Bitdeer Technologies Group (BTDR) Q4 2024 Earnings Report, Transcript and Summary

Good day, and thank you for standing by. Welcome to Bitdeer's Fourth Quarter and Full Year 2024 Earnings Conference Call. [Operator Instructions] Please note that today's conference is being recorded. I would now like to turn the conference over to your speaker host, Yujia Zhai of Investor Relations. Please go ahead.

Thank you, operator, and good morning, everyone. Welcome to Bitdeer's fourth quarter and full-year 2024 earnings conference call. Joining me today are Jihan Wu, Chairman and CEO; Matt Kong, Chief Business Officer; Haris Basit, Chief Strategy Officer; and Jeff LaBerge, Head of Capital Markets and Strategic Initiatives. Haris will begin today by providing a high-level overview of fourth quarter and full-year 2024 results and then cover the Company's strategy and a detailed business update. After that, Jeff will cover Bitdeer's fourth 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 of 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. For more detailed information on non-IFRS financial measures, please refer to our earnings release that was published earlier today, which can be found on Bitdeer's Investor Relations website. Thank you. I will now turn the call over to Haris. Haris?

Thank you, Yujia, and good day, everyone. Welcome to our Q4 and full-year 2024 earnings conference call. I'm excited to share the many developments happening at Bitdeer and walk you through the progress we've made since last quarter. Before diving in, I'd like to briefly highlight our Q4 financial results for which Jeff will provide more details in a few minutes. Starting on slide 3. For Q4 2024, total revenue was $69 million, gross profit was $5.1 million and adjusted EBITDA was negative $3.8 million. This lower performance compared to Q4 2023 was primarily driven by the impact of the April 2024 halving, increased global network hashrate, lower hosting and cloud mining revenue, and higher R&D costs. These negative impacts were partially offset by higher year-over-year average self-mining hashrate and higher Bitcoin prices. Over the last year, we made a deliberate decision to prioritize resources on the development of our own ASIC technology. This limited our hashrate growth, but gives us massive advantages going-forward that differentiates our business from the rest of the sector. As our mining machines become available in volume in the coming months, we will be able to rapidly increase our self-mining hashrate at a significant cost advantage as well as sell our machines to external customers to begin penetrating the $4 billion to $5 billion annual ASIC market. A core pillar of our strategy is to develop internal technologies and capabilities, driven by our belief that this is the best way to maximize long-term shareholder value. To secure long-term success, we are committed to building a fully vertically-integrated business. This includes developing our own power generation assets, globally diversified data centers and leading-edge mining hardware. As seen on slide 5 of our supplemental presentation and in pursuit of our vertical integration initiative, we successfully acquired a 19-acre site that is fully permitted for construction of a 101 megawatt gas-fired power plant. This project is near Foxcreek, Alberta and was acquired earlier this month for $21.7 million in cash. It also includes approval for a 99 megawatt grid interconnection with Alberta Electric System Operator. We plan to develop and construct the power plant in partnership with a leading engineering procurement and construction firm for the completion and energization expected by Q4 2026. In parallel to the gas fire generator, we plan to develop a 99 megawatt data center for Bitcoin mining as the first phase of this site's development. We estimate the total capital expenditure for constructing the gas-fired power plant to be approximately $90 million, with an additional $30 million allocated for electrical and data center infrastructure. Upon completion, we intend to deploy approximately 9 exahash of our SEALMINER A3 mining machines, which are expected to deliver industry-leading efficiency of 11 to 12 joules per terahash. We estimate energy production costs at this gas-fired power plant to range between $20 to $25 per megawatt-hour based on current gas prices. Additionally, as part of the acquisition, we plan to implement our carbon utilization system that captures CO2 making this project a net zero carbon producer. This initiative is expected to help offset Canadian carbon tax obligations while also creating potential future revenue opportunities through the sale of carbon credits. Further, as part of our energy optimization strategy, we plan to curtail and sell power back to the Alberta grid during periods of peak demand. This approach is expected to enhance cost-efficiency while contributing to grid stability with minimal impact on Bitcoin production. We are extremely excited about this acquisition. This newly acquired site and power generation