Bitdeer Technologies Aktie

Good day. Thank you for standing by. Welcome to Bitdeer's Second Quarter 2026 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to turn the conference over to your first speaker today, Pretesh Dahya, Head of Investor Relations. Please go ahead.

Thank you, operator, and good morning, everyone. Welcome to Bitdeer Technology Group's Second Quarter 2026 Earnings Conference Call. Joining me today are Jihan Wu, Founder, Chairman and Chief Executive Officer; Haris Basit, Chief Strategy Officer; and Michael Potter, Chief Financial Officer.

Today's call will begin with Haris providing a review of our recently announced Tydal Norway colocation lease agreement, followed by Michael, a review of our business segments and second quarter financial results.

Before management begins their formal remarks, I would like to remind everyone that during today's call, we may make certain forward-looking statements. These statements are based on management's current expectations and are subject to risks and uncertainties, which may cause actual results to differ materially. For a more complete discussion of forward-looking statements and the risks and uncertainties related to Bitdeer's business and industry, please refer to the company's filings with the U.S. Securities and Exchange Commission.

I also want to note that we will be discussing certain non-GAAP financial measures and operating metrics today. A reconciliation of these measures to those directly comparable GAAP measures is included in our earnings release issued earlier today. These non-GAAP measures should not be considered in isolation from or as a substitute for the most directly comparable measures paired in accordance with GAAP.

As a reminder, changes to the fair value of our digital assets are reflected in GAAP net income and may introduce noncash volatility into our reported results. With that, I will now turn the call over to Haris.

Thank you, Tesh, and good morning, everyone. This is our first earnings call since we announced the execution of our $4.7 billion colocation lease at Tydal, Norway, and I would like to begin by putting that agreement in context. Over the past several years, we have worked to build a power infrastructure portfolio that we believe is well positioned to support both our Bitcoin mining operations and our expansion into AI infrastructure. Tydal represents an important step in converting that portfolio into long duration contracted revenue, and it establishes AI infrastructure colocation as an additional pillar of our business alongside our AI cloud, Bitdeer Mining and ASIC development and manufacturing operations.

We are pleased with the terms of this agreement and with the caliber of Volta as our counterparty. We view Tydal as an important proof point for this strategy, and we intend to pursue additional opportunities of this kind as they arise. Agreements of this scale and complexity reflect the work of our team over the past several years, and I want to thank everyone involved for their efforts in reaching this milestone.

With that, let me walk through the terms of the Tydal transaction in more detail. On August 4, we announced that our subsidiary, Tydal Data Center AS has executed a 16-year colocation lease and services agreement with Volta at our Tydal campus in Norway. Under the agreement, we are delivering 121 IT megawatts, supported by approximately 133 gross megawatts. This will be spread across 4 existing data halls and will be configured to run NVIDIA GPUs. The lease is expected to generate approximately $4.7 billion in contracted base term revenue over the initial 16-year term. An renewal option, if exercised, would increase the potential contract value to approximately $8 billion over 24 years with the tenant termination right at the end of year 10.

The leases structured has a modified growth arrangement. Volta pays a combined base rent and service fee with a 3% annual escalator and electricity costs are fully reimbursed on a pass-through basis, which protects our margin from energy price volatility and provides a highly predictable cash flow. Over the 16-year based term, we expect this agreement to generate average annual revenue of approximately $2.4 million per IT megawatts.

Importantly, Volta's obligations are anticipated to be backed by an institutional grade credit structure. This anticipated credit enhancement meaningfully reduces our counterparty risk and improves the financeable quality of these contracted cash flows, which supports our plan to access attractive debt financing terms to fund the remaining development as Tydal. Our remaining capital expenditure is approximately $500 million, which we believe is significantly more capital efficient than a typical greenfield data center build.

We expect to raise project level financing to fully fund our remaining Tydal capital needs and to provide significant additional capital. Delivery is structured across 2 equal-sized phases targeting December 31, 2026, for the first phase and March 31, 2027, for the second. A word on our tenant. Volta is a compute infrastructure developer focused on large-scale AI and data center deployments in power advantage markets. Volta has announced a $10 billion strategic partnership with an AI lab and a broader development pipeline exceeding 1 gigawatt.

In selecting Volta as our partner at Tydal, we evaluated their technical ability to execute a large-scale GPU deployment, the quality and enforceability of their anticipated credit support and their ability to move rapidly to match the anticipated Tydal construction time line. It is important to note that Tydal will incorporate leading-edge NVIDIA GPUs has one of the highest reliability grid connections in Europe, is 100% powered by renewable energy and has an extremely high energy efficiency with a PUE of approximately 1.1.

Our broader power and infrastructure portfolio stands at approximately 3 gigawatts of total global electrical capacity at the end of the second quarter, up approximately 12% year-over-year. Furthermore, we continue to evaluate opportunities for additional grid connected and behind-the-meter expansion sites globally across both new and existing sites.

Our objective is straightforward. Continue acquiring, building and converting powered infrastructure. We will share updates on our progress here when appropriate. I will now turn the call over to Michael to walk through our business segment updates and second quarter financial results.

Thank you, Haris, and good morning. I'm happy to join everyone for the first time as Bitdeer CFO. Execution remains our top priority. While the Titolease that Harris described has now been executed, there is significant work ahead to prepare for the first RFS date. We will update as appropriate as those milestones are achieved. The focus in our view is on the quality of altered demand, robustness of the financing structure and contractual protections. We believe this transaction compares very favorably on these dimensions, and we look forward to demonstrating that through continued execution.

Looking ahead, we see plenty of interest in various sites within our portfolio that could potentially be used as AI/HPC data centers. A key advantage of our model is the ability to utilize capacity for crypto mining and to an AI data center is ready, ensuring our power assets remain productive and secured.

Turning to our Bitcoin mining business. Self-mining hash rate reached approximately 73 exahash per second at the end of the second quarter, up approximately 342% year-over-year, supported by approximately 243,000 active self-mining rigs. This is an increase of approximately 113% year-over-year. We mined 783 Bitcoin in April, 921 Bitcoin in May and 990 Bitcoin in June. Total second quarter production was approximately 2,694 Bitcoin, an increase of approximately 377% year-over-year.

Our co-mining hash rate also continued to accelerate meaningfully during the quarter. This reflects our continued deployment of seal miners into third-party facilities. It grew over 260% sequentially. We believe our; combination of self-mining, co-mining and hosting gives us multiple channels to monetize our growing seal minor production. We have the flexibility to allocate hard work for the channel that offers the best returns as market conditions evolve. Our Sealminer platform continues to reinforce our structural cost advantage. The commercial launch of our A4 Ultra hydro unit operating at 9.45 jewels per terahash at the chip level continues to lower our internal cost per exahash.

During the quarter, we also launched the Sealminer Hydral our first machine designed for script algorithm mining. This broadens our product line beyond Bitcoin focused hardware. Our internal manufacturing capability means that we're not subject to third-party markups when deploying Sealminer rig into our own fleet. This remains a structural cost advantage relative to other mining operators.

In July, we broke ground on our first U.S.-based manufacturing site a 187,000 square foot seal minor manufacturing facility in Sparks, Nevada. This is expected to be completed by the end of 2026 and we'll be capable of producing 10,000 units. It is expected to create approximately 70 high-quality local jobs. At the Massillon, Ohio site, we have 174 megawatts of capacity currently online for mining. With reconstruction of 2 previously fire damage buildings underway, a significant portion of that cost has been recovered through supplier insurance coverage. We believe the associated capacity to be energized in phases during the third quarter.

In June, we broke ground on our Fox Creek, Alberta site, a $155 million investment includes a fully permitted 101-megawatt on-site natural gas power plant with grid interconnection. The site uses a closed-loop dry cooling system. Alberta's bring-your-own generation framework gives us the flexibility to curtail compute workloads and sell power back to the grid.

Our AI business continued to scale during the quarter. AI cloud annual recurring revenue reached approximately $76 million at the end of June, an increase of approximately 77% quarter-over-quarter, utilization was approximately 95% across 4,248 deployed GPUs. We also signed a new 10-year lease for 21.7 IT megawatts of capacity in Malaysia with expected in the first quarter of 2027. It is designed to support 128 NVIDIA GB300-NVL72 systems.

On the product side, we deployed NVIDIA's Nemotron-III model onto our bitter AI model studio on the first day of its launch.

Turning to our financial results. Second quarter revenue was approximately $228.8 million, an increase of approximately 47% year-over-year and approximately 21% sequentially. The year-over-year growth was driven primarily by the continued expansion of our self-mining hash rate and the associated increase in bitcoin production, along with accelerating contribution from our AI cloud business which contributed $14 million, an increase of approximately 284% sequentially.

Total gross profit was negative 8.5 million with a gross margin of approximately negative 3.7%. Importantly, this represents a $30.5 million sequential improvement that demonstrates the operating leverage of our vertically integrated model. This recovery was driven by 2 key factors: normalized seasonal power costs, which dropped approximately 15% sequentially and an improvement in our blended fleet efficiency to 15.8 jewels per terahash.

On a year-over-year basis, our gross margins reflect continued but moderating pressure from the additional depreciation expense from our expanding mining fleet and the still challenging hash price environment. Adjusted EBITDA was approximately $31.1 million, an increase of approximately 575% over the year and approximately 116% sequentially. This sequential improvement illustrates the operating leverage of our vertically integrated model as both hash price and power cost dynamics continue to stabilize. Operating loss in the quarter was $101.7 million and net loss per share was $0.37.

Turning to the balance sheet and cash flow. Net cash used in operating activities was approximately $158.5 million, an improvement of approximately $188 million sequentially. This was driven by the capitalization of Sealminer related inventory to PP&E as we're allocating those rigs for internal use. We ended the quarter with approximately $496 million in cash, cash equivalents and restricted cash compared to approximately $298 million at the end of the first quarter. Total long-term debt was approximately $1.8 billion, a reduction of approximately $78 million sequentially. The increase in cash primarily reflects $457 million in proceeds from our at-the-market equity program during the quarter. We view this as a prudent capital raise, establishing the liquidity reserve necessary to execute on our AI/HPC pipeline, power land acquisition priorities and to help ensure that our Tydal site progresses on schedule.

Today, we filed a new shelf registration statement since our previous shelf, which was about 1.5 years old, had largely used up its capacity. We now qualify as a well-known seasoned issuer, which makes us eligible for automatic shelf registration. Considering the increased size and scale of our business, we also filed a prospective supplement to move our existing ATM program to the new shelf and take down $1 billion for offering under that program.

Looking ahead, we expect to broaden our sources of capital with much of our remaining 2026 financing needs to be met through project level debt financing anchored by our previously discussed goals of accessing the debt markets for Tydal. We expect this debt financing will also unlock significant incremental liquidity to support our broader AI/HPC pipeline.

Our preference is to prioritize non-dilutive project level financing over equity issuance, wherever the underlying contracted cash flow support it. Consistent with this, in Q2, our Bitcoin wave per spend was funded entirely through cash generated by our mining operations and debt collateralized by our wafer bank, not equity issuance.

In terms of guidance, we are revising our full year crypto mining infrastructure capital expenditures to $200 million to $280 million driven by additional infrastructure development opportunities we see in North America. As a reminder, this guidance excludes CapEx for Sealminer hardware GPU, AI cloud or colocation development.

On expenses, we continue to actively manage our overhead as we scale the organization to support our expanding co-location and AI cloud businesses alongside our core mining and ASIC manufacturing operations, and we expect our general and administrative expense run rate in the second half to reflect the incremental head count and infrastructure needed to support that growth.

