we think about the next few years, if we enter into a domain where CapEx, let's say, from hyperscalers starts to plateau, how do you think about the business and managing the business in that scenario? And how coupled should we be thinking is your growth to cloud CapEx directly, because data is different than compute and CapEx intensity could be different in different areas. So how do you think about that? And if it plateaus, then how do you manage that?

Yes. To me, I mean, there is a certain link, right? The more data centers are being built and the more compute power is being built that results in more data being generated and the need for more data storage. And so there is somewhat of a link there. But when you look at -- I mean, the spend on hard disk drives by the hyperscalers is probably on about, what, $30 billion or so out of their $700 billion of CapEx. So it's relatively small. Having said that, right, sometimes I worry as well. I mean, are they going to continue to spend $600, $700, maybe soon, $1 trillion on mostly compute power and so on.

And what will happen if that starts to slow down? Well, again, there is a difference between compute and data storage, right? Compute power is being reused all the time. I mean it takes whatever, 3 seconds to generate a token, you generate a token, the compute power is available again for another 6 seconds or 10 seconds of compute to generate something. That's not the case with storage, right? Storage keeps compounding. Every time you generate a token and you generate an output, that gets stored. And the next 6 seconds that compute power is being used to generate something, that gets stored. And so every time you use compute power, you get more and more storage out of it.

And so in my mind, even if the CapEx on compute and memory, which is really closely tied to the compute cycle there, slows down over time, who knows, that does not indicate to me that the spend on storage will slow down, right? It's decoupled.

And so as you think about managing that, right, there is an underlying assumption of this exabyte growth that sort of is also predicated somewhat loosely maybe on this growth in CapEx. But if, let's say, that we were to see a change in the trajectory in the exabyte growth, what are the levers that you can use? And the visibility, I think, that you're trying to establish in this pricing sort of more deterministic pricing in some ways all sort of feeds into that, right? So as you think about a potential maybe deceleration in growth, the ways to manage that, like what are some of the levers and you're sitting in your seat as CFO you would take?

Yes. So we're definitely a lot more disciplined than compared to 3, 5 or 10 years ago on multiple factors, right? First of all, discipline from a pricing point of view. Again, we want to get paid based on value, right? It doesn't matter what the cost is of the product. It's based on value. And wherever you are in the cycle, you have to remain very disciplined on pricing.

Second, of course, is on capacity, right? In the past and other industries have made the mistake, right? The demand is strong. You go spend a couple of billion dollars and add more capacity. We are not doing that. We, as WD are not doing that. We are not spending CapEx dollars to support unit capacity expansion, right.

But we do believe that we can support the strong demand growth, which I talked about it earlier of plus 25% exabyte growth, right, through our technology and product road map by moving to higher capacity drives. That will require some CapEx because we need to invest in our head and media operation, right? Because we might -- also if we not only work on areal density, but on more platters per box, you need more heads and you need better heads and you need different media and better media and maybe more media as well.

But we are not spending CapEx to add unit capacity, right? And so that, I think, sets us up really well. I mean, again, I don't see it slowing down in the next 3 to 5 years based on the inputs I get from my customers. But mean every business on the planet is somewhat cyclical, right? Although it's a secular growth business, but it will be somewhat cyclical. If things slow down or speed up, we will adjust that by working on our technology and product road map, and we are not adding unit capacity. And that, I think, is a lot different compared to what happened 3, 5 or 10 years ago.

Yes. And I think of maybe an analogy like I often get asked this question about like, well, when the transition happened from 2D to 3D NAND, it brought a lot more bits on there and that kind of killed it, is HAMR going to bring a ton more capacity online. I guess the difference is you don't need to run at the same utilization rate of your fabs like the NAND does. So you do have more flexibility in managing your output. So maybe to close, and I know there's lots of talk, but it's already -- we're out of time.

Maybe to close, Kris, I would love to get your thoughts from your seat on you're generating tremendous amounts of cash. You've kind of gone through this liquidation of or disposition of your stake in SanDisk. It's been amazingly great for you guys. So as you think about putting all of those things together, how should investors think about capital return allocation of capital? What do you think we should look forward to when you're throwing off so much free cash flow?

Yes. First of all, I mean, the free cash flow is extremely strong. We are approaching 30% free cash flow margin, which is on or about $1 billion of cash per quarter, right? So very strong free cash flow. Second, the balance sheet is clean and strong, right? At the end of last quarter, we were down to $1.6 billion of debt. And we have $2 billion of cash. So we are actually in a positive net cash position. And at the end of last quarter, I still had 1.7 million of SanDisk shares. And so -- which we still are working on the monetization as well.

So strong cash flow, strong and healthy balance sheet. What do we do with all the free cash flow we generate? We return it back to the shareholder, right, through a combination of our dividend program and share buyback program. The dividend, we've already increased it twice, and we will, in the future, continue to increase it. But the vast majority of the free cash flow is being returned through the share buyback program. And there, there is no hesitation.

I know we have seen a nice run in the stock. But I do intrinsic value. I look at the intrinsic value of the company. I look 3, 5 years out based on my projected revenue growth, my gross and operating margin expansions, my strong free cash flow that will continue to expand. And so there's no hesitation to return all the cash back through share buyback.

Excellent. Unfortunately, we have to end it over there. Kris, thank you so much. Really appreciate you being here.

Thanks. Thanks for having us. Thank you.

Thank you.

Good afternoon, everyone. Welcome to the next fireside chat here. I'm Samik Chatterjee, and I cover hardware and networking names at JPMorgan. With me, I have the pleasure of hosting Western Digital and from Western Digital have Kris Sennesael, who is the Chief Financial Officer. Kris, I hope I pronounced that right, but please if I didn't. Let's talk about the industry first before we get into any Western Digital questions. But do you want to read out the safe harbor first?

Yes. And Samik, thanks for hosting us here. Great conference so far. But before we go, I have to make some disclosures that today, I will be making some forward-looking statements based on management's current assumptions and expectations, including to our product portfolio, business plans and performance and market trends.

These forward-looking statements are subject to risks and uncertainties. So please go read the 10-K and other SEC filings that provides more information on risks and uncertainties that could cause actual results to differ materially from expectations.

We might also be talking about some non-GAAP financials and the reconciliation you can find on our Investor Relations website.

Okay. I saw that Ambrish just walked in and saw you reading the safe harbor, so he's probably happy as well. Let's start with the demand drivers for the industry. I think genuinely, everyone has been surprised of the demand drivers and how much it has exceeded expectations consistently. Just talk about what are the biggest drivers you're seeing right now in terms of driving the exabyte growth and why we could potentially have this exceed expectations more consistently going forward as well?

Yes. It's a great time to be in the hard disk drive business. And when Irving, our CEO or myself or anybody of the company goes and talk to the customers, they come back with good news and better news. It looks like every time we talk to them, the demand is stronger for longer, which is good.

And so just to put it back a little bit on -- in February of 2025, we indicated that we expect the market to grow mid-teens on an exabyte compound annual growth rate over the next 3 to 5 years based on what we see in the cloud. And we indicated that if and when AI kicks in, the growth could accelerate into the low 20s.

And so fast forward to today and at our last earnings call, we've actually indicated that we now expect exabyte growth to be higher than 25% compound annual growth rate over the next 3 to 5 years. And so what's driving that? First of all, the base is still there. So the cloud, right, 8 billion people loading pictures and video up into the cloud, sharing it to social media and so on, that continues to be very strong.

In addition to that, every enterprise on the planet, including WD, we understand the value of data, and we're not throwing away any more data. We store all the data we can capture in our manufacturing side, in our engineering operations. throughout the company. And I think every other company on the planet is basically doing that. In addition to that, we definitely now get a better understanding of AI. And AI has different facets as well.

First one, of course, is training, training of the multimodel models that is out there that is not only text-based but also pictures and more and more video-based. The training is still ongoing because all the big players continue to develop their next-generation model and store a vast amount of data.

They also work on retraining, relearning of the models and leverage all the data. But now in addition to that, we've definitely moved in inferencing. And we estimate that roughly 2/3 of the compute power in 2026 will actually be used for inferencing. And so what -- how do we benefit from inferencing? Well, what we noticed is that all the chat box all the way down to Agentic AI that is using inferencing, the output of that is being stored and the vast majority of that output is being stored on hard disk drives. They want to know what's being generated. They want to know the history, and they want to know as many characteristics around that.

And so that's driving a lot of more data storage. In addition to that, and that's the last one, and that's still early days, you have the physical AI, right, which is autonomous cars, robotics and in the future, humanoids. What do those devices have in common? They have multiple cameras attached to it, and they want to try to capture video as much as they can and store that and feed that into their training models as well.

There is actually not enough data available there to train the models that they now use AI to create synthetic data to complement the real-world data to further optimize their algorithms. And so when you put this all together, we now are convinced that exabytes will grow more than 25% on a compound annual growth rate in the next 3 to 5 years.

Got it. Very comprehensive. So maybe then how is that driving the engagement with your customer change? So for example, you've talked about in the past, long-term agreements, build-to-order? Like what did those historically look like before you started to see the AI tailwinds?

What do those look like now? How is AI and the demand that you're seeing changing the engagement with the customer in relation to these agreements?

Yes. So first of all, we definitely have changed our customer engagement model, and we do have a lot deeper customer engagements right now, especially with the hyperscalers, right? And so the relationship has changed from a supplier customer relationship 3-plus years ago to a really strategic partner.

And all our customers, the hyperscalers fully understand the critical value of data and the critical part that hard disk drives and storage plays in their overall AI data center build-out. So that is very important.

We also have, again, much deeper customer engagements. They share their technology and product road maps. They share and give us much more visibility about their demand for storage multiple years out. In return, we have done the same, right?

We share our technology and product road maps with them. We've invited them to our factories. We share our capacity and manufacturing road maps with them. And they fully understand that it takes 52 weeks to produce a hard disk drive, 9 months to produce the wafers that become the heads and then 3 months to put it all together, including testing.

And so we have told all our customers or the larger customers, we expect 52-week order lead time for purchase orders because it takes 52 weeks to produce a hard disk drive. And so in addition to that, our customers also wanted to secure more supply further out. And that's why we have entered into LTAs with our larger customers for calendar year '27, with some of them for '28 and even less of that '29.

For us, again, those LTAs are frameworks, right? Those are not take-or-pay hard agreements. They are there to create more visibility from the key customers to us and commitments that we make to them. Those LTAs have a volume component and a price component in it. It's slightly different by customer. And again, there is still some flexibility there.

The most important thing it's creating better predictability, better visibility on both sides.

Got it. going to the 52-week lead time that you referenced, I mean, one of the questions, and I'm sure you've got this from investors as well is you start investing in capacity today, you're shipping out 52 weeks later. What happens if the customers' demand profile changes, right? And you're getting into these LTAs, but not necessarily that these are take-or-pay contracts.

So how do you get the confidence about adding capacity when you have lead times that extend to 1 year almost at this point? And are you -- how do you get confidence around not -- that you're not adding capacity at the peak?

Yes. So first of all, again, for us, it's very important to continue to get that visibility multiple years out, right? That helps us to plan our business. And so far, as I just explained, the demand signal is stronger for longer, right, which is good. Now when we look at that, we look at that, do we have to add unit capacity to support plus 25% compound annual growth rate for the next 3 to 5 years? The answer is no. We know how to support the strong growth in exabytes of our customers, not by adding unit capacity, but through technology and product transitions, right?

As you probably know, currently, our average terabyte per unit that we ship in near line is only about 23 terabyte per unit. Despite the fact that we do have the highest capacity we have available today is 32 terabyte units, and we ship a lot of them already. And we've indicated that we are currently in qualification for the next-generation ePMR drives that go up to 40 terabytes, right?

Qualifications are going really well, and we are going to ramp volume production in the second half of calendar year '26. And then in parallel, we have been working on HAMR for 10 years. We're now in qualification with 4 customers. Qualifications are going really well.

And that goes up to 44 terabyte drives and the ramp will happen in calendar year '27. In addition to that, we have a technology and product road map that will get us to 50, 60, 70, eventually plus 100 terabyte drives, right? And so we have plenty of room to -- as we execute on our technology and product road maps, as we collaborate with our customers on qualification and adoption of those higher capacity drives, we can continue to execute and deliver and supply 25-plus percent more exabyte on a compound annual growth rate basis.

Got it. So maybe that's on the volume or the storage side, storage exabyte growth side. Let's talk about pricing, but maybe let's start with a different angle, which is pricing has -- pricing structure has changed for the industry. It's a much more favorable pricing environment. Can you go into the competitive landscape of the industry structure today? And how much do investors need to bank on rational competition between the players to see a favorable pricing environment continue in the industry?

Yes. So I can't speak for my competitors, but you know the competitive landscape. It's basically 2 large suppliers and a third smaller player out there. So that's the competitive landscape. The way we go at pricing, it's a value-based pricing, right?

As we execute on our technology and product road map as we continue to innovate, right, and move to higher capacity drives, we provide more value to our customers. Higher capacity drives give them better rack density, lower real estate cost, better power consumption and more exabytes, right? And so there is more value for our customers, and we want to get paid for that, right?

And so that's why we have continuously increased our pricing per terabyte. If you look at last quarter, our average -- blended average ASP per terabyte was up 9% on a price per terabyte year-over-year. And again, as we continue to further deliver more value to our customers, we see further opportunity to increase the pricing on a price per terabyte basis.

Okay. Okay. Maybe just talk about the technology approach you have relative to some of your competitors or your primary competitor. You have a multipronged approach on the technology front. You're doing -- you're ramping HAMR, you're ramping ePMR, which is different from your competitor where there's much more of a single focus on the technology front.

Why the multipronged approach? And how does it impact eventually financials when you think about OpEx spending or R&D? How are you sort of thinking about the strategic rationale as well as the financial implications of it?

Yes. First of all, our customers, rightfully so, are very careful to store your data, my data, everybody's data. They are very conservative about that. For them, quality, reliability, they don't want anything to go wrong with your data. And so they are somewhat conservative. And so we listen to them, we work with them.

And let me go back a little bit. I think there was maybe a misconception in the industry that ePMR drives will max out at 28, 30 terabytes per unit. We clearly proved that was a misconception that was wrong. We are shipping in high volume 32 terabytes right now. We are in qualification with the 40 terabyte that will ramp in production in the second half of calendar year '26.

And we believe we can further improve ePMR all the way up to 50 or 60 terabytes in the near future. For now, we think it will cap out there, but who knows, maybe engineers will find ways to further stretch that out. In parallel, for the last 10 years, we have been working on our HAMR technology. It's in really good shape. We are, as I said before, in qualification with 4 customers, and that comes up to 44 terabyte per drive when we launch that in volume production in 2027.

And then, of course, we will continue to expand the capacity per drive into the 50, 60, 70 all the way up to 100-plus terabyte there as well. So that's what we do from a technology transition. As you probably know as well, ePMR and HAMR comes in 2 flavors. We have the CMR and the UltraSMR flavor.

UltraSMR gives us 20% more exabyte versus CMR, mostly done in firmware, software adjustments. And again, our customers want more exabytes. Two large customers are fully adopted UltraSMR. The third one is now ramping UltraSMR.

But we have been talking to all our other customers and say, you want more exabytes? Let's collaborate on the adoption of UltraSMR. And we have a plan now to convert almost all our customers or the larger customers to UltraSMR with qualifications finalized by the end of 2027.

Okay. I mean -- so maybe just trying to think about the financial implications here. You migrate most of your customers to UltraSMR. You have a cost benefit there. But on the flip side of it, how should we think about R&D expenses to support these multiple platforms that you're working on?

Yes. So of course, if we would only focus on one platform that would be a little bit result in lower operating expense. We choose to support both of them because, again, we believe there is more value for our customers there. They love the ePMR proven technology, which has been out there for 10 years.

They understand the quality, the reliability. We know how to scale it in a very condensed way. And so yes, we have to deal with some higher operating expense right now. By the way, we're also not only investing in higher capacity drives, we're also investing in improving the overall performance of the drive, increasing the bandwidth as well as the speed of which the data comes in and out of the hard disk drive.

We're getting really good feedback from our customers on that with our high bandwidth drives as well as the dual pivot technology. Of course, that requires some operating expense as well to support that. But we believe, again, that the value creation we can realize by investing, focusing on innovation is the right strategy to do that. Operating expenses going forward, there's definitely leverage in the model, right?

I mean revenue is going to grow a lot faster than our operating expense. We don't need much more operating expense. Yes, there is some inflation and merit and some other adjustments. But the operating expense will continue to come down as a percent of revenue at a good pace.

Okay. So maybe let's move to focusing a bit on EPM. Even as you're ramping HAMR, ePMR clearly is a differentiation for you because when investors look at gross margins, your gross margins are actually ahead of your peer despite ramping HAMR more recently compared to your peer company.

Why can't your competitors replicate what you're doing with ePMR? Like what does that competitive moat look like? And what prevents competitors from copying or replicating what you've done?

Yes. I mean I think it's an architectural and strategic product road map and technology road map decision, right? Our competitor decided not to continue to invest in ePMR and invest on HAMR, and that's their decision. Again, there is I mean it's only 2 and 1/3 smaller player, right? We don't see new entrants. We don't see new investments into the hard disk drive business. We have our technology road map. Again, we believe that is a really smart, good decision.

HAMR is a great technology, and we are all in on HAMR, right? And for us, HAMR is the path to plus 100 terabyte drives. But HAMR does come with some incremental cost. You have to add a laser to each and every head. So there's some incremental cost. You have some incremental cost on the substrate.

Again, it's the right thing to do. We're all in on HAMR. We're making good progress. We're ahead of schedule. But we do believe that there is still a little bit of longer life with ePMR in a very cost-effective way. Again, a product that our customers knows very well, scalable, reliable, and we will continue to execute in parallel, ePMR and HAMR.

Okay. Okay. That's good. So that's a good segue. When you talk about the 100 terabyte drives, you've outlined the road map to it from a technology perspective, which will be driven by HAMR. What are you looking at in terms of internal milestones to track progress to it? And what are the hurdles that you think you need to overcome if you -- in terms of the path to that 100 terabyte drive...

Yes. Again, we're feeling really good about that. And when you think about how do we continue to increase the capacity per drive, there's 2 ways of doing that. One, which is the most important one is to increase the aerial density, right, and find ways to store more terabytes per platter, right? Currently, we are approaching 4 terabytes per platter, but we have a path to get that to 5, 6, 7, eventually 10 terabytes per platter, right?

We know in the lab how to do 10 terabytes. It's still going to take us multiple years to ramp that in high volume, but we know how to do that, right? And so aerial density is the most important factor. In addition to that, we know how to add more platters per unit, right? And that's not an excuse for aerial density.

Aerial density has to be at the industry-leading levels for moving all the way up to 10 and more eventually. But adding more platters per unit also gives you a cost advantage as well. And so we will continue to find ways to add more platters as well. And so now how are we getting to better aerial density? You need to make investments on your head technology, right?

The wafer nano manufacturing and head technology that does the read and write functionality as well on the media, right? And media is substrate. Do you use aluminum or glass, right? And currently, in ePMR, we use aluminum in HAMR, we will use glass substrates. And then how do you condition your substrates and turn it into media for magnetic recording and work on the recipe there.

And so we -- there's a lot of science, chemistry, magic that happens there, but we have very smart engineers, which, again, have figured out most of that. Now it's just about making sure we continue to execute on it, deliver on it. In many cases, you just need time to prove it out, make the improvements over time, make sure the quality, reliability is sustainable. And then, of course, you have to work on manufacturability in your factories as well.

And all of that is being done in parallel. It's really hard to do that, right? Again, that points to the high barriers to enter in the hard disk drive space, but we know how to do it and keep focusing on execution.

Got it. So you have these 2 parts, ePMR, HAMR, you have the 40-terabyte ePMR being qualified, you have the HAMR being qualified. When you go to a customer with both, what are you likely to see in terms of customer allocation of share?

How do they pick and choose between ePMR and HAMR? What do you think likely will happen when you go to a customer with both of those options?

So customers don't think about recording technology. They want scalable, reliable exabytes. Right? And however you can get that to them, right? Obviously, with ePMR, again, it's a technology they have been using for more than 10 years. They understand the quality, the reliability.

We have gone through multiple ramps of previous and next generation of ePMR. They feel very comfortable around that. HAMR, it's a new technology. Again, they don't have any experience with HAMR drives that are 5 or 6 years old in their fleet, right? They don't -- nobody has, right?

And so they are probably a little bit more careful there. They want to go make sure as you go through qualifications, -- they spend a little bit more time making sure they understand the quality, the reliability, how the drives behave.

And so that's why it takes a little bit more time. But again, to answer your question, they don't care about the recording technology. By the way, again, back to the question of the LTAs, the LTAs we have are exabyte based. They don't specify the recording technology. They, so to speak, couldn't care less. They want high-quality, high reliable exabytes.

Let me just pause here and see if anyone in the audience has any questions.

You said earlier that you feel confident you don't have to add unit capacity to support the 25% plus CAGR in exabytes. At what level would you feel like you do need to add unit capacity in order to support that?

Yes, that's a great question. And so we've indicated that we expect exabyte to grow now at plus 25%. And now the question is what is plus 25%, but I'll leave that up to you. And so -- but I'm comfortable with the plus 25% that we don't have to add unit capacity. Again, I'll give you some examples, right?

Today, we ship 23 terabytes. That's the average, right? If I can get to 44 terabytes, right, that's almost 44 over 23, almost 100% more exabyte per unit, right? Now again, you need to qualify the 44. We need to ramp the 44.

There is a certain adoption curve. You have to move customers over and over. And 44, shortly after that, we will get to the 50s and the 60s and the 70s and the 100 plus, right? And so if I look ahead over the next 3, 5-plus years, and I look at my technology and product road map, I don't see a need to add more unit capacity and still support very strong exabyte growth year-over-year.

Great. And just a quick follow-up, if I may. whatever that number is much higher than 25%, is it fair to say that the hyperscalers haven't shown you that level of stronger for longer yet?

No. Every time we talk to them, they come back with stronger and longer, right? Again, like a year ago, we were -- they gave us a certain view of the demand for the next 2, 3 years.

And it was like, yes, but we don't know exactly year 4 or 5, right? And as we have multiple discussions on a continuous basis over the last 12, 18 months, they now single a much stronger demand each and every time we talk to them, and they are willing to almost commit further out 2031, 2032.

One more question in the front. Can you just talk about some of the innovation on the high bandwidth technology? Just interested kind of how that shifts hard disk drives potentially in the data center and whether it can expand the kind of share of the total storage spending?

Yes. So today, like if you look at data centers, right, roughly 80% of the data is stored on hard disk drives and all or about 20% is stored on SSDs. Despite the fact that SSDs is a lot more expensive in terms of cost of acquisition and even in terms of total cost of ownership, that was the case 12 months ago.

And today, that's even a lot worse, right? Their cost of acquisition has gone up a lot. Their total cost of ownership has gone up a lot. And so our customers don't like that, right? But there is a reason why on or about 20% is stored on SSDs because they have a performance advantage in terms of throughput and speed, right?

And so our customers came to us and say, can we work on this? Can we collaborate on this? Can you help us? Can you improve the performance of your hard disk drives? And we have smart engineers that looked at it and studied it and found solutions. And we've basically found a combination of 2 solutions. One is high bandwidth drives, which is mostly a firmware, software solution to create better performance. And then in combination with the dual pivot, which is a hardware solution. And when you combine both of that, you will have more throughput and a bigger pipe, more bandwidth that will, over time, increase the performance 2x, 4x and eventually 8x compared to what we have today. And customers are very excited about that.

We are already sampling with 2 customers, the high-bandwidth drives, and we see a lot of traction there. And so as we execute there, right, there is a possibility for us to, so to speak, encroach into that SSD segment. We're never going to wipe that out completely, but it creates an opportunity for us to move into a warmer layer of data.

Let's move to -- for the last few minutes, just discussing HAMR a bit more and really trying to focus on HAMR yields. So in the past, you mentioned when you were ramping 18 terabyte or 20-terabyte platforms that you start with about 50% yields that ramps when you get to a more mature phase that ramps to about 80%, 90%.

Is that a good trajectory to think about HAMR that you ramp starting with 60 and you get to a maturity of 80%, 90%? And how much time before you get to that mature level of yield?

Yes. No, I think the story is slightly different. I mean a couple of years ago, yields -- overall yields were in the 60s, right? But we've really focused on automation in the factories. We captured a lot of data and initially used machine learning applications on top of that.

And now we use more advanced machine learning, call it, AI applications that has really helped us to drastically increase the overall yield from somewhere in the low 60s to somewhere now in the low 90s. Now of course, every time you go through a technology transition, you go a little bit through a learning curve in terms of yield, although in -- if it's moving from one generation ePMR to the next generation of ePMR, that has gone really smoothly. It's a little bit more challenging if you move from one recording technology, ePMR to the next generation HAMR of recording technology. That is not easy to do. That is challenging, but we know how to do that. Again, the good news for us, based on our strategy and technology and product transition, we are not in a rush. We want to take our time.

We want to make sure, again, that the customers are comfortable with the quality, the reliability of the product and ourselves are comfortable with that. And then we are going to ramp it in the factories, not feeling rushed because we have the 40 terabyte ePMR, right?

But we will take our time, go through the learning curve and make sure that if and when we launch HAMR right, it is kind of neutral to accretive from a gross margin point of view.

Got it. Got it. Okay. So I was going to follow up on that, but let me just wrap up with a question in terms of pricing then. You did report 9% pricing, as you said.

Now when you think about the go forward, obviously, you reprice these LTAs, you get that pricing benefit. Why shouldn't we expect pricing to accelerate from here on given that generally waterfall would indicate you have more LTAs coming up with better pricing over time. Why shouldn't pricing accelerate from the 9% that you've seen last quarter?

Yes. So again, from a pricing point of view, we feel good, right? The demand is strong. The supply is increasing, but we think over the next 3 to 5 years, demand supply will remain tight. And it's up to us to execute on our technology and product road map, provide more value to our customers by moving to higher capacity drives and then eventually as well, increasing the overall performance through the high bandwidth drives and the dual pivot. And as we provide more value to our customers, get paid for that, right?

We're not going to give it away. We want to continue to increase the price on the price per terabyte. And while we do that, provide more value to our customers, potentially lower their total cost of ownership, it's a win-win for them. It's a win-win for us, and it's a win-win for our shareholders. So that's how we approach it.

Great. I'll wrap it up there, but thank you to the audience, and thank you to you as well for coming to the conference. Thank you.