project represents a big step for Bitdeer, positioning us as the world's first fully vertically integrated Bitcoin miner at scale. With this development, we are on a path to achieve one of the lowest Bitcoin mining costs in the industry. In terms of our ASIC business, which is highlighted on Slides 6 through 8 of our supplemental presentation. We continue to execute on our advanced chip roadmap, positioning ourselves for a transformative 2025. Owning and deploying our own mining ASICs is an integral part of our full vertical integration strategy. It will provide us distinct advantages such as a lower-cost structure, enhanced capital efficiency, and a dramatically improved supply chain compared to the broader industry. In addition, commercializing SEALMINER ASICs allows us to diversify our revenue streams into the rapidly growing ASICs market, where we see a strong demand for alternative suppliers of ASIC solutions. I will now provide a detailed update on our ASIC roadmap. Starting with SEALMINER A1. To date, we have energized 0.4 exahash of our SEALMINER A1 miners and they have demonstrated solid performance. The mass production of the remaining 3.3 exahash remains on schedule and is expected to be completed in March 2025. For SEALMINER A2, this miner represents a significant advancement in our technology roadmap, leveraging our proprietary SEAL02 chip. In Q4, we commenced mass production at TSMC to deliver approximately 35 exahash of SEALMINER A2s by October 2025. This represents a delay of approximately one month compared to our original expectation due to a 6.4 magnitude earthquake that struck Taiwan on January 21, 2025. The SEALMINER A2 is our first commercial miner available for sale to external customers. So far, we have allocated 7 exahash of the current 35 exahash for external sale. Initial customer demand has been extremely strong with pre-orders for the 7 exahash being oversubscribed by a factor of six. We have already received 20% of the total price as down payments on all 7 exahash. Volume shipments to these customers will commence in March 2025. The remaining 28 exahash of A2 capacity is currently intended to be used for our own self-mining. The first batch of air-cooled machines have been delivered to our mining data centers for testing and are running stably. Building on the success of A2, our R&D efforts for SEALMINER A3 are progressing smoothly. This next-generation model will feature the SEAL03 chip, which is expected to deliver industry-leading energy efficiency of 10 joules per terahash at the chip level. Finished wafers from the initial tapeout are expected in March 2025 with production readiness targeted for later this year. Achieving 10 joules per terah chip efficiency would mark a major milestone for the industry, positioning SEALMINER A3 as the most advanced and energy-efficient ASIC on the market. As energy efficiency remains the most important single metric influencing buying decisions, we believe halving the most efficient ASIC is the key factor to winning market share. Unlike other industries such as smartphones where switching costs include significant ecosystem frictions, Bitcoin mining ASICs operates in a highly fluid market where transitioning to alternate machines creates little or no friction. In this space, the most efficient and reliable machines win. With that in mind, we are eagerly awaiting the return of CLO3 sample wafers scheduled for March. Looking ahead to the second-half of this year, our SEALMINER A4 project is a testament to our commitment to maintaining technological leadership and pushing past perceived limitation of ASIC design. The SEAL04 uses a revolutionary new digital chip architecture that significantly enhances energy efficiency and is projected to have a chip level energy efficiency of 5 joules per terahash. Tape-out is planned for Q3 2025. We believe SEALMINER A4, along with our third-generation chip, will position Bitdeer as the leading supplier of the world's most energy-efficient mining machines, significantly strengthening our market position and unlocking substantial value for our customers and shareholders. As we look to the remainder of 2025, we are fully committed to executing a successful entry into the multi-billion dollar ASIC market while also rapidly ramping our self-mining hashrate. On slide 8, of our supplemental investor presentation, we have laid out our expectations for self-mining hashrate into Q4 of this year. Given the significant amount of power capacity we have coming online, our plan is to initially prioritize our current ASIC production towards self-mining. We plan to energize the remaining SEALMINER A1 and 28 exahash of SEALMINER A2s on-top of our existing 8.7 exahash of self-mining hashrate as of January 31st, 2025. Therefore, upon deployment of a combined 31.3 exahash of SEALMINER A1 and A2 machines, we expect our total self-mining hashrate to be approximately 40 exahash. We expect to be able to deliver this with machines fully racked and energized by Q4 2025. Please note that this forecast does not