To summarize, the second quarter was a solid quarter for Bitdeer. In the past few months, we have clearly demonstrated our ability to deliver against our HPC/AI objectives. We executed our first major AI infrastructure colocation lease agreement at Tydal, officially launching our colocation data center business as a core pillar of our company. We reached this milestone on compelling terms and with the partner, Volta, we are excited to work with. Our underlying power portfolio continues to scale. Our AI cloud business is demonstrating strong momentum, and our Bitcoin mining and Seal miner platforms illustrates the benefits of vertical integration. Thank you for joining us this morning. Operator, please open the call for questions.

[Operator Instructions] Our first question coming from the line of Nick with B. Riley Securities.

Guys, congrats on getting the first deal done here. You spoke to the pipeline and looking at new opportunities globally. I was just hoping you could frame up kind of some of the opportunities you're seeing outside of the U.S., how near term of these opportunities could be?

I want to make sure I understand your question. You want to know what our opportunities are outside the U.S.?

Exactly. Yes. Just hoping to get a better sense for the pipeline ex U.S.? How many megawatts are you assessing today? And what's the earliest that some of these megawatts could be brought online?

So I'm sure Jihan could give some additional color on this. But we do have a lot of large Asia-based potential customers that are slowly starting to turn online. Our existing GPUs in Asia are quite used up, and we're expecting to add more capacity as customers come online. I don't know if Jihan has anything additional you want to talk about.

On the [indiscernible] deployment, I think the near future, mostly in Malaysia. We have data center already signed up. And we also have other opportunities in active discussion. And the demand side is also very strong. And the ,it is multiyear and highly profitable contract. And now is our execution. So we are quite busy on executing those AI contract ends in Malaysia. And then Norway will be in the next. We reserved like 15 megawatts a little than 50 megawatts in out in Norway ourselves, that will be a little bit later than in Malaysia. And it is expected within next year.

And then we will have some of our U.S. data center activated like in Tennessee and Washington.

That's helpful. Would you consider any new sites outside of the U.S. for co-location purposes? Or are you seeing any opportunities for expansion in your European footprint on the colocation side?

We are actively looking for other part asset opportunities in Europe. But since we haven't done any deal yet. So I think it's too early to really talk about Europe. And right now, our focus should be on execute the deal in Tydal. I think that's very important for us. If we can execute that, we will generate revenue and cash flows. I think that's -- and this will also generate kind of credibility of our data and execution. So I think that's super important for us right now.

Understood. And then maybe one more, if I could. I saw, if I read correctly, just at the Knoxville site, there was a full redesign of the project, and I believe ready-for-service time was slightly pushed out. Can you just talk about -- maybe provide some additional background on what drove that? Was that potential customer driven? Was that just kind of better fit in video architectures? What was the reason for that?

I don't know if you want to talk about that, Haris or you want me to jump in on it.

I think probably best for you, Michael. .

So when we have been going to market and looking at the requirements around our Tydal Norway side, originally, we had expect to Tennessee to be 2 different projects. but the market demand is more for a single, more monolithic amount of megawatts offered in one place. So we redesigned it instead of having 2 separate smaller sites inside our bigger site to dispute be one data hall complex. So that was the redesign we did, and it's reflected on -- from the discussions we've had with potential customers and what their requirements are.

And the expectation actually aligns with the second test. So that will be low Phase I, Phase II, it's only 1 phase, and they were all the online and at the same time. So if we look at the expected Phase II, I think that's actually not pushed down a lot, but there will be low Phase I early activation.

Our next question in queue coming from the line of Mike Condal with Broadman Capital.

Could you spend a minute on how you decided on Volta and kind of what maybe their ultimate demand is? And then secondly, Rockdale and kind of Clarington, what are next steps at both of those sites?

Maybe I'll start with the Volta side and then have Michael or Jihan speak to the other sites. So why we chose Volta was really they have done a great job actually in finding innovative ways of addressing this business in both the capital markets, the customer that they have obtained and their contracts for the customer. They also were able to move rapidly, which is something that we wanted to see at the Tydal site since that site is coming online in just a few months.

And in general, the commercial terms were also very favorable. So we thought that they were the right choice for us at the Tydal side. Their ultimate demand can be much larger, but of course, they're just getting started. And -- so we -- the initial leases for 121 IT megawatts.

Got it. Then just an update on next steps at Rockdale and Clarington.

I don't know if Jihan wants to add some further color. But we're continuing to build out the power infrastructure as per our previously discussed time line in Rockdale, that was something that was predating the or the allocation of the power in the site. And we're preparing in the background what we believe we need to do if we want to do an AI data center at that site. Clarington, we're developing it for crypto mining right now. The power is available coming up soon, and we'd like to be able to make sure we use it fully. There's no big update on the lawsuit in Clarington now. The motion to dismiss that we had filed was turned down by the judge, which is pretty common in these pretrial things, and it's gone into a discovery now. We continue to believe that the lawsuit doesn't have any merit, and we continue to work on that.

Our next question coming from the line of Kevin Cassidy with Rosenblatt Securities.

Congratulations on landing Volta. Two things on that. Well, maybe with the Texas government putting a pause on the new data center grid, maybe Harris, you might have touched on that, but how does that affect your colocation opportunities in Texas?

Michael, do you want to answer that one?

It's still a little bit early to make strong comments on that because the actual criteria haven't come out yet. The one big site we have in Texas, which is Rockdale, that side -- all the activity there is pre-batch zero and not related to that, that we're working on. As the actual news gets out and the criteria come out, I think we'll be in a better position to be more specific in our comments.

Okay. Great. And with Volta is large cloud and worldwide, is there a chance to just expand your relationship with them to other sites? I mean there is that potential, of course, with Volta, but we haven't released any information on that yet.

Okay. Maybe I'll ask one other. How should we be model Sealminer rigs going forward? And for external sales?

So right now, the existing wafer inventory that we have, we've designated for internal use with the very challenging hash rate and our desire to expand our ability to do self mining and coal mining. We made the decision to use it internally. We have the capability and the capacity to deploy it, and that's the best way to get that working and earning us some cash. It is an example of our model being extremely flexible in that we do have a fair amount of powered land that's in earlier stages of development for other uses that we can quickly deploy cryptocurrency mining into it and make sure we generate cash off the power.

Also, if you keep using the power it's less likely that the utility will try and move it away from the land. So it's important that we can quickly react when we do get land into our portfolio and use it in the crypto mining that we do is a big advantage there.

Our next question in queue coming from the line of John Todaro with Needham.

Congrats on the lease. First question, just as it relates to the lease it was 133 gross megawatts signed. I think we had that site going for 225. Just trying to understand why Volta didn't go for the full amount, I mean are you keeping some for cloud? And if so, I guess, why and the strategy there? And then I have a follow-up.

So the full amount of power there is actually 180 gross megawatts. And so of that 180 gross megawatts, so we're leasing 133 gross, which would be 121 IT megawatts to Volta and then retaining 47 megawatts growth for our own AI cloud use there. So as you know, this market is dynamic. There's a lot of activity in the -- both the colocation and AI cloud space. We think there's still a lot of opportunities for us in Norway with the 47 megawatts that we've retained. And we haven't made final decisions on exactly how that would be deployed in for whose benefit in terms of the ultimate tenants there.

So we think that's significant potential upside for us, and we felt it would be beneficial for us to retain that power for ourselves.

Understood. And then just one on the credit guarantee. We've seen some in the past where you have kind of a neo Cloud and your offtaker is like an anthropic-like entity but they're still either the chip manufacturer or a hyperscaler involved in the credit guarantee. I guess, can you just walk us through a bit more how that process works and what maybe some of the other avenues you were looking at before ultimately settling on the way it's structured now?

Yes. So I don't want to speak about the alternatives that we looked at since it was our proprietary in many cases. But the letter of credit structure here, I think, is quite useful, and it might be quite useful to the industry as a whole because for the first time, it has backing or credit backing from someone other than, as you said, someone who's chip vendor or a hyperscaler. So it opens up additional funding opportunities. And it is a very high level of high rating for letters of credit from the types of banks that we expect these letters of credit to come from.

So we know this is innovative and new to the industry, but we think it's going to be something that catches on.

Our next question moves you coming from Brett Knoblauch Cantor Fitzgerald.

And I just have a few, Haris on the Tydal lease, I kind of talked about an additional $500 million of CapEx needed. Curious how much money you guys have put into that site, call it, excluding the additional CapEx you need.

Actually, let me ask Michael to answer that question for how much has actually been put in.

I mean, we've put hundreds of millions of dollars of development into the site already some of which has actually been spent around the AI data center and originally developing the site to be used as a cryptocurrency site. So we made a reasonably significant investment in there in the past which prepared us so that we could act very quickly when this opportunity came up. This is more of a brownfield or semi-developed sites in a complete greenfield. So it allowed us to act very quickly and meet the time lines that Volta was hoping for.

Understood. And then high level, how should we think about how you're going to allocate capital between maybe leasing some of the tower land you have and building your own AI or expanding your AI cloud business. I guess if we think about Volta and your AI cloud, to some extent, they probably compete against each other, you're both selling compute. Is there a preference for you to want to be bigger on the cloud side? Is this going to be a mix approach? I guess how are you guys thinking of it internally?

I think Jihan can give some deeper color on that. But we're remaining flexible. If there's a good colocation deal that we feel it's in the best interest of the company to do, we can still execute on something like that. And if we look at what we can do around owning the GPUs ourselves or providing the GPUs ourselves and serving a customer ourselves, it's something that we strongly consider as well. We have said that we're going to do one or the other as the exclusive way to do it Automate.

And then maybe if I can just follow up with one more. As I look at upon 3 gigs of kind of efficient power and power coming online over the next several quarters, I guess is there a site that we should be focusing on that is kind of most ready or the next catalyst to either be leased or either to be converted? Or anything we should think about the future catalysts ramp year.

I think Jihan talked about this a little bit earlier that in Asia, we have ability to grow quickly in Malaysia because of the new site that we're signed up in Malaysia. So in the very short term, I think Malaysia is the one you see the most activity on. If you look at the table, we provide Washington State and Tennessee are the 2 that are sort of designated specifically for that that we're actively working on and doing construction and such.

So in terms of any like soon news, those are 2 that are most likely to come out. The other site is really depending on where we are with our discussions with customers and what we decide to turn on that. And as that happens, we'll make sure we'll update everybody.

Our next question in the queue coming from the line of Mike Colonnese with H.C. Wainwright.

Congrats on the deal. Great to see. I have 2. First, on the AI/HPC business. So Harris, you touched on this a bit earlier, but I just wanted to follow up on it. So if you could just talk about the preferred use for the additional 47 megawatts of gross capacity at Title at this stage. It sounds like you're considering either the GPUs and service type model or another colocation deal. If you were to do a co-location type of contract, would Volta be interested in this incremental capacity? Or would you be marketing that out to new prospective tenants?

I mean I don't really want to speak for Volta here on this call, but there is a lot of interest from both in expanding our relationship, and we will consider it along with other options.

And then just given where GPU pricing is, obviously, you guys have had some success in that business as well. would you say you're more or less likely to deploy your own GPUs to use that additional 47? Or do you think at this juncture colocation model would be the preferred way.

We haven't made any final decisions for the 47 megawatts. But if you're looking for a forward-looking comment there, I don't know, Jihan, if you want to add some color there.