Thanks for having me.

Hello, everybody. Thank you for joining. Tom O'Malley here, just kidding. Tim Long sitting in for Tom O'Malley. I'm the IT Hardware, Comm Equipment analyst at Barclays. Very happy to be hosting Western Digital. Kris CFO, thanks for coming with us. I have covered the stock for a fair amount of time in the past. Tom covers it now. So it would be good to catch up. Obviously, a lot of good things going on with the stock in the industry. So before we get into it, Kris, I think you got to start off with the Safe Harbor.

Yes. And thanks for hosting us here, Tim, at this great conference. We have a lot of good meetings already. But before we start my compliance department asked me to make sure that today, we will be making some forward-looking statements based on our current assumptions and expectations as it relates to our product portfolio, business plans and performance, market trends and future financial results. Those forward-looking statements are subject to risks and uncertainties. So please go to our SEC reporting, Form 10-K, and other filings for more information about the risks and uncertainties that could cause actual results to differ materially from expectations. We also will be talking about some non-GAAP measures, and you can find a reconciliation between GAAP and non-GAAP on our investor website, investor.com or investor.wdc.com.

And I will give an opportunity with about 10 minutes left for folks in the audience to ask a question. So first, maybe just mark-to-market here a little bit. This year, obviously, there's been a lot of success the last few years with your business. We're starting to see AI industry transferring from a little bit more of a training-based market to inferencing. Could you talk about that transition, how that relates to WDC and HDDs? And do you see that as a positive? Or how do you see that playing into the business?

Yes. Obviously, we are very well positioned as a strategically-focused hard disk drive company in this data economy. It hasn't been that long -- in February of 2025, just after the separation of our Flash business. The management team was indicating that we expect the exabyte growth to be mid-teens year-over-year CAGR for the next 3 to 5 years, with a potential upside if and when AI kicks in. While in February of 2026 at our Innovation Day, we've actually already started seeing AI kicking in, and we were comfortable indicating that we now expect or in February, expect the exabyte growth to be 25% CAGR over the next 3 to 5 years.

Now, in the meantime, we have further ongoing discussions with all our customers. We do get long-term visibility from our customers based on our deep customer engagements that we have and last week at our earnings call, we were actually very confident to say that we now expect greater than 25% exabyte growth over the next 3 to 5 years on a CAGR basis. Maybe something trending that starts with a 3. And so what is driving that? First of all, the key growth is still there, right? More and more data is being generated by, I don't know, 7 billion or 8 billion people on the planet.

And every business and enterprise on the company in the world now continues to store all their data instead of even not capturing it or throwing it away, all data is being stored. Now in addition to that, of course, we do have the AI kicker. And there is 3 main blocks there. First of all, training of the multi-model LLM is still ongoing. A lot of training has been done, but all the big players continue to relearn and retrain their existing models, making continuously improvements and which will result in the next generation of LLM. Some of them are also training very specific dedicated language models for specific tasks and that will continue as well.

And that requires a lot of data. But we've also moved already into inferencing, as you indicated. On or about this year, 2/3 of the compute power that's installed will be used for inferencing. And what we learned now with inferencing that it is actually really good for hard-disk drive and data storage because every time you have an interaction with a chat box or even for more complex tasks by agentic AI where agents now execute complex tasks. All that output is being stored and not just the output is being stored, but also a lot of data along the way, the reasoning, how they get to certain of those answers, all of that is being stored, so there is a history.

And then third, what we also see now is the physical AI. And that's just beginning, physical AI, robots, humanoids, autonomous driving, autonomous cars. They use a lot of video, all those devices have video. They use a lot of video capturing that's then being used as well to train and optimize the algorithms. There is actually not enough real-live data to do that. And now they use AI to create synthetic data set that's being used to further optimize the models. And so when you put this all together, it's really compounding, right? More data for training, more better models, more inferencing that creates more data that's being fed into the loop and that all requires storage in the most reliable economical way, and that's Hard Disk Drive.

Great. Yes, there was a lot in there, and I think it leads into the second question pretty well. I think one of the things that people struggle with is modeling this business. It's not like there's a direct-attached per GPU or when you're thinking about forecasting this exabyte growth, do you think about per gigawatt, per accelerator, how are you trying to formulate what your growth rate can be relative to what you're seeing from the deployments from your customers?

Yes, it's -- you're absolutely right. There is -- it's hard for us to make a direct link to the number of Gigawatts of compute that's being installed. For memory, which is closely attached to all of that, there is a more direct link as I explained it, right, Hard Disk Drive storage and Data Storage follows different economics, right? It's actually -- so again, it's a compounding loop, versus compute it's being reused, right? You use a couple of seconds of compute power and it generates something; the GPU frees up and can be reused, but the output of your inferencing is being stored and will be stored forever, right? And so it's slightly different.

And so that's why we -- the last 12, 18 months, have really established those deep customer engagements with all the large hyperscalers in the world, in the U.S., in Asia, and other places. And we've really worked with our customers to get long-term visibility and ask them the right questions, what are their technology and product road maps, what are their road map in terms of data center build-out how much capacity or exabyte do they think they will need over the next 3, 5-plus years. And based on that, we are developing our own technology and product road maps to support that.

Okay. Great. You're leading me right into the technology stuff here. You laid out this road map where you're going to ramp both the ePMR and HAMR up to 60-terabytes. Talk to us a little bit about that decision and what that means for customers and your ability to supply multiple technologies. What are the benefits of that type of strategy?

Yes. We have an industry-leading technology and product road map that is, again, really developed in close relationship with all our customers. And for the last 10 years or so the dominant recording technology was ePMR, right? And that's what we use today. Initially, the industry was thinking that we could only get to maybe 30 terabytes on ePMR. While we've broken that logic, right? And today, we're already shipping in our latest generation of ePMR up to 32 terabytes and we found a way to continue to stretch the ePMR technology to 40 terabytes, 50 terabytes, and 60 terabytes in the future. The 40 terabytes we already have; it's in qualification with 3 customers right now, and we are getting ready for a pretty steep ramp in the -- of the 40 terabytes ePMR in the second half of calendar year 2026.

Now in parallel, for the last 10 years, we have been working on the next-generation technology, which is HAMR, and we are in a really good spot there. We have -- we're going to launch the product and start ramping it in calendar year 2027 with 44-terabyte solutions. That is currently in qualification with 4 customers. We are getting really good feedback from our customers; they're really happy with the performance, the quality, the reliability of the product and how it performs into their qualification systems. They're really happy with the areal density, as I said, up to 44 terabyte per drive.

And we still have a little bit more work to do, but we're not rushing it because we have our ePMR technology that goes in parallel to our HAMR technology that kind of de-risks it for our customers and de-risk it for ourselves, and we think that's the right way to play it.

Okay. And maybe just digging into the HAMR a little bit, talk -- so there's still a little room to go, positive feedback so far. But what are kind of some of the technical milestones that you think you need to hit? And what's the current view on kind of availability and revenue time line for some of the HAMR products?

Yes. Again, we're -- we've been working on it for 10 years. It's a new technology, right. It's heat-assisted magnetic recording that require some redesign of the head, including the integration of a laser. It requires some modifications on the substrates and the media to do the magnetic recording. But again, we figured it out, and we're finally getting to a point here now that we are in qualification with 4 customers and getting ready for the ramp in 2027.

By the way, HAMR is not just about getting to 44 terabytes per drive. We believe that with HAMR, over time, we will get to 50, 60, 70 and eventually more than 100 terabytes per drive, right? And that is the technology. We invest in that. By the way, we're also investing already in the next-generation technology that will probably come to market 10 years down the road, which is called a HDMR, heat-dot magnetic recording.

Okay. Yes. I hope I'm not doing this in 10 years to ask you about it, but maybe I will -- we'll see. Some other new product developments you talked about addressing bandwidth and power -- the high-bandwidth drives and dual-pivot actuators. Could you talk a little bit about those technologies time lines? Was this customer push? Was it or customer pull or WD push, and it's obviously addressing some bottlenecks. So, what do you think the outlook or the opportunity for WD in these areas are?

Yes. If you look today, in data centers in the cloud, roughly 80% of the data is being stored on Hard Disk Drive, right? And the remainder, 20%, is stored on SSDs or NAND, or Flash or whatever you want to call it. And so despite the fact that is a lot more expensive in terms of cost of acquisition and in terms of Total Cost of Ownership, Flash SSDs is a lot more expensive. But why is it not 100% on Hard Disk Drive? Well, Flash, SSD, they have a performance advantage. They have faster in and out and higher throughput, and better speed than Hard Disk Drive, but our customers, of course, they asked us, Can you improve your speed and throughput.

And so, we have smart engineers who looked at it, and we found ways to drastically, up to 8x faster, right, increase the speed and the throughput of the current Hard Disk Drive. It's something that we've announced on the Innovation Day. It's called High-Bandwidth Drives, right? Most of that is being done in firmware and software. And that, in combination with a dual pivot, which is a hardware solution, right? Will get us to those solutions. Currently, we are shipping samples to 2 of our customers. So far, there again, the feedback is very positive. They really like what we see. And it's something, again, the customers want it, the customers need it. It's important to potentially for us to expand a little bit into the more warmer and/or hot data with higher throughputs, but we also need it as we move to higher capacity drives, 50, 60, all the way up to 100 terabytes.

Okay. I want to get into some of the business-type questions here. Maybe just starting structurally here, HDDs are now vast majority, vast, vast majority going to the Hyperscalers. Could you talk about how the relationships are changing, whether it's contract terms, prepayments guarantees, or LTAs. Maybe just give us a sense of kind of how things are structurally changing because there's such tight supply and such high demand?

Yes. I mean, again, our business has changed drastically for many reasons, right? Even if you look back a couple of years ago, more than 50% of the business was still Client consumer. And then we started moving into the cloud, with the Hyperscalers and some of the Enterprise Cloud providers as well. Fast-forward to today, 90% of our business is with the -- is Cloud, mostly with the Hyperscalers. We still have 5% of our revenue in Consumer and 5% of our revenue in Client with the PC business. But the vast majority is Cloud. We have the Hyperscalers. And we've completely changed the business engagement model, right? I mean, it used to be in the past that we were -- it was turns business.

We were taking orders for shipments in the same quarter. That is no longer the case. We went to, again, our major customers, we explained them our technology and product road map. We invited them to our factories and we showed them. It actually takes almost a year to produce a Hard Disk Drive; 9 months for the wafers that eventually turns into heads and then 3 months to put it all together and test it. And so we again, we've completely changed the business engagement model. We get a lot more information from our customers. We provide a lot more information to them.

And we've learned from each other that creating better visibility in the long-term benefit both of us, right? And so now we do have better visibility from our customers, in some cases, always stretching out 3, 4, 5 years, right? And so while we're having those discussions with our customers, -- they -- we made it clear that, first of all, you have to place your orders 52 weeks in advance. So we have purchase orders for the next 52 weeks, right? Of course, that has volume price and the exact SKU that we're going to skip to them are fixed, but they also say, hey we have to secure our supply for longer period of time.

And that's where we entered with a lot of the large customers into LTAs covering all of calendar year '27. With some of them, we actually covered calendar year '28, and with some of them all the way going into calendar year '29. Those LTAs, they have a volume and a price component in baked in. But the volume there, of course, there is some flexibility there, right? In some cases, they might need a little bit more, and we will try to get them more at potentially different pricing. Also, in some cases, maybe some customers need a little less. Well, the sooner you tell us, the better maybe we can move that supply to another customer. We might -- of course, there might be some different pricing as a result of that. And so that's all baked in those relationships.

I think for me, LTAs, it's a big word. For me, the most important thing is, again, better visibility, partnership, collaborating with technology and product road maps, creating more supply because for us, back to maybe your other question, right, we see very, very strong demand. We believe the next 3 and 5 years, we will be able to supply that strong demand growth, not by adding unit capacity, but by working on our technology and product road map moving to those higher-capacity drives and strong collaboration with our customers, right, because they have to adopt those new technology and they have to adopt those higher-capacity drives. We're moving more and more customers to UltraSMR, which gets from CMR, which gives them a 20% boost. And that needs to all happen together in collaboration and that's more important to me than a piece of paper, which is called LTA.

Right, just to dig into it a little bit, 2 follow-ups. One, I think you talked about this year mid- to high-single-digit year-over-year ASP increase each quarter. Is it safe to assume when you're looking at the LTAs that go to '27, '28, '29, we're factoring in similar-ish type of -- or ASP increases, number one? And number two, I know you made the point before, and I agree, is very different than NAND and DRAM. But obviously, those prices are going up much more dramatically. So maybe just bridge that for us? And does that mean that at some point there is even more opportunity on the pricing side because some of the other related semiconductor areas are seeing more?

Yes. So we obviously feel really good about the pricing environment. But we do play completely different than the memory side, which is more of a commodity and a commodity trade, right? We play the mid- to long-term relationship with our customers and our partners. And our pricing is based on value, right? The more value we provide to our customers, the more we want to get paid for that, right? And how do we provide more value? Well, by moving to higher-capacity drives, right? and moving to higher-capacity drives have all kind of benefits for our customers, better rack density, lower real estate cost, lower power consumption.

And so we can actually charge them a little bit more on a price-per-terabyte basis and still delivering a lower Total Cost of Ownership, which is a win-win for all of us. Obviously, operating in a very strong demand environment with supply that comes online as we execute our technology and product road map. That obviously helps. And so currently, we are seeing price-per-terabyte increases in the mid- to high single digits. Last quarter, it was 9% up year-over-year. At Innovation Day, I have indicated that I do expect for all 4 quarters of calendar year '26. ASP per terabyte to be up mid- to high single digits. And again, we are executing to that. And yes, I do believe that, longer term, there is further opportunity to improve the pricing as we deliver more value to our customers.

Okay. We did have a question. I'm assuming it's related. Can we get the microphone there, please?

And I just have 2 follow-up questions on the LTA. And some customers are willing to -- are eager to sign LTAs, if another group of customers, they don't sign LTAs then what would happen to those customers who don't want to sign LTAs?

Nothing. Again, we have a set of -- it's a limited set of customers, right? You have 4 or 5, maybe 6 large Hyperscalers in the U.S. You have a couple of large Hyperscalers in Asia. And then you have some smaller and upcoming Cloud players out there. That's the set of customers. And we have relationships with all of them. So we don't actually care too much who is a winner and who grows faster than the other one because we deal with all of them. Our market share with each of those players is on or about the same. And so we want to make sure, of course, that we have supply allocated to each and all of those players. Even if you have an LTA, if you don't have an LTA, it doesn't really matter, right? For me, LTA is a piece of paper. We have customer relationship, customer engagements. That's the most important thing.

My second question is -- if 1 customer -- because the company mentioned that for capacity increase is really for -- through a product upgrade. And if one customer is willing to say, co-invest with Western Digital, let's say, put a 50% CapEx down on the table and then just physically to increase production capacity with you and will you be willing to take that offer and to physically increase production capacity or the factories to push volume up?

So the answer is no. Again, we're -- in memory, we start seeing that, but memory is a totally different business than Hard Disk Drives, right? Memory has really high CapEx requirements. They need to go build new factories to go support the incremental demand. In Hard Disk Drive, the CapEx is very low. it's 4% to 6% of revenue, right? So -- and we don't need to add more unit capacity. We can increase our supply somewhat in line with the strong demand that we see by moving up to higher-capacity drives, and that's the answer.

Maybe I'll see if there's any other before while the mic is out. Any there's one over here.

This is a question on potential new or upcoming competitors. This is, of course, by history, a very, let's say, oligopolistic industry. We've seen some competitors emerging over the years in the DRAM and NAND space in China. Can you share something maybe a bit, say, from your side of the business, what you're seeing out there?

Yes. So this -- today, in the Hard Disk Drive, there's basically 3 players, 2 large players, including Western Digital and then a third smaller player. We don't see other new entrants in that business because it's really hard and difficult business with really high barriers to enter, right? I mean you need to know how to make heads that does write and read, right? It's a very complex specialized technology that goes through a 9-month wafer fabrication process, very complex and hard and difficult to do.

It's not your typical CMOS or other semiconductor thing. In addition to that, you need to know about magnetic recording, the substrates, the media, and how you condition that and then being able to bring that all together in high-volume production. Again, we know a little bit about Hard Disk Drive because we've been doing it for 55 years. It still takes us 10 years to develop the next-generation technology. So we been working 10 years on HAMR. Now, HAMR is ready. We start working on HDMR, which will probably take us another 10 years. And so we don't see new entrants trying to get into this space because it's really hard.

I'll go back to the list here. I want to talk a little bit about gross margin. You kind of touched on some peripheral areas now. I stop covering the stock 5 years ago, it was in the low 20s, and now you're more than -- well more than double that. So maybe just talk a little bit about gross margin sustainability. Obviously, upping your exabyte growth rates helps. Talk a little bit about introducing HAMR into the mix. Not necessarily adding physical capacity, adding higher products. So maybe just break us down all the moving parts as you look at the gross margin sustainability over the next few years here.

Yes. Yes. I feel really good about the gross margin. Again, this is a totally different business than it was 3 or 5 or 10 years ago, where margins for a very long period of time were 20%, right? Here, we're mean the value of data has increased a lot, and we add more and more value to our customers by moving to those higher-capacity drives. And you look at it, gross margin has moved from 20s to 30s to 40s. Last quarter, we were the first one crossing into something starting with a 5-handle at 50.5%. We guided to 51%, 52%.

For this quarter. And we do see further gross margin expansion as ASP per terabyte continue to go up. And as the cost-per-terabyte continues to go down both driven by moving to higher-capacity drives, which as already explained, we will move from 30s to 40s to 50s to 60s, eventually 200 terabytes, right? HAMR is an important part of that because, again, with ePMR, we think we can only get to 50 terabytes, 60 terabytes. So we need to get HAMR going because that gives us the path to get to more than 100 terabytes over time. HAMR, again, we are not rushing into it because we have this dual track with ePMR and HAMR. HAMR by itself. There's some structural difference there that adds a little bit of cost, right, because you need to add a laser per head and you have a little bit more expensive media to deal with as well. But you get the cost-benefit by moving to higher capacities-per-drive. And combine all of that, again, I feel good about where gross margin is and further progression and expansion of the gross margins over time.

Okay. Yes, it could be all that. It could be the hardware guys stopped covering and the semi guy started covering you, and everything is better in life. I get that a lot. You mentioned lasers and HAMR last earnings call, you did talk about acquiring some IP and talent for HAMR internal development. So could you talk a little bit about that decision? You mentioned it's obviously an incremental-cost item and complexity. So maybe talk about how you think that what -- how big of a role will WD take in that vertical integration relative to merchant and what do you think that could mean for margin structure and, quality of product, and everything else?

Yes. No, it is important because, again, HAMR is heat-assisted magnetic recording. And basically, what we do is a laser that's attached to the head that before you flip the bit from 1 to 0, you hit the media, and that helps to drive further areal density. We do have an external source of the laser, but we thought it is a critical component in our overall HAMR technology road map. And so we wanted to do it internally as well. So we acquired some IP and some know-how and some smart people.

And so now we're working in parallel, working with our external source and an internal source that will further help us to improve the areal density in HAMR that will also eventually allow us to more closely integrate the laser into the head itself, which will also then create some more space and help us to put more platters into the box as well. And so there's a lot of performance advantages, integration advantages. Eventually, that will also then result in cost advantages by doing it in-house and integrated.

And what would the time line for this type of integration be?

Yes. I mean this is something that we are working on in right now, but that is -- I mean, we will continue to make it smaller and further integrate it into the heads over the next 2, 3, 5 years. It's a long-term project.

Okay. Follow-up on lasers. You spoke about vertically emitting lasers with a shorter-term form factor, allowing you to compress the space between platters and ultimately more disks inside a drive. Talk a little bit about the advantages there and the time line there and how that helps scale capacity.

Yes. Again, for us, the most important thing is areal density. We're really focused on increasing the areal density per platter. And there's a lot that needs to happen. We need to get better heads for that. We need to get better media for that. And -- but we are on track. I mean, today, we figured out how to get to 4 terabytes per platter and we have a path to get to 5, 6, 7, and eventually 10 terabytes per platter, right? So -- that is the most important thing. Now in addition to that, we also look at how -- is there a possibility to put more platters in a box without changing the physical structure of the box, right? And we again, are industry-leading there. We already have 11 platters in a box. We know how to get to 12 and eventually to 14 and potentially more after that. But that's the 2 items that we focus on to continue to drive to higher capacity drives.

I just want to give one more chance for anybody. Yes, one more.

I'm just asking question again from the angle of what maybe disrupts, let's say, this very optimistic future in a way. And if we look at a logic or the discussions around GPU today, there are new chip parts architectures coming to the market, which maybe do reference more efficient, hence, more -- less computing power. But on the storage side, you mentioned the amount of data, which needs to be put away and used again. Is there -- if you would have to put a black hat on, where is the vulnerability in a way of this, let's say, positive projection going forward? What is the vulnerability or the weak spot?

To be honest, I don't see it. I mean we are the most efficient, scalable, reliable with very strong economics for our customers. Very strong economics for our customers, right? A way of storing massive amounts of data. And we don't see any hesitation from our customers. I mean, AI is a data-driven economy, right? And it all is based on massive, massive amounts of data and Data Storage. And we provide that to our customers in a very lucrative, economical, scalable way to them.

And I don't see that based on all the conversations I have with my customers, I don't see that. I mean in the past, there was this idea about like all SSDs are going to replace Hard Disk Drive in the cloud, right? I mean, -- if you go back a couple of years ago, that was the case in the PC, right, where Hard Disk Drive in PC did get replaced by SSDs in the PC, fair. All right, right? And maybe 1 or 2 years ago, some folks believe that was going to happen in the Cloud as well. I don't see that. Just based on total cost of ownership differential, cost of acquisition differential, and just the overall performance of technology and product innovation that we continue to do not only by going to higher capacity drives, but also improving the overall performance, throughput and speed of our Hard Disk Drives. I don't see anything else that could disrupt that.

And if people say that a lot of inference is going to the edge or on the device? Is that a negative or necessary?

No. I mean eventually, data is being stored and the vast majority of data is being stored in the Cloud. Even if there is some local compute or local inferencing on the site, in many cases, that got uploaded into the Cloud as well.

We'll try to be quick with these 2.

And if China decided it was a national strategic priority to disrupt your orders and place you on a 10-year view, they don't have a chance even if they throw everything at it.

They've already tried it twice and couldn't get it done and we don't see them trying again, but maybe tomorrow, they can and then maybe 10 years down the road, they might succeed in that or not.

Yes, And partly linked to that, just what are the biggest obstacles to you increasing supply? What are the physical limitations you have in your own supply chain and your own ability to build as you compare to the memory example being, obviously, heavy CapEx fab build-out. What's the equivalent for you in that regard? And I guess linked to that, but not wholly linked to that is what therefore, prevent at these levels of situation and margins, someone else saying actually we can also do that. What's the biggest barriers to that?

Right. And so I don't need to increase my capacity or let me be clear. I don't need to increase my unit capacity. I don't need to produce more Hard Disk Drives. I can supply the very strong demand growth by moving to higher-capacity units, right? So that's why my CapEx will continue to stay in the 4% to 6%.

And what is the biggest areal density you expect to [indiscernible]?

Again, I mean, making heads is a very complex, hard thing to do, right? You need years and years and years of experience. You can't go to TSMC and say, make me some heads. They -- it's a very specialized unique technology. And then the substrate and media magnetic recording is something that requires years and years and years of technology and innovation, being able then to bring that all together in a box, it's just hard to do.

Okay. I think we are up on time. Kris, really appreciate it. Thank you very much. Thank you, everybody for joining.

Thanks.

Good afternoon, and thank you for standing by. Welcome to Western Digital's Third Quarter Fiscal 2026 Conference Call. [Operator Instructions] As a reminder, this call is being recorded. Now I'll turn the call over to Ambrish Srivastava, Vice President, Investor Relations. You may begin.

Thank you, and good afternoon, everyone. Joining me today are Irving Tan, WD's Chief Executive Officer; and Kris Sennesael, WD's Chief Financial Officer.

Before we begin, please note that today's discussion will contain forward-looking statements based on management's current assumptions and expectations, which are subject to various risks and uncertainties. These forward-looking statements include expectations for our product portfolio, our business plans and performance, ongoing market trends and our future financial results. We assume no obligation to update these statements. Please refer to our most recent annual report on Form 10-K and our other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations.

In our prepared remarks, our comments will be related to non-GAAP results on a continuing operations basis unless stated otherwise. Reconciliations between the non-GAAP and comparable GAAP financial measures are included in the press release and other materials that are being posted in the Investor Relations section of our website at investor.wdc.com. Lastly, I want to note that when we refer to we, us, our or similar terms, we are referring only to WD as a company and not speaking on behalf of the industry.

With that, I will now turn the call over to Irving for introductory remarks. Irving?

Thanks, Ambrish, and good afternoon, everyone, and thank you for joining us today. WD started calendar year 2026 with great execution, driving strong sequential and year-over-year revenue growth in our cloud, consumer and client businesses while expanding gross and operating margins. Gross margin exceeded 50%, driven by our continued innovation and focus on improving total cost of ownership for our customers through higher capacity drives and increased adoption of our UltraSMR products with strong operating leverage lower interest expense and an efficient tax structure. These efforts resulted in nearly doubling of our EPS compared to last year.

These results underscore our commitment to leading-edge innovation and strong execution. It is an exciting time to be part of WD, a focused HDD company and a strategic partner to hyperscalers and cloud service providers in this AI-driven data economy. We are well positioned with business momentum building across our entire portfolio with greater visibility into long-term customer demand.

Looking at the bigger picture, it is clear that data and data storage are becoming more critical and valuable as AI workloads extend from training to large-scale inferencing, data generation is at an inflection point. This year, inference is expected to account for roughly 2/3 of all AI compute. This larger focus on inference increases the amount of data generated, which in turn increases the need for data storage. The scale of what is happening is also considerable. One leading hyperscalers [ LLM ] processes over 16 billion tokens per minute via direct API used by their customers. While another AI company processes over 2.5 billion prompts every single day from 900 million active users.