include anticipated additional wafer allocations for SEAL02 or for SEAL03. Depending on the exact manufacturing schedule, these anticipated additional wafer allocations could increase our self-mining hashrate well above the 40 exahash guidance for Q4 2025. In addition, I'd like to address the recently published U.S. Department of Commerce Bureau of Industry and Security ruling regarding advanced computing integrated circuits. Based on our preliminary review, we do not expect that the application of the BIS rules will have any impact on the delivery of SEAL chips. As the Outsourced Semiconductor Assembly and Test, OSAT, companies for SEAL chips are approved OSAT companies under BIS regulations. Next, I'd like to talk about our energy infrastructure and pipeline that's highlighted on Slides 9 and 10 of our supplemental investor presentation. We have secured one of the largest globally diversified power portfolios in the industry with over 2.6 gigawatts of total power capacity. About 1 gigawatt is scheduled to be newly energized over the course of 2025. This massive power portfolio allows us to deploy our SEALMINER machines for self-mining and also capitalize on the significant demand for HPC and AI data center power. Based on a third-party feasibility study, many of our sites are suitable for HPC and AI data centers. In particular, our Clarington, Ohio site is a strong contender given its near-term access to power, scale and fiber, and water resources. We are actively working with leading data center developers on long-term partnerships for selected sites. The shortage of reliable power for AI data centers is a critical challenge for the industry and we believe Bitdeer can play a significant role. We look forward to providing an update to our shareholders in the near future. In summary, we made significant strides across all of our strategic initiatives to close out 2024, and we couldn't be more proud of the entire Bitdeer team. We expect 2025 to be a pivotal year as our efforts start to bear fruit and we look forward to sharing updates on our progress. I'll now turn it over to Jeff LaBerge, our Head of Capital Markets and Strategic Initiatives to go over financial results for the quarter.

Thank you, Haris. Before I go over Bitdeer's fourth quarter and full-year 2024 results, I'd like to remind everyone that all figures I refer to today are in U.S. dollars and that all comparisons are on a year-over-year basis, unless stated otherwise. I would like to also note that these results are unaudited and preliminary, and we are working to file our 2024 annual report and 20-F with the SEC by the end of March 2025. Starting with Q4 results. Consolidated revenue was $69 million versus $114.8 million. Self-mining revenue was $41.5 million, down 11.5%, primarily due to impact from - the April 2024 halving and higher global network hashrate. This was partially offset by an increase in our average self-mining hashrate to 8.4 exahash from 7 exahash and higher year-over-year Bitcoin prices. Cloud hashrate revenue was $2.3 million versus $16.2 million. This decline was primarily due to long-term cloud hashrate contracts rolling-off and our decision to reallocate nearly all of this hashrate to our self-mining operations over the course of 2024. General hosting revenue was $8.5 million versus $25.2 million. Membership hosting revenue was $12.4 million versus $23.4 million. The decrease in hosting revenue was mainly due to the expiration of certain hosting contracts as well as the removal of older and less-efficient machines by other hosting customers following the halving in April 2024. Compared to Q3 '24, total hosting revenue and gross profit improved by 7.2% and 106.7%, respectively, due to higher margins of profit-sharing in Q4 and more efficient mining rigs. Total gross profit for the quarter was $5.1 million versus $27 million and gross margin was 7.4% versus 23.5%. The decrease in our gross margin was primarily a result of the April 2024 halving and the expiration of contracts from our high-margin cloud hashrate business. Going-forward, as we begin to rapidly grow our hashrate with new, much more efficient SEALMINERs, we expect our margins to improve significantly, assuming all else equal in terms of Bitcoin price and network difficulty. Total operating expenses for the quarter were $42.5 million versus $27.4 million. The increase was primarily driven by higher engineering staff and other costs related to the R&D of our ASICs roadmap and non-cash amortization expenses of intangible assets related to the acquisition of FreeChain. Sequentially versus Q3 '24, total operating expenses decreased by approximately 1%. This was primarily due to no tape-out costs in Q4, which was almost entirely offset by increases in the non-cash amortization expense-related to the acquisition of FreeChain, year-end bonuses and consulting expenses for capital markets activities and compliance activities. Other net gain loss for the quarter was a net loss of $479.8 million versus a net gain of $1.1 million due to a $55.8 million