Actually, there are already some interested customers talking with those capacities. So for any GPU deal, we need to consider about the financing and the deployment technical together. So I think right now, it's not -- it's not quite the right time to talk about too much. Our execution I think [indiscernible] right now is more in the front of us. So which means if we got some interest from customers. And we will focus to discussing with Malaysia share capacity. And after that, we will start to execute those T 47 megawatts of air cloud.

And then one more for me. Just on the Bitcoin mining side. So [indiscernible] has been one of the few miners that is continue to aggressively expand capacity in this environment, really position you well for the next stage of the cycle here. Just curious how we should think about hash rate growth in the back half of the year? It sounds like you have the wafer inventory to develop, deploy, manufacture your own steel miners for your proprietary use. Just trying to think through the best way to model hash rate expansion from here.

I think in general, for modeling purposes, if you look at the steady increase we've had over the last few quarters, it will be similar to that where we being about the same amount of mining machines per month going forward. So it'll be -- I don't think we'll be greatly accelerating from where we are, but I think steady deployment at a reasonably high rate. as we've shown over the last few quarters.

Our next question coming from the line of Brian Kinstlinger with Alliance Global Partners.

This is Kevin for Brian. For the Tydal, what tasks does the Bitdeer need to accomplish between now and the beginning of both Phases 1 and 2?

Well, I mean, Phases 1 and 2 are full RFS, right, that will be completely ready for the customers' GPUs at that time. And so effectively, we have all of the long lead time items have been ordered. They will arrive -- the scheduled arrivals are before the respective RFS dates. We have to finish installing those there's a commissioning part of it that is also very critical. So it's in very good state right now. We've had an engineering analysis and it looks like it's we're going to hit our target dates. There's probably 100 things to do off the top of my head, I can't name what they are, but there's -- there's 100 people on site working on this, and it's moving ahead -- there's no specific item that stands out that is more critical than the others that could potentially cause a problem. So we feel like we're in a pretty good position to hit those dates.

Got it. And then Bitdeer recently announced it will lease 21.7 megawatt data center in Malaysia for AI cloud. Can you share some more details like the cost and how you finance the GPUs and any other infrastructure needed. When you expect this data center will be ready for your customers? And then -- maybe could you share a range at full capacity of what the annual revenue run rate would be for this data center?

Yes, it's a little bit early to talk about specifics and forward forecast on that. I'm sure that Jihan can give some additional color, but we do have several customers that are interested in it. And the GPU financing will depend on whatever contract we have and how it needs to be done based on that contract. When you have a good quality customer backstopping the use of the GPU. It's normally much more straightforward to find financing for GPUs.

Last question are coming from the line of Ben Summers with BTIG.

So Harris, you mentioned some exploring behind the meter opportunities. And I know you guys are doing some on-site generation in Alberta. But any additional color you can provide on preliminary conversations or steps you've taken to develop some behind-the-meter power moving forward?

I think it's -- other than I was talking about Alberta in the past, I think it's too early to talk about that at any other site. So if any of those things develop further, we can mention it in the future. We have talked about this being the first site, but that we do have the ability to expand it -- so this is sort of a test of concept for us in Alberta underneath the program they have there to bring your own power and the availability of stranded gas in quite a few locations in Alberta, which sort of ensures you get reasonable supply at reasonable costs. It is a site that is Fox Creek works well for us, we can expand in using the same model and the capability we've been developing internally. We have looked at other sites that we can do more in. And if we decide to do it, we'll update everybody at that time.

Super helpful. And then just on the AI cloud business quickly. We touched on the expansion there. Just curious what you're seeing in terms of term length for new GPU cloud contracts and maybe how that's compared or how that's changed over the past 90 days.

Haris, Jihan, do you have any comments on that?

Well, right now, what we are discussing with our customers on the contract mostly focusing on those long-term contracts, I mean, 5 years. We are not spending very much time on talking contracts shorter than that. And we will -- our expansion will vocally be supported by this kind of long-term contract. -- thank you.

[Operator Instructions] I will now turn the call over to Michael Potter for any closing comments.

Thanks, everyone, for joining us, and we look forward to speaking again soon. Operator, you can end the call.

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

Good day, and thank you for standing by. Welcome to the Bitdeer Technologies Business Update Tydal AS Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to turn the conference over to your speaker today, Tesh Dahya, Head of Investor Relations. Please go ahead.

Good morning, and thank you for joining us today. We appreciate your time and your continued interest in Bitdeer.

Before we begin, I want to note that today's call is a transaction-specific investor update. We are not discussing quarterly financial results and will not be taking questions at the conclusion of prepared remarks. A replay of this call, together with the press release and investor presentation we published this morning will be available on our Investor Relations website at ir.bitdeer.com.

Please note that this call may contain forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These statements involve risks and uncertainties, and actual results may differ materially from those expressed or implied. We encourage you to review the risk factors discussed in our most recent annual report on Form 20-F filed with the SEC as well as subsequent filings.

We will also reference certain non-GAAP financial measures and operating metrics on this call. Please refer to this morning's press release for additional disclosure. Joining me on the call today are Haris Basit, our Chief Strategy Officer; Michael Potter, our Chief Financial Officer; and Haakon Bryhni, Chairman of the Tydal Datacenter.

With that, I will now turn the call over to Haris.

Good morning, everyone, and thank you for joining us today for what is a historic milestone for Bitdeer. Today, we are officially announcing our entrance into the AI infrastructure colocation market. We have executed a 16-year data center lease and services agreement at our Tydal campus in Norway that represents approximately $4.7 billion in contracted base-term revenue.

With an embedded 8-year renewal option, the potential contract value increases to approximately $8 billion over 24 years. We are pleased to be working with Volta, Dell Technologies, NVIDIA, JPMorgan and their partners to provide one of the largest and most advanced AI data centers in Norway.

We have built what we believe is one of the most strategically positioned power portfolios in the world, including approximately 3 gigawatts of capacity across multiple continents in locations that are highly suited for AI and HPC infrastructure.

Today's announcement is a critical proof point that our portfolio can be converted into long-term contracted revenue streams with high-quality partnerships. Given the depth of our 3-gigawatt power portfolio, there are many opportunities to replicate and build on what we have achieved here today. We view this agreement as the first of a series of agreements that will progressively convert our power assets into contracted AI and HPC infrastructure revenue at scale.

Now I will walk you through the specifics of the partnership. We are leasing 121 IT megawatts supported by 133 gross megawatts to Volta Tydal AS, a subsidiary of Volta. Volta is a fully integrated AI infrastructure platform focused on building the utility of compute. Founded by former Brookfield executives, Volta has already established a global footprint with committed capital and top-tier talent across London, New York and Palo Alto. They are backed by institutional capital with notable investors, including Azora Capital, Andreessen Horowitz, NVIDIA and the family office of Michael Dell.

Volta's sole end customer for this site will be a leading AI lab. The entire 121 IT megawatts are being configured in 4 existing data halls to run NVIDIA's Vera Rubin GPUs with Dell Technologies serving as the technology provider.

The lease is divided into 2 equal-sized phases. Phase 1 is targeted to begin December 31, 2026, and Phase 2 is targeted to begin March 31, 2027. Bitdeer is in early phases of development for 2 additional data halls in Tydal that are outside of this lease, totaling 47 gross megawatts.

The target date for these 2 data halls is the second half of 2027. Given the level of interest in our power infrastructure and overall strength of the AI/HPC market, we plan to retain flexibility for the remaining 47 megawatts for AI/HPC use cases so we can optimize the strategic value of the Tydal site.

Volta's obligations are anticipated to be supported by a credit backstop issued separately by JPMorgan and top-tier global financial institutions totaling approximately $1.3 billion. This credit enhancement is structured without dilution to Bitdeer shareholders. It preserves our full ownership of the Tydal campus while enabling Volta with a durable, well-capitalized counterparty commitment behind the lease.

Bitdeer and Volta are well aligned in our vision and complementary in our capabilities and expertise, bringing together Volta's platform and customer relationships with Bitdeer's power and infrastructure position. This is the kind of long-term alignment we believe benefits both companies. The strategic significance of this agreement has the following 5 points. First, it converts power infrastructure into long-term contracted revenue. This is a core financial pillar of Bitdeer, own, develop and expand superior power assets and then convert them into long-duration contracted cash flows through high-quality relationships.

Second, it diversifies our revenue mix. The Tydal agreement accelerates the process of shifting a significant portion of our revenue from Bitcoin into AI. Third, it establishes a repeatable template we intend to carry forward as we expand our AI portfolio into future agreements, including for our key North American sites. The modified gross lease structure, the credit enhancement framework structured without shareholder dilution and the phased delivery approach are all features that we expect to carry forward into future agreements.

Fourth, it demonstrates our ability to develop and operate a leading -edge AI data center at scale for the most demanding customers. Fifth, we plan to raise additional debt capital to fund our infrastructure growth. Given our relatively limited remaining CapEx needs at Tydal, we expect this approach to drive greater capital efficiency and generate significant excess capital to help accelerate additional AI/HPC projects, particularly for our key North American sites.

I will now turn the call over to Haakon Bryhni, Chairman of our Tydal Datacenter, to discuss the site's unique operational advantages, delivery time line and our community commitments in Norway.

Thank you, Haris. At Tydal, we have created a purpose-built and purpose-operated infrastructure for high-performance computing applications. Located in the Kirkvollen industrial area of Trøndelag, Norway, Tydal stands out in the European market for many reasons. Key among them, the infrastructure is fully energized today.

Time to power is among the key bottlenecks facing AI developers globally, and Tydal addresses that constraint directly. We are delivering the 121 IT megawatts to Volta across 2 rapid, equal-sized phases. Phase 1 covers 60.5 IT megawatts across data halls 2 and 3 and a portion of data hall 1 with a target commencement date of December 31, 2026. Phase 2 covers the remaining 60.5 IT megawatts in data hall 4 and the remaining part of hall 1, targeting commencement by March 31, 2027. The site is deeply integrated into Norway's renewable energy ecosystem. Tydal is connected directly to Norway's hydropower grid with access to renewable power in a jurisdiction where power availability is both reliable and competitively priced.

Our power stability is backed by a highly redundant regional network of 16 hydropower stations and the wind farm. This allows us to enable zero carbon baseload power, providing a meaningful advantage as hyperscale and AI tenants face increasing scrutiny around the carbon profile of their infrastructure.

To support the demanding requirements of NVIDIA's Vera Rubin and NVL72 racks, we are upgrading the electrical infrastructure to meet the strict 4N/3 redundancy. The entire 180-megawatt gross capacity is protected by 4 independent 60-megawatt hour battery energy storage systems. Our cooling infrastructure is equally advanced. We utilize 2 independent water supplies, enabling a closed-loop hydro cooling system that achieves a highly efficient PUE of approximately 1.1.

Connectivity is resilient and carrier neutral, featuring redundant fiber optic infrastructure with existing providers, including Telenor and NTE Telekom. Importantly, Bitdeer's commitment extends beyond our facilities' walls. We are actively creating high-tech jobs in Norway and fostering a circular economy. We plan to export waste heat from our data halls to an adjacent business area of approximately 112,000 square meters, which is being developed for businesses that can utilize the heat, including potential local food production. This ensures that our rapid growth provides substantial, sustainable local value creation for the Tydal municipality.

I will now turn the call over to Michael Potter, our Chief Financial Officer, to discuss the financial structure, unit economics and capital raising plans.

Thank you, Haakon. Good morning, everyone. The economics of this modified gross lease start Bitdeer firmly on our journey to become a significant player in the AI infrastructure industry.