While the resources that are used to create tokens are recycled, the data that is being created must be stored. Every token, every prompt and every query answered and checkpoint state creates data that requires persistent scalable and cost-efficient storage. And the majority of this data is stored on hard disk drives. As we look ahead, we see the rise of agentic AI. The next wave and arguably the biggest yet. What we are seeing with agentic AI frameworks represents a structural shift from AI that answers questions to AI that continuously execute workflows. That transition materially increases data generation and extends data retention cycles. Every hour of autonomous agent work and every action and agent takes creates data that must be stored. As a result, we expect agent AI to drive a step function increase in capacity-oriented storage demand, particularly in cloud and enterprise environments.

Beyond agentic AI, two more ways are building simultaneously, synthetic data, the primary fuel for physical AI is by design orders of magnitude larger than real world input step seeded. Across industries, physical AI data factory frameworks are being designed to transform limited training data into larger synthetic data sets at scale for robotics, autonomous vehicles and [ Vision AI ]. And physical AI itself, robots, industrial systems, autonomous fleets generates continuous streams of video, sensor and [ ocean ] data that must be stored, versioned and set back into training loops. These forces are not additive. They are a compounding loop. Inference creates data, agents consume and generate more data, physical AI creates data and train synthetic models that create more data. And ultimately, the loop accelerates.

We are truly seeing that the AI-driven data economy is creating an unprecedented demand for high-capacity reliable and high-performance storage on HDD's. This reinforces our conviction that long-term data storage growth will be greater than 25% CAGR. WD's technology and product road map is purpose-built to meet this growing demand. As we shared at our Innovation Day in February, we continue to innovate to meet our customers' needs through a combination of capacity leadership and performance innovation. Our high-capacity drive road map now extends from our 44 terabyte HAMR and 40 terabyte ePMR drives that are currently in qualification to a road map that goes beyond 100 terabytes.

On HAMR, we are accelerating our development, and we are now in qualification with 4 customers. We are qualifying our 40 terabyte ePMR drives with 3 customers and are on track to start volume production in the second half of calendar year 2026. With UltraSMR technology, which works across both ePMR and HAMR drives, we are expanding our customer base significantly. Three of our largest customers now have adopted the technology. Two are already meeting nearly all of the [ exabyte ] demand with UltraSMR, while the third is rapidly ramping in that direction. We plan to have all of our major customers qualified on UltraSMR by the end of calendar year 2027.

We are delivering on major aerial density improvements, along with our focus on performance, innovation with our high bandwidth drives. Customer response to our innovation has been very positive. Our high bandwidth drives are currently sampling with 2 hyperscale customers with an additional customer scheduled to start this quarter. Our dual pivot technology is being built specifically for new AI workloads with an [ OpenAPI ] approach aimed at simplifying deployment at scale. Based on our industry-leading technology and product road map, we are well positioned to support growing customer capacity, demand and address their AI workload needs. Our long-term visibility continues to improve with the duration of our agreements now extending into calendar year '28 and calendar year '29. We continue to see strong demand from across our client consumer and OEM enterprise customers as well.

In summary, the tailwinds shaping our industry today are both exciting and dynamic. And in WD, we remain focused on meeting our customers' needs while enhancing the value proposition and delivering long-term shareholder value to our investors.

With that, let me now hand it over to Kris to share our Q3 results and outlook for Q4.

Thank you, Irving, and good afternoon, everyone. The WD team delivered strong results, making solid progress against our strategic priorities with continued focus on innovation and disciplined execution, while advancing key initiatives and remaining tightly aligned with our customers' growing exabyte demand. As we move forward, we are encouraged by our momentum and remain confident in our ability to deliver sustainable revenue growth expand gross and operating margins and create long-term value for our shareholders.

During the third quarter of fiscal 2026, revenue was $3.3 billion, up 45% year-over-year, driven by strong demand across all our end markets and an improved pricing environment. Earnings per share was $2.72, almost double compared to a year ago. Revenue, gross margin and earnings per share were all above the high end of the guidance range. We delivered 222 exabytes to our customers, up 34% year-over-year. This includes over 4.1 million drives or 118 exabytes of our latest generation ePMR, with capacity points up to 32 terabytes demonstrating our ability to quickly ramp new technologies and products in support of strong customer demand growth.

Cloud represented 89% of total revenue at $3 billion up 48% year-over-year, driven by strong demand for our higher capacity nearline product portfolio and a stronger pricing environment. Consumer represented 6% of revenue at $186 million, up 24% year-over-year. Client represented 5% of total revenue at $179 million, up 31% year-over-year. Both client and consumer segments saw strong year-over-year exabyte growth and improved pricing.

Gross margin for the fiscal third quarter expanded to 50.5%. Gross margin improved 1,040 basis points year-over-year and 440 basis points sequentially. The drivers of strong gross margin performance include continued mix shift towards higher capacity drives, along with ongoing execution of our pricing strategy and tight cost control.

Operating expenses were $397 million or 11.9% to revenue, a 40 basis points sequential improvement, demonstrating further operating leverage in the model. The sequential increase of operating expenses was driven by the [ angulation ] of R&D project expenses as we continue to expand our HAMR qualifications with more customers. Strong top line growth, expanding gross margin and leverage in the model drove operating income to $1.3 billion, up 116% year-over-year translating into a strong operating margin of 38.6%, up 1,260 basis points year-over-year. Interest and other expenses were $24 million, and our effective tax rate in the fiscal third quarter was 16% taking into account the diluted share count of 385 million shares earnings per share was $2.72, an increase of 97% year-over-year.

During the third fiscal quarter, we significantly strengthened our balance sheet by monetizing 5.8 million shares of SanDisk, which led to a $3.1 billion reduction in our debt. As a result, only $1.6 billion of convertible debt remains outstanding and with $2 billion in cash and cash equivalents, we ended the quarter in a net positive cash position of $450 million. At quarter end, we still owned 1.7 million shares of SanDisk. Additionally, during the quarter, we received an upgrade from Standard & Poor's and Fitch to investment grade level.

Operating cash flow for the third fiscal quarter was $1.1 billion and in combination with a disciplined approach to capital expenditures with CapEx of $145 million. This resulted in strong free cash flow generation of $978 million for the quarter and a free cash flow margin of 29%. During the quarter, we made $43 million of dividend payments and increased our share repurchases to 752 million repurchasing 2.9 million shares of common stock. Since the launch of our capital return program in the fourth quarter of fiscal 2025, we have returned $2.2 billion to our shareholders by way of share repurchases and dividend payments.

Also, given the Board and management confidence in the business, the Board has approved a 20% increase of the cash dividend from [ $12.5 ] per share of the company's common stock to $0.15 per share, payable on June 17, 2026, to shareholders of record as of June 5, 2026.

I will now turn to the outlook for the fourth quarter of fiscal 2026. As we continue to operate in a strong demand and pricing environment, with longer-term visibility across our cloud, consumer and client businesses. We anticipate revenue to be $3.65 billion, plus/minus $100 million. At midpoint, this reflects a growth of 40% year-over-year. Gross margin is expected to be in the range of 51% to 52%. We expect operating expenses in the range of $385 million to $395 million. Interest and other expenses are anticipated to be $10 million. The tax rate is expected to be 16%. As a result, we expect diluted earnings per share to be $3.25 plus/minus $0.15 based on a non-GAAP diluted share count of 385 million shares.

In summary, this quarter's results and outlook highlight our commitment to disciplined execution, focus on innovation and deep customer engagements. Our strengthened balance sheet and robust free cash flow empower us to invest with confidence in the business. With strong momentum and a clear capital allocation framework, we are well positioned to drive durable earnings and free cash flow growth and create long-term shareholder value.

With that, let's now begin the Q&A. Ambrish?

Thank you, Kris. Operator, you can now open the line to questions, please. And to ensure that we hear from as many analysts as possible. Please ask one question at a time. After we respond, we will give you an opportunity to ask one follow-up question. Operator?

[Operator Instructions] And today's first question comes from Erik Woodring at Morgan Stanley.

And Irving, congrats on the really nice results. I would love if you could maybe go into a bit more detail on the specific tailwinds that HDDs and Western Digital are seeing from agentic AI. Meaning exactly what parts of the workflow in agentic are ripe for HDDs? And again, just tying that back to your comment on greater than 25% long-term exabyte growth, where does that go as a result of agentic AI.

Thanks, Erik, for the question. We really see 3 core drivers of HDD growth going forward. One that we've seen for quite a while right now, which is the ongoing storage requirements that's associated to training. So that's not going to end, training will continue. Relearning, reinforcement, retraining is going to happen and what we are seeing from our customers as they retrain and reinforce learning with these models, the quality of the model results that they get are improving. So they continue to store all the data they're generating to enable improve quality of the model. So that's one continued driver that we see.

The second driver that we're seeing is obviously the rise of agentic AI and inferencing with every inference that happens, new data is getting generated. And what's happening is that all that new jet data that's getting generated is getting started as well to both feed back into trading models and be start to support future inference references as well. So that's the second key driver of what we're seeing, both in terms of agentic AI and inferencing.

The third driver that we see for data storage for HDD is obviously physical AI. As we've highlighted before, physical AI with the limited data sets they have as well as autonomous vehicles, robotics, is using AI to generate a lot of synthetic data to further train and enable physical AI as well. And obviously, that needs to -- any data that generated out of it gets stored and feeds that whole training and synthetic data development look as well. So those are the 3 big drivers of growth that we see going forward, Erik. That's why we have the confidence to see exabyte growth growing beyond 25% CAGR going forward.

Thank you, Erik. And did you have a follow-up?

Yes. Just a quick follow-up, Kris, for you. Over the last 4 quarters, you guys have really shown a lot of gross margin expansion. I think it's 260 basis points on average over the last 4 quarters and you just did 4.5 points of gross margin expansion. For the June quarter, guidance implies about 100 basis points of gross margin expansion. So just curious if there is conservatism baked into that forecast or if there are any emerging headwinds we need to consider just given you should be accelerating cost down and you're seeing really nice pricing growth. So I just want to get some context around the June quarter gross margin, please.

Yes, Erik. First of all, I'm really pleased that we delivered strong gross margin in the third quarter and breaking into the 50% gross margin range with 50.5%. For Q4, we are guiding to 51 to 52 so some further good improvement in the gross margins. If you look at the incremental gross margins on a year-over-year or quarter-over-quarter basis, for 3 quarters in a row now and including a fourth quarter -- the quarter that we guided to, you see very strong incremental gross margins in the plus 70%, plus 75% range on a year-over-year and quarter-over-quarter basis. And so we believe that we will be able to continue to further improve gross margins. Obviously, we're only guiding 1 quarter at a time. But we definitely based on the strong pricing environment that we operate in, which is based on more and more value that we provide to our customers right, as well as a better mix as we move to higher capacity drives with ePMR, and later on, of course, moving to HAMR drives as well as more and more moving and driving adoption of UltraSMR will give us further gross margin uplift.

And then, of course, we continue to execute well from an operations point of view. And so when you put it all together, very pleased with the gross margin and the gross margin trends going forward.

And our next question today comes from Amit Daryanani with Evercore.

Congrats on a nice set of numbers here. I guess my first question is just on the pricing side. And I think on a per terabyte basis, pricing was up like 8%, 9%, high single digits in March. It's a big step-up from the flattish trends that you've seen in the last few quarters. So could you just help us understand what is sort of enabling this step up? And is it reflective of some of the LTA contracts that you've engaged in? Just trying to get a sense on [indiscernible] Irving this is the new normal on pricing as we go forward?

Yes. So Amit, thanks. Yes, pricing was up 9% year-on-year. It is like a couple of things. Obviously, the ongoing value that we're creating for our customers better TCO value as we said, our whole pricing philosophy is to be able to enable better TCO value for our customers and to be able to share in that value creation to pricing. As we highlighted at Innovation Day, as Kris highlighted, also, we said that as we move forward towards the latter part of calendar year '26, we would see pricing increase more towards the high single-digit range. And so that's what you're seeing from us.

And that's really reflective of the timing of new LTAs coming on board as far as we move forward to new periods of LTAs, obviously, we're about to deliver our next-generation of ePMR in the second half of this calendar year, there will be a step-up in capacity point that would deliver more TCO value. So therefore, we are able to share better pricing as a result of that as well.

You have a follow-up Amit?

I do. Thank you Ambrish. Maybe on just on the other side of this, cost per exabyte, I think, was down, again, 10% give or take in the quarter. Can you just talk about what is the white framework for us to think about [ copper ] exabyte declines? And should we expect a bigger step down as you transition towards the higher [indiscernible] next-gen ePMR into this year?

Yes. I think if you look at the costs down, we delivered 10% year-on-year, and that's probably the right way to look at it. Going forward, we continue to be focused on delivering higher aerial density. So that's a big cost driver. As I mentioned, we will be introducing and ramping up our next-gen ePMR in the second half of the year. We also have an increasing uptake of customers on UltraSMR, which is a good cost driver for us as well. Obviously, we get a 20% uplift on capacity without the associated cost by the end of fiscal year '27, close to about 60% of all the exabytes that we ship will be on UltraSMR.

Our teams continue to work on platforming the products to drive further cost downs and also ongoing value engineering to see how we can further reduce elements -- high cost elements of our bill of materials. And obviously, the ongoing just supply chain efficiency that we have, both from our procurement and manufacturing operations. We -- if you put that all together, we feel confident that we can continue to deliver the trend that we've just mentioned.

Absolutely, our next question comes from Aaron Rakers at Wells Fargo.

Congrats on the results. I want to go back to the 25% growth rate and think about as you see agentic AI drive incremental structural demand, maybe you can help us appreciate how you're thinking about the capacity to fulfill that demand? Is it a continued ability to just mix higher? Or is there a point in time where some capacity investment might have to play itself out?

Yes. Thanks for the question, Aaron. So maybe just to hit it at this juncture, we still do not see any need to increased unit capacity. So we have no plans for that. Our focus is really to continue to improve aerial density. So as we introduce our next-gen ePMR, which is a 40 terabyte drive, there will be a 25% step up already from our current drives, which are at the 32 terabyte capacity range. And then there's this opportunity to further mix up our customers as we've highlighted in the past as well. So we've seen an acceleration of mixing up. And as we introduced high capacity drives in the next 2 quarters so we will see an acceleration of that going forward as well.

A follow-up, Aaron?

Yes. Thanks, Ambrish. Maybe on the capital structure now with the debt for equity transfer behind you, you've still got 1.7 million shares of SanDisk. And I know that you increased your dividend, I think it was 20% with this release. First, I'm kind of curious, any updated thoughts on how you're thinking about capital return, building cash on the balance sheet versus maybe just returning what appears to be very strong free cash flow generation going forward?

Yes, Aaron. And I agree with you. We do have a very strong free cash flow and free cash flow margin the free cash flow margin last quarter was 29%. And so we're approaching our plus 30% or above 30% free cash flow margin. In terms of capital allocation and capital return to our shareholders. We are not changing our policy or our framework here. We are returning all the excess free cash flow back to the shareholders through a combination of our dividend program and share buyback program. As you have seen what we've done last quarter, we will continue to do so going forward. We are making an increase to the dividend with a 20% increase to $0.15 per quarter, and we will continue to execute on our share buyback program.

Our next question today comes from Tom O'Malley at Barclays.

Congrats on the good results. So I'm looking at SanDisk results with [indiscernible] and I'm seeing over 100% sequential pricing increases? I know you guys got asked on pricing already in the sense. But just from a [ 30,000-foot ], maybe you could talk about with the gap kind of exploding between NAND and some of the hard disk drive players. How much appetite do customers have to keep on taking pricing increases? And what's your strategy there about how much you could push given the gap is moving higher?

And then secondarily, you're hearing about some of the industry potentially doing long-term agreements where you have [indiscernible] upfront, could you maybe talk about your appetite to do that and what that would mean for the industry if you saw some of that.

Thanks, Tom, for the question. So in terms of pricing, our pricing philosophy is really to provide predictable pricing to our customers. The one thing that they appreciate and want to avoid is volatility in terms of pricing. So our whole focus is really to provide predictable pricing, as we said, as we deliver higher more value to a better TCO to higher capacity drives as we deliver more innovation in terms of performance, whether it's throughput or bandwidth enhancements, as we laid out in innovation, that gives us the opportunity to create more value for our customers to be able to share in that through better pricing.

Our hope philosophy is to ensure that we do that in a very predictable way. Why do we want to do that? Because predictable pricing enables our customers to make long-term architectural decisions. And that's really our focus, and that gives us the confidence why we are putting forward the [ roman ] that we have while we're making the investments that we're making. So we're not looking for it to be optimistic from a pricing standpoint, but really provide that predictability of pricing, enable our customers to make those long-term architectural decisions that support the structural change in the hard drive industry that we've been talking about going forward.

To your question around LTAs, we continue to make progress in LTAs, we now have LTAs to extend into calendar year '29 as well. Obviously, as we shared in the past, those LTAs are exabyte based with a degree of pricing associated with it. Obviously, the LTA volume that we are putting together for our customers does not meet the full requirement that they want and anything that we can deliver above and beyond what we call the base volume requirement that we've agreed with our customers. That's subject to a different pricing regime that gives us an opportunity to drive some incremental upside from pricing as well.

All right. No following-up for Mr. O'Malley. We'll go to the next caller please, operator. Thank you.

Absolutely. Our next question today comes from Asiya Merchant from Citi.

This is Mike Cadiz for Asiya Merchant at Citi. Congratulations on the quarter. So my first question is, would you be able to provide any color or additional color on yields and reliability. And as a result, are there any implications to the cost declines that you can think of?

Sure. I mean if you look at our ePMR products that we're shipping today and what we're anticipating as we go into volume when the second half of the year for our next generation ePMR, they continue to be in the 90% range, right? Quality also, which has been one of our key considerations remains very high. This is the hallmark of who we are as a company, high yields, known quality products, and that's something we'll continue to focus on, both in our ePMR products, which is and in our HAMR products, which is the focus of our HAMR qualification right now to ensure we provide reliability, we have the right quality, we have the right manufacturing yields as well. So that remains constant.

And immediately, in terms of current yields and quality, we don't see any changes.

You have a follow up, Michael?

Yes, I do. Thanks. So given the price differential currently between [ hard disk ] flash, would you be able to attribute the strength in HDD demand because of that? And are you seeing any architectures changing, which I think you said at this point is not?

Yes. Look, I think flash is a great technology. It has a specific role in the storage stack. We will play in slightly different spaces, right? If you look at large-scale object storage, which requires long-term retention. That's where HDD really comes to the [indiscernible] 80% of all data that stored within the hyperscale data center. If you look at workloads, that we require high [ IOPs ], high throughput. That's where flash really comes to the store -- to the fall. And even in inferencing, right, is a symbiotic relationship. The new data that's created from inferencing typically will get stored on HDDs. The vectoring data that's required for inferencing, that's actually start on flash. So it's a very symbiotic relationship. Obviously, some of the new innovations that we are delivering, our high bandwidth drives as an example, in our dual pivot technology that will improve throughput and bandwidth will continue to the performance of our HDDs and continue to deliver more value to our customers going forward.

But we don't see any, at this point, any major structural changes to architecture. But that's why, again, we want to make sure there's predictability in the pricing we provide so customers can make decisions not 1 year out, but they're making architectural decisions 2, 3 years, 5 years out as well.

Our next question today comes from Samik Chatterjee with JPMorgan.

Maybe for the first one, if I can just ask you about the quarter, you obviously had a strong set of numbers you are including both on revenue and gross margins coming in above the high end of your guide. But when I look at it, the outperformance on gross margin was a lot more rated the outperformance on the high end of the revenue, is there something more specifically going on with gross margins maybe in terms of like cost reduction, what really outperformed relative to your expectations is probably where -- what I'm trying to get to in terms of the magnitude of the outperformance is on those 2 metrics?

Yes. And again, on gross margins, there's 3 major drivers. The first one is pricing and the pricing environment, and that obviously continues to be very strong and was a little bit better during the quarter than we expected when we provided the guidance. As we've indicated, not all the pricing going into the quarter is locked. And so we do have some opportunities. By the way, not only in our cloud business, but also in our client and consumer business where we see further continued opportunities in terms of pricing.

Secondly, mix, and there, again, we're making good progress driving to higher capacity drives and more adoption of UltraSMR, and that's playing out really well. And then the teams continue to execute really well on driving down costs across the board throughout the supply chain. And so great execution during the quarter. And I expect going forward, similar levels of execution.

You have a follow-up, Samik?

Yes, please. So maybe just sort of then looking at the cost [ petabyte ] and sort of a follow-up to Amit's question earlier. You're doing this sort of 10% decline in cost per exabyte right now. As you start shipping the 40 terabyte ePMR and then eventually the HAMR drive, why shouldn't we expect that cost per exabyte to maybe declines to [indiscernible]? I'm just trying to think about the trajectory and as you ship those sort of lower cost over profiles, why shouldn't that trend sort of accelerate from where it is today?

Yes. So we -- first of all, we're only guiding one quarter at a time, but I have confidence that the teams will, again, continue to execute on those 3 levers that we have that I discussed just a moment ago. We're are ramping the next-generation ePMR in the second half of calendar year '26 so that's not that far out. As Irving already talked about that, we're feeling good about that ramp, the manufacturability and the yields there. The HAMR ramp, we're making really good progress on the qualifications. Now with 4 customers getting really good feedback from the customers, and we expect to ramp that in 2027, still a little bit of work to be done in terms of yield and reliability and quality, but good progress being made by the operations teams. There is going to be an adoption curve, right? We're not switching overnight to those new products that are being launched. And so the improvements will be phased in over the ramp period.

And our next question today comes from Wamsi Mohan with Bank of America.

This is Aisling Grueninger on for Wamsi. Congrats on the results. And just one question from me. You mentioned the UltraSMR JBOD platform as a way to broaden adoption beyond your current target base. Can you just talk about whether that primarily expands your reach into Tier 2 CSP customers or enterprise customers? And just how material could this opportunity become over the next 1 to 2 years?

Yes. We definitely see it as an opportunity to expand our reach into Tier 2 CSPs, even some of the hyperscalers in the Asia region as well. And that's one of the enablers where we are forecasting by the time we get to end of calendar year 2027. The vast majority of our key customers will all be on UltraSMR either fully adopted or materially underway in terms of qualification. And that gives us also the confidence, as I mentioned, as we get to the end of fiscal '27 close to 60% of the exabytes that we ship will be an UltraSMR.

And our next question today comes from C.J. Muse at Cantor Fitzgerald.

Curious on the agreements, particularly as they extend out into '27, '28 and beyond. How should we think about pricing? And what is embedded inside there? Is there a fixed kind of variable kind of percentages or what?

Yes. So C.J., thanks for the question. So the construct of the LTA is broadly, obviously, there's an exabyte volume tied to it. There's a pricing type to depending on the duration there may be periods of pricing adjustment as we introduce new capacity points as we introduce new capabilities. That gives us the opportunity to adjust pricing going forward?

The follow-up, C.J.?

Curious on the remaining SanDisk position now that it's beyond 12 months, is that something that is now taxable? And any sort of implications of beyond that window. And how are you thinking about time frame in terms of monetization?

Yes. So we still have 1.7 million SanDisk shares after we did the debt for equity monetization in Q3 of fiscal '26. It's our intention to monetize the remaining 1.7 million shares in an equity for equity transaction. We've indicated it's our intention to do that before the end of calendar year '26, and this will be in a tax-free manner.

And our next question today comes from Karl Ackerman at BNP Paribas.

I have one for Irving and one for Kris, if I may, but I'll ask Irving first. Irving, when would Western Digital consider adding internal head and media capacity to support these multiyear commitments from customers? For example, have you had discussions regarding prepayments for future capacity adds?

Yes. We definitely are looking ahead in media investments. As we said in the past, we're not adding -- we're not making any investments in terms of adding a unit capacity. But when we talk about aerial density improvements or increasing capacity per drive, that does involve technology investments to support new media recipes, new media substrate, new hit designs as well as the potential to increase this over time, as we've highlighted in our Innovation Day where we are able to get to [ 14 ] this over time.

Now our #1 focus is to increase the terabytes per disk to make sure that it's very competitive within the industry. And then beyond that, we were able to be able to add more [indiscernible] to the drive as well. So that's the most cost-effective way to deliver incremental capacity to our customers. So we're definitely looking, and if it makes economic sense, we'll look to add head and media capacity in terms of investments, but not unit capacity investments.

Karl, you had a follow-up for Kris?.

Yes, I may. Kris, when you note that you have agreements extending into 2028 and 2029 with your major customers, could you delineate that with respect to build-to-order and LTAs. For example, do you have build order contracts addressing much of your nearline capacity this year? Or does that extend into 2027 as well?

Yes. And so the manufacturing lead times is not about a year. And so most of the purchase orders are being placed a year in advance. And then if you look beyond the first year, we are going into those LTA frameworks that has been explained by Irving before. So there is still a little bit more variability beyond the first year.

Absolutely. Our next question today comes from Krish Sankar with TD Cowen.

This is Eddy on for Krish. Irving, when you look across your 4 largest hyperscale customers, are you seeing demand patterns that are broadly similar? Or is there a meaningful divergence in how aggressively different customers are scaling based on their AI road maps I'm just wondering if there's anything specific outside overall tax growing that is driving demand for HDDs?

No. I would say, in general, the profile is quite similar. Obviously, as I've highlighted, the demand for storage is increasing because storage is persistent, right? If you look at it, if you talk about inferencing, the resources that are used in inferencing, whether it's compute, or whether it's memory, they can get recycled, but the data that's getting generated for inferencing is not being recycled. All that data is [indiscernible] generated is getting stored and that's storage, the data that's being stored is persistent, and that is consistent with what we see with all our top 4 customers, whether their business model is in such or did their business models in advertising or in the enterprise software space is pretty consistent.

It's really this ongoing data storage requirements to support training improvements and training to support the demands of inference and to support synthetic data being driven by physical AI.

Thank you, Eddy. Operator, will you go to our last caller, please?

Absolutely. Our last question for today comes from Jim Schneider at Goldman Sachs.

I was wondering if you could maybe talk about at some point time in the future, let's say, at the end of calendar '27. When you -- what level of coverage you would expect to be shipping in terms of HAMR in terms of exbyte shipments?

Yes. We don't have a number that we are putting out there. Right now, Jim. Obviously, our focus has been to, as we stated repeatedly, is to ensure we derisk the transition for customers to HAMR, we have taken a slightly different path where we have a dual track process. We continue to deliver a density improvement in high-capacity ePMR drives even as we introduce HAMR. And that gives the customers both the confidence in the transition, but at the same time, to be able to enjoy better TCO through the higher aerial density. And when we get to the right reliability, we get to the right yields, we'll make that transition accordingly to HAMR to become -- to make it the mainstream of exabytes that we ship.