Our starting combined rate over the 16-year base term is approximately $160 per kilowatt per month, inclusive of both base rent and service fees. Electricity costs are fully reimbursed by the tenant on a pass-through basis, entirely protecting us from energy price volatility. Because the agreement includes a 3% annual escalator on both the base rent and the service fee, our average realized rate over the 16-year base term is approximately $202 per kilowatt per month, consistent with market escalators for long-duration data center leases.

To put that into aggregate numbers, over the 16-year base term, we expect to generate an average annual revenue of $2.4 million per IT megawatt. At the site level, this drives an expected average annual revenue of approximately $290 million. These economics reflect the durability of a long-term, largely pass-through lease structure with a well-capitalized tenant, translating to an estimated NOI margin of approximately 90%.

From a capital expenditure perspective, we expect to significantly benefit from our previous development because we are developing from an existing powered infrastructure at Tydal rather than building a greenfield position, our remaining CapEx is approximately $500 million. This equates to approximately $4 million per IT megawatt with 121 megawatts of critical IT power.

To fund this build-out, we plan to raise additional debt financing. Given the strength of the expected credit backstop, we are actively engaged with leading financial institutions. We expect this additional debt financing to fully fund our remaining capital requirements at Tydal as well as provide significant additional capital for future projects. We expect to provide further detail on our financing approach as the development plan is finalized.

I will now hand the call back to Haris for brief closing remarks.

Thank you, Michael, and thank you all for joining us this morning. We are proud of this agreement and what it represents for Bitdeer and for our partnership with Volta. Today's announcement demonstrates that Bitdeer's global power portfolio is the foundation for a scalable AI infrastructure business.

We look forward to updating investors as we reach Phase 1 commencement, advance our development plan and execute across additional sites in our portfolio. Thank you for your continued support of Bitdeer. We are just getting started.

Thank you for your participation in today's conference. This does conclude the program. You may now disconnect. Thank you.

Hello, and welcome to Bitdeer Technologies First Quarter 2026 Earnings Conference Call. [Operator Instructions] I would now like to hand the conference over to Tesh Dahya. You may begin.

Thank you, operator, and good morning, everyone. Welcome to Bitdeer Technology Group's First Quarter 2026 Earnings Conference Call. Joining me today are Jihan Wu, Founder, Chairman and Chief Executive Officer; Matt Kong, Chief Business Officer; and Haris Basit, Chief Strategy Officer. Today's call will begin with Haris providing a review of the company's first quarter results, operational progress and strategic direction, and I will close with an update on our financial performance.

To accompany today's call, we have provided a supplemental investor presentation available on Bitdeer's Investor Relations website under Webcasts and Presentations. Before management begins their formal remarks, I would like to remind everyone that during today's call, we may make certain forward-looking statements. These statements are based on management's current expectations and are subject to risks and uncertainties, which may cause actual results to differ materially. For a more complete discussion of forward-looking statements and the risks and uncertainties related to Bitdeer's business, please refer to the company's filings with the U.S. Securities and Exchange Commission.

I also want to note that beginning with the first quarter of 2026, Bitdeer has transitioned from International Financial Reporting Standards to U.S. Generally Accepted Accounting Principles. As part of this transition, Bitdeer has adopted FASB ASU 2023-08, which requires digital assets held to be measured at fair value each reporting period. Changes in the fair value of our digital assets will flow through GAAP net income and may introduce noncash volatility into reported earnings. We will discuss this further during the financial review.

With that, I will now turn the call over to Haris.

Thank you, Tesh, and good day, everyone. The first quarter of 2026 demonstrated Bitdeer's fundamental strength and resilience. In a challenging environment for the broader mining industry, our vertically integrated platform advanced across our 4 strategic businesses: Bitcoin mining, ASIC development, AI cloud and colocation data center infrastructure. We are making significant progress in each area.

First, our Bitcoin mining production has grown almost 500% year-on-year. Second, we launched the industry-leading SEALMINER A4 series. Third, we are rapidly growing our AI cloud revenue. And fourth, we are well on our way towards converting our Tydal Norway facility into what is expected to be Norway's largest AI data center with a lease tenant in advanced stages of negotiation. The combined strengths we see across our portfolio create optionality that is genuinely differentiated within our industry.

We remain committed to Bitcoin mining and see significant opportunity ahead. At the same time, our 3 gigawatt global power capacity is a strategic asset that is increasingly relevant to AI and colocation customers. In Q1, we delivered total revenue of $188.9 million, an increase of approximately 170% year-over-year, with an adjusted EBITDA of $14.4 million, an approximate $60 million increase year-on-year. Tesh will cover additional details of the financials here shortly.

But first, let's turn to a review of our power and infrastructure portfolio, which remains the foundational asset underlying everything we are building. We continue to make meaningful progress across our global infrastructure footprint during the quarter. As of the end of March, we had approximately 1.7 gigawatts of electrical capacity online and a total global power pipeline of approximately 3 gigawatts. We believe this represents one of the largest and most AI suitable power portfolios among publicly listed companies in our sector, and it continues to provide us with strategic optionality as demand for large-scale compute infrastructure intensifies.

Our core sites are sizable, dispersed across multiple continents and regulatory jurisdictions. They include access to renewable energy with attractive economics, featuring infrastructure designed to support intensive continuous operations. These are characteristics that are difficult and time consuming to replicate, and they are increasingly what large-scale AI customers are looking for as they pursue power-constrained deployments.

Over the past several months, we have seen the demand dynamics for AI data center capacity continue to sharpen. The supply and demand imbalance for AI compute has widened, and we expect this shortage to persist well into 2027 and beyond. Time to power remains a critical variable, and we are positioned to serve customers seeking both near-term and midterm capacity in a way that very few operators can match. Against this backdrop, we are prioritizing colocation arrangements for our larger sites, which are best suited to serve hyperscale, Neo cloud and enterprise tenants seeking substantial committed capacity.

For our smaller facilities, we continue to pursue AI cloud opportunities, deploying capacity on a contract-backed basis. This tiered approach reflects a disciplined allocation of capital across our portfolio, matching the appropriate commercial model to the scale and characteristics of each site.

Let me walk through where we stand on key development sites. Tydal Norway remains our highest priority colocation opportunity. On March 30, 2026, our subsidiary, Tydal Data Center AS, entered into a formal agreement with Data Center Installation AS, a specialized Norwegian contractor, to develop and convert the Tydal facility into an AI data center.

The project will deliver 180 megawatts of gross installed capacity and the first phase is expected to be completed as early as December 2026. Upon completion, the Tydal facility is expected to be Norway's largest operational AI data center and one of the largest in Europe by installed capacity. This facility is being built primarily for colocation usage. Designed in accordance with NVIDIA guidelines and closely following NVIDIA reference designs, it is intended to support deployment of both GB300s and NVIDIA's latest Vera Rubin AI technology.

What makes Tydal particularly compelling to prospective tenants is a combination of attributes that are genuinely rare, stable baseload power enabled by 100% renewable sources and an excellent power usage effectiveness, or PUE, of approximately 1.1 enabled by the cold climate and chilled water available from a nearby lake. Furthermore, the site was built such that it substantially reduces retrofit capital requirements relative to a greenfield build. We expect our remaining CapEx costs to complete the Tydal site to be significantly lower than typical greenfield data center development costs.

Orders for most long lead equipment have been placed and decommissioning of Bitcoin mining rigs at the site is already underway. Upcoming near-term milestones include finalization of key equipment installation contracts and technical installation work in several of our data halls. Lastly, we have also begun technical due diligence work on behalf of our future tenant. We are in advanced stages of negotiations with a potential colocation tenant for Tydal. These discussions, when completed, would result in highly regarded and well-recognized end users.

Morgan Stanley has been retained as our financial adviser for this project. Signing the Tydal lease agreement is management's highest priority. At Clarington, Ohio, we have 570 megawatts of power under contract with AEP. This is one of the largest AI data center development opportunities in the United States among publicly listed companies in our sector. Design and preparation work is continuing for the site. As we have disclosed, litigation filed by a neighboring company could affect the timing of construction.

Our attorneys feel strongly that we have a well-founded case and that the litigation has limited merit. On the business side, we are evaluating plans that can mitigate the impact on our overall development time line. We remain optimistic about the potential for the site, and we continue to build strong relationships with the local community and government officials at all our Ohio sites.

At Rockdale, Texas, we are pursuing a dual-track strategy that maintains our existing Bitcoin mining operation while developing new AI infrastructure on adjacent land. In addition, we are working with ERCOT on incremental power capacity of 179 megawatts targeted for energization by year-end. This will bring our total power capacity at Rockdale to over 740 megawatts. We are actively engaged in discussions with several prospective colocation tenants for this site.

The Rockdale site benefits from its location in the ERCOT market and will be designed from the ground up to support AI workloads. This approach allows us to maintain revenue-generating mining operations throughout the development period rather than interrupting them. Beyond these 3 primary sites, conversion projects are advancing at Wenatchee, Washington, and at Knoxville, Tennessee. Both sites are undergoing design and permitting work for AI data center conversion with the Wenatchee site and first phase of our Knoxville site targeted for completion in the fourth quarter.

At Niles, Ohio, we are actively working towards the development of our 300-megawatt grid interconnected site with a target energization time line of the fourth quarter of 2028. We also plan to break ground on our 101-megawatt Fox Creek, Alberta, Canada site in June of this year. Furthermore, we continue to aggressively look for additional opportunities to invest in land and power capacity, and we will share these updates as appropriate.

The U.S. continues to be the primary hub for Bitdeer's global operations, bolstered by our confidence in pro-business, pro-innovation policies that support the growth of AI and digital assets. We remain firmly committed to scaling our presence in the U.S.

On the Bitcoin mining side, the expansion of our self-mining platform continued throughout the quarter. Self-mining hash rate grew from 55.2 exahash per second at the end of December 2025 to approximately 65 exahash per second exiting March. 65 exahash per second represents a year-over-year increase of more than 400%. We mined 668 Bitcoin in January, 705 Bitcoin in February and 661 Bitcoin in March. The modest decline in March relative to February reflects seasonal factors at our Norway and Bhutan facilities rather than any underlying deterioration in fleet performance as witnessed by our April production of 783 Bitcoin. We expect to see continued momentum in the months ahead.

Our mining operations are not plateauing. The SEALMINER A4 series officially launched on April 7, 2026, represents the most efficient mining rigs anyone has delivered. The flagship A4 Ultrahydro model operates at 9.45 joules per terahash. The A4 series also includes the A4 Pro Hydro and the A4 Pro Air at 10.9 joules per terahash. These machines provide deployment flexibility across different site configurations and cooling environments. The A4 Pro Air is one of the most efficient air-cooled mining rigs in the world.

SEAL04-2 chip development continues at our U.S.-based design center. Importantly, our internal manufacturing capability means we are not subject to third-party hardware markups on these rigs when deploying them into our own fleet. This is a structural cost advantage over other mining operations. The inclusion of these new machines will continue to improve our fleet efficiency of approximately 16.4 joules per terahash as of March 31, 2026. That efficiency improvement, combined with our advanced chip design and supply chain resources, translates directly into lower cost per unit of hash rate produced, which means better mining margins at any given hash price level.

Over the next several quarters, we plan to leverage our growing fleet of SEALMINERS beyond our existing mining data center capacity and work with third parties to deploy incremental co-mining capacity at their facilities. This will allow us to maximize mining economics in the near term while maintaining flexibility to opportunistically drive SEALMINER sales into the second half of the year, depending on market conditions. On the SEALMINER manufacturing front, preparations for our Reno, Nevada factory are progressing. The facility lease has been signed and construction permit applications have been submitted to local municipal authorities, and we anticipate starting construction by Q3.

U.S.-based manufacturing is a core component of our vertically integrated strategy and aligns with both our operational resilience objective and the evolving trade and supply chain environment. Our AI cloud business has matured from a pilot service into a commercially distinct, structurally attractive business segment with rapidly growing revenue and a deepening enterprise customer base. For the AI cloud business, annual recurring revenue, which was approximately $10 million at the end of January, grew to approximately $21 million by the end of February and reached approximately $43 million at the end of March.

GPU utilization climbed from 41% in January to 94% in March. At quarter end, we had 2,128 GPUs deployed, including H100s, H200s, B200s and GB200s with 1,948 under active external subscription. More recently, in our April production update, we announced annual recurring revenue has now reached approximately $69 million with over 4,000 GPUs deployed.

Customers are committing to longer durations, which improves revenue visibility and cash flow stability. Since late 2025, we have seen hourly pricing of H100s increase by approximately 40%. This is in direct response to demand levels, and the market is absorbing this increase without meaningful friction. This pricing power reflects the strong fundamentals of our AI cloud business. In January, we deployed our initial NVIDIA GB200 NVL72 infrastructure at our Cyberjaya, Malaysia facility. This marks the first phase of an accelerated expansion designed to support enterprise-grade training workloads on the Grace Blackwell architecture.

In February, we launched a managed Kubernetes service with GPU-native orchestration, providing enterprise customers with scalable infrastructure for AI training and inference. Our model studio platform now supports more than 50 leading open source models, enabling clients to deploy everything from basic inference to advanced multimodal applications through a single managed environment. In March, we showcased our integrated AI solutions at the NVIDIA GTC Conference, generating incremental business opportunities and strengthening our brand presence within the AI infrastructure ecosystem.

We are actively evaluating U.S. data center leasing opportunities and expect to bring GPU capacity and AI cloud services online for U.S. customers in 2026. Consistent with our stated approach, any large-scale U.S. GPU expansion will be backed by committed customer contracts.

Turning to our balance sheet. In February, Bitdeer successfully priced an upsized offering of $375 million in 5% convertible senior notes due in 2032. We ended Q1 with cash, cash equivalents and restricted cash of $298 million. We expect that the bulk of our fiscal year '26 total financing needs will be addressed through project-level debt financing following a signed lease agreement for our Tydal, Norway site.

Now I will hand it back to Tesh to go over the detailed financials.

Thanks, Haris, and good day, everyone. It's great to be here, and I look forward to meeting many more of our shareholders in the coming months. Let me walk through our detailed financial results for the first quarter. Before I begin, I would like to remind everyone that all figures are in U.S. dollars. And as noted earlier, this is our first quarter reporting under U.S. GAAP.

In addition to discussing results calculated in accordance with U.S. GAAP, we will also reference certain non-GAAP financial measures, including adjusted EBITDA. Adjusted EBITDA excludes noncash fair value changes on our digital assets and convertible note derivative liabilities, along with certain other items, and we believe it provides the most consistent basis for assessing core operational performance. For a full reconciliation of non-GAAP measures, please refer to our earnings release published earlier today on Bitdeer's Investor Relations website.

First quarter consolidated revenue was $188.9 million, an increase of approximately $119 million year-over-year. The year-over-year growth was driven primarily by the significant expansion of our mining hash rate and associated Bitcoin production, reflecting the continued SEALMINER deployment throughout 2025 and into 2026. Sequentially, revenue declined from $224.8 million in the fourth quarter of 2025, reflecting lower average Bitcoin prices during the first quarter relative to the fourth quarter as well as a larger portion of our manufacturing output going towards self-mining deployment rather than external SEALMINER sales.

Total gross profit was negative $39 million, reflecting a gross margin of negative 20.7%. Three converging factors drove the outcome. First, Bitcoin prices remained under pressure throughout the quarter. Second, our mining fleet carries substantial noncash depreciation expense amounting to $70 million, given our rapid expansion. As a reminder, we now depreciate mining rigs on a 3-year straight-line basis, and the pace of SEALMINER deployment throughout 2025 and into 2026 generates a significant concurrent charge. Third, a seasonal power cost dynamics at our Norway and Bhutan facilities weighed on energy costs in the first quarter.

Looking ahead, the path to gross margin recovery is straightforward. A4 Series deployment lowers our cost per Bitcoin mined, spring and summer rate normalization reduces electricity costs and the scaling of AI cloud revenue improves margin composition as that segment grows.

Adjusted EBITDA was $14.4 million for the quarter, an increase of approximately $60 million year-on-year. The sequential decline from $24.3 million in the fourth quarter of 2025 reflects the gross margin dynamics described earlier. Operating loss in the quarter was negative $159.5 million and earnings per share was negative $0.68. Net cash used in operating activities was $346.9 million, a 42% reduction versus the Q4 net cash used in operations of $594.7 million. The primary drivers of the sequential reduction were lower SEALMINER supply chain and manufacturing costs, partially offset by higher electricity costs.

Turning to the balance sheet. We exited the first quarter with $297.7 million in cash, cash equivalents and restricted cash compared to $177.9 million at year-end 2025. Total borrowings at the end of Q1 were approximately $1.92 billion. For the full year 2026, we reiterate our guidance for total infrastructure capital expenditures in the range of $180 million to $200 million for crypto mining data center construction. This guidance covers crypto mining infrastructure only and does not include CapEx for SEALMINER hardware, GPUs, AI cloud or colocation development. Additionally, we anticipate a continuation of growth in our mining hash rate, albeit at a more moderate pace than we have seen throughout the prior 2 quarters.

In summary, the first quarter of 2026 was a quarter of execution and strategic advancement. Gross margins were under pressure from a combination of low Bitcoin price, the depreciation accounting impact of our fleet expansion and seasonal power costs. These factors are transitory and the forward catalysts for margin recovery are tangible and progressing, A4 deployment, power cost normalization, colocation and scaling our AI cloud. Against that backdrop, we delivered on the key elements that will define the value creation we expect to deliver over the coming quarters.

We launched the SEALMINER A4. We grew AI cloud ARR by 105% in a single month. We engaged a construction partner for Norway's largest AI data center, and we strengthened our balance sheet with $375 million in new capital. The colocation pipeline ahead of us is substantial, and we are pursuing it with full organizational focus. We entered the second quarter with strong operational momentum, a differentiated asset base and a team that has demonstrated its ability to execute at scale. We are energized about what lies ahead and remain committed to delivering long-term value for our shareholders.

Thank you. Operator, please open the call for questions.

[Operator Instructions] Our first question comes from the line of Greg Lewis with BTIG.

Haris, I appreciate we're in advanced discussions on Tydal in Norway. That being said, kind of curious how you're thinking about that. I noticed in the comments, we talk about the design and planning. Like how much design knowing that there is some similarities between certain customers and what they expect from a data center, but there are some differences. How far in the process of the final design can we get? Is that something that we then need to wait for the customer to kind of move forward?

And just as we think about the opportunity in Norway, like I know we're talking about hyperscalers. But like -- how important is that? I know there's some big tech scandy companies maybe that we wouldn't think are traditional hyperscale -- that some people might not think are traditional hyperscalers, but are kind of big tech companies in Northern Europe. Just kind of curious if you could provide any color around some of those questions on that opportunity.

Okay. Sure. Thank you, Greg. So with regards to the exact technical specifications, and there are differences between customers because different customers want to put in different machines versus GB300s versus Vera Rubin and the mix of those machines. And so -- but we have, I would say, the vast majority, almost entirely of the design in hand. We're still communicating with the most likely tenant here, the one that we're very close to signing, to make sure that all the design elements meet what their requirements are, which turn out to be very close to what the NVIDIA reference designs are.

So we think we have that well in hand. There's ongoing discussions, but just over very detailed type of stuff at present. And then with regard to the type of tenant, the 2 most important things here are that there would be a very sound credit, an investment-grade client or a very good credit wrapper. And so that's important. And then, of course, the economics of it are important. And we're focusing on those 2 things. We think we've -- if it goes through the way we expect -- in the time frame we expect, I think investors should be relatively pleased with both of those issues. And I can't say too much more about the tenant, but it won't be too long before I think we can announce that deal.

Understood. And then I did want to touch on the Clarington. In the press release, you mentioned -- and then actually in the prepared remarks, we mentioned the -- maybe some of the delays that are going on, realizing that, that is active. Could you kind of at least provide like some broad strokes around what is actually happening? I mean, yes, just kind of like that was news to us. So I just want to understand...

I'm sorry, which...

Some of those headwinds are that you're going to have to deal with?

Which location are you referring to?

I'm sorry, Clarington.

Clarington. Well, we announced earlier about the litigation at that site. And we're still working through that, and we expect that, that will have an impact on the construction schedule. There's not really a lot more I can say about that. We are looking at ways of mitigating those impacts, but...

I mean I guess what I would ask is the power is approved, so it would have to be something more around like the land user. Is that how -- is that kind of...

Yes, it's not really a question of the power...

Our next question comes from the line of Mike Colonnese with H.C. Wainwright.

Nice to see all the progress across your business lines here. So it sounds like SEAL is progressing nicely. I was wondering if you could provide a little bit more color around Rockdale. It sounds like you're going to simultaneously construct a new AI data center alongside your Bitcoin mining operations there. Can you talk about the level of client demand for that specific asset? It sounds like a really unique opportunity given the power capacity and really what the development time lines could ultimately be for that part of the portfolio for an AI colocation opportunity?

Yes. I think it's a little early to predict the exact development time line for that. It's a very attractive site for AI. And one of the things to make it even more attractive would be to have more land, which is what we're working on at that site. But the power is there, and it's going to be expanded to even a larger envelope of power over 700 megawatts. So it's a good location for AI. We're speaking with several potential tenants there. They span from hyperscalers to Neo clouds and even some others. But the level of demand, I think, is very high. I think the -- I can't really put a good time frame on the execution of that site yet. But we're moving forward on at least making sure that we have an appropriate land space where we can develop the AI data center while the Bitcoin mines are still operating.

Got it. Very helpful color, Haris. Appreciate that. And then just sticking on the AI side, but more on the cloud business, that is seeing really strong growth here between GPU deployments, utilization rates. Just curious to get a sense as to how durable that revenue stream is here. Obviously, the utilization rates are helping, but to the extent you could share more information around the contracted element to it? And then also, if there are any sort of internal benchmarks you guys are looking to grow that business this year? Obviously, you have multiple business lines you're working through, but thinking about GPU expansion from the around 4,000 that you guys have today, the best way for investors to think about that?

Yes. I think there's tremendous demand for GPUs. And it's really on our part, limited by how quickly we can bring up these GPUs and AI cloud sites. But the demand is there. It's across the board. It's -- we mentioned we were able to raise the rates on our H100s by 40% and have no problem booking those. So we're also starting to get longer-term contracts. I don't know, Jihan, did you want to add anything to that? Maybe...

Okay. I muted myself. Okay. Right now, most of our contract is a long-term contract right now. It takes majority of our machines in long-term contract. And right now, customers will need to agree with us on such terms in 3 to 5 years.

Our next question comes from the line of Mike Grondahl with Northland Capital Markets.

This is Logan on for Mike. First, can you just provide some insight into the conversations around pricing and terms at Norway and also some color on just what the remaining hurdles are to getting a lease signed at the site?