Yes. And just to add there, we indicated we have now 4 customers in qualification with HAMR. We are somewhat ahead of schedule compared to our initial plan. The feedback that we are getting from all our customers is very positive. And so our HAMR development is going really well.

And that concludes our question-and-answer session. I'd like to turn the conference back over to Mr. Tan for any closing remarks.

Thank you. As we shared today, we're really excited about the opportunity ahead of us. And the road map that we put forward in WD really positions us well to address our customer needs and the demands that they have going forward. I want to take this moment to really thank all our WD drivers, business partners for their commitment to our customers and all that they do for WD.

So thank you again for joining us here today and hope all of you have a great rest of the day.

Thank you. This concludes today's conference call. Thank you for joining. You may now disconnect your lines.

So we're going to get started guys. Welcome to day 2 of the flagship TMT conference here. My name is Erik Woodring. I lead the U.S. IT hardware coverage here.

I am pleased to be joined by Western Digital today. Irving Tan, CEO; Kris Sennesael, CFO. Before we get into introductions and whatnot, quickly, before we begin, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures for important disclosures. If you have any questions, please reach out to your sales representative. And then, Kris, I'll let you do the safe harbor agreement and then...

Yes. Thanks, Erik, for having us. And just housekeeping, today, we will be making forward-looking statements based on management's current assumptions and expectations, including with respect to our product portfolio, business plans and performance. These forward-looking statements are subject to risks and uncertainties, and so please refer to our Form 10-K and other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations.

We will also be making references to non-GAAP financials and a reconciliation to GAAP and non-GAAP results can be found on our Investor Relations section on our website.

Perfect. Thank you, Kris. So, Irving, first time together at the TMT Conference, same with you, Kris, but obviously, mainstays for everyone here. I would love if you guys could maybe, Irving, just start with maybe importantly, in an all-encompassing look back on the first year at the helm, post-split with SanDisk. I know the year was tremendously strong. But like what did you learn from the last 12 months? And how are you applying that as we look forward, call it, 1, 3, 5 years down the line?

Thanks for the question, Erik. I think the main thing we took away and what we've really seen is a structural change in how our customers really view the value of data and by association, the value of storage that comes with it, right, really driven by what we're seeing with AI boom and also just the ongoing secular growth of the cloud as well. And that's really translated into a very different dynamic that we have with them. A change in how they view hard drives as not a commodity, but a strategic element of the AI and cloud stack going forward.

That also means it's a great responsibility for us because we need to be able to deliver quality, scalable, reliable exabytes to our customers going forward. And that's really shaped the strategy that we've laid out for the company when we spun out SanDisk, which is really focused around the customer really getting much closer to the customer, just not in our traditional engagement with their supply chain organizations. We're much deeper with their engineering and technical organizations as well.

This has really helped us shape the innovation roadmap that we have going forward, which we shared most recently at our Innovation Day just last month beyond just the higher capacity drives that we're delivering to deliver better TCO, but also better performance capabilities and energy efficiency as well.

So that's really translated, that deeper customer insight has translated into faster innovation, better aligned product roadmap, I would say, an industry-leading product portfolio. And then obviously, underneath that, a lot of focus has been just really getting the fundamentals of the business right to be able to support where we want to go going forward in terms of making sure we are operationally excellent across every aspect of the internal organization. Kris has done a great job driving just the financial discipline within the company. And then obviously, we're also doing a lot to reshape our human talent capabilities to better prepare the company for where we need to be going forward to be able to better take advantage of technology advancements and automation as well.

Kris, anything you want to add on the finance side?

Yes. Maybe so as a result of this deep customer engagement and focus on innovation and execution, we have now much better visibility in the business. And what used to be a cyclical business is now a long-term secular growth business. Again, with longer-term visibility. And that also has translated in much better and stronger financial output. If you look in terms of our gross margins, our operating margins and our free cash flow margin, and there is still more room to grow from here as we continue to execute as well.

In addition to that, we have now transformed the company into a strong balance sheet company with some of the monetization of our SanDisk shares and in combination with strong free cash flow. So to me, this is a totally different business with much better visibility and stronger financial output.

Perfect. So we'll get into all of this. But I'd love to maybe start with the question that I get very often which is, we want to understand kind of exactly how AI is acting as a tailwind for HDDs. And what I mean by that is it's clear that leveraging more data, storing more data is going to be a broad tailwind. But what are the specific use cases for HDDs in an AI world? Can you give a few emerging examples, how do we size those opportunities? And is there a rule of thumb for us to do exabytes per megawatt, exabytes per CapEx dollar, exabytes per GPU, something like that. Just would love if you could dig into that.

Yes. I think providing a rule of thumb, it's always a bit of a tricky one, Erik, because it's very much use case dependent, right? What we do know for a fact is that obviously, as LLMs transition more towards multimodal, being multimodal in nature, especially as you transition to video, which requires order of magnitude, more storage requirements than still images or tech. That's going to be a big growth driver for storage going forward.

But if you look at AI workload value chain here, if I simplify it, there are really 4 pieces to it, right? That's data ingestion. That's just storing all the metadata that's being created, whether it's in the cloud, or whether it's in AI. And that's where I would say HDDs really come to the fall because of the superior economics and the reliability. So a lot of data ingestion, the basic storage of metadata is all being done on HDD.

And then the next piece of the value chain is really data preparation, right? As you prepare for model training, what do you need to do with data? So model preparation and model training, they're sort of symbiotic in nature. What do our hyperscale customers do? They take all that metadata that store on HDDs in order to prepare for the models that they want to run, they cash that data into SSDs to be able to deliver the velocity of data flow to feed the GPUs. But again, HDD still play a very prominent role in that, right? And once those models are developed, they require that those models and associated data that was used to train those models are required to be stored as well. And what is the primary storage media that they use, again, it's HDD.

And then we go to the last piece of the AI sort of workload value chain, which is inference. And does this belief that inference is purely about SSDs as well. That may be true of today because the look back in inference is relatively short in terms of time. But as you go forward, as inference look backs get longer and longer and longer, you can't store that amount of data on SSDs because it's just not economically viable.

So the similar behavior that we see in model training, right? Again, having all that data for inference stored on hard drives, prepping that metadata into SSDs to be able to support inference workloads is what we are seeing happening going forward as well. So HDDs play a role across the entire value chain. Obviously, a much stronger footprint in the ingestion and storage piece and then it gets a bit lighter as we move towards inference side. But it plays across the...

I want to touch on video. This feels or it seems like a really exciting emerging opportunity. 3 years ago, YouTube was doing 2 million uploads per day. Today, it's 20 million uploads per day, over 20 billion videos on the platform. It still feels like things like Google Veo or Sora are early days. But can you maybe just help us understand how important you see video, and again, kind of an extension multimodal as kind of the key driver as we look forward as maybe the most exciting part of what's emerging from AI.

Yes. I mean video itself, as I mentioned, requires an order of magnitude more storage, somewhere between 100 and 1,000x more between tech and still images, right? So it's definitely going to be a big driver of storage. And as you mentioned, Veo and Sora still in its infancy. We're talking about 15, 20-second clips. You've seen some of the interesting capabilities that have come out, right? If you guys have seen that sort of fight between -- I think it was Tom Cruise and Brad Pitt, as well, right? So that actually you start to see AI-driven video formats getting longer and longer, and there is potential for it to be short films or even full blown films down the road. So that's definitely a big driver of growth.

But we're talking a lot about the consumer aspect. I think what's sometimes forgotten is industrial applications as well. If you take autonomous vehicles, as an example, right? Literally, the video feed from every car right of an autonomous vehicle that's flying the roads here in many parts of the U.S. and the world is getting stored both for compliance reasons, just in case the vehicle meets into an altercation, there's ability to go back and reference what's happened.

But also that video is being stored to be able to create synthetic data to train models as well because you don't have a real life situation very regularly to be able to train a model if you have a power outage where the traffic lights go down or if you have a police stop, what does the vehicle do? You're able to simulate a lot of that by using the addition -- the video that you stored using AI to generate synthetic data to create these corner cases to be able to train these models, right? And so we think -- and that applies to whether it's autonomous vehicles, I just came from visiting our facilities in Thailand, where our latest facilities are very automated. Again, we're using a lot of the video feeds to train the tools and the robots and the machinery that we have to run a fully automated facility that flights out. These industrial applications that are also very video-rich are requiring tremendous amount of storage requirements.

Okay. That's really helpful. As a follow-up on this, obviously, we've heard you guys demand is extremely strong. Can you maybe just help us give an update on visibility where it stands today? And underlying that, a question that I get is always about overordering in this type of environment. How do you -- as you think about that elongation of visibility also protect yourself from that risk of overordering and what it could do in the future?

Sure. In terms of visibility, we have 1 of our top 5 hypescalers that has given us orders all the way through to calendar year '28. Two of them have given us orders all the way through to calendar year '27 and we're pretty much for our top 7, we have firm POs for all of calendar '26. So I would say visibility is very strong. This is a big change from even just probably a year, 2 years ago where our agreements typically span 2 quarters, right?

Beyond that visibility, the commercial constructs that we have for these orders are much more robust financially as well. So there's obviously a lot of downside protection in terms of the financial risk that we are exposed to as well. And we also don't just use our customer demand signal as the only driver of the supply decisions that we make. We use it just as 1 data point, right? We also look at our own internal assessments. We run our own ML models to determine what we think that supply is. We look at our supplier base, and we have a triangulation of factors that determine sort of what supply base we should be creating to support that customer demand.

And just given the tightness of the market and obviously, the kind of bullishness behind your customers' views on the long-term growth of data, can you maybe just help us understand the top customers that don't have these commercial agreements beyond calendar '26. Is that a WD-driven decision? Is that a customer-driven decision? Why do some customers decide to give you 3 years of commercial agreements why others might not?

I would say, we're in active discussions with the remainder of them. So it's -- those are the ones that I've highlighted that we've closed commercial contracts with them, obviously. The rest of them are in active discussions with us to be able to secure supply in the outer years as well. I didn't address one of the important points that you raised. Is there a concern about customer hoarding or inventory buildup. We actually do not see that at all. This is quite different from the situation that we had sort of post-COVID where there was a big buildup of inventory during the COVID period. Our customers undertook a lot of software enhancements to improve the utilization of their storage asset. All that's been taken into account, really pretty much everything that we're shipping to them is getting deployed as quickly as they can as well.

Okay. And maybe last question on this topic is, again, just longer term, when you have conversations with your large DSP customers, I think you've talked about a bit of a different relationship now it's less transactional. It's more partnership-based. Do you believe that this kind of demand planning and visibility is different now? Has it fundamentally changed if we look back versus kind of that '22, '23 era?

I think the first thing to qualify is no demand planning purchase perfect. So will there be a degree of error in it, absolutely right? But I think the big difference is, one, we are talking about a much longer-term horizon. What's giving us a lot more confidence in the demand signal of our customers, it's also their willingness to actually sit down and show us not just a demand signal for storage, but the roadmap of applications that they are looking to launch into the future and how those applications are translating into storage requirements, specifically for hard drives. That's giving us a lot more confidence.

And that reflects the fundamental changes as I mentioned from the very onset of the discussion, where there's a structural change in how our customers view the value of data and storage. And they're beginning to realize if we -- they don't have the storage capability, they're able to monetize, they're able -- the ability to compete in cloud and AI is going to be severely impacted. And that's a big change that we are seeing.

The other change is also we've done a lot of education to help our customers understand that as it drives that we produce increasing capacity going forward on improved performance and energy consumption. The lead times for these drives are getting longer as well. There's a lot more sophistication and complexity that we're building into it. So they are aware that if you take the longest item in the lead time, which is the head wafers we produce, it takes us roughly about 9 months to produce one hit wafer and another 3 months to convert that into a head reader, right? So even if you give us a demand signal today, you're not going to see any change until 12 months down the road.

So a lot of that education has really changed their perception. And again, it goes back to what I mentioned at the very onset, this fundamental shift in terms of how they see the strategic importance of hard drives, particularly around associated with the value of data in the world of AI and cloud going forward.

And Erik, maybe just to add there, keep in mind that 90% of our revenue today is linked to cloud, right? If you will look back 3 or 5 years ago, that was probably less than 50%, and the other 50% was our consumer business and our client PC business which is now down to 10%. Obviously, there is more cyclicality and volatility in consumer and client than there is in our cloud business.

So I want to touch on the point, Irving, that you made about cycle times, and I know this is a topic that has been well covered. No plans to add greenfield unit capacity. And just making sure that still stands today, that's still the view. And I think some people maybe try to tie looking to external vendors for head and media, additional head and media capacity as something that could imply greenfield unit capacity additions from your end. Would just love if you set the record straight about what you're doing with unit capacity as you look out right now?

Yes. So thanks for the question, I get that a lot. Let me be very, very specific and explicit about it. We are not adding any drive unit capacity, right? What we are doing is to see how we can debottleneck our existing investments, how we can leverage more automation to obviously improve throughput, which is just part and parcel of running a very efficient business that we're doing. And that's kind of why we were in Thailand looking at what opportunities that exist. We have said that we are going to continue to make investments into our head and media capabilities because. That's a big driver of areal density capability going forward, where a lot of the investments are more targeted at how we can launch the next generations of head and media, which drives higher areal density at the end day.

In relation to our approach to using third-party component providers in head and media. That is actually part and parcel of our -- what has been our ongoing innovation framework, right? Yes, we are vertically integrated as a company, but we actually do utilize third-party head and media for 2 purposes. One, we are able to leverage core innovation with them that is beneficial for us and also challenges our internal teams to do better. So we're constantly always benchmarking our internal assets. This is what we can get from the market, looking at whether the combination of those assets to drive higher areal density. That's something we can do on an ongoing basis. We've always had that.

We also have used them to some degree in our supply chain, right? Even some of the -- a portion of the drive that we use today already incorporates some of those third-party head and media. And that gives us supply chain resiliency as well and create some degree of buffer for us even as we drive some degree of innovation in our supply chain that takes some capacity.

And last maybe question before we get into kind of pricing and innovation is, so we take all this together, your Innovation Day, maybe 3 weeks ago, we talked about -- or you guys talked about how that Nearline outlook has changed. You now see mid-20 exabyte CAGR. That was kind of your bull case assumption, maybe slightly higher than your prior bull case assumption. Just help us understand what has now the kind of math that you've gotten to get there, the thought process around driving there? And could we be sitting here in a year and discussing the new bull case, the new outlook for Nearline exabyte shipments?

Well, I'd never say never. I think Kris would probably agree, we feel quite comfortable with the 25% CAGR on exabyte growth. Again, our focus is to be able to support that. Exabyte growth through areal density improvements, not by adding more unit capacity. And I think Kris did a great job at Innovation Day, really showcasing what can be done.

So if you take what we are delivering to the market today. Last quarter, our average near line capacity is 23 terabytes. We are shipping today a 32-terabyte drive. If we can shift all our customers to that 32-terabyte drive, that's a 40% increase in exabytes into the marketplace without having to add any unit capacity. If you fast forward to the end -- the second half of this year when we're going to introduce our 40-terabyte SMR drives on the ePMR recording technology. That gives a 75% capacity increase versus that 23-terabyte average that we're shipping today.

So we have that capability, right, by getting customers to adopt technology faster. We've also seen a very rapid adoption of our UltraSMR capability. We are consistently over 50% of the bids that we're shipping every quarter on UltraSMR already, and we see that increasing, over time, we have our top 3 largest customers fully onboard on UltraSMR, and we have 3 more in the pipeline coming up as well.

Okay. Great. Just, Kris, I'm going to turn it to you just on pricing. At the end of the Innovation Day, I thought maybe one of the most surprising kind of financial updates we got was that as we look to kind of the remainder of calendar year '26, as we think about price per -- blended price per terabyte, that can grow mid- to high-single digits. And that's kind of inclusive of the pricing that you've been able to set in these commercial agreements to date.

You -- after that, you mentioned stable pricing beyond calendar '26. I'm just going to put you on the spot a little bit and just say, why can't that be better than stable, just given how strong demand is, where supply stands, what the substitute technology, potential substitute technology is doing from a pricing standpoint. Just the pricing outlook beyond calendar '26 as you see it today? Why could that be stronger?

Yes. And so just to reiterate, right, for calendar year '26 for all 4 quarters, we do see mid- to high-single digits year-over-year ASP per terabyte increase. And then beyond that, I indicated that the pricing environment is stable, meaning that we will continue to see flat to slightly up pricing on a price per terabyte from the higher levels that we reached in calendar year 2026.

Now can it be better than that? Sure, right? But the further out there, I'm not going to commit on further price increases. Yes, we are looking at pricing from competing products, but guess what are those prices going to be in '27 or '28.

All right. Okay. Let's turn to the Innovation Day. One of the questions -- main questions I got after Innovation Day was just the desire to dual track ePMR and HAMR up to at least 60 terabytes, right? Obviously, HAMR will go well beyond that. What are the benefits that you get from kind of selling both platforms? And just help me understand, are there added costs that we need to all think about as you ramp these 2 platforms simultaneously.

I think, first and foremost, it's very important to qualify that we are firm believers in HAMR. HAMR is important for us to be able to scale capacity points to 100 terabytes and beyond, as we've laid out in innovation day. But we're in this transition period as we're moving from, one, a very established, very reliable technology to a new one. And with any technology transition, there's always a period of learning in optimization that needs to go on, whether it's yields, whether it's reliability. And as I mentioned from the very onset, our customers depend on us to be a reliable quality supplier that can deliver exabytes at scale and support them with our TCO. So we take that very seriously, and that's kind of why we sort of dual track this, right?

And if you look at our ePMR portfolio we are -- we've announced the next generation of it, which we are qualifying with 2 of our largest hyperscale customers right now. Historically, we only take 2 quarters to qualify it. So we anticipate that we will start to ship in volume in the second half of the year. This is at the 40-terabyte level. And as we laid out in Innovation Day, we can probably get by 2028 to 60 terabytes. So it's roughly on a 12-platter platform So that's roughly 5 terabytes platter. And then we will also introduce HAMR at the same capacity points in parallel. This gives our customers the flexibility to transition from one technology to the other.

The other important point that we made in Innovation Day that may be lost sometimes is that we are making the drives interchangeable. So our customers are able to plug and play a HAMR drive from WD or ePMR drive from WD into the same rack without any software changes or operational changes on the other side. So the whole focus is to make the whole transition seamless, reliable and to be able to deliver scalable exabytes to the customers to support their needs.

We don't see it being an operational cost adder because the approach that we're taking is all the mechanical designs for the 40 terabytes and above ePMR will be the same mechanical designs that will be used in our HAMR platforms, right?

The only difference in the HAMR platforms will be the fact that we have a laser of which we have a third-party provider, and we have our own internal capability and the transition to glass. But the mechanical design, the firmware for both will be identical as well. So it's really leverageable across 2 platforms.

Okay. And just touching on HAMR quickly. You announced at the Innovation Day, another customer, another large CSP customer qualifying on HAMR. I think the time line is still we should expect volume shipments in calendar -- the first half of calendar '27. Just provide the latest update. I just want to make sure that's the latest for all of us.

Yes. So we have 2 CSPs, 2 of our top 5 customers that have really started HAMR qualification. Again, we're anticipating the ramp will be in -- which we pulled forward 6 months earlier. So we were anticipating to start qualification in the second half of '26. We pulled that forward to the first half of '26. In fact, we pulled it forward to the first quarter of '26. So we have 2 CSP customers qualifying HAMR with 2 CSP customers qualifying the 40-terabyte UltraSMR EPM platforms as well.

In terms of the ramp for HAMR, we're still targeting that to be first half of calendar '27. We are very comfortable with the areal density improvements that we've been making in HAMR. The focus on the qualification is really ensuring that we get the same reliability, quality and yields that we can get for ePMR.

Because ultimately, the approach that we've taken is, one, to derisk the transition from our customers, give them the right reliability, give them the highest capacity points in the industry at scale to give them the best TCO but ultimately also gives us as a company the best economics. So they were able to transition from one recording technology to the other and ensure that transition is margin neutral to accretive going forward.

Okay. And maybe last innovation question before we turn to Kris on some of the numbers is, you announced a number of other kind of technology innovations, high-bandwidth drive, dual pivot. You announced power-efficient HDDs. Maybe just collectively, the thoughts behind all of this innovation. I know they maybe address different parts of the market what are the intentions as you roll out these different technologies? What are they addressing that is so critical?

So the introduction of the high-bandwidth drives really address the throughput challenges that hard drive SaaS vis-a-vis SSDs, dual pivot addresses the IOPS limitations that HDDs have versus SSDs. And as we position HDDs more and more towards AI workloads, these are important factors that our customers have told us, and again, goes back to the whole pivot that we've made as a company to be much more customer-centric.

So you need to solve these issues for us, especially when you get to the 50-terabyte capacity level, bandwidth and IOPS becomes an issue. And so we've been very busy working on it. We actually have already engineering samples for a high-bandwidth drive in a customer. We anticipate that for high bandwidth drives. On average, about 20% of customers will take it up, some were much higher. Some will be lower depending on the workloads that they run.

But again, it's all software related. So anything we do there will be accretive to the business. It's not factored into Kris' financial model that we shed in the Innovation Day. And then on power efficient drives, again, it caters to a specific segment of the market where there's this opportunity to say I can trade off some performance because I don't need the throughput. I don't need IOPS, but I can save energy, which is more important to me. So what we're doing is to give customers a much more granular ability to sort of segment their storage workloads by creating these capabilities that really are attuned for the AI workloads going forward.

So we're obviously focused on areal density, which clearly is a cost benefit for you, cost PCL benefit your customers you've been able to kind of cost down per terabyte at a rate of around 10% per annum is -- maybe, Kris, this is for you, but is that how we think about cost downs annually going forward? Could those accelerate as you move into 36s and 40s and beyond? Just how do we think about the trajectory of cost per terabyte as you bring these new higher capacity drives to market?

Yes. You think about it the right way. We have been executing really well on driving down the cost per terabyte we probably have the lowest cost per terabyte in the industry, which translate, of course, in leading gross margins in the industry as well. But obviously, we're not going to stop there. We are focused on driving down the cost in our global manufacturing footprint. We're also collaborating with our supply chain partners and taking out costs through value engineering. And then probably most importantly, as we move to higher capacity drives through areal density improvements, you will see a further improvement in the cost per terabyte.

Okay. Perfect. So we bring all of this together, the model, obviously, is extremely impressive, 50%-plus gross margins, 40%-plus op margins, 30%-plus free cash margins, kind of a target of $20 of earnings. I realize this is going to be -- it sounds like an aggressive question, but rational oligopolies in my mind, are extremely powerful market structures. What's the time line as we think about that path to that $20? Is there a path beyond $20?

Yes. So the time line was the next 3 to 5 years but you explained it really well. So those targets were not ceilings, right? Those targets were almost like floors. We believe we -- once we hit those targets, we will be able to continue to operate the business at or above those targets.

Okay. Perfect. So just to be very clear, and I think that's a very powerful message, which is, as we think about history, and maybe through cycle margins, through cycle earnings power, the kind of point that you're making is there's been a structural change in kind of demand signals. And as you've been able to innovate, as you've been able to kind of refocus on operations as a stand-alone HDD focused company, we believe that these are kind of the floor as we think about them as we move forward. That's kind of the message.

Yes. Yes. I think that's fair. And so again, we don't see the cycle turning anytime soon based on the strong visibility that we have from our customers. Now is there in the future going to be some periods where the growth is going to slow down, I think that is possible. But again, it's all about being disciplined from a pricing point of view, continue to innovate and drive down the cost per terabyte, and as a result of that, I do believe we will continue to be able to continue to operate the business at or above those targets.

Awesome. So I want to make sure we touch on capital structure because I think it's incredibly important. You monetized some of your SanDisk holdings. You've retired, I think it's $3.1 billion of debt, that's your term loan and senior unsecured debt. You're left with about, I think it's 1.7 million SanDisk shares, and you have your $1.6 billion convert. What's the path forward for here when we think about the capital structure, either addressing the convert? What do we want to do from a leverage standpoint? What -- where do we go from here?

Yes. So the only thing that's left right now in terms of debt is the $1.6 billion convertible debt, which is matured in 2028 with callable in November of 2026. We still have 1.7 million of SanDisk shares, which is on or about $1 billion, give or take, in value. And we've indicated that we want to continue with the monetization of the remainder of that stake potentially through an equity-for-equity transaction which basically will further reduce our share count.

We will evaluate once that is all done, we will evaluate the balance sheet and the cap structure and see how we will operate the business going further. But in the meantime, of course, we will continue to generate very strong free cash flow. And all that free cash flow is being returned back to the shareholders through a combination of our dividend program where we still have plenty of room to grow. We are committed to our dividend program and plenty of room to grow as well through our ongoing share repurchases and at Innovation Day, we announced a new $4 billion program that sits on top of the prior $2 billion authorization. And we will -- there's no hesitation there. We're definitely in the market today and continue to return the free cash flow back to the shareholders.

I love that. So just with the remaining time we have, maybe Irving, Kris, you can tag team this question. But -- just what's the final message you want to leave us with as we end the session, maybe things that could be undervalued or underappreciated as you think about Wall Street's perception of the story, I just want to give you the kind of final word for everyone here.

Yes, maybe from my perspective very quickly, I really do feel that, as I said, there's a renaissance of structural change in terms of how storage is perceived and the value of it as a result of the value of data, right? And that -- you see that from the visibility that we have.

Second is, what I think that's underappreciated is actually the innovation that we have, right? There's been a lot of focus on some of the innovation in our industry. But what we've shown is we can innovate on capacity, we can innovate on performance, we can innovate on power efficiency. But equally important, what you've probably seen from us over the last 12 months is not only about innovation. It's about this relentless focus on execution, right? And we've been delivering on everything that we said that we would do since the day on the spin, and that's something that we continue to be very focused on not just laying out an innovation roadmap, but making sure we're consistently executing towards it. And I think we do that well. We'll bring a smile to Kris' financial model and maybe ask something beyond as well.

Thanks for hosting us Erik.

Awesome. Thank you, guys.

Thank you very much.

Good morning. And Chuck, you had me at Ambrish, but you just reminded me how I should pronounce my own last name. So thank you. Look, what a fantastic attendance, especially in this cold fridgeid in New York. So we really appreciate all of you, not just those who are here. We also know, of course, some of you have made the trip from outside of town. So thank you for doing that.