Yes. I don't think we're going to give you satisfaction on the pricing other than we think it's at the -- near the top of the market of what we've seen announced. So it's -- we think it's going to be quite good. Let's see. For what's left, there's just a lot of detailed work. We are in the late stages or advanced stages of negotiating the lease, and there's just a lot of small details. There's no one big thing that stands in the way. But -- all these small things do have to get handled before we can have a finished signed lease. So there's no one thing that's standing in the way here. We're trying to move as fast as we can. It's our highest priority within the management chain here. And we're applying a tremendous amount of resources to it. There's just a lot of detail that needs to get covered here.

Got it. And that's great to hear on the favorable pricing. And then one more. In your April update, you mentioned that various other sites outside of Norway, Clarington and Rockdale are in advanced stage negotiations. Are you guys at a point to be able to formally call out those sites by name? And if not, can you just provide some color around how demand for your sites has changed over the last 90 days?

Yes. I mean I think the last 90 days has stayed. It's pretty much -- it stayed very strong. I don't know how to quantify whether it's gotten a little stronger or not, but we haven't seen any diminishment, that's for sure. And yes, we are not in a position to announce the schedule for any of the other sites in terms of the colocation. The AI cloud sites, we have announced Q4 of this year for Wenatchee and the first phase of the Knoxville, Tennessee site. So is there something different than that you were looking for?

No, that's all good. Got it. Congrats on an impressive start to 2026.

Our next question comes from the line of Kevin Cassidy with Rosenblatt.

Yes. Congratulations also on all the progress you have. Just going back to the AI cloud, very impressive that you're able to raise hourly rates by 40% on the H100. What's the trend as you go to the higher performance GPUs? What kind of rate increase should we expect on the hourly rate?

Maybe I'll ask Jihan to answer that question since I don't have a good feel for that.

Well, because previously, we have mostly signed with a short-term contract. So after those contract ends, we have -- we had an opportunity to raise the rates. But right now, most of our GPU cost is in kind of a long-term contract. So the rate will be relatively stable from now.

I guess your [Technical Difficulty] how the rate differs from the high-end machines to the 1,800, right, is something like that or...

Yes. And as we...

Yes. If we go into higher-end GPU, that means we are deploying new GPU. And right now, all those contracts are in negotiation where we are preparing the data centers and the installation work. Generally, we can feel that the customers are quite competitive on the demand side, and we needed to carefully choose our clients that can be stable and also profitable. That's what where we need to wait. So I think generally, I'm quite optimistic about the profitability of the GPU renting AI cloud business because the customers are quite willing to pay good price to get the GPU.

Okay. And just maybe it wouldn't be [indiscernible] your conference call if I didn't ask about the SEAL04 second version. You mentioned you're still working on it, but do you have any timing? And are we -- what the targeted joules per terahash is?

No, we haven't changed the target, but -- and we're not ready to announce new timing on that yet. So I sorry about that.

Okay. Just obliged to ask that question every...

[Technical Difficulty]

Our next question comes from the line of Nick Giles with B. Riley Securities.

Maybe just a follow-up from an earlier question around to what stage across some of these other sites would you be willing to kind of build for colocation purposes? And what level of CapEx would be associated with that? And then do you have a rough estimate of how much CapEx you've deployed to date towards colocation conversions across the platform?

We haven't announced or revealed our CapEx for colocation conversion other than to say that the amount of capital required in Norway is remarkably less than the normal amount of CapEx required. We expect that the CapEx requirements at other sites, the U.S. sites will be closer to the typical amount needed to build an AI data center that we will be able to get some of the savings that we had in Norway at other sites, but not to the same extent.

Understood. I appreciate that, Haris. Maybe switching gears, just on the Reno Nevada site, the facility, the ASIC facility, any kind of preliminary estimates on what CapEx could be there? And then how would this change your margin profile in that business, if at all?

Yes. I think just to remind everyone, that's the site where we're assembling ASIC mining rigs. Jihan, do you have an answer for that question?

[ Probably, no. ]

Yes. So I don't think -- we haven't reported the amount of capital required for that site. It's significantly smaller than the amount of capital required for like a data center or even a Bitcoin mining site. What was the other part of that question?

Really, I was just curious...

The margin.

The margin profile of that business, yes.

Well, it will be a little bit more expensive to build in Reno than it will be in Asia or to assemble there. But most of the cost of our mining rigs is really embedded in the silicon itself, which is still made by TSMC in the same location. So we think that the incremental cost of assembling in the U.S. will be covered by, for example, tariffs and things like that. So we think it will be a very good location for us and within a reasonable price increment of building in Asia, especially if you account for tariffs.

Our next question comes from the line of John Todaro with Needham.

Congrats on the progress so far. I guess just going back to Rockdale and Clarington, obviously, some pieces need to still be completed there to get development moving along. I think my understanding is it's mostly acreage. I guess just what are some of the limiting factors there? Are we just kind of in negotiation processes for that? Do you need some additional cash to get those items done? I guess just trying to understand that a little bit better to see how far along we can be.

Well, I think one way to think about it is that the amount of power, say, in Rockdale is much larger than the amount of land, right? So that's something that we want to rectify. And it's really around those kinds of issues that we want to make sure we're able to fully utilize all of the power that we have at those locations. And of course, in Clarington, there's the additional complication of the litigation.

Okay. Understood. And then shifting to Bitcoin mining machine sales. So obviously, a lot of the U.S. public miners are pulling back as they shift towards AI HPC. You guys have had a little bit more external sales in international markets. So wondering how does that change the sales strategy? Is it actually a benefit versus a negative that there's more public U.S.-focused miners pulling back and maybe that shifts more opportunity to do sales internationally? I guess just trying to frame up how that shifts for external sales a bit longer term here.

So one thing I just want to remind you is that we're not really pushing that hard for external sales at this point. We -- almost all of the output is being used internally by our own data centers. We do, as you mentioned, sell internationally. So the fact that a lot of the U.S. mining companies are pulling back is not -- has a little mitigated impact based on that. Then -- I don't know, Jihan, do you want to make any additional comments on Bitcoin mining ASIC sales or demand?

Right now, because of the constrained supply of the semiconductor fabrication service, we intend to do more self-mining. And self-mining is also -- I believe it's a profitable business. So we still have some Bitcoin mining sites. We haven't filled it. And we also have a lot of partners that want to do co-mining partnership with us, which means that we provide the Bitcoin mining rigs and they provide the Bitcoin mining farm, so we can share the Bitcoin mining hash rate, and we will get the majority out of it.

And the electricity bill is to be transparent and no markup from the mining partnership side. It's quite scalable. So we are not very aggressive on selling the mining rig right now. Right now, the Bitcoin price is still in its bearish situation. And if we sell the mining rigs, we will have to do it at a very bad price, I think. So I think to expand our self-mining is the best economical decision for our company.

Understood. So the focus is, yes, almost primarily on internal.

[Operator Instructions] Our next question comes from the line of Brian Kinstlinger with Alliance Global Partners. All right. I don't have a response from Brian.

All right. Ladies and gentlemen, I'm showing no further questions in the queue. I would now like to turn the call back over to Haris for closing remarks.

Well, I think actually, it's Tesh, do you have some closing remarks, Tesh?

Yes. I think we just want to thank everyone for joining the call. We're exiting the first quarter with clear operational momentum here, a focused strategy, and we're really executing decisively on our AI infrastructure pipeline. Thank you for joining us today, and we look forward to driving sustainable long-term value creation.

Thank you, everyone.

Ladies and gentlemen, this concludes today's conference call. Thank you for your participation. You may now disconnect.

Thank you for standing by, and welcome to Bitdeer's Third Quarter 2025 Earnings Conference Call. [Operator Instructions] I would now like to hand the call over to Yujia Zhai, Investor Relations. Please go ahead.

Thank you operator, and good morning, everyone. Welcome to Bitdeer's Third Quarter 2025 Earnings Conference Call. Joining me today are Matt Kong, Chief Business Officer; Haris Basit, Chief Strategy Officer; and Jeff LaBerge, VP of Capital Markets and Strategy. Haris will be in today by providing a high-level overview of Bitdeer's third quarter 2025 results and then cover the company's strategy and a detailed business update. After that, Jeff will cover Bitdeer's third quarter financial results in more detail, and then we will open the call for questions.

To accompany today's earnings call, we have provided a supplemental investor presentation. This presentation can be found on Bitdeer's Investor Relations website under Webcasts and Presentations. Before management begins their formal remarks, we would like to remind everyone that during today's call, we may make certain forward-looking statements. These statements are based on management's current expectations and are subject to risks and uncertainties, which may cause actual results to differ materially. For a more complete discussion on forward-looking statements, and the risks and uncertainties related to Bitdeer's business, please refer to its filings with the SEC.

Further, in addition to discussing results that are calculated in accordance with International Financial Reporting Standards, or IFRS, we will also make references to certain non-IFRS financial measures, such as adjusted EBITDA and adjusted profit and loss. For more detailed information on our non-IFRS financial measures, please refer to our earnings release that was published earlier today, which can be found on Bitdeer's IR website. Thank you.

I will now turn the call over to Haris. Haris?

Thank you, Yujia, and good day, everyone. It's great to be with you today. Since our last call, we've made significant progress across all of our strategic priorities, and I'm excited to share how Bitdeer is growing from a global leader in bitcoin mining, into a vertically integrated bitcoin infrastructure and AI platform.

Let's start with the numbers on Slide 3. Q3 marked a period of rapid growth and strong execution. In the third quarter, total revenue reached $169.7 million, up 173.6% year-over-year and up 9.1% sequentially. Gross profit came in at $40.8 million, and adjusted EBITDA increased to $43 million, both substantially improved from Q2. This performance reflects continued execution in our self mining business. Mass production of our SEALMINER ASICs drove 273.1% year-over-year and 105.4% sequential increase in our average operating self-mining cash rate to 29.1 exahash per second. As of the end of October, we achieved 41.2 exahash per second, surpassing our 40 exahash per second target that we set out at the beginning of the year. Looking forward, we plan to continue deploying our SEALMINER ASICs to fill our substantial global power pipeline. Through a combination of decommissioning both degeneration rigs and adding new SEALMINER ASICs, we expect meaningful growth throughout 2026, which will ensure we remain one of the world's largest bitcoin miners. In addition to hash rate growth, we expect continual improvement in our fleet-wide energy efficiency to drive increased margins and profitability in the quarters ahead.

Our investments in chip design, data center infrastructure and global power portfolio are paying off, not only in strong financial performance, but also positioning us for the next major wave of demand for compute. Our bitdeer.ai computing business continues to scale, reaching at an annualized revenue run rate of $8 million at the end of October. As of October 31, we operate 584 GPUs with an 87% utilization rate. Our newest B200 systems installed late in Q3 are being tested by customers and will drive additional revenue. We are finalizing deployment for NVIDIA's GB200 systems and expect to reach 1,160 GPUs operated by end of the year. We are also in the process of procuring NVIDIA's next generation GB300 and B300 systems. And as we look ahead, Bitdeer's growth will be anchored by 3 strategic pillars Bitcoin mining, ASIC development and HPC/AI. Together, these represent a vertically integrated highly dependable platform that leverages our technology expertise and extensive power portfolio.

To accelerate our AI footprint, we are taking bold deliberate steps to simultaneously pursue both colocation and cloud services. With respect to co-location on our last earnings call, we guided the market that we intended to pursue a joint venture model with a development partner. In September, due to a significant increase in market demand, we let the exclusivity period under our LOI with this development partner expire. This was a strategic decision that gives us greater flexibility and allows us to take a more direct role in the HPC/AI data center market and retain more of the economics. We intend to develop data centers using our own internal development team, which will be significantly augmented through strategic hiring alongside highly experienced [ DBC ] and general contractors on a fee basis. Regarding our cloud services business. Over the past 18 months, we have developed a vertically integrated AI infrastructure platform in Singapore. That includes bare metal GPUs as well as orchestration, networking and managing services. These additional services are highly sought after by small and midsized enterprise customers who require more than just bare metal offering and are being underserved in today's market. Now that we have a proven concept in Singapore, we are ready to expand this business line into Malaysia, the U.S. and Europe. Our current customer discussions range from early to mid-stage start-ups in the biomedical, robotics and gaming industries to more traditional U.S. enterprise customers seeking to expand their footprint. We will provide additional details as this business model develops over the coming months.

Moving now to our HPC/AI infrastructure plans. For our 570-megawatt Clarington, Ohio site, we have already begun the design and procurement process for an HPC/AI's suitable substation, which is expected to be energized in the first half of 2027. The local utility at Clarington has confirmed the full 570 megawatts will be available by the end of Q3 2026, nearly a year earlier than expected. Given its size, the Clarington site could be utilized for our cloud services business or colocation. At the same time, we have made the decision to convert our 175-megawatt Tydal site in Norway into an AI data center by Q4 2026. Given the announcement of Stargate in Norway in July, we have seen a significant increase in inbound interest from potential tenants. So we believe this site could be used for either our cloud services business or colocation. The site was designed by our local Norway team with HPC/AI in mind as the end use. So it already includes liquid cooling capabilities and a more robust electrical infrastructure.

Furthermore, it utilizes a substation that is powered by 18 hydro power generators and 1 wind farm giving it a high degree of reliability. Our current analysis indicates that the site could be ready to accept GPUs by the second half of 2026 with conversion costs well below U.S. and European AI data center benchmarks. Additionally, we plan to expand our Singapore cloud services business into Malaysia. Through a combination of loan and lease opportunities, we anticipate activating up to 15 megawatts of cloud services capacity in Malaysia during 2026, with the ability to expand significantly beyond that if demand persists. We are also upgrading our 13-megawatt Wenatchee, Washington site using a proprietary modular data center technology. This conversion is expected to be completed by Q4 2026. Further, we have initiated the conversion of 10 megawatts of power capacity at our Knoxville, Tennessee side into an AI data center with targeted completion in Q4 2026. We are evaluating potential U.S. data center rental opportunities to bring our AI cloud services online domestically as early as Q1 2026. In summary, the supply and demand imbalance for AI compute continues to widen, and we impact this shortage to persist well into 2027. Based on our estimates under the most optimistic scenario, converting 200 megawatts of our power capacity fully towards AI cloud services could generate an annualized revenue run rate exceeding $2 billion by the end of 2021.

Turning to our ASIC business. When we launch our aggressive ASIC road map last year, our goal was clear. Industry leadership in performance and energy efficiency. Our R&D team has delivered on that promise. In September, we launched the Wenatchee SEALMINER A3 series, now among the most energy-efficient products in the market. Mass production has started and initial shipments are expected this month. We anticipate the series will generate meaningful revenue in 2026. Looking ahead, our focus shifts to our SEAL04 chip. To derisk the development and ensure success, we are pursuing 2 distinct design approaches. The tape-out for the first SEAL04 design was completed in September, and latest sample verification demonstrated approximately 6 to 7 joules per terahash power efficiency at the chip level under local ultra power-saving mode. We are targeting mass production to begin in Q1 2026. In the meanwhile, development of the next-generation SEAL04 is significantly delayed.

Next, let's turn to our energy infrastructure shown on Slides 7 through 11 in the supplemental investor presentation. In Q3, we continued our rapid build-out of our global power and data center portfolio. As of October 2025, we fully energized the Tydal, Norway site and the full 500 megawatts in Jigmeling, Bhutan. This brings our total available electrical capacity to approximately 1.6 gigawatts and our total global power pipeline to approximately 3 gigawatts. For our AI cloud services and colocation strategy, we believe we have one of the most attractive power portfolio in the industry. Across our sites in Clarington, Ohio, in Norway and Renato, Washington, we will have over 1.3 gigawatts of HPC suitable power by Q3 2027. This gives us a significant advantage in time to power and the ability to deploy massive GPU capacity rapidly. In September, we announced a new 300-megawatt site in Niles, Ohio. The project remains on track for energization in Q1 2029. The site spans 41.8 acres and includes an interconnection agreement with First Energy. It is located about 75 miles from our Massillon, Ohio site and 125 miles from our Clarington, Ohio site. We continue to secure low-cost power sites globally, reinforcing our competitive advantage in both mining and AI infrastructure. In summary, we are proud of our team's execution this quarter. These efforts are already reflected in our financial results and have established a scalable foundation for long-term growth.

I'll now turn it over to Jeff LaBerge, our VP of Capital Markets and Strategy to go over our detailed financial results for the quarter.

Thank you, Haris. [indiscernible] over to Bitdeer's third quarter financial results I'd like to remind everyone that all figures I refer to today are in U.S. dollars. Q3 consolidated revenue was $169.7 million, up from $62 million in Q3 2024 and $155.6 million in Q2 2025, or up 173.6% year-over-year and 9.1% sequentially. Self mining revenue was $130.9 million versus $31.5 million in Q3 2024 and $59.3 million in Q2 2025, or up 315.6% year-over-year and up 120.7% sequentially. These results were primarily due to a 273.1% year-over-year and 105.4% sequential increase in self-mining hash rate as well as higher bitcoin prices. These increases were partially offset by higher mining difficulty. SEALMINERs sales revenue was $11.4 million compared to $0 in Q3 2024 and [ $69.5 million ] in Q2 2025. Total gross profit for the quarter was $40.8 million versus $2.8 million in Q3 2024 and $12.8 million in Q2 2025. Gross margin was 24.1% versus 4.5% in Q3 2024 and 8.2% in Q2 2025.

The year-over-year and sequential increase in our gross margin was primarily driven by higher self-mining revenue and improved fleet efficiency. We expect to continue gross margin improvements over the coming quarters as our hash rate ramps up and overall fleet efficiency improves. Total operating expenses for the quarter were $60.5 million versus $42.9 million in Q3 2024 and $42.3 million in Q2 2025. The year-over-year and sequential increase was primarily driven by the one-off R&D costs for the SEAL04 chip development and tape-out and noncash amortization expenses of intangible assets related to the acquisition of free chain. Other operating income was $26.5 million, primarily due to a $22.2 million mark-to-market adjustment to our cryptocurrency receivables. As a reminder, under IFRS, Bitcoin and other cryptocurrencies are classified as intangible assets and are measured at cost less any accumulated impairment losses with no subsequent upward revaluation permitted.

However, during the quarter, we entered into a $100 million Bitcoin-backed loan facility, pledging approximately 1,400 bitcoin as collateral. As a result, these bitcoin were reclassified as cryptocurrency receivables. IFRS requires that any cryptocurrency known as a receivable or payable to market, which led to this adjustment. Other net loss for the quarter was $238.5 million versus $14.7 million in Q3 2024 and $108.5 million in Q2 2025. The net loss was due to the noncash derivative losses on the convertible senior notes issued in August 2024, November 2024 and June 2025, which I will discuss in more detail in the liability section. IFRS net loss was $266.7 million versus $50.1 million in Q3 2024 and $147.7 million in Q2 2025. Adjusted loss was $32.8 million versus $25.6 million in Q3 2024 versus $24.4 million in Q2 2025. The increase in loss was primarily due to higher operating expenses and interest expense related to the increased borrowings, partially offset by the year-over-year higher revenue and gross profit margins.

Adjusted EBITDA was positive $43 million versus negative $7.9 million in Q3 2024 and positive $17.3 million in Q2 2025. The year-over-year growth was primarily driven by significantly higher self-mining cash rate as a result of the company's mass production and deployment of SEALMINERs A1 and A2 during 2025. Note that both the adjusted loss and adjusted EBITDA figures for the quarter do not include the $22.2 million favorable mark-to-market gain from Bitcoin pledged as collateral. This quarter's higher year-over-year and sequential top line and non-GAAP bottom line performance was mainly driven by higher self-mining cash rate, SEALMINER sales and higher Bitcoin pirces. these were partially offset by higher global network cash rate and higher R&D costs as previously described. Net cash used for operating activities was negative $520 million, primarily driven by SEALMINERs supply chain and manufacturing costs, electricity costs from the mining business, general corporate overhead and interest expense. Net cash generated from investing activities was $27 million, which was driven by $60 million of capital expenditure, of which $32 million was related to data center infrastructure and related construction. Proceeds from disposal of cryptocurrencies from our primary business was $89 million. Net cash generated from financing activities for the quarter was $388 million, which resulted primarily from approximately $320 million of borrowing from a related party and $91 million of proceeds from shares sold under our ATM program, partially offset by $48 million of repayments of borrowings.

Moving to our 2025 infrastructure spend. We expect CapEx for the continued buildout of our global power and data center infrastructure to be in the range of $210 million to $240 million for calendar year 2025. This range includes reported infrastructure CapEx from the previous 9 months of approximately $168 million. The remaining projected CapEx is extended on the completion or near completion of our data centers in Tydal, Norway, Jigmeling, Bhutan, Massillon, Ohio and Ethiopia as well as the partial completion of the 101-megawatt gas-fired power plant in Alberta, Canada. Please note that this guidance only factors in power and data center and does not include CapEx for SEALMINERs and GPUs.

In terms of our balance sheet, we ended the quarter in a strong financial position with $196.3 million in cash and cash equivalents, $82.2 million in cryptocurrencies held at cost less impairment $163.9 million in cryptocurrency receivable held at fair market value and $824.3 million in borrowings, excluding derivative liabilities. Please note the $82.2 million in cryptocurrencies is accounted for according to IFRS rules and is currently below its market value. Derivative liabilities were $672.5 million which related to the November 2024 and June 2025 convertible senior notes, representing a $234.6 million increase compared to the last quarter. This is a noncash fair value adjustment driven by the increase in our stock price and does not impact our liquidity or operations. Under IFRS, certain derivative instruments such as warrants and convertible debt are required to be revalued at fair market value in the reporting period.

As our stock price increases, the fair value of these instruments rises, resulting in a higher reported liability and vice versa. The reported liability will ultimately be netted at settlement either upon conversion to equity or expiration and does not represent an actual cash outflow. Finally, regarding our outstanding ATM facility, we've sold 6.2 million additional shares during the quarter. Thank you, everyone. That concludes the prepared remarks section of the our earnings call.

Operator, please open the call for questions.

[Operator Instructions] Our first question comes from the line of Greg Lewis of BTIG.

Haris, thanks for the update on kind of the progression of the HPC opportunity as you guys think about it. I wanted to talk a little bit about that. You called out a few of the sites. I mean it sounds like that Washington and maybe Tennessee could be maybe move up in the queue just given their size. So I guess, one, as we think about beyond -- I guess, first, if you could talk a little bit more on how you see the opportunity in Asia progressing. And then as we kind of continue to gain momentum in Asia, how we're able to kind of expand that into the U.S. Jeff, if you're on the line, I guess, I'd be curious to know about -- it looked like we added a site in Ohio, Niles, Ohio. Are you there?

Okay. Can you hear me okay?

Yes, sir, please to proceed.

Yes. Okay. So sorry about that. I'm not sure what was going on. So Greg, let me answer your question, I answered it once, I think nobody heard me. So it's not that we're going to do Malaysia first and then go to the U.S. We're doing both simultaneously. We're definitely moving forward in Malaga and also in a number of locations outside of the U.S. So that I think I answered your first question. Do you have a second question? I can't remember what it was.