And so on behalf of the entire WD team, I would like to provide a very warm welcome to everybody, also to those who are joining us on our live webcast. So now you all know that before I say another word, you know what I'm going to do. Yes. I'm going to turn to our safe harbor statement. So please bear with me and do pay attention to what we have on the slide here. Today's discussion includes forward-looking statements. I've not memorized it. These statements are based on current expectations and involve risks and uncertainties that could cause actual results to differ materially. Additional risks are described in our SEC filings, including our Form 10-K, which was filed on August 14, 2025. This presentation today also includes non-GAAP financial measures. Reconciliations for historical periods are available in the appendix and on our Investor Relations website, plug for that website, www.wdc.com. So great. We got that covered. Let me now turn to the agenda. So you probably did not expect us to be here soon because we were only here about a year ago.

The New York City winters cannot keep us out. And we also have quite a few new developments that we want to share with you, and we thought best to do that in person. So here we are. Today, we'll start off with Irving Tan, our CEO, who will really set the stage for the discussions today. We will then have our Chief Product Officer, Ahmed Shihab.

And we, from IR, from day 1, we have been waiting to get Ahmed in front of everybody. So no pressure, Ahmed. She got it. And after that, we have Kris Sennesael, our CFO. I think Chris needs no introduction. Everybody knows Chris. Following the Q&A, and this is important -- sorry, after the sessions, we will have closing remarks. After that, we'll have a Q&A session. So all your questions will be covered. So please hold on to your questions until we get to that session. And very important, once we are done with the Q&A, we will have lunch concurrently, we'll have our demo room open, which is right behind me. And so we have some of our latest innovations there, and these are working products that are in our customers' hands. So this is not a demo per se.

These are products in our customers' hand. Equally important, we have WD, some of our brightest folks here. So please avail of this opportunity to talk to them. Now you all talk to Irving, Kris and I and Amitesh all the time. This is your one chance, and I say only for today, expand your reach. There will be no business cards being handed out there. But -- so without further ado, I would like to welcome on to the stage. It is my honor and my pleasure to welcome our CEO, Irving Tan.

Thanks, Ambrish. And it's wonderful to see all of you again. It's been just a year since we were last together. But what a year it's been. If you recall back in February of 2025, when we had our Investor Day, we shared with you the strategy of the company, what we were going to do. And over the last 12 months, we've been singularly focused on driving execution against each one of those 6 pillars that we laid out in our strategy. And I thought I'll spend a moment just giving you some examples in terms of what we've achieved.

We've not completed it, but we are well on the journey. In terms of really focusing on our customers, we've been focused on building trust with them. We've reorganized our teams to better engage with them across sales, engineering, product management. That's translated into our customers having a lot more confidence into our products, our road maps.

And you can see that from the long-term agreements that they signed with us, some extending all the way through calendar year '27, one even extending all the way through calendar '28. But beyond that, they've given us deep insight into what are their needs for AI and the cloud, not only of today, but for tomorrow as well. And a lot of that insight is going to be translated into the products that we're going to deliver both now and into the future. Ahmed is going to give you some insight into that. We continue to execute on delivering industry-leading technology, not just the highest capacity drives, but capacity at scale.

And just last quarter, we shipped our industry-leading 32 terabytes, 3.5 million units of them. This quarter, we're projecting to ship just under 4 million units. That's one of the fastest ramps of a high-capacity drive ever. And we've also pulled forward the qualification of our HAMR products, and Alper will give you more insight into what's happening there as well.

But it's not just about delivering products. We are also innovating in terms of what we want to do to deliver greater innovations into the drives that we are producing for today and tomorrow, but also innovating to deliver growth -- new growth vectors for the company going forward. We shared a bit about what we're doing in lasers to better improve manufacturability, performance in our HAMR products going forward. But also, we're working on new areas of growth in terms of what we can use our magnetics capability and HAMR fabrication for the next wave of technology growth like in quantum computing as an example.

We've also been executing in terms of what we need to do to be able to deliver the exabytes that our customers want at scale, reliably. Our operations teams consistently deliver drive yields above 90% and continue to do amazing things in terms of improving productivity to eke out even more capacity and units within the same investment footprint that we already have. And one of the biggest things that's sometimes underappreciated is the cultural change that we've delivered through the company.

We've really harnessed the deep experience that we've had in building hard drives over 55 years, but combine that we're bringing in new talent, engineering talent that's come in from our customers who really understand what hyperscalers need in terms of their architectures today and tomorrow, feature functionality that's really critical to them. We've also brought in new leaders across different parts of the company that are experts in their respective areas within HR and finance. But most importantly, when you bring all this together, what have we been able to achieve? Really strong financial results.

If you look at the revenue growth that we've delivered over the last 12 months, the margin appreciation, ultimately, what is done is to translate into very strong free cash flow. And in the last 2 quarters, we returned 100% of the free cash flow that we've generated. And we also were able to bring our net leverage well below the 1 to 1.5x that we laid out last year.

But it's not just executing on our strategy and what we feel are the right milestones that we've been hitting. We've also been recognized externally. We got added to the NASDAQ 100 in December of last year, a great milestone for us as a recognition of the financial performance we've been delivering. But beyond the financials, we are also very focused on being a good corporate citizen. We continue to deliver on our sustainability goals. We've also been recognized with the cultural change as one of the best places to work in America as well and something near and dear to my heart. We continue to do it in a very ethical way. Yesterday was also a very important day. We launched our new brand and logo. It's not about a brand and logo. It really signifies our transformation to becoming a data-centric company. So I'll take a moment to just explain to you a bit of the imagery. If you look at the lines on the left of you, that signifies hard drives, but more importantly, hard drives being deployed into data center racks.

That's who we are, a data center company that's at the heart of AI and cloud today and going forward. Secondly, if you look at the colors. This represents a company in motion, strategically, operationally, financially, culturally as well. And the simplification of our name to WD just reflects what all of you and our customers call us. We're simple, we're easy, we're WD. So with that, we're very much in the heart of cloud and AI. And cloud and AI is not only a big opportunity today, it's an opportunity that's accelerating at a very rapid rate. It's being driven by AI that's transitioning to large language models that's multimodal in nature. Video is going to be a big driver of storage requirements going forward. But that's just one part. That's been driving a lot of the growth with model with training and development. As we move into the next phase of AI where AI is going to be monetized through inference, inference is going to generate significant amounts of data. And guess what, every query that you put in, every prompt, that history has to be sought.

That's going to not only generate significant amounts of data, it's going to also generate significant amount of storage coming out of that data as well. That's just the second growth driver. There's a third one. If you look at some of the innovations coming out in terms of autonomous vehicles in robotics that's harnessing the power of AI, a lot of that is also going to be powered by multimodal LLMs that's going to be able to deliver the intelligence, the ongoing learning that autonomous vehicles and robotics need to bring to the market. And if you put this all together, what we see is a demand for storage exabytes growing at a CAGR of 25-plus percent over the next 5 years. And even with this growth, we feel very confident that HDDs will represent 80% of the storage media that's deployed within a hyperscale environment, the continuing superior economics that we deliver and some of the performance that you'll hear from Ahmed today is further going to strengthen that capability. So it's a very exciting time for us.

And as more data gets generated as a result of AI, it's very clear that HDDs will be the predominant data storage media for raw data, content storage, new content generation. And with some of the new capabilities that we intend to bring forward through performance, extend maybe into the opportunities into disaggregated storage as well. But don't just take it from us. Let's take a moment to listen to Karthik from Meta in terms of what they are looking for from a HDD technology partner to support their business, not only today but into the future as well.

[Presentation]

So you heard from Karthik, what do they want? At level, they want capacity to be able to meet the growing demand for storage. What does that mean for us? The ability to deliver the highest capacity drive independent of recording technology. Second, they want us to deliver that high-capacity drives in a reliable manner that's consistent with the high quality that they've experienced from us in the past. Third, it needs to be done cost efficiently, as he mentioned, because HDDs are the bulk of storage.

So having the right TCO model for them is really fundamental and important to their business model. Fourth, it's not just enough to deliver the highest capacity drive per unit. You need to be able to do it at massive scale because they're betting their ability to meet storage demands on us as a technology partner.

Fourth, it's not just about capacity and scale. That's hygiene. We're just meeting the needs of what they need today to support storage requirements. As we move into the world of AI workloads and new capabilities that they're innovating on today, what they also need from us is how are we delivering greater performance in our drives that are suited to the AI workloads of the future. And also, one thing we know about AI, it's pretty power hungry. So how can we contribute to power efficiency that's helping to offset some of the energy requirements that AI requires as well.

But ultimately, what they want is when you bring this all together, how do we give them the capacity that they need, the quality that they're used to, the cost efficiency that they derive from HDDs and new capabilities in a way that there's no disruption. No disruption to their operations, no disruption to their software stacks and the innovations that they're driving within their business.

So they can count on every drive that we produce to be plugged and played in very reliably. So what are we doing as a company to deliver to our customers' needs? You're going to hear a lot more from Ahmed. But just let me set it up for him across 4 key things. First and foremost, what you're going to hear is that we're going to deliver the highest capacity drives at scale with the right reliability and quality that our customers expect from us. Second, we're going to ensure that we get them smooth transitions across recording technologies. There's a lot of talk about ePMR and HAMR. The fact is, if you ask our customers, they don't really care about recording technologies. It goes back to those 6 things we mentioned, high capacities, reliability, scale, cost efficiency, performance and efficiency, power efficiency and then ultimately, minimal disruptions.

And that's what they want. Third, we're going to focus on delivering innovations that are required to power the AI workloads of the future and to address some of the limitations that HDDs have. Higher bandwidth requirements, higher throughput requirements are what AI applications need going forward.

And you're going to hear a lot more from Ahmed in terms of how we're going to address that. Last but not least, how do we make it easier for our customers to adopt technology? Not everyone is blessed with some of the big hyperscale resources, big storage teams that are able to build storage racks and to be able to design storage architecture. So we're also making it easier for the next generation of AI players, the Neo clouds to be able to adopt our hard drive technology in a very easy way without having to have this massive investment into building out storage teams.

But we're not -- what you're going to hear today is not the end of our innovation journey. It's just the beginning. We have a rich set of core capabilities and IP that we're continuing to build on. Just the new laser capability we just announced is one aspect. of it. Well we're continuing to focus on our rich IP capability across magnetics, material science and now photonics has been added to that list.

And we have deep systems expertise that we are able to bring together hardware, software and firmware to deliver reliable quality systems to our customers. And this is something we're going to continue to innovate on, and we'll be sure to have innovation days into the future where we can keep you updated in terms of what we're doing. So ultimately, what can you expect from all of us going forward? It's 2 things, 2 very simple things. If you forget anything I said, just remember these 2 things. First, the focus on execution that you've seen from us over the last 12 months, that is going to continue. That is going to be the mainstay of what we continue to do. We're going to stay laser-focused on continuing to execute for our customers and for our shareholders.

Second, we're going to combine that strong execution focus have a massive acceleration of innovation. So you're going to see a lot about what we are delivering in terms of innovation today. Ahmed is going to walk us through a lot more about that. But ultimately, what are all of you here interested in?

How are we going to bring that continuous execution focus and speed of innovation together and what does that mean to the returns and the value that we can create for our shareholders. And Kris is going to walk us all of us through the updated financial model that we see that's derived from all of these things coming together. So with that, I'm very pleased to be able to introduce our Chief Product Officer, Ahmed Shihab on the stage.

Thank you, Irving, and good morning. Good morning. There you go. That's better. We need to warm up in this cold weather. I'm Ahmed Shihab, the Chief Product Officer here at WD, which has been a culmination of a long journey for me that has started when I was actually a little boy at my father's carpentry workshop. Through him, I learned that turning design ideas and concepts into products customers loved is a wonderful thing. Delivering what customers wanted, kept them coming back for more. And sometimes they brought their friends.

It's really ignited a passion for me, building amazing products that customers needed and they kept coming back for more. Because when you delight customers, they like you. They buy from you. They keep coming back for the product. It's good for them, good for us. At AWS, Microsoft, that passion became customer obsession. It was rooted in scale and disciplined execution. At WD, we are customer obsessed. And we make amazing products, marvels of science and engineering. For me being here, it's being home. We care about our customers. We care about our products. What a moment to be here. AI, much like the Internet and mobile and cloud has shaped our lives for the last 30 years. AI is set to do the same thing and transform everyday experiences. We all use it. I know I do. It's in your car. It's in the factory. We ask it questions constantly. It's become indispensable. It's fast, it's growing, and it's here to stay. But AI is not one thing. It's not magic. It's built of many workloads.

To answer our questions, AI needs data. That data is ingested to ingestion workloads in the cloud that collect data, whether it's sensor data, whether it's data from your car, whether it's data that has been stored in the library of Congress. It's data that just needs to come in with documents. That data goes into towards customers who use object stores and parallel file systems to store all this data called the cool tier.

Lots of data gets stored on that. But that data by itself has to be prepared, has to be ready. You've heard words like vectorizations and a bunch of other technical terms. What it means is that we clean up that data and make it ready for training. We read the raw data, we process it. So that warms up the workload. It needs more transactions, it needs more work. And all that is handled through the object store software.

Then comes the big workload. We have to train the models. We've heard a lot about training. It's where the GPUs come in. And all that data is pipelined from the object stores through SSDs that sit in the GPU boxes to then teach the model how to think about the world, to extract all that data. You need performance in that space.

But AI becomes profitable through inference. That's the part we all interact with. It's the part that answers our questions. needs a lot of performance because it's dealing with millions and millions of users all at the same time.

But it also needs to understand who you are because the more you use it, the more data it generates about those interactions, the more it gets to know you. And that data is then generated and then stored again. So the more we use AI, the more data is generated and the cycle continues. I'm sure you have the experience that it feels like AI is learning over the last few years. I certainly had that experience. I had the privilege of being an early user of Copilot at Microsoft. And you can literally watch it learn day by day, get better day by day. We started to think of these workloads like a brain. We all learn and it was learning from data. But much like we learned and trained our brains at school and universities, we typically went to a library to learn.

One way to think about how storage and all these complicated technologies work is that AI's brain goes to storage to go to the library. This library in the cloud is 80% hard drive. It's just a fact based on our experience. Customers want to grow that library, want to add to it. Our job is to grow with them.

But to do that, we have to understand what our customers need. We have to understand them. They have to be the center of how we think. It's not just what they want, but also what they need now into the future because when we understand their needs, we can build the technology they're going to need for the future, so it's ready when they are. Customers trust us. They trust us because of our 55-year history of delivering innovation with quality and reliability. They trust us and they have deep conversations with us about what they need, how their future is evolving so that we can think ahead of their needs.

And you heard from Irving's presentation, what Karthik said about how he views storage, how he sees the essential role of the hard drives. But what really resonates with me is what Andy Warfield, AWS is architects for all storage and one of the main brains behind driving all their storage technologies. He said this at re:Invent last year. To me, what Andy is saying is I need innovation. I need innovation from the hard drive industry.

I love that innovation. But please don't let it break my business while you're delivering that innovation. Back to the point of smooth, no disruptive changes. And this demand at that scale is just not negotiable. And it's not just Andy, we're hearing that from or AWS. We hear that from every customer. They know how to pick up the phone, so you tend to call me. And we're hearing that, we literally hear that from every customer. Because -- so this is what AI needs, what customers are asking for. And what Andy's words are saying has really shaped and continues to shape our road map. This is more than you've seen in the road map for the 5 years in WD in the past.

Capacity delivers the exabytes customers need to drive AI to build and continue to evolve and grow AI. Performance is an important part of storage in the cloud, storage for AI. We need more of it. And because power is so important to store that enormous amount of data that is being created through inference, we need to deliver lower power drives.

And as Irving pointed out, there is a new class of customers with large data estates that don't have the resources or the time to build a storage stack from scratch to take advantage of UltraSMR, all the capacity points and the performance and the power that we are delivering. For them, we'll simplify it. We'll give them a shortcut through our platforms. There's a lot here. But you're asking, can we get going? All right. I'm sure top of mind for most of you is what is our HAMR story. Anyone care? All right. So this is our HAMR drive. It's real. You can see it working next door. But most importantly, it's in our customers' hands. They're qualifying it. We're focused on the reliability of the new technology, and we're focused on increasing the aerial density of those drives.

Last week, we announced we have customer already qualifying our HAMR drives. Today -- thank you. But wait. Today, I am very excited to say we have our second customer. it's not just having the drive. It's making sure it delivers what customers expect and need to so we don't break their business. Here, this chart shows 5 HAMR drives in a box, running 3 different qualification workloads. Each line is a different drive.

Each color is a different drive. What you see here is the performance and the consistency of that performance across all those drives. This is just a sample. We have thousands of drives running. The key for us is to make sure that the drive performs and behaves and is reliable enough such that the customer does not have to change their software and by making the performance of a HAMR drive virtually identical to an ePMR drive, they don't change their operational practice. That's a painful thing for a hyperscaler. But more importantly is that we can put an ePMR drive next to a HAMR drive in the same box, same infrastructure for the same workload. Simplify the transition, make it a smooth transition and a smooth ramp for the customer.

So you ask, how do you do this? Or maybe you don't, what I do. We based our design on our trusted 11 platter ePMR platform, same design that customers have been deploying in the millions for the last 10 years. It's 11 platter mechanics. It's the same firmware as the ePMR drive so that we can focus on the reliability and the aerial density increases of the recording technology, the HAMR part.

Today, in our labs, we already have 4 terabyte per platter. I know it's early, but to do the math, that's 44 terabytes per drive. So it's the same architecture as the ePMR drives that smooth the transition, that means you could use the same boxes, the same software, simplifies the transition. And we deliver the areal density increases to get us going on the ramp. Our HAMR drives start at 40 to 44. And as you saw in the road map slide, we showed 100. So how do we get to 100 terabytes from here? As you all know, I'm sure, HAMR requires lasers. Lasers provide the heat that makes the recording possible. It sits on every recording head. This is a micrograph of the laser sitting on top of the HAMR drive.

Today, everybody in the industry, us included, use what's called an edge emitting laser. It's the gray box that you see sitting on the top. They're great. They work really well. They are the ones powering HAMR today. But they have 3 challenges. The light produced sometimes is wasted. So they waste a bit of energy. That could be better used elsewhere. They're quite tall.

So we have to make sure the platters are some distance apart from each other. And during the hard drive manufacturing process, the yields are not as big as good as we'd like them to be. So for the last 6 years, we've been working on our own patented laser technology. It solves for those 3 problems. By emitting more light, harnessing more of that light into the recording technology, we will increase the aerial density of the HAMR platters from 4 terabytes all the way to 10 terabytes by 2028 per platter. We have 11 platters. It's one of the reasons I'm confident about 100 terabyte HAMR drives by 2029. This technology is not theoretical. It's actually already in the labs.

We've watched it do in the recording. The other part of it, as you can see from the micrograph, they're shorter. So it allows us to add yet more capacity per drive by packing up to 14 platters into the same 3.5-inch form factor. 10 terabytes, 14 platters, that sounds like 140 terabytes. Well, thank you. So this is us inventing ahead of the customer on capacity.

The third part, I haven't forgotten, is the lasers emit lights vertically, not from the edge, which means we can test the laser and the head independently, so we can increase our yields. Very excited to see this in the lab, not just in the lab, they're actually in drives recording HAMR physics. So we want to run ahead of the customers. We have all this technology in our labs. And it is how we get to the road map. That means we can accelerate the pace with which we introduce new capacities. But that pace has to be matched by the pace our customers can take that capacity and put it into their production fleets. As I'm sure you know, qualifying a hard drive is a long process. It used to be we waited until we get to the end of the manufacturing process where we can produce drives and volume, then we start the qualification.

As a customer, I didn't like that. It just took too long for us to get the capacity into our fleets. Being at WD, I'm pleased that we actually did something about it. Together with our customers, we started to introduce the hardware into our labs and the qualification of the hyperscaler software so that we can start the qualification process while we're still developing the drive.

That cuts out months from a qualification process. So by the time we're ready for volume manufacturing, the drive is ready to ramp. As Irving said earlier, we've ramped up our latest generation of drives very quickly. That's one of the reasons. The other part with rapid generation of more capacity points, customers will have a lot of qualifications. So instead of qualifying every single capacity point, they qualify one set of capacity point, let's say, 36 to 41 terabytes, and we will just ship them more capacity as we make it available. One qualification, many capacity points.

And the next one is going to be at 42 to 56 and so on. So that innovation, not just in the drive design, but also in the processes we do gets us faster time to capacity in customers' hands in their fleets where they need it the most. So putting it all together, our HAMR capacity goes from 40 terabytes to 100 terabytes by 2029.

And as you saw, a little bit beyond that. That's what gives me confidence. We're working with our customers to prove the reliability of the technology. We want them to feel very sure of our reliability is as good as they used to.

And to deliver a smooth transition on these capacity points, we're extending our ePMR road map. So our ePMR road map, ePMR is the workhorse of the industry. It has been delivering capacity for the last 10 years, and AI has also changed the rules. Customers are no longer focused about technology transitions, recording technology transitions. They want capacity now. They want a proven technology so that we can scale it quickly. Customers challenged us to deliver more capacity. But the industry was told this was not possible without HAMR, right?

Well, we did it anyway. I'm very proud to show you the world's first 40 terabyte ePMR drive in customer qualification today. This drive is 11 platters. It's the existing design we have been building and using. The way we got to 40 terabytes is through material science, changing recipes, engineering the heads differently, using our triple actuation technology and a lot of determination from our engineers.

They wanted to deliver for the customer. This is already -- I think we talked about -- it's already in one hyperscaler, and we are now adding our second. But it's not enough to have the 40 terabytes. Does it deliver the experience customers want. Same as HAMR, we tested during development to make sure that we are ready to qualify that drive for customers. What you see here is the same view of the workloads, the rewrite workloads that our customers use to deliver that capacity. The short answer here is it's a proven technology. They qualified it quickly, and it's already at high yield. And it meets the workload demands without them changing their software. It's exactly what the experience we want for customers is. But there's more. Why stop at 40? Our customers want the overlap between HAMR and ePMR.

We continue to drive media recipes, head designs to drive aerial density. We're increasing the aerial density per platter. We're also borrowing some of the ideas from HAMR. So to get to 60, we'll go to 12 platters. We're packing more capacity inside the same 3.5-inch form factor, all without changing the power profile of the drive.

So customers don't have to spend more energy doing what we do. So the overlap is designed not to break the customer business. They can transition to HAMR as and when they want. The capacity points give them a choice between where to go to ePMR and HAMR, and they will qualify both, so they have the choice in their infrastructure. These drives are made on the same manufacturing lines because they're the same design. In fact, we're so confident in our road map for HAMR. From last year, we said back in the '27 for the ramp, we pulled it in by 6 months. So it's ramping in the first half of 2027. Our road map goes from 40 terabytes all the way to 100. And you can see all this next door. But while capacity itself is important, we also said that AI needs more performance and power efficiency.

This performance has led some customers to think about using QLC flash to provide that performance in addition to hard drives. It's very attractive because -- but QLC has a problem when the data is constantly moving as it does in the AI workloads and in the cloud workloads. QLC wears out. And that means you have to make a lot of changes in your software, so it doesn't wear out.

Otherwise, you end up with silent data corruption. And that is just as scary as it sounds. Hard drives, on the other hand, just don't. They don't wear out. They'll operate for years without wear out and customers like that and it simplifies their code. The other reason to consider QLC flash is that the headline performance of a QLC drive is 6 gigabytes per second. That's way more than the 200 to 250 megabytes per second of a hard drive. But that's a headline number. It's only true when that drive, that QLC drive is attached directly to the GPU with a great big bus. In the real world, in object stores deployed in massive scale, that's not how it's done. Hard drives and QLC drives are connected to the network via a thing called the SAE interface.

It's a thin pipe that takes data from the drive to the network. It can only support 530 megabytes per second. So customers would get less than 10% of the performance of QLC for 10x the cost of a hard drive. Do you think that's a good deal? I don't now or as a customer. But you don't have to take my word for it. Aaron Ogus, who runs the storage world, the storage technical leadership in Microsoft, all of it past and present shares our opinion.

[Presentation]

With Aaron asking for without asking, is, hey, can you give me a hard drive that fills the pipe? That sounds incredible, not something that we've seen before. Well, this is how we did it.

[Presentation]

Isn't that just fabulous? This is not theory. Customers have this in their hands. They didn't trust us when we said we could do this. They just said, hey, show me. So we did. We put them in their hands and they tested them and they see the same results. And you can see the results next door. Typically, customers have 2 questions for me in rapid succession when we talk about these drives. One, when can I have them?

Because they fit in my existing infrastructure. I don't have to change anything. I just get more. And the second, it doesn't matter what I say, go faster. The thing they really love about this is that we can go to 4, 6 and 8x the performance. It is scalable. By the time we get to 100 terabytes, we could be 8x the performance of today's drives. We already have the technology to do it, and we're developing it, so we're ready for when customers are ready for it. We'll introduce this capability at the 50-terabyte mark to meet the customers' demand so that they are ready for us to consume and take advantage of all this performance. But bandwidth, the megabytes per second from a drive is not all the story. Yes, we can saturate the link at 500 megabytes per second, but we also want more transactions per second.

As you know, the transactions per second has to do with the actuators, the way they move. So we have to double that just to match the bandwidth, right? What's exciting about it is not the change in the software. But I know you're thinking, haven't we seen this before? No, it's a simple answer. What you've seen before required hardware changes, software changes and more power.

Other than that, it was great. But that didn't meet what customers needed. This design will fit in an existing customer chassis with that change. It can be made on the same manufacturing lines. And they just see more performance. So customers are really excited by this. double the transactions smoothly for customers. By the way, you can see it next door. It's really cool to see the arms moving. Dual-pivot technology helps customers focus their software effort on improving more performance for AI versus having to deal with how the hard drives are working. And we'll introduce this at the 60-terabyte mark. We are inventing ahead of when customers need this technology. So we're ready for them when they are. So we're starting to put high-bandwidth drive technology in our customers' hands today.

My biggest problem is finding them enough material so they can start testing. Dual pivot technology will be in their hands in late '27 and '28. So all the performance that the customer's hardware can support their existing boxes, their existing software, their existing networks without having -- can be delivered from these drives with the capacity that we are building without having to use QLC. We deliver performance 10x cheaper than QLC can.

But both technologies in next door, please take a look for yourselves. And equally important, this technology applies to ePMR and HAMR equally. It's independent. As we said earlier, inference generates a lot of data. That data has to go somewhere. Not all that data is useful. My kids keep generating bird videos. They're obsessed. I don't think the world needs that many bird videos. So not all that data is going to be going to the cool tiers. We need something new where we still want to store that data, but we want to store it with a lower TCO. One way to do that, the way we think is sensible, is to reduce the amount of power the drive consumes, optimize -- we optimize the drive differently. So by spinning the drive slower, we can reduce the power by 20%.

But we only trade 5% to 10% of the sequential I/O, not something that the customers have seen before or even thought was possible. It's because of the combination of technologies that we talked about and a few other things under the hood that allows us to make that trade-off. And the icing on the cake is you got 10% more power -- sorry, more capacity, more power. So at 100 terabytes, that's a whole 10 terabytes more. Pretty meaningful. Same drive, same infrastructure, same software, same manufacturing line.

So these drives increase our manufacturing velocity. They work for customers as they need them today. And we'll start qualification for this drive in 2027, waiting for customers to get -- take advantage of the capabilities and build the cooler storage tier they've been wanting to build for some time. So all these technologies that we just described extend the reach of hard drives into our storage library on the performance side, more towards the door, accessing more of the data that is fast paced.

And on the other side, we got closer to the archive where data that needs to be accessed in a few hundred milliseconds or a second don't have to be thrown to tape or it takes 8 hours to retrieve. We're confident through this reinvention, this work that we're doing that, that 80% share of the library will increase.

So customers get bigger, faster, more power-efficient drives. And outside the hyperscalers, as we just described, customers would like to take advantage of hard drive economics, not because they want to, but because it's using flash for their business in the beginning made sense. They went quickly, so they use flash, build up the business, build up the workload.

But that agility trade for cost is becoming a barrier to their profitability as their workloads get bigger. Their data estates grow fast as their businesses become more successful. So they need hard drive economics. We are at least 10x cheaper. We have demonstrated that we can deliver the performance and we can deliver the cost efficiency and we can deliver Ultra SMR.

But these are features that are quite complex for customers to do from the beginning. So we want to offer a shortcut. The shortcut is a simple, open API that allows customers to integrate that API to their existing file system, their existing object store. We'll make it available on flash so that they can continue to use what they have and what they're used to. But we'll also extend it to abstract all the hard drive features. So they can take advantage of UltraSMR. They can take advantage of high-bandwidth drives, dual-pivot technology, 100 terabyte drives.

So we do the qualification work behind the API. So they don't have to. So we get faster time to capacity and technology, and this will be available for them in 2027. So this road map you see here, it really changes how we see storage. For years, it was treated like plumbing, necessary, invisible, a cost to be avoided. AI changed the rules. Storage is really how where AI goes to learn. AI becomes a storage, becomes a foundation for AI.

And the hard drive is the building block of that foundation. This road map here is not just about capacities and performance, it's about the reinvention of the hard drive. It's about making it easier and more accessible to use both the performance capacity and economics of hard drives.

So we're delivering seamless capacity growth to 100 terabytes per drive, more exabytes for customers, more performance and the cost structures they need. So customers can focus on building the AI brain. a great moment to be at WD. Because while our customers are building that brain, we can focus on building the beating heart of AI, hand-in-hand with those customers. With that, thank you.

It's now my pleasure to introduce Kris Sennesael, our CFO, to talk about the updated financial model.

It's amazing to see all this innovation in WD. And thank you, Ahmed, for your leadership, and thank you to Ahmed and all the engineers at WD at picking up the pace of the innovation inside the company. It's remarkable. So good morning, everybody. It's great to see you all here in New York, Ice cold New York, exposing yourself to this Arctic blast. I think it's time to pick up the heat a little bit, right? So so far, you've heard from Irving and Ahmed about our strategy, our focus on execution and innovation.

And especially from Ahmed, you've heard all the innovation that is fueling our industry-leading technology and product road maps, which is driving a lot of positive momentum in WD. And so now it's time to connect the dots between all the innovation and what it does to our business model, our financial model and what it does for you as a shareholder, current shareholder or prospective shareholder and how we create long-term shareholder value. Before I go into the details, I hope that everybody realizes today that the WD you see today is a structurally different WD compared to Western Digital of 3, 5 or 55 years ago, right? Today, 90% of our revenue is tied to data center build-outs, cloud and AI, 90% of our revenue.

And more and more data centers are being built out as more and more data is being stored in the cloud. And now we see an acceleration because of AI, AI workloads. More data is being used for training of the models, more data is being used for inference. And that by itself creates more data that needs to be stored. And Irving already told us 80% of all the data in the world is being stored on hard disk drives. That actually has changed our business and our business model a lot. right? We used to be a seasonal cyclical business. But today, that's no longer the case.

Today, we are a long-term secular growth company tied to a strong secular growth in data center, cloud and AI. And that's a big change. In addition to that, we are focused a lot more on innovation, right? Innovation in our technology road maps with HAMR, PMR, increasing the performance of our drives, working on power-optimized solutions. And all of that is really in sync with our customers. And it's all about us providing more value to our customers. That's good for WD. That's good for our shareholders. And last but not least, today, we can do this from a position of strength. We have a strong balance sheet. We have a great business model and financial model, and I'll talk more about that.

We generate a lot of cash, and we can deploy and invest that cash as well as turn it back with confidence to our shareholders. And all 3 of that is a great recipe to create long-term shareholder value. And that change, this major change is very recent. It happened all in the last 12 or 18 months. And the team here really made a lot of changes. It started on or about a year ago with the separation of our flash business and the hard disk drive business.

And we created a strategically focused hard disk drive company with a new management team, a great new leader with Irving, a much stronger focus on innovation, as we talked about, and much deeper customer engagements with leading companies in the data center and storage and AI space. And let's have a look at our scorecard, what this new management team in the last 12, 18 months has done. And on the chart here, you can see fiscal '24, '25 and '26. '26 is not done yet, as you all know.

Fiscal '26 ends in June of '26. And the number shown here is 2 quarters of actual 1 quarter, the third quarter at the midpoint of the guide and the fourth quarter is even consensus estimates. So this is the expected number for fiscal '26. And look at what we've done. Revenue in 2 years, we've doubled or is expected to double from $6 billion to up to more than $12 billion, which is a 40% CAGR over 2 years.

Of course, we are very much focused on growing revenue, but that's not the only thing. For me, as important is the quality of the revenue, right? It's how much value we provide to our customers because that will drive the long-term success of the company. And the value we provide is clearly reflected in our gross and operating margins. And look at what we've done with gross profit and operating profit.

The growth is even faster than revenue and gross margins has improved from high 20s in fiscal '24 to high 30s in fiscal '25 and is expected to further improve in fiscal '26 to mid- to high 40s. And because of the leverage we have in our model, operating margins have even faster improved now with operating margins expected to be in the mid-30s in fiscal '26, right? This is a remarkable change.

And this is just the beginning, and I'll talk more about that. later on. So the strong revenue growth, the expanding gross margin also translated into very strong cash generation, right? Look at what we did just in the last 3 quarters. Free cash flow margin was greater than 20% in the last 3 quarters. And again, it's a result of driving revenue growth, expanding margins, but also focus on working capital management.

Just in the last couple of years or 2 years ago or so, the cash conversion cycle was more than 100 days, right? Today, in the last quarter, cash conversion cycle is approximately 40 days. So major improvement there. And we also have a very disciplined approach to our capital expenditure, right? We've guided and indicated that CapEx is expected to run on or about 4% to 6% to revenue.

Actually, in the last couple of quarters, it was below the low end of that range. And true to our commitment that we made in February of 2025, we have returned all the free cash flow back to the shareholders through a combination of our dividend program and share repurchase program. So let me go back for a second here to February of 2025. And Irving and Ambrish and the team, I think, did a great job at the Investor Day back in February 2025.

And at that time, they had a set of assumptions, which made a lot of sense at the time. And let's just do a quick recap, right? The team was expecting nearline exabyte growth compound annual growth rate in the mid-teens, with potentially an uplift if and when AI kicks in. They also were expecting pricing to go down in the mid- to high single digits year-over-year, in line with historical trends.

And they were expecting innovation in the company and a gradual ramp in areal density and a gradual ramp to higher capacity drives. Well, in just 1 year time, a lot of things has changed. And here are the new set of assumptions that we see in our business. On exabyte growth, nearline exabyte growth, we are now expecting over the next 3 to 5 years, nearline exabyte growth on a compound annual growth rate of mid-20s.

And that is we have a lot better visibility than we used to have because of the deeper customer engagements. We basically have the POs for '26 in hand. We have long-term agreements with some of our customers, 2 of them covering all the way until calendar year '27, additional customer for '27 and '28. So much better visibility, much higher confidence in nearline exabyte growth of mid-20s for the next 3 to 5 years.

By the way, with a lot of our customers, we talk -- we have deeper insight and we talk about '29 and '30, right? We don't have commercial agreements there, but we do have much better visibility. Pricing environment has changed drastically as well. We no longer expect price declines. We actually, over the next 3 to 5 years, expect what I call a stable pricing environment.

Now probably I have to explain that a little bit more. For me, a stable pricing environment is ASPs per terabyte kind of flattish to slightly up low single digits, right? And just to illustrate that, if you look at what we did in the September and December quarter of 2025, ASP per terabyte was up 2%, 3% year-over-year, right?

Actually, if I look at calendar year 2026, 4 quarters of calendar year '26, I do expect ASP per terabyte to go up mid- to high single digits year-over-year for all 4 quarters, mid- to high single digits year-over-year for '26.

And then beyond '26, I expect us to continue to operate in what I call a stable pricing environment, of course, from the higher level that we established in '26. And in terms of product mix, well, if you listen to Ahmed, we definitely have been able to accelerate the pace of innovation, and we will now drive a more accelerated improvements in areal density and accelerated improvements to higher capacity drives. And all of that gives us confidence that we will be able to supply to the higher demand in the mid-20s that we see. It also will help us to expand profitability because those higher capacity drives will result in a lower cost per terabyte. And last but not least, we will be able to do that without having to add more unit capacity, right, and without having to spend a lot more on CapEx.

And let me explain that a little bit more. Last quarter, the average nearline drive capacity was on about 22, 23 terabyte per unit per drive. And despite the fact that we have a 32-terabyte drive available, UltraSMR, ePMR. We actually ship that product in high volume. Last quarter, 3.5 million units, the vast majority of that at the highest capacity point of 32 terabytes.

And this quarter, expected to be close to 4 million units. If we can convince all our customers to adapt the highest capacity, we can ship 40% more exabytes. It's that simple, right? And we are working with all our customers to qualify the higher capacity points. We're working with all our customers to qualify UltraSMR. And the story, of course, keeps getting better, right? Ahmed talked about pretty soon, we will have 40 terabytes available. That actually will allow us to ship 75% more exabytes as over time, as we qualify and ramp the new product with all our customers. And then soon after that, we will have 44 terabyte, which basically will allow us to double the amount of exabytes that we can ship to our customers without having to increase the unit capacity, without having to spend a lot more on CapEx.

And the story, of course, keeps getting better if you think about 100 or 100-plus terabyte per drive. So all of what it does, it creates more exabytes that fuels the revenue growth that also is a key element in our gross and operating margin expansion as well. And with that, let's update our long-term financial model. I mean, last year, less than a year ago, the team provided a financial model. But again, so much has changed in the business.

The business has really transformed strategically focused hard disk drive company with a lot of acceleration in our innovation road map. So here is the new model. You can have a look at it. But before I go into the numbers, a couple of clarifications here. First of all, this is a long-term target model, meaning we're not operating at those profitability levels yet, right? But we have clear line of sight in the next 3 to 5 years to hit those targets. Second, you can see on all of those numbers, they have a greater than sign in front of it. That's on purpose.

There is no ceiling to those numbers. And we are working hard, and we have line of sight to go and hit those targets in the next 3 to 5 years and then continue to operate the business at or above those target levels. So let's look at the model. We are now expecting revenue growth on a compound annual growth rate of more than 20%, right? And that is based on nearline exabyte growth in the mid-20s and a stable pricing environment. We're now expecting and targeting gross margins greater than 50%.

And that is based on a stable pricing environment on one hand and further cost reductions on a cost per terabyte basis, driven by a move to higher areal density and higher capacity drives as well as great execution by our operations team that will continue to drive down the cost in our global manufacturing footprint as well throughout our global supply chain. And with leverage in the model, we target now operating margin greater than 40%. That means that our operating expense, which is on a glide path to 10% or less than 10% to revenue, right?

And so revenue growth, expanding gross and operating margin in combination with good working capital management and disciplined CapEx, again, within the 4% to 6% to revenue, this translates into a free cash flow margin target of greater than 30%. And there is one more thing. When you add this all up and put it in your model, we target over the next 3 to 5 years to run the business with earnings per share of greater than $20.

And when you combine all of that, this is how we create long and durable shareholder value. This is our new model. So let's talk a little bit about capital allocation. Our priorities in terms of capital allocation have not changed. We will continue to invest in the business. We will reduce the debt and return all the cash back to the shareholders. And I'll talk about each of them a little bit more.

First of all, we will continue to invest and reinvest in the business, right? We've increased our profitability targets without starving the business. I mean innovation is so important. Through innovation, we create a lot of value for our customers, for WD and for our shareholders. And so there is no hesitation. Our HAMR road map, the ePMR road map, getting to higher performance drives, power optimized drives, there's no hesitation.

Innovation is key, and we will continue to drive innovation in the company. Second, we will continue to reduce our debt. And as most of you know, before the separation, our net debt position was on or about $5.1 billion. Since the separation, we've been able to reduce our net debt position to $2.7 billion, which is basically we still have $4.7 billion of debt.

We have on or about $2 billion of cash, you get to $2.7 billion of net debt. But don't forget, we still have 7.5 million of SanDisk shares that we had since the separation. 7.5 million of SanDisk shares at today's share price is only about $5 billion, right? So really happy with that. And I've said it before, it's our intention to monetize that share, right, and to monetize that share on or about the 1-year anniversary of the separation.

I have to clarify that a little bit because we do -- and we typically do that in a debt for equity type of transaction, right? And I just said we have $4.7 billion of debt. However, only $3.1 billion of debt qualifies for a debt for equity transaction. I have $1.6 billion convertible debt that doesn't qualify for this transaction.

So -- so we will -- and we are exploring alternative ways to monetize that. One option is that we do an equity for equity swap for the remainder part, which could lead to further share count reduction. So stay tuned for that. But assuming good execution on all of that, we could actually move from a net debt position today of $2.7 billion to a net positive cash position once all that is done. So strong balance sheet, good progress on the debt reduction.

And then last but not least, right, we are committed to continue to return the free cash flow back to the shareholder through a combination of our dividend program and the share repurchase program. The dividend, we are committed to this program, and we are committed to grow over time our dividend and dividend payout. On the share repurchases, you all know in May of 2025, the Board approved a $2 billion share repurchase authorization.

We immediately switched it on. And as of today, we've already used $1.5 billion of the $2 billion, right, and repurchased 14 million shares. And because of the confidence that the Board has in our business model and the management team has in our business model, yesterday, the Board approved a new $4 billion share repurchase authorization, $4 billion that sits on top of the $2 billion that was approved, of which still $0.5 billion is left.

And this really underscores our commitment to continue to generate a ton of cash and also our commitment to return that free cash flow back to the shareholders. while, of course, we continue to invest in the business. So in summary, it is a really exciting time to be part of W&D. W&D that is a structurally different company that is very well positioned to drive long-term growth, to continue to expand our gross and operating margins, to continue to focus on cash and cash generation, cash that can be returned back to the shareholders. And when you put it all together, drives a lot of earnings per share growth, and that is how we create long-term shareholder value. So thank you. And with that, I'll turn it back over to Irving for the closing remarks.

Thank you. Well, I thought I'll just make it very easy in the wrap up and just say 2 simple things. So if you're thinking about WD, why WD and why WD now? First of all, I hope you taken away from today's session, we've really pivoted the company to really be focused on our customer.

That's building trust, that's building confidence in the product that's giving us insight that's fueling the innovation pipeline of the company today and going forward. Secondly, we have transformed to being a data-centric cloud AI-focused hard drive company. front and center in one of the fastest-growing, most exciting segments of any industry in the world. Third, you heard from Ahmed, the industry-leading technology road map that we are delivering, not just about capacity, which this industry has been very focused on, and we are focused on it, but also on performance, on efficiency, making transition simple and ensuring that a new class of AI cloud players can adopt our technology easily as well.

We'll continue to leverage on the resiliency, the innovation, the laser focus on productivity of our global manufacturing footprint and the vertical integrated nature of what we do. Fourth, you can have our commitment that we have a world-class team, all 40,000 people, WD drivers that are laser-focused on continuing execution that you saw from us over the last 12 months and going forward.

And obviously, Kris really brought it home with just a strong financial model that we've laid out and the focus on ensuring robust capital returns going forward. So if you had any doubt about why WD and why now, I hope we've dispelled all of that. And our focus is on ensuring that for all of you, our business partners, our shareholders, we're here to deliver long-term value creation for all of you. So we're going to transition to Q&A. So just give us a moment as we set up the stage. And so I'll invite back Ahmed and Chris and Ambrish, who's going to help us facilitate the discussion as well.

All right. Great. I didn't anticipate too many questions.

Amit was staring over my shoulder.

I guess we have to go first with Amit in that case. So just a quick, we have mic runners or 4 of us, 4 of WD mic runners. We'll start off with Amit, and then we'll queue up Aaron Rakers right next to Amit. Name and firm name.

Amit Daryanani, Evercore. Thanks a lot for the presentation. I appreciate the $20 road map makes the model a lot easier.

Greater than $20.

Fair enough. Greater than $20. I guess the 2 things that stood out, and I'd love to get your perspectives on this. One is on the cost per bit decline. Can you just talk about how are you thinking about cost per bit declines over the next 3 to 5 years in the road map? Is that different in ePMR versus HAMR? Just if you can walk through that math a little bit. And then, Irving, for you, it seems like you want to commit towards the ePMR and a HAMR road map at least through calendar '28. If all customers care about is the highest capacity, the most reliable at scale, why have a dual technology road map when it seems like HAMR is going to work out? Just walk through that dynamic given what customers are asking about.

That sounded like more than one question in one question. So let me try to address the cost down question first. So obviously, the cost down that we've delivered dollar per terabyte has been about 10% over the last few quarters. Obviously, we're not guiding to anything beyond that. But more importantly, what you've heard from Ahmed, even as we deliver a road map that has both HAMR and ePMR, the common platforming that we've done, right, to ensure it's common mechanicals, the fact that we've internalized our own laser capability gives us the confidence that we'll be able to maintain a very competitive cost down trajectory going forward. In relation to the road map, and I think Ahmed, feel free to join in. I mean, obviously, what we want is to give our customers really smooth transitions. And you heard how risk adverse they are to transitions given the size and scale and the importance of storage to their business.

And so we'll continuously work with our customers and to make sure that the economics work for them, the economics work for us. But the reality is as we go beyond, say, 60, 70, we will have to transition to HAMR. As we get to 80, 90, 100 terabytes days, at some point, the laws of physics will require us to transition to HAMR, but we want to do it in a way that makes economic sense, manages risk, gives our customers smooth transitions and equally important, makes economic sense for us as well. The benefit we have is we're able to do both whilst delivering the strong financial performance that Kris laid out.

We'll go to Aaron. And after that, we'll come to Eric over here. Eric.

Ambrish. Aaron Rakers with Wells Fargo. First question on the laser, the integration of the laser technology. I'm curious, and I apologize if I missed this. Can you help us appreciate when that inserts itself in the road map? Is that a point of gross margin leverage? Just any kind of further details on how we should think about that technology path as we go to the hammer.

And then, Kris, for you, I just want to ask, I mean, first of all, congrats on the SanDisk ownership. It's $5.075 billion. I'm curious as we pivot now to this significant free cash flow generation that appears to be supported by this model is how do you actually structurally think about your capital structure? Are you wanting to build cash in the balance sheet? What's the appropriate level of cash to run the organization? I'm just trying to get a frame like how do you think about excess cash generation? Because you clearly are going to have a lot of capacity for capital return. I'm curious how you longer term think about that.

[indiscernible] You take the laser question?

Yes. Well, I was hoping that Kris would answer the laser question, too.

I can, but please go ahead.

Well, thank you. The laser technology, it is smaller, it is cheaper for us to produce. So that is going to be useful and accretive to the business. And we are starting a slow ramp in introducing it. So we -- it's not going to be a light switch transition because we want to be careful about when we transition that. So we'll start the transition around about the 48, 40 terabyte range and even a little bit before that.

Yes. In terms of capital structure, so assuming a successful monetization of the SanDisk shares, we will end up with $1.6 billion of debt on the balance sheet, which is the convert. At the appropriate time, I want to get rid of the convert, but still keep on or about $1.6 billion of debt on the balance sheet. And so yes, we have strong free cash flow. The intent is to return all that free cash flow back to the shareholders.

Great. We're going to go to Eric, and then we'll tee up Asiya after that over here. and then Wamsi right after that.

Awesome. I'll [indiscernible] Aaron and say, remember, you guys once had 23 million chairs of SanDisk.

Hindsight is 20,20 years.

Very true, very good.

No, thank you guys for all the content here. I just want to kind of break it down, which is we've gone through a lot of different configurations and platforms that you're introducing. They all have 10 -- 11 platters on them. Your competitor is coming out with HAMR with 10 platters. How do you remain price per terabyte, price per exabyte competitive? How do you continue to push down cost per terabyte when we have that kind of differential? What are you doing differently that allows you to remain competitive despite that kind of 1 extra platter per drive?

So maybe I'll just make one quick comment, and I'll leave it to Ahmed. I think when you saw his presentation, he had 2 very important points. One, is we're going to continue to focus on increasing rural density. -- right? So if you look at it, if we're going to deliver a 40-terabyte ePMR, as an example, that's on 11 pads, that's close to 4 terabytes per platter, right? When we get to 60, we'll introduce another plan. So that's 5 terabytes per platter. So we're increasing areal density, right? Similarly on HAMR, we said we will get up to 10 by 2029.

'28.

'28. Sorry, he's even ahead of me, right? So if you talk about what the rest of the industry has been saying, they're saying we're going to get up to 10 terabytes per platter. So we're equal or potentially time line-wise, going to be ahead of them in areal density. Now what we're doing is to increase capacity drives beyond aerial density.

So we're going to be at parity or even a hit on a density per platter. But we have that ability because some of the innovative work that the teams have been doing, the laser technology to -- within the same drive to increase the number of platters, right? So that's how you get to potentially 140. So we feel actually from a drive economic standpoint, actually, our dollar per terabyte is going to be even more competitive going forward.

Okay. Helpful. And then maybe just a quick follow-up on going back, whatever it is, 12 months ago to the last Analyst Day, I think we were setting a floor for op margins and gross margins at 24% and 38%. I know we're kind of dreaming the dream now, which is exactly what we are all asking for. But if we do think about the other side of that, like where do we think about the potential floor in margins? Is it any different than a year ago? Or has that changed as well?

No. So the financial model that I put out there, as I said, it's a target model. So we still have some time to go and some work to do to get to the targets. But I have confidence that once we hit those targets over the next 3 to 5 years, we will be able to continue to operate the business at or above those targets.

We'll come to Asiya and then after that, Wamsi.

Asiya Merchant, Citigroup and thank you for all the content. That was great. Just one for Kris, and if I can, one for Irving as well. Kris, when you talk about stable pricing, I know right now, it's a great environment. Just walk us through why stable pricing makes sense beyond maybe the next few quarters where you have great visibility, if there is a down cycle, why do you still think stable pricing could probably prevail?

And then 1 for Irving. Irving, on earnings call, you also talk about quantum computing. I didn't get to hear anything about it. Maybe it's just the time frame here that we're talking about. Just walk us through that innovation that you've talked about? And how does that fit into the broader road map.

Yes. So as it relates to pricing, for me, pricing is fully tied to how much value do we deliver to our customers, right? It has nothing to do with tightness in the supply chain or strength in the cycle or weakness in the cycle. It's how much value do we provide to our customers. And moving to higher capacity drives with better aerial density with better performance is what our customer wants. It's what our customers need and delivers a tremendous amount of value. right? That's why we are going to remain disciplined in terms of pricing, no matter what we are in the cycle. And that's why I have high confidence in this stable pricing environment.

Yes. Maybe I'll just add on to that. I think I've been in regular communication with our customers, one of the big ones just 2 weeks ago. I mean what they really want from us is predictability in pricing. It's very hard for them to be designing their data center architectures 2, 3 years, 5 years when you have the volatility that we are seeing in some of the storage tiers, right? And so what we're really focusing is on continuing to deliver value to them. And as Chris said, be able to share that value so we can deliver that stable pricing. Obviously, in the near term, we're seeing mid- to high single digits, but in the long-term model, a stable pricing environment. In relation to your second question on quantum and we didn't want to distract from all the exciting things. We have to stay on the drive itself. And obviously, as I mentioned, we hope to have regular Innovation Day, we'll keep you posted on what we're doing. But the reason I mentioned it because we are not building a business just for today.

We're also incubating new growth vectors for the company for tomorrow. And that's why I took a moment to highlight the rich core capabilities and IP that we have. And if you look out what's the next probably tech growth driver beyond AI, quantum represents a huge opportunity going forward.

And our magnetics capability, our magnetics junction technology capability, the nano fabrication technology by which we build our drives positions us very well to deliver cost economical qubits to support quantum computing going forward.

And that's kind of why we made the strategic investment into collabs, not because it's a financial investment, but we're also a key technology partner to them and to be able to bring quantum computing economically at scale, just like we have done for the hard drive industry.

Great. And while you get ready for a question, I want to widen my horizon and see who's in the back because I've been sticking with the front ventures, but we'll go with Wamsi and...

Wamsi Mohan, Bank of America. I appreciate you guys doing this, and Kris love your comment on no ceiling to numbers. That's definitely encouraging here. I guess the 2 questions I have are, one is you noted smooth transition across recording technologies.

And it looks like relative to last year, your ePMR road map is significantly different. I think last year, you said ePMR might end at 36% though that we're saying 60-plus, so very different. How do you that impacts the adoption of HAMR at your customers, just given the fact that so far, you've seen tremendous success with ePMR, and that might be an easier transition at higher capacities.

So do you think it impacts the transition to HAMR from a WD perspective? So that's my first question. And second question is, there was a lot of innovation you shared here in terms of performance, improving performance in terms of power, in terms of buyouts. So just curious how you think about this could expand the TAM for Western Digital? And does that -- when this comes around, does it accelerate your growth rate even further?