Yes. I mean -- and I guess I was curious, like are these -- are all these sites going to be NVIDIA, are you only looking at it sounds like we're focused on the is that kind of kind of -- or could we say it seems like some of these other data center [indiscernible] are looking beyond NVIDIA. I know at 1 point, you were also looking to develop something beyond just an ASIC chip as well.

So right now, everything that we're doing in AI is largely NVIDIA-based. So the -- we're not looking at developing our own AI at the moment, if that's what your question was.

Okay. And then my other question was around the I guess we acquired a new site Niles, Ohio. I'd be curious what was kind of the process in that? Was that a site that would have been looking at? I'm trying to understand, I guess, a couple of things. One is realizing every site is different. How should we think about the timing of incremental site allocations? And I'd be curious about that.

Yes. Great. This is Jeff. So the Niles, Ohio site was actually acquired in a few months ago. We just finalized it and made the announcement last month. So strategic acquisition, 100 miles from both of our sites -- other sites in Ohio. Energization does not come until Q1 of 2029. So it's a [indiscernible]. So we think it's like us additional optionality depending on how we -- what direction we go with the Clarington and the Massillon site long term. So just long-term optionality. That's how we're thinking about that. And in general, we are in a mode of actively looking for sites that might be useful. So we are -- we have a group that's actively doing that.

Our next question comes from the line of Kevin Cassidy of Rosenblatt Securities.

This is Chris Meyers on for Kevin Cassidy. And I'm just looking if you guys could provide some additional specifics on the reason for the delay in the more 04 SEALMINERs shift?

Yes. So just to be clear, it's the new generation of architecture Well, we're very comment in the technology still. It's really just that the implementation of it is quite a bit more difficult than we originally anticipated and does involve significant changes to the design flow of EDA tools. And so we're just working through those things to do that. And really the first of the delay.

Okay. And if I could ask a follow-up, are there any additional R&D amendment that those come to market as the SEALMINER 04 comes to market?

I mean I don't think there's going to be any especially unusual R&D expenses. Of course, our R&D expenses associated with any chip.

Right. And then, I guess, if I could ask one last follow-up, and you could describe the cost difference between assembling these SEALMINERs in the U.S. versus outside of the U.S. provide a little detail on that would be great.

So there'll be -- obviously be a slight increase in the production costs for [ $0.01 ] in the U.S. So if you think about 3 months in the seal minor, about 70% to 75% of the cost is the chip. So the remaining 25% to 30% is the balance of the manufacturing is really where the delta will be on. So yes, we would expect the debt to be higher in most cases. But again, depending on what the tariff is more exporting countries, we don't have an exact number for you at this time.

Our next question comes from the line of Mike Grondahl of Northland.

Two questions. One, how will you decide whether you'll go to cloud service provider route versus the colocation route? Any insight there would be helpful. And then two, could you talk a little bit about the demand you're seeing in Norway and how that compares to the demand for some of your U.S. sites?

Yes. So with respect to versus co-location, we are definitely putting our emphasis on the AI cloud space. we will opportunistically use colocation where didn't mind makes sense, but our primary focus is on AI cloud. With regards to Norway, yes, I mean, the site the demand is largely driven by that site being one of the few places in Europe where you can get a large site with low power. And in very specific case of our side, it's very much developed towards already being a Tier 3 type of data center. So it doesn't take much effort to -- or much resources to change it, is it would a normal Bitcoin mining side. So there was a lot of interest. There is a lot of interest in for those reasons.

Is the demand greater for that site as compared to maybe a U.S. site?

I don't know -- I mean perhaps because it's so close to being a fully developed Tier 3 site, the answer is probably yes. It just takes less effort to turn it into an AI data center than our U.S. sites.

Yes. I think time to power has really been what's been driving the demand for it.

Obviously, the site, as Haris said, has a lot of attractive attributes that make [indiscernible] power much shorter. We think we could have this up and running sometime in the second half of next year possibly. So I think that's driving a lot of it.

Our next question comes from the line of Nick Giles of B. Riley Securities.

I wanted to ask a question regarding your financial options for development of the HPC/AI capacity. So kind of to what extent do you expect that CapEx to be funded by your GMV partner and additionally, is there any impact on your part of capital from your PTC holdings given that it could potentially have collateral rates?

So you've got a lot of background noise there. I'm not sure we asked that. Were you asking about a development partner?

Yes, Jeff, apologies [indiscernible]. So the question is basically -- on the financial motion of the development of HPC/AI. So I would wonder to what extent do you expect total CapEx to be funded by your JV partner? And is there any impact on your [indiscernible] capital from this point holding because I would assume they could potentially be you as collateral and help bring the rate down?

So at the present time, we don't have a JV partner in this development of HPC/AI, so we had initially guided that we were going to approach using a joint venture partner, development partner. We did have an LOI with a group previously this year. We looked at LOI expire in favor of basically pursuing the opportunity more on our own using HPC contractors and augmenting our own internal development team. So we are -- not to say we may not pursue it in the future, but at the present time, our strategy is to pursue it more on our own.

And my follow-up is about regarding the most optimistic scenario way of 100 megawatts of AI cloud capacity. -- could you provide more detail on how this capacity will be allocated across the site as possible.

I'm having trouble understanding the question. Sorry about that.

No. Can you hear me? I mean, this is about 200 megawatts of 12 capacity, but if you offline as -- so yes, you can provide the fleet in a side would be helpful.

Of the 200 megawatts, right, that includes in Singapore includes Wenatchee, includes Tyal in Norway. And we expect the vast majority of that to be an AI cloud as opposed to colocation. Is that question -- did I answer your question?

Our next question comes from the line of Mike Colonnese of H.C. Wainwright & Company.

I appreciate all the updates today. Just a couple for me. First one, as an ASIC manufacturer with better visibility into chip availability, do you foresee any procurement risks with regards to carrying some of the latest gen GPUs from NVIDIA for the broader infrastructure industry?

We haven't seen anything so far. I mean, it's I'm not sure that Jeff and I are well positioned to answer that question right now because we're not directing in the procurement. We're getting what we require for mid-year, but our initial quarters are relatively small. So I'm not sure if they're indicative of the industry as a whole.

Got it. Got it. And Haris, you mentioned in your prepared remarks that you intend to continue to deploy your still minor ASICs across your sites. Can contain build out the decline in mining business in 2026? I guess how should we think about growth coming out of this year and looking into next year for Bitcoin Mine?

We haven't released any estimates on that yet, but it will be significant. We're not plateauing at the level that we're at now. We're still on a very steep upward trajectory.

Yes, Mike, I would say, I think we've got a lot of levers to pull there. We've got new capacity coming online in Massillon, Ohio, 221 megawatts in [indiscernible], 50 megawatts. We have some unused capacity in some of our current sites right now. And we have -- also have a couple of 100 megawatts that is dedicated to or generation miners that we'll be looking to replace in the coming months. So it really needs a capacity and ASIC manufacturing will be sort of the bottleneck not power.

Our next question comes from the line of Dylan Hessman of ROTH Capital Partners.

To follow up on some of the cloud things about the $2 billion ARR. How do you sort of expect your customer base to be across that $2 billion? I know you mentioned some of the industries you're working in, but do you expect it to be multiple customer -- like multiple big customers? Or are we talking 10, 20, 30 or 30-plus small contracts?

So I think it will be a combination of both. So in places like Asia, we are seeing a lot of demand from that, I would say, small and middle market companies that are looking to stand up some AI models in some of those industries that we mentioned. We're also seeing demand from larger, call it, and large-sized enterprise customers that are looking to deploy high double, low triple digit megawatts in GPUs. So I think to get to that number, as we kind of laid out as our most optimistic scenario would likely involve a combination of both solid middle market and maybe 1 or 2 larger enterprise customers.

Got it. And as a follow-up, on the DAS and build side of things, how are you guys seeing build cost and supply chain trending as you sort of get closer to the finishing, I guess, a lot of sites that are in the pipeline?

Yes. Look, I think the long lead time items are still the same ones as we know from before, it's a lot of the electrical equipment, transformers, switchgears, breakers, things like that. Look, I think they're out there to be found. We've got a very experienced procurement team. This is not the first time where this type of equipment has been in high demand and shortages and we've had success sourcing in the past. I think some of the potential tenants and other EPCs that we may be working with, there's a possibility that they have some of that stuff kind of queued up as well. So the avenues to get it. But I think the supply chain still remains stressed to some extent.

Our next question comes from the line of John Todaro of Needham & Company.

I just wanted to confirm stuff from earlier. So one, it does seem like the focus is more on AI cloud, so GPU as a service versus HPC colo. Are you seeing a lot more demand in that area? And should we also expect the Ohio sites would be more geared for AI cloud versus a colo lead? And then I have a follow-up.

The Ohio side, yes, I think the answer is yes, it could be more and more AI cloud, but it's a large enough site where it could be divided as well. So final determination of that site is not clear at the moment. You do have a follow-up question you said.

Yes. But also just within that, I guess, is there -- are you seeing more customer demand for AI cloud versus colo. And then I guess my follow-up would be just you think about returning more megawatts out there. We have heard from some peers that there is kind of a stranded power available. Is there kind of a guardrail of how much a capacity maybe annually you guys think you could procure?

We're seeing a lot of demand for both colo and AI cloud. So I start to quantify which one is more or less, but we're seeing a lot for both.

Yes. And then yes, I agree with that. I mean I don't -- I think it's just different demand. I mean, obviously, different core customers are looking for some same customers are. But I think the demand right now is robust. Time to power. I think is the most critical, it was recurring thing we're hearing. So 2026, early '27 power is desirable right now. So I think for both business models. On the power procurement side, yes, I mean, look, I think we're seeing that, too. I mean there is some stranded power. I think we're seeing also more of a willingness to -- for tenants and end users to be more accepting of that type of power behind the meter power potentially. And so we're seeing that, and we're -- like I said, we've got a great procurement team that's been able to acquire low-cost power sites in the past, and we're confident in their ability to continue to do so.

[Operator Instructions] Our next question comes from the line of Bill Papanastasiou of KBW.

With respect to the delayed development of the SEAL04 miner, where would you say the confidence level sits today with respect to this ASIC having industry-leading specs when it gets released? Just trying to understand the extent of the delay.

The confidence is high that the architecture is the right one to go core in future design. So I don't think there's any bill lack of confidence in that. And that for subsequent designs, we would be using it in where the confidence is low, it is the exact timing of the release.

Bill, did you have a follow-up question?

No, just that question.

I think we have one more...

Our next question comes from the line of Brian Kinslinger of Alliance Global Partners.

I'm just wondering if you can help provide some sort of estimates or how you see return on invested capital for cloud service provider versus colocation services?

Yes. I think there -- it's a different return profile. So obviously, we've seen some of our peers out there putting numbers out. And I think those numbers are largely depending on how you're looking at it with I would say both the -- if you're looking at both the GPUs and the data center cost in there and ARR pacific load, I think they can be obviously different. We're seeing co-location. Obviously, the rates very similar to what others have reported recently. So that's typically a yield on cost. So I think it's just going to depend on your construction costs. we're constructing. And I think the end user. I think we're seeing a lot of variance in end user requirements as far as infrastructure needs, where the backup power is required, where they'll be providing some of kind of key infrastructure there. So that on the colocation side, especially, I think it's going to really affect our return on invested capital or any other return metrics.

Ladies and gentlemen, that does conclude the Q&A portion of our call and Bitdeer's conference for today. Thank you for participating. You may now disconnect.