Maybe Irving, you take that?

Both Question?

Both of questions.

So the -- when we talk to our customers, one of the things that's really very clear about it is they really don't like the sub transitions. And they want to have that smoother ramp in terms of equivalent capacity so they can choose when to ramp the capacity points for themselves. And that's really the genesis of why we pushed the ePMR technology further.

We don't think it's going to decelerate HAMR adoption because customers understand to get to the beyond 60 terabytes, the 80 to 100, they really need to adopt HAMR, but they want to be sure of the quality and the reliability of this new technology, new recording technology that hasn't been in the data centers for 2, 3, 5 years. So they get comfortable with it. So they appreciate us creating that road map for them. And since it's manufactured on the same lines, we're not -- we've created a fungible capacity from that perspective. So we feel that it's the right approach for the customers with certainly the feedback we've heard from all the customers is they appreciate us the thoughtful approach that we're taking to the transition. So they will transition at their own time. There's a deadline, but there is a transition.

Yes. I guess in terms of the performance features, Wamsi, that you're referring to, obviously, we anticipate the take-up rate within the existing fleet of drives that we are shipping to customers on the high bandwidth drives say 20% drives would want that -- customers would want that in their fleets, right? And then obviously, the energy-efficient drives that actually opens up a new TAM for us because that actually is an unfilled need between hard drives and archival storage TS. Obviously, all these new capabilities will be reflected in pricing that's above and beyond what we deliver for pure capacity today.

Great. We'll go with Steve Fox over there. And then...

Thanks for all the great information. really an eye-opener for someone who's followed this industry for many years. One question, it seems like maybe it's an older question, but thinking about your areal density curve, it looks like it's going to be a lot faster over the next 3 years than you're talking about for bit growth.

And you just mentioned new TAMs, et cetera. So can you sort of talk about -- and I recognize you're not going to be 100% at 10 terabytes per platter in 3 years. But can you sort of talk about how that you manage that sort of additional supply through this areal density and where else it could go?

So increasing the areal density, the acceleration of the areal density that we're seeing, it's really the focus of our engineers on the media recipes, how we design the media recipes and how we increase the -- change the head design. So we're taking further advantage of technology we've had in our drives for some time. We're just being a lot more bullish about what they can do, and that's how we get the higher densities that we're looking for right now.

But I guess I'm wondering, it compares to like a 25% bit growth CAGR and you're increasing aerial density by 2.5x over like 3 years. So how does -- how do we equate those 2 numbers together?

So we will continue to innovate as fast as we can and Ahmed has lined out the timing of that -- of course, on top of that, you have to lay an adoption curve at the customers. And the combination of all of that for us, we believe we will be able to supply in line with the nearline growth of mid-20s that we expect over the next 3 to 5 years.

We're going to go to call Tom O'Malley.

Tom O'Malley, Barclays. Thank you guys for doing this. Appreciate it. Not much to pick at here. Just 2 questions on the technology side. You spent a little time talking about the bottleneck of SATA for flash. Can you talk about industry improvements in memory bandwidth and if you see that getting better over the next couple of years and if that may help flash providers?

And then second, you talked about a common API for flash and hard disk drives. A big gating factor for more flash use in the past has been the difficulty with moving the software over. As you guys integrate that, is that going to hurt you in customers being more readily able to move back and forth between flash and HDDs?

Improvements in -- so improvements in memory interfaces is just going to accelerate the pace with which data moves around. So that's just going to require more and more data to be transferred in and out of storage. So that's why we feel our bandwidth drives help us really meet some of that demand. So we see that trend is going in parallel with each other. And I think that's an important thing for us to have delivered to our customers.

On your second question, which kind of -- will the API hurt us versus help us? We feel that those customers that have already started on flash, we want to meet them where they are. So we don't want to make an abrupt transition for them. We want to make a smooth transition for them as well. So by offering the same API and making the transition on the same media, they get the same performance capabilities. And then when they're ready, they can take advantage of the hard drive economics without disrupting their business. And I think that's an important transition.

I don't think it will hurt us going the other way because we're starting with hard drives, you're going to like the economics, you're going to like the performance. There will be a small amount of workloads that may benefit from flash, but being customer-centric is the right thing to do for the customer.

Yes. If I may add, I think it's ultimately down to the workloads. And specifically, what we're looking at in the platform with those open APIs is to really focus on the Neo clouds where they're starting from a flash first perspective to really open up the HDD TAM that we don't really actively participate today. As they start to grow and the amount of the storage requirements increase, the economics will have -- require them to move into that direction.

We go right next to Tom.

Ananda Baruah, Loop Capital. With the road map for aerial density that you provided, which is super impressive, is there a useful way to think about mitigation of the supply-demand gap that's in place today? Is there a time frame that you think supply/demand can start to normalize.

It sounds like, regardless, you guys have your paradigm on economics and value add that you have. So it sounds like that won't change based on supply/demand sort of equilibrium, but any context there, given that your aerial density road map seems like it's going to move pretty quickly over the next couple of years?

Yes. I mean, as we know, the supply environment is very strong and the demand side of the house is trying to catch up. Obviously, from a road map standpoint and from a production standpoint, there's limited opportunity in what we can do for calendar year '26.

But as we get -- as you see the road map as we're able to ramp up as customers, again, as Chris mentioned, depending on the adoption cycle, as we move to the latter part of '27, '28, we anticipate that gap to narrow. But again, something that we've seen is the demand for storage continues to grow at a rapid rate. So even the mid-20s, who knows what it's going to be going forward.

Can we come to the front here, please?

[indiscernible] on for Karl Ackerman at BNP Paribas. So I have 2 questions. First to Ahmed about the high bandwidth drive, the idea of dual actuator drives have existed for years, but never really took off. Why is that? And why do you believe it can be successful today? And I have a follow-up.

I think as I said earlier, the -- these ideas have been around for a while. But typically, the way they were implemented required a customers to change their interfaces, the software, the hardware, and they typically cost more money and more power. That's been the history of trying to introduce these technologies. What we've done in our designs is really paid attention to keeping the customer experience the same. So we took all the complexity away from the customer and kept it inside the drive. And that's a testament to our engineering capabilities and the way that we're able to work through the drive in the small form factor we have, but keep the customer experience the same. That's what's different about these. So I've seen many of these ideas as a customer come to me over the years. And typically, it suffered from poor experience, poor power or it's more expensive. And in this case, we solved for those problems very quickly based on the history of technologies we've been incubating for quite some time.

Yes. I hope you see the big change in our focus as a company. We talked about a year ago being really focused on our customer. And everything you probably have about a customer close to 100 times over the last 2 hours, and everything they are doing starts with the customer at the center and having Ahmed and more and more of our talent that comes from the hyperscale environment really gives us good insight into what we need to be doing from our customers. It's not a technology-first view. It's a customer see view that technology enables the outcome that they want.

Sorry, I have a follow-up. So it's -- so just an overarching question. Just long term, look at where we are in the cycle right now. Demand is through the roof and the supply is constrained. The industry is very consolidated.

So I feel like it's hard to apply any of our previous experience of the historical cycles on to this one. So how do you think this cycle would unwind? What would make you -- maybe like anything make you worry about like in 3 to 5 years, anything would change? Or are you all very well at night? I know no one has a crystal ball, but any of your long-term perspective would be very helpful.

Well, we'll let you guys predict the cycle, right? We don't do that. You got some experts at it. I think our focus is on staying close to our customers. The fact that our Three of our top 5 customers have entered into long-term agreements with us to extend 1 and 2, all the way up to calendar '22 to '27.

As Kris mentioned, we're having active discussions with the rest for '28, '29 and some of them for even 30. That visibility gives us the confidence of where things need to go. And the fact that we are not doing this business with high CapEx unit capacity focus. We can toggle technology if we need to meet exabyte supply and demand balance.

Mehdi Hosseini, Susquehanna. Two follow-ups, one for Ahmed. Should I assume that most of the drives installed today and over the next couple of years are going to be on a SaaS bus internet infrastructure?

I know most of the drives that exist today are on SATA interfaces. So all the hyperscalers use SATA interfaces, they're sufficient for their capabilities and they have done for the last 10 years.

Is that going to be sufficient, especially as for inferencing and assuming that inferencing some is done in the cloud and some on the edge. How would that change the SaaS or SATA requirement?

We don't see the -- the customers have been really good at building the software layer, the objects layers [indiscernible] system layer such that they can aggregate the performance for many thousands of drives into serving those workloads. But as those workload how are they looking for that more performance from us. so that they can serve even more capacity and more performance to their end users. And so the interfaces and the design of the boxes will evolve to meet the performance needs that we have in the future.

Okay. And just on free cash flow question. Should I assume that the capital intensity will remain in the same range?

Yes. Absolutely. So the CapEx intensity is 4% to 6% to revenue and all the innovation that Ahmed has talked about it, especially because it's even recording technology agnostic will not require a step-up in CapEx.

By the way, I guess, if we were trying to figure out the downside is we have to figure out when is the next flood hitting Malaysia. God forbid.

So you used the 4 or 5 questions for the next several earnings calls as did Amit. But we're going to go over there to Matt.

Matthew Schaeffler, Investment Strategies Fund. This -- the amount of innovation sort of belies this question. But how do we know we're spending enough on R&D?

I think the results show that we have been very frugal about how we get to market. We test the ideas, the way that makes sense to us. We incubate them. We test them and only develop the ideas that are really resonating for our customers' experience, take them forward. So there's a lot more ideas than what we are looking at behind the scenes. It's just we do it in a very responsible and very frugal way.

Yes. Maybe I'll just add on to that, and Chris wants to add as well. We have definitely not styled R&D at all, right? And even a year ago at the Investor Day, we said very clearly, we would return 100% of excess free cash flow to our shareholders. And that's after the investments that we make into R&D.

So Again, we are building a business not only for delivering our core drive business for today, we're also making investments into incubations for the future. These new growth vectors are something we are and will continue to make investments in going forward, both R&D investments and also financial investments into whether it's startups or M&A that we think is necessary because our innovation framework has 3 pillars to it. We build internally, we partner, right? We co-innovate with our partners, and we also have a buy model of it. So that will not change going forward. Kris?

Yes. And so again, we're not starving the company. We are focused on accelerating the pace of innovation. I will allow him to spend more R&D dollars provided he does it in the most effective, most efficient way and there is a strong return on investment. And I'll stop it there.

We have time for one more question. We're going to go to C.J. over there.

C.J. Muse with Cantor Fitzgerald. Two questions. First one, I want to push you a little bit on gross margin. You're just a couple of bars away, I think, from 50% if you keep pricing stable and you maintain that 10% cost down. So curious, is there something that's occurring going forward in the transitions that's reducing either your cost down or yields not going to impact things as well.

We'd love to learn a little bit more on that. And then second question would be probably a question you haven't received an ever, which would be cannibalization going the other way with NAND pricing moving the way it is, I use talking with customers and starting to see demand perhaps pick up that had traded away over to SSDs. Thanks so much.

Yes. So I'll take the gross margin question. So you think about it the right way. We are making good progress at improving gross margins. We are getting closer to the 50% and 50-plus gross margins. Obviously, we have a technology transition in front of us. But we've always stated, and I'm happy to repeat it today that if and when we transition to HAMR, this will be neutral to accretive to the gross margins.

Yes. And C.J., to your question, look, I think what you see from the innovation that we shared today from a capability standpoint, from a performance standpoint, we are reinventing the hard drive to make it much more competitive, much more suited for the AI workloads for today.

At a minimum, it's definitely going to preserve that 80% of bikes that stored in the hard drive. Do we think there's opportunity for us to expand on that given pricing? Again, it's highly volatile environment, we'll see. But at a minimum, I think it really gives us the confidence that we will continue to maintain the 80% share with some upside opportunity to grow it over time.

Great. Thank you. So this -- we come to the conclusion of our first session, but to paraphrase, Ahmed, what's next? We're not stopping here. We're going to break for lunch. Lunch is a floor down. And simultaneously, we'll have the demo room open. So as I said earlier, please avail of this opportunity and meet some of our brightest folks back there.

I would really encourage you to see demos because it's not slight way. You're going to see the 40-terabyte drive. You're going to see the dual actually the dual pivot and you're going to see the high bandwidth drive and as well as the HAMR drive.

Thank you again.

Good afternoon, and thank you for standing by. Welcome to Western Digital's Second Quarter Fiscal 2026 Conference Call. [Operator Instructions] As a reminder, this call is being recorded.

Now I will turn the call over to Mr. Ambrish Srivastava, Vice President, Investor Relations. You may begin.

Thank you, and good afternoon, everyone. Joining me today are Irving Tan, Western Digital's Chief Executive Officer; and Kris Sennesael, Western Digital's Chief Financial Officer.

Before we begin, please note that today's discussion will contain forward-looking statements based on management's current assumptions and expectations, which are subject to various risks and uncertainties. These forward-looking statements include expectations for our product portfolio, our business plans and performance, ongoing market trends and our future financial results. We assume no obligation to update these statements. Please refer to our most recent annual report on Form 10-K and our other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations.

In our prepared remarks, our comments will be related to non-GAAP results on a continuing operations basis, unless stated otherwise. Reconciliations between the non-GAAP and comparable GAAP financial measures are included in the press release and other materials that are being posted in the Investor Relations section of our website at investor.wdc.com.

Lastly, I want to note that when we refer to we, us, are or similar terms, we are referring only to Western Digital as a company and not speaking on behalf of the industry.

With that, I will now turn the call over to Irving for introductory remarks. Irving?

Thanks, Ambrish, and good afternoon, everyone, and thank you for joining us today. The growth and impact of AI continues to accelerate across numerous industries. As generative AI models become the norm and agentic AI scales to drive business productivity, it is clear that AI is becoming a true strategic enabler of business transformation. AI inference has also begun to take hold in many ways becoming the true AI workload with deployment to chat bots and virtual assistants and customer relationship management tools. Further innovations in physical AI are also accelerating quickly, generating increasingly larger multimodal models propelled by advancements in autonomous vehicles and robotics.

In all cases, it is data that is needed to fuel the entire AI process from training to inference to enable stronger models and sharper inference results. And as more data is generated and the value of data increases, the demand to store it is expanding at a rapid rate.

As AI capabilities expand, cloud continues to grow as well, and both are driving the surge in demand for higher-density storage solutions. In this new era where AI and cloud dominate, Western Digital has taken a customer-focused approach to managing the strong demand by working closely with our hyperscale customers, ensuring that we deliver reliable, high-capacity drives at scale to give them the best performance and total cost of ownership.

We are doing this by continuing to focus on increasing our drives areal density and accelerating our HAMR and ePMR road maps as well as upshifting our customers to accelerate adoption of higher capacity drives and UltraSMR technology. This last quarter, we shipped over 3.5 million units of our latest generation ePMR products, offering up to 26 terabyte CMR and 32 terabyte by UltraSMR capacities, representing strong confidence and adoption by our customers.

We have also started qualification of our HAMR and next-generation ePMR products, each with a different hyperscale customer. These drives will offer our customers the higher capacity and improve total cost of ownership that they are looking for. In addition, we continue to accelerate our HAMR innovation. To support this, we recently acquired intellectual property assets and talent that will help us in the development of our internal laser capabilities. Also this past quarter, in partnership with software ecosystem partners, we announced our UltraSMR-enabled JBOD platforms, expanding UltraSMR adoption to a broader customer set. These platforms deliver significantly higher storage density compared to conventional drives. giving customers hyperscale-like performance and make mass scale data analysis more sustainable and efficient.

We are truly seeing our approach resonate with our customers, and this is reflected in longer-term agreements and better visibility into their requirements. We have firm purchase orders with our top seven customers through calendar year 2026. We also have in place robust commercial agreements with three of our top five customers, two through calendar year 2027 and one through calendar year 2028. These agreements indicate a strong trust that we have built with our customers and confidence in our ability to meet their exabyte needs. We are hosting an Innovation Day on February 3rd in New York next week, where we will share updated road maps for our HAMR and ePMR products as well as further details on core innovations that we are developing to improve our drives performance, energy efficiency and throughput. We will also provide an update on our financial model.

In keeping with our strategy to incubate new growth vectors based on our intellectual property and core capabilities, last month, we announced a strategic investment in Qolab, which combines our expertise in material science and precision manufacturing with Qolab's breakthrough approach to quantum hardware design. Working with Qolab, we aim to advance next-generation nanofabrication processes that improve qubit performance, reliability and scalability. Looking ahead, we see our positive momentum continuing and we will remain focused on supporting our customers' exabyte storage requirements while completing qualifications and launching our next-generation HAMR and ePMR drives.

I will now hand it over to Kris to share our Q2 results and outlook for Q3.

Thank you, Irving, and good afternoon, everyone. Western Digital delivered another quarter of strong financial performance, reflecting disciplined execution across our organization and our ability to meet the customers' growing demand in the AI-driven data economy.

During the second quarter of fiscal 2026, revenue was $3 billion, up 25% year-over-year, driven by strong demand for our nearline drives. Earnings per share was $2.13. Both revenue and EPS were above the high end of the guidance range. We delivered 215 exabytes to our customers, up 22% year-over-year. This includes over 3.5 million drives or 103 exabytes of our latest generation ePMR with capacity points up to 32 terabytes. Cloud represented 89% of total revenue at $2.7 billion, up 28% year-over-year, driven by strong demand for our higher capacity nearline product portfolio. Client represented 6% of total revenue at $176 million, up 26% year-over-year. Consumer represented 5% of revenue at $168 million, down 3% year-over-year.

Gross margin for the fiscal second quarter was 46.1%. Gross margin improved 770 basis points year-over-year and 220 basis points sequentially. The improved gross margin performance reflects continued mix shift towards higher capacity drives and tight cost control in our manufacturing sites and throughout the supply chain.

Operating expenses were $372 million. As a percentage of revenue, operating expenses declined 120 basis points sequentially, primarily due to operating leverage in the model. Operating income was slightly above $1 billion, translating into an operating margin of 33.8%. Interest and other expenses were $45 million, and our effective tax rate in the fiscal second quarter was 15.1%. Taking into account the diluted share count of 378 million shares, EPS was $2.13, an increase of 78% year-over-year.

Turning to the balance sheet. At the end of our fiscal second quarter, cash and cash equivalents were $2 billion and total liquidity was $3.2 billion, including the undrawn revolver capacity. Debt outstanding was $4.7 billion, translating into a net debt position of $2.7 billion and a net leverage EBITDA ratio of well below 1 turn. Operating cash flow for the fiscal second quarter was $745 million, and capital expenditures were $92 million, resulting in strong free cash flow generation of $653 million for the quarter, which reflected a free cash flow margin of 21.6%.

During the quarter, we made $48 million of dividend payments and increased our share repurchases to $615 million, repurchasing 3.8 million shares of common stock. Since the launch of our capital return program in the fourth quarter of fiscal 2025, we have returned $1.4 billion to our shareholders by way of share repurchases and dividend payments. Also, today, we announced that our Board has approved a quarterly cash dividend of $0.125 per share of the company's common stock, payable on March 18, 2026, to shareholders of record as of March 5, 2026.

I will now turn to the outlook for the third quarter of fiscal 2026. We anticipate revenue to be $3.2 billion, plus/minus $100 million. At midpoint, this reflects a growth of approximately 40% year-over-year. Gross margin is expected to be between 47% and 48%. We expect operating expenses in the range of $380 million to $390 million. Interest and other expenses are anticipated to be approximately $50 million. The tax rate is expected to be approximately 16%. As a result, we expect diluted earnings per share to be $2.30, plus/minus $0.15 based on a non-GAAP diluted share count of approximately 385 million shares.

To wrap up, Western Digital achieved another strong quarter with performance ahead of expectations. Our guidance for the next quarter underscore continued favorable trends in our business alongside our disciplined approach to free cash flow, capital returns and long-term value creation for shareholders.

With that, let's now begin the Q&A. Ambrish?

Thank you, Kris. Operator, you can now open the line to questions, please. To ensure that we hear from as many analysts as possible, please ask one question at a time. After we respond we will give you an opportunity to ask one follow-up question. Operator?

[Operator Instructions] Operator? Our first question today comes from Aaron Rakers with Wells Fargo.

And I will stick to one, Ambrish. On the gross margin line, the guidance that you're giving for 47% to 48%, I guess the back of the envelope math would suggest that you're maintaining what looks to be like a 70%, maybe 75% incremental margin flow-through. So, I guess, my question is, how do you think about the durability of that incremental margin? Or maybe taken another way, how do you think about the cost curve down on a per terabyte basis as we look out over the next, call it, several quarters?

Yes, Aaron, thanks for your question. And so, first of all, I'm really happy with what's going on with the gross margin. We delivered 46.1% gross margin, up 220 basis points quarter-over-quarter, up 770 basis points year-over-year. And we are guiding to 47%, 48%, so 47.5% at the midpoint, which is up 740 basis points on a year-over-year basis. And Aaron, I think your math is working. The incremental gross margin is on or about 75%, depending on how you look at it on a year-over-year basis or a quarter-over-quarter basis. So I've stated before, I'm very comfortable with an incremental gross margin higher than 50% and definitely 75% is higher than 50%.

I mean in gross margins, there's two sides to the equation. On one hand, you have pricing environment. On the other hand, you have the cost environment. In pricing, I've talked about that before. We see a stable pricing environment with prices on a price per terabyte kind of flattish to slightly up. Actually, last quarter, it was up 2%, 3% on an ASP per terabyte basis. So that clearly demonstrate the value that we continue to deliver to our customers.

And on the cost front, the teams continue to execute really well. We continue to upshift our customers to higher capacity drives, which gives us a cost benefit. And then there is great execution as well on driving down the cost in our manufacturing assets as well as throughout the supply chain. And when you look at it last quarter, the cost per terabyte was coming down on or about 10% on a year-over-year basis. And so when you put this all together, we continue to drive further gross margin expansion. And we believe in the next couple of quarters and beyond, we will continue to be able to do that.

The next question is from Erik Woodring with Morgan Stanley.

Irving, just given the tightness of the HDD market and kind of the significant inflation that NAND is going through right now, can you maybe just talk about maybe your patience in being able to sign purchase orders further into calendar '27 to extract better economics just relative to maybe how you were approaching signing POs last year? Is that making any difference in the economics you're able to extract? And then -- thank you.

Yes. Thanks, Erik. As we highlighted, we're pretty much sold out for calendar '26. We have firm POs with our top seven customers. And we've also established LTAs with two of them for calendar year '27 and one of them for calendar year '28. Obviously, these LTAs have a combination of volume of exabytes and price. And in relation to pricing, I think first, it's important to recognize that our customers actually have value that there's actually a structural shift in the value that we deliver to them, especially in the impact that we have to their total cost of ownership as the business moves more and more towards inference where monetization is happening.

So, in this case, the pricing that we've provided there reflects the value that we're delivering to them. And so as Kris mentioned, we continue to see going forward, a stable pricing environment that gives us an opportunity to continue to extract more value as we deliver both better TCO value to our customers and to better support their supply-demand needs as well through higher capacity drives.

Do you have a follow-up, Erik?

Sure. Just very quickly, Kris, would just love to know how you're approaching the SanDisk share ownership. Do you still plan to monetize before, I think it's the February 21 deadline? And more importantly, how do you expect to leverage those proceeds?

Yes, Erik. As you probably know, we still have 7.5 million SanDisk shares, and it's our intention to monetize those shares before the one-year anniversary of the separation. likely in a similar transaction that we have done before, meaning it's a debt for equity swap. And so the proceeds will be used to further reduce the debt.

The next question is from C.J. Muse with Cantor Fitzgerald.

I guess could you speak to how customer engagement and contracts are evolving in this very tight environment?

Yes, C.J., this is Irving. Thanks for the question. One of the things that we've been very focused on over the last year is really develop a much more customer-centric approach. As we've shared in the past, we've really pivoted our organization to be centered around our big hyperscale customers with dedicated teams for each of them. That's really deepened the relationship that we have with them in terms of both technology road map development, in terms of getting better visibility of their demand requirements, and you see the result of that in the longer-term LTAs we've been able to structure with them.

We're also looking forward to sharing with all of you the innovations that we are going to be delivering to support the AI needs -- workloads needs going forward at our Innovation Day next week. But definitely, the relationship has improved, as I highlighted, they definitely see the value and the structural -- that's resulting in the structural change that we're seeing in terms of pricing with them that's also resulting in the longer-term contracts that we have.

Ultimately, what we want to do is to be able to ensure that it's a fair value exchange, deliver predictable pricing to them because one of the things that they are concerned about is the high volatility of some tiers of the storage space, right, and to ensure that there's sustainable value creation, both for them and for us along the way.

Do you have a follow-up, C.J.? Let's go to the next. Sorry, C.J., go ahead.

Yes, sorry about the numbers. I guess just to follow on the SanDisk share comment. Can you talk about your plans thereafter? Are you going to focus more so on share repurchase or other?

Well, we are already focusing on share repurchases since we've announced the $2 billion share repurchase authorization in May of 2025. We already have repurchased $1.3 billion or we have used $1.3 billion of that program, repurchasing on or about 13 million shares, and there is no hesitation. We will continue to use that program.

The next question is from Wamsi Mohan with Bank of America.

This is Aisling Grueninger on for Wamsi. Congrats on the results, guys. Just one question for me. minds on the mix of UltraSMR. Just given your order book LTAs, how is this mix trend on UltraSMR trending? And how does this mix shift play a role kind of as a driver of gross margins moving forward?

Yes. That's a really great question, Aisling. Thank you for that. Well, last quarter, we crossed on the nearline portfolio, 50% mix on UltraSMR, and we actually see that increasing. As we've highlighted, one of the things that we're doing to better support the growth in demand from our customers is really to upshift them to higher capacity drives. A big part of that is the upshift to UltraSMR-based drives, and we see more and more customers adopting UltraSMR. We have our top three customers fully on board. with UltraSMR drives already today, and we have another two to three more that are moving into a process of adopting UltraSMR. So we are likely to see the UltraSMR mix of our total nearline exabyte base continue to increase going forward. That's actually very important for us because, one, we are better able to serve our customer demand needs.

As you recall, UltraSMR gives a 20% capacity uplift over CMR and a 10% capacity uplift over the standard -- industry standard SMR. But equally important from a gross margin standpoint, UltraSMR is a software-based solution. So it's very accretive for us from a margin standpoint as well. So a higher shift higher mix of UltraSMR is definitely going to be beneficial both to our customers and to our ongoing profitability as well.

And Aisling, one thing in Irving's prepared remarks, we mentioned the JBOD that we have launched, which also expands our UltraSMR customer reach beyond what we have been targeting so far. So thanks for your question. Maybe we go to the next question, please.

The next question is from Asiya Merchant with Citigroup.

It's Mike Cadiz at Citi for Asiya today. So I have a question and perhaps a follow-up. So the first is, could you provide any color or additional color on yields and reliability? I know that is -- those are a couple of points that Irving has brought up over the past couple of quarters in relation to the multiple rollouts. Is there any implication to cost per bit declines that we can think of?

Yes. Thanks for the question, Mike. So our yields on our ePMR products continue to be very, very they continue to be yielding very well in the low 90s percentage yield range. And obviously, from a reliability and quality standpoint, we received very good feedback from our customers. The fact that we've been able to, last quarter, deliver over 3.5 million units of our flagship ePMR drives is a testimony to the confidence that they have in terms of the reliability and the quality.

In terms of the cost related to the cost down, obviously, as we get yields up, cost continues to decline as the UltraSMR mix goes up within those new products as well, that's also going to be a driver of cost down as well.

Okay. Do you have a follow-up?

I did. So can you talk more about any progress or the progress from your Rochester test and integration site how you're leveraging perhaps those efforts to accelerate maybe in the existing customer transitions?

Yes. One update that we shared in the prepared script is actually we -- last quarter, we indicated that we would start HAMR qualification. We pulled it forward to the first half of calendar '26. We actually have started qualification of those drives already this month for HAMR. On top of that, we've also started qualification for our next-generation ePMR drives as well. And obviously, our Rochester SIT Lab plays a critical role in ensuring that we have a very smooth, quick qualification. And equally important, as they move into production environments that they deliver the same reliability and quality. that our customers have been used to our previous generations of products.

Again, on this one, we look forward to sharing a lot more on the 3rd of February in our Innovation Day, we'll be highlighting the updated road maps for both our ePMR and HAMR portfolio. And so we look forward to sharing more of that exciting news next week.

The next question is from Amit Daryanani with Evercore.

This is Hannah on for Amit. I was just wondering, are there any notable investments related to HAMR that are currently flowing through COGS or operating expenses? And should we expect those costs to roll off or normalize as HAMR begins to ramp?

Yes, we have been working on HAMR for the last 10 years, and the engineering teams are making good progress. So there is no change there. We will continue to work on those programs, and we will, in general, continue to innovate and make performance improvements to our programs, continue to drive higher capacity drives and those investments will continue.

As it relates to the gross margin, we haven't started the HAMR ramp yet, but we are confident once we start ramping HAMR that will be neutral to accretive to our gross margins.

Yes. Maybe just to add on to what Kris said, even with the HAMR ramp that we anticipate will happen at the start of calendar year '27, our CapEx as a percentage of revenue on a run rate basis will still be within the 4% to 6% range.

Did you have a follow-up?

No.

The next question is from Karl Ackerman with BNP Paribas.

Rose mid-teens in 2025 and is projected to advance double digits again in 2026 as agentic AI is supposed to drive a cyclical recovery in front-end conventional servers. But in your case, because hard drive units are highly correlated to demand for conventional servers, and you're also seeing a content uplift from these new drives. Do you believe Agentic AI demand can enable you to exceed your long-term exabyte growth CAGR of low 20s?

Yes. Thanks for the question, Karl. Well, I think we've definitely seen exabyte growth over the last few quarters in the low 20s, as you've highlighted. Actually, we see as the AI value changes from model training to inference, more data is going to get created as a result in order to enable the inference delivery, more data needs to get stored as a result of that data getting generated as well.

And if you look at the economics of being able to deliver inference at the right cost structure to drive mass scale adoption, again, a lot of that data that's getting generated and require storage will be delivered -- will be stored on hard drives as they are, as we've highlighted in the past, where hyperscalers really are masters of managing the economics and moving data across the different tiers of SSDs HDDs and tape as well. So from our perspective and the conversations that we've been having with our customers, inference is definitely going to drive a significant amount of data storage requirement, and that's really positive for HDDs going forward.

Do you have a follow-up, Karl?

If I may, Ambrish, just going back to HAMR, it sounds like you've pulled in the progression of HAMR, at least your first primary customer. Can talk about the interest beyond your initial customer given the robust hyperscaler demand for exabyte capacity?

Yes. Thanks for the question, Karl. As we've mentioned, we are starting qualification in the first half of this year. We've already started that with one hyperscale customer already this month, and we will be initiating another one -- initiating qualification with another hyperscale customer relatively soon.

The next question is from Thomas O'Malley with Barclays.

Just a follow-up on some of the comments from the preamble about acquiring some IP, I think, on the laser side. Could you maybe give us a little more detail on that, maybe the size of the purchase? And then what in particular you needed to add on the laser side that you felt like you need to go out and do a deal?

Yes. Thanks for the question. Well, unfortunately, the terms and conditions of the deal are confidential. So we can't really share too much about that. But we are excited about acquiring this technology. We'll share again more of that next week at our Innovation Day.

But what I will say is it's definitely going to give us the benefit of taking much less real estate in the drive, right? And that will actually help with manufacturability in terms of reliability. And we also see that with this innovative technology, energy requirements to support the lasers will also be reduced compared to the conventional laser diodes. So we're quite excited about the -- both the IP and the capabilities that we've acquired.

Do you have a follow-up, Tom?

I do. With NVIDIA's addition of the KV cache offload and the NAND attach that's thought to go with that, I was curious if you guys have been engaging with any large hyperscalers or any large procurers of storage for any kind of solution that would maybe attach on to custom silicon deployment, say, something that brings the hard drive a little bit closer to some of the accelerators, if there's a road map for those or if you're engaging in that way with any customers?

Yes. No, thanks for the question. Again, I think the initiative that NVIDIA has been driving is really to help accelerate inference capability. And as I've highlighted, as a result of that, the velocity and the volume of data is going to get generated much more rapidly. And the benefit from us, obviously, will be able to require -- it will require a lot more storage, which obviously HDDs are well suited with the superior economics.

We are working on, as we've highlighted in the prepared remarks on interesting innovations to improve both our bandwidth and throughput of our drives. Again, something we are looking forward to be sharing with all of you next week as well.

The next question is from Vijay Rakesh with Mizuho.

Kris, pretty phenomenal numbers here. Just a quick question on the HAMR side. Are you expecting to pull in the HAMR road map time line given how tight supply is, et cetera?

Yes. We've pulled in the qualification already by half year. And we've started the qualification process with one customer. As I just mentioned earlier to Karl's question, we will be starting a qualification with a second customer imminently on qualification. Obviously, getting HAMR and higher capacity drives to our customers are a key part of the approach that we're taking to meet the strong demand for exabytes from our customers on HDD.

But it's also very important to remember, we also have started the qualification of our next-generation ePMR drives. And those products have shown the ability not to only deliver very high capacity per drive, but also to be able to support a high degree of scalability and manufacturability where we are able to deliver large volumes of drives to our customers. So, last quarter, we delivered over 3.5 million drives. And this quarter, we're looking to deliver close to 4 million drives.

Got it. And then on the gross margin trajectory, I guess, with the incremental 50% drop-through, when you look at HAMR ramps, any thoughts on how we should look at those margins? I guess you might call it on the Innovation Day, but any preliminary thoughts there?

Yes. I mean, as we've consistently highlighted, we see the transition once HAMR gets to the same scale as our ePMR portfolio, the gross margins for HAMR will be neutral to accretive from what we have with ePMR.

The next question is from Steven Fox with Fox Advisors.

I was wondering if you could maybe talk about the revenue per exabyte in the quarter compared to last quarter and a year ago in the sense that how much of the change quarter-over-quarter and year-over-year is related to change in mix? And then I had a follow-up, if I could.

Yes. Maybe I can start off here, and Kris might want to add in. Look, the big driver of our sort of revenue per exabyte both year-on-year and quarter-on-quarter is related to our cloud segment. So, our big hyperscale customers, we see very strong demand from that segment. So, obviously, that's driving a lot of the bits and the revenues associated with that.

And in that segment, as Kris has highlighted, the pricing is stable. So, in fact, it was up single digit last quarter and year-on-year as well. So that's a positive trend that we continue to see, and that's going to be a growth driver for the business for the year and probably for the next two years as well.

And if I could just quickly follow up. Can you just -- is there any commentary on how successful you were in terms of maybe getting out more exabytes during the quarter than originally planned for or whether through quicker customer qualifications or your own efficiencies? Any update there would be helpful.

Yes. Well, last quarter, we delivered 215 exabytes, right? That was up 22% year-on-year. And again, a lot of it is being driven by our cloud portfolio. As we've highlighted, we shipped over 3.5 million units of our current ePMR products that go up to 32 terabyte. So it's a clear recognition of the stability, quality and scalability of that product. So we will continue to do that, and we look forward to ramping the next generation of ePMR and HAMR in the coming quarters to better support the customer demand.

The next question is from Ananda Baruah with Loop Capital.

On cost down, you mentioned that I think, Kris, December quarter was 10% year-over-year down. And with UltraSMR becoming a larger portion of the ship and then with HAMR coming on sort of margin neutral to positive, do you think that cost down, I think it's classically been about 10% year-over-year. Do you think that can increase in coming years, cost out increasing per exabyte ship?

Yes. So, currently, it's on or about 10% cost per terabyte or exabyte reductions. Obviously, we will continue to innovate, continue to push to higher capacity drives, continue to upshift our customers to adoption of those higher capacity drives, including UltraSMR. And all of those actions will lead to further cost reductions on a cost per terabyte.

I think it's fair to say that's on or about 10% is a good number. And as we potentially accelerate our road maps, we could potentially drive that higher.

That's super helpful.

Do you have a follow-up, Ananda?

Yes, quick. This may be one for next week actually. But just interested in understanding how far up the areal density stack do you think you can get CMR and UltraSMR before you really get HAMR going?

Yes. I think that's something we are looking very much forward to sharing with you next week. So we look forward to seeing you there.

The last question is from Krish Sankar with TD Cowen.

This is Eddy for Krish. You mentioned you had three LTAs for 2027 and 2028 that are volume and not price based. I do wonder what is the reason these contracts are not locked in price, especially given the very tight supply? Is it the customer who prefer not to lock in price? Or is it you guys who prefer to have the flexibility? Any high-level color would be helpful.

Yes. Thanks for the question. Just to clarify, we have two customers that have LTAs through calendar year '27, one customer that has an LTA through calendar year '28. These LTAs have both price and volume conditions in them.

Okay. Noted. That's great color.

This concludes today's conference call. Thank you for joining. You may now disconnect.

Good morning, everyone. I am -- my name is Erik Woodring. I lead U.S. IT hardware research based out of New York for Morgan Stanley. Delighted to kick off day 1 here at the Nasdaq Conference with Western Digital.

Before we get into things, first, please see the Morgan Stanley research disclosure website at www.morganstanley.com/research disclosures for important disclosures. If you have questions, please reach out to Morgan Stanley's representative.

So I'm pleased to be joined today by Irving Tan, CEO of Western Digital; Kris Sennesael, CFO of Western Digital. Before we start, Kris has a safe harbor statement, and then, we'll get into things.

Yes. Thanks, Erik, for hosting us here. And so before we start, just a couple of remarks today. We will be making some forward-looking statements in our discussions based on management's current assumptions and expectations, including with respect to our product portfolio, business plans and performance, market trends and dynamics and future financial results.

These forward-looking statements are subject to risks and uncertainties. So please refer to our most recent financial reports on Form 10-K and other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations. We will also be making references to non-GAAP financials, and a reconciliation to GAAP and non-GAAP results can be found on our website as well.

Perfect. Awesome. Thank you, guys, for joining today.

I figured we'll start at the very top, and Irving, demand for HDDs is very strong right now. You've kind of conveyed a lot of messages around demand strength and visibility. I'd love to understand at a more granular level why demand is so strong, meaning are these some of these new AI video platforms supporting incremental strength in HDDs? Is it companies projecting out kind of token growth? What is it exactly that it feels like over the last 3 months, demand has really kind of ticked up? Why is that?

Yes. Thanks for the question, Erik. Actually, we've seen demand probably tick up probably for the last 3, 4 quarters. Now, at the highest level, there are 2 primary drivers, right? The first is just the ongoing transition of more and more on-prem storage to the cloud. So it's sometimes forgotten, right, that just basic cloud storage services continue to grow at a reasonable pace, somewhere in the mid-teens exabyte growth CAGR is what we're seeing.

Then, obviously, that's been turbocharged more recently by AI. And within the AI, we see 2 -- again, 2 key drivers. One is the value of data is increasing because of AI. So more data is actually getting stored even as more data is getting generated, so we traditionally are looking at data generation rates that are going to increase over the next 3 years by 3x.

Historically, the percentage of data that gets stored is roughly in the 2% to 3% range. We're seeing that more move up to the mid-single-digit range. So just the value of data to train models, to drive inference is increasing.

Second, I would say it's video, right? And the big driver of storage has always been in video, that's whether it's in the basic cloud or in AI as well. And obviously, there's been a lot of talk around AI-enabled videos for consumers or for marketing purposes like Sora and the likes. But people tend to forget also there are major industrial applications that are driving a lot of video. Take autonomous vehicles for a while.

We just came from Arizona where you have tons of Waymos driving around, right? All those vehicles are storing the telemetry of the vehicle. They're recording all the videos of their journeys, both for regulatory reasons, and they're storing the videos to actually train the models, right, to enable the vehicles.

What's interesting, these autonomous vehicle companies are running out of roads to train their models because they keep plowing down the same routes. And so they don't learn anything new. And they're using that video and AI to generate synthetic video to further train the model, and so there's a self-fulfilling prophecy of more video getting used, more video getting generated as well to train these models. And we see these are the primary drivers of growth.

Okay. Okay. And then, when we talk about AI as an incremental HDD demand driver, we don't necessarily hear or see a number of, like, what I would call, AI-native vendors or neo clouds or AI labs coming to market necessarily with huge HDD orders. So just what do you hear from them in terms of their storage architecture? Because seemingly, that is a large kind of new market, clearly, AI-centric. What's the message you'd send on demand coming from those players?

Yes. You're right in the sense that the bulk of the demand is coming from what I would say the more established players, like the top 4, 5 hyperscalers that have really dominated the market. And if you take the neo clouds, the CoreWeaves of the world, xAIs of the world, I mean, their primary focus has to be -- has been putting their investments in GPUs and in memory, right, which makes sense because that's the business model. They are trying to drive inference. They're trying to drive compute resources. But what they do, do is they connect back into the typical hyperscale storage environment, right? So as an example, CoreWeave is a big user of Amazon Web Services S3 as an example.

And so you're still -- we are still seeing the benefit of that, just not necessarily with the neo clouds, but they're leveraging the hyperscalers who have really perfected the infrastructure management and the economics of storage. And we are seeing -- we are having conversations with them of late because I think what they're also figuring out, as they evolve their business model purely from compute to memory, storage has to become a key component of that just in terms of the service they provide and the economics of their business.

Okay. Okay. So maybe Kris, we will bring you into the conversation as well with you, Irving. And what does -- ultimately, my question is, what does that mean if we think about exabyte growth in kind of the time frame that you laid out at your Analyst Day earlier this year? You kind of outlined an exabyte growth CAGR of between 15% and 23%, 15% with your CSPs, kind of 8 points of AI potential upside. Is there an upward bias to this metric? Just how do we think about it, again, just given what we hear about HDD demand being so strong?

Maybe I can start off, and Kris can jump in. So, at our Investor Day back in February, we laid out a base exabyte growth case of 15% CAGR. I know it's a very conservative case to be very frank, really just driven by the ongoing organic growth of the cloud. And then, we laid out an upside case, what we call the AI uplift case, which was a 23% exabyte growth case, right? So an 8-point delta between the 2. And that was driven by what we thought would be the AI use cases at that time, right?

I would say we've been pleasantly surprised. We are seeing the trending probably more in the mid-20 range. Last quarter, our growth rate was 30% exabyte growth. And we're getting better and better visibility in terms of the growth rates going forward. So as we've shared in our last earnings call, we now have firm purchase orders from our top 5 hyperscaler customers that extend throughout the entire calendar year '26. One of those 5 have given us orders for all of calendar year '27 as well. So we definitely see the trajectory continuing.

Okay. And I want to also make sure we touch on supply, a hotly debated topic in the market. You guys -- the industry alongside yourselves, have been very disciplined with supply. Maybe for the benefit of the audience -- I know you get the question of would you expand supply, would you expand unit supply. Maybe the way I want to ask it is, is there something that could convince you to expand unit supply above and beyond where the run rate is? And what would that be if there was anything?

Yes. I mean, we've been quite clear and explicit to say as things stand, we are not adding any unit capacity. Our focus is really on improving aerial density. Maybe for the benefit of the folks in the room who may not be that technical, aerial density is about increasing the amount of exabytes we can store on a single platter within the drive. And so as we increase aerial density, we can go from our current flagship drive that's 32 terabytes. We can go to 40, 50, 60, eventually up to 100 terabytes per drive. So our focus is really on supporting our customers' needs by driving aerial density.

The reason, Erik, is those contracts and POs that we've -- I just shared with you, they are giving us orders based on the amount of exabytes we can support them with. They're not unit-based, right? What customers really want is can you supply me the amount of exabytes? And the higher density we can provide in terms of exabytes per drive is good for our customers because it's better TCO for them. They have better rack density. They have less power consumption. There's less real estate that's required. So there's less need to build out new facilities as well.

Okay. Which makes a ton of sense, and we'll kind of get into aerial density and technology innovation. I'd love to better understand -- so we've kind of talked high level demand. We just addressed supply, at least unit supply. When it comes to the totality of the interaction of those 2 sides of the market, how do we think about supply growth into calendar '26 and into calendar '27? Because you are coming out with new technology, you're moving up the aerial density curve. So seemingly, you should be able to meet the shortages better as you move along through this technology innovation curve. So just -- before we get into the specifics of the actual technology, just how do we think about the ramp in supply from your perspective from this year into the next 2 years?

Yes. I would say that, obviously, we are in a very tight supply-demand environment that will probably persist a bit into calendar year '26. And as you rightly pointed out, as we deliver higher capacity drives through aerial density improvements, we will catch up over time. So we are making investments in our head and media facilities to support the next generation of technology that we're bringing out to support the increase in head and media capability. And those are the big drivers of aerial density that will feed into the units.

Okay. And maybe just to add there, Erik, last quarter, our average terabyte per drive that we shipped was approximately 22 terabytes despite the fact that we are shipping in high-volume 32 terabytes. So in addition to that, the 22 terabytes was up 21% on a year-over-year basis. And this is for our nearline drives. So this clearly illustrates that we can increase our average capacity per terabyte on or about in the mid-20s in line with the expected demand growth.

Right. Okay. Okay. So I don't -- I was going to ask you a question on that, but I'll stay to my track, which is going back to maybe the comment you made earlier, Irving, about visibility. So you generally have nearline bits accounted for from your top 5 customers. You have 1 customer through calendar '27. Can you maybe just elaborate, one, is that kind of the totality of the visibility that you have today? And really, the second part of that is how do you protect yourselves from the risk of kind of over or double-ordering from those customers? Because clearly, if they want bits, they'll try to get the bits from you, but you don't want to create an industry glut. So how do you protect yourself?

Yes. I think it's a really good question. I mean, the industry has been quite notorious for not doing a good job. So we're very mindful of that. As I mentioned, the orders that we have for the 5 customers in '26 and '27 are pretty much firm. There are specific commercial terms in them if the -- with financial teeth associated with them.

We are in discussion in '27 for not only the remaining 4, but extends probably to the top 10 customers that we have to look at longer-term contracts that extend all the way up to calendar '29. One even is talking to us -- it's '28, sorry. One of them is talking to us about extending the contract to calendar year '29 as well. So that gives us a lot more visibility.

Now, is there a risk that may be double-ordering? I mean, that's always on people's mind. Right now, we are quite confident that isn't because we have a good sense of the inventory that they have. And pretty much everything we're shipping to them is being racked and stacked and shipping to data centers almost immediately. It's a just-in-time model. That's how tight things are.

But if you extrapolate forward, obviously, there's always a risk in that. But the way we look at it is we use our customers' demand signal, what they're asking us to supply to them in terms of exabytes. There's just 1 data point in how we run the business and how we make supply investment. So we do use our own internal analysis. We look at how many data centers are being built over the next 4 to 5 years, when are they going to come on stream.

We have some internal proxies in terms of the gigawatts of a data center, how much that would translate into HDD capabilities and various other metrics that we use as well. So we form our own opinions, and that's basically what forms the investment decisions that we make.

And the other side of that is it's just even fascinating to me that your customers are giving you that level of visibility. Historically, you could argue the CSPs weren't always the most accurate forecasters of their own demand. What do you hear from them? Why are they giving you this level? Is it purely a supply-driven dynamic? Or is it more to that?

Well, I would say they're getting better in terms of forecasting. They still have ways to go, just like we do as well. I think what has fundamentally changed both for us and for them. I mean, if you take us as an example, first of all, today, 90% of our business is in the data center. So we are a data center company. I mean, if you go back 5, 6 years ago, only 50% of what we did was in the data center, and it was primarily through OEMs, right? Today, it's 90%. The vast majority of that was large hyperscalers both in the U.S. and in Asia. So it's a very -- we are a very different company today.

We are -- we basically are aligned fully to the hyperscaler data center environment. And so we've sort of restructured our organization as well to specifically cater to the hyperscale environments of our top 5 customers. As an example, we have dedicated resources across sales, product management, engineering, customer technical support, firmware to specifically support the architectures that they're building for today and to tomorrow.

If you take the customer viewpoint also, with AI, in particular, they've realized that actually storage is a critical piece of the AI infrastructure stack, right? And if you don't have enough storage, you're not going to be able to continue to train models. You're not going to be able to deliver inference. You're not going to be ultimately be able to monetize AI. So the relationship between us and the hyperscalers have really changed a lot, where it used to be very transactional or it used to be predominantly true to OEMs. It's a much more direct relationship.

They're giving us greater visibility around the use cases they're going to bring out over time, how that's going to translate into storage requirements and giving us a lot more conviction in terms of the exabyte demands that they're projecting. It's really underpinned by a degree of fact.

Okay. Okay. No, that's helpful. So let's move to the technology innovation side of things. It's a kind of exciting time. We've -- for the entirety of my coverage, we kind of -- the time covering this space, we've heard about HAMR coming, and now, it's here/on the cusp. And so I believe at Western Digital, you are qualifying up to 36 terabyte UltraSMR in the first quarter of next year, launching that in the second half of next calendar year and then qualifying HAMR next year to launch in volume in the first half of calendar '27. Have any of those targets shifted? Any change in how to think about the qualification process?

And lastly, when you say in volume, I think there is a specific focus when you say in volume because you want to ship at least 1 million units per quarter at least. So just unpack all of that for us.

Yes. So we've taken a slightly different tack in terms of how we're introducing technology into the market, especially given how critical our technology is to the business of our key customers, the hyperscalers. And again, without overcomplicating things, there are 2 primary recording technologies in the HDD industry. One is called PMR, perpendicular magnetic recording, and the other one is called HAMR, the acronym is heat-assisted magnetic recording.

And while some of our peers have gone head on into one particular approach, i.e., HAMR, we've taken a slightly different track by derisking the transition. We are committed to HAMR, as you pointed out. We've accelerated our HAMR qualification and pulled it forward by 6 months. And HAMR is critical to enable the industry to get to 50, 60, 70. We have been able to see that it will deliver a 100-terabyte drive in the not-too-distant future.

But in parallel, we continue to deliver the highest volume, highest capacity drives to our PMR portfolio as well. And again, there, we pull forward the qualification of our next-generation PMR drive to Q1 of calendar year '26. As you pointed out, the current roadmap is for it to be a 36-terabyte drive. We have brilliant engineers within the team. I think there's an opportunity we can eke out a bit more. So we'll share a bit more at the innovation data we are hosting in New York on February 3rd, so if you have an opportunity to join us live or watch the webcast as well. But the whole approach is to really derisk the transition of any technology to our customers.

And secondly, it goes back to the point I made at day 1 as much exabytes as they can get, right? And therefore, our view is how do we deliver the most amount of exabytes to them at scale. And we feel that for the time being and for a period of time, that's true doing -- using our PMR portfolio. The data point on that is that our current generation of PMR drives, which is now 3 quarters into production. The first quarter after qualification, we shipped 800,000 units. In the second quarter, we shipped 1.7 million units. Last quarter, we shipped 2.2 million units and delivered 70 exabytes.

So if you do the math, it's north of 30 terabytes per drive. As we shared in our investor call this quarter, we will be shipping north of 3 million units, probably somewhere in the 3.4 to 3.6 million range. So it shows you the scalability, the confidence in the customers having the reliability of the products and its ability to deliver exabytes at scale. And we'll continue to do that. And eventually, we will transition to have both PMR and HAMR in the market, and then, over time as we get to higher and higher capacities, we'll transition fully to HAMR.

Right. Okay. And maybe as a follow-up, this is a question I get fairly often. You might have just answered it, but I want to give you the dance to answer this specifically, which is there's an investor debate whether your competitors coming to market with HAMR before you guys put you at a competitive disadvantage because you couldn't technically deliver the same TCO. What would be your rebuttal or help us better understand why not being first to market in HAMR doesn't put you at a competitive disadvantage?

Yes. I would think the first thing I would say is if I talk -- go back to those contracts that we mentioned customers have given us for '26 and '27, they are exabyte-based. They are technology agnostic. Customers don't care whether you're shipping them -- I mean, none of them wake up in the morning and say I need to get a HAMR drive or PMR drive, just give me exabytes. That's the focus right now. But what they do want is a partner that is able to deliver to them exabytes at scale that's reliable. That's not going to have any issues 1 year down the road, 2 years down the road, 3 years down the road in their hard drive fleets where they have millions and millions of units.

So we actually don't feel we're at a disadvantage. Obviously, we get the benefit of that any fast follower gets because we avoid some of the pitfalls, and therefore, we've been able to compress the qualification cycle. But for us, it's really a function of how do we support our customer TCO value and the economics of our business. And we actually think that for the near to medium term, our ability to do that through PMR, and then, eventually through a combination of 2 before we transition for you to HAMR is the most optimal for our customers and for us.