non-cash derivative loss on the Tether warrants and a $413.7 million non-cash derivative loss on the convertible notes issued in August and November. This loss was caused by the significant increase in our stock price during the quarter. IFRS net loss for the quarter was $531.9 million versus $5 million. Adjusted profit and adjusted EBITDA for the quarter was negative $36.9 million and $3.8 million, respectively. This quarter's losses were primarily due to the year-over-year revenue declines in our cloud hashrate and hosting businesses and lower gross profit margins in our self-mining business as a result of the April 2024 halving, and higher R&D expenses as described previously. Quickly touching on our full-year 2024 results. Revenue was $349.8 million, gross profit was $66.4 million, and adjusted EBITDA was $39.4 million. A detailed breakdown of our full-year results by business line can be found in our earnings press release that was issued today. In terms of Q4 2024 ending balance sheet, there were several notable items that I would like to highlight. Prepayments and other assets were $310.2 million, up from $97.1 million. This change was primarily driven by our advanced payments to TSMC for our SEAL02 mass production. Inventories were $64.9 million, up from essentially zero. This mainly included wafers, chips, WIP and finished steel miner inventory. Intangible assets of $83.2 million and goodwill of $35.8 million are mainly from our 2024 acquisition of Norway and FreeChain. In liabilities, derivative liabilities were $763.9 million, which relates to the tether warrants and August and November convertible senior notes. These are non-cash fair value adjustments driven by the significant increase in our stock price and do 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 each reporting period. As our stock price increases, the fair value of these instruments rise, resulting in a higher reported liability. The record liability will ultimately be netted at settlement either upon conversion to equity or expiration of the underlying securities, and do not represent an actual cash outflows. With respect to liquidity, we ended the year in a strong financial position with $476.3 million in cash and cash equivalents, $77.5 million in cryptocurrencies, and $208.1 million in borrowings, excluding derivatives. In Q4, we began to hold a portion of our Bitcoin mined and we anticipate to continue this strategy for the foreseeable future at management's discretion. Please note, our P&L and adjusted EBITDA does not include any fair-value gains or losses related to cryptocurrencies on our balance sheet due to IFRS accounting rules that require cryptocurrencies holdings to be accounted for on a cost basis, net of impairments rather than fair-value basis. Moving on to our cash-flow statement. Q4 2024 cash used in operations was $325.1 million compared to $67.1 million. This material change was primarily driven by payments of $190.6 million to TSMC for SEAL02 wafers, which represent more than half of the required wafers for the previously-announced 35 exahash and payments of $52.8 million to TSMC for of tape-out of SEAL03, including initial risk wafers. Please note that while the SEAL03 tape-out costs were paid in Q4 2024, it will be expensed on the Q1 2025 P&L. Further, I think it's important to highlight that our cash inflow from the sale of Bitcoin related to our Bitcoin mining business are classified in the investing section of our cash-flow statement. Net cash used in investing activities was $10 million, including $48.4 million of capital expenditure for infrastructure construction and mining rigs, offset by $38.8 million of proceeds from the sale of cryptocurrencies received from our principal business. Net cash generated from financing activities for the quarter was $522.8 million and was primarily related to the proceeds from our convertible note issued in November and our ATM program. In terms of our 2025 capital expenditures, we anticipate CapEx including Foxcreek, Alberta, to be in the range of $340 million to $370 million and could be fully-funded by the existing balance sheet. It should be noted that this infrastructure spend assumes the sites are developed for Bitcoin mining and does not include CapEx for SEALMINERs used for self-mining. Finally, I'd like to provide some context on the $1 billion ATM shelf registration that we filed on January 3rd, 2025. I want to emphasize that we will continue to utilize our ATM in a disciplined manner that we believe will add value to existing shareholders. Our management team's ownership in the company is among the highest in the industry, thus aligning our interests with our shareholders. On slide 12 of our supplemental investor presentation, you can see our ATM usage for 2024 in January of 2025. Notably, we did not begin meaningfully utilizing our ATM until the fourth-quarter of 2024 when our stock price had risen significantly. As a result, our total ATM usage in 2024 plus January 2025 accounted for only 14.4% of our shares outstanding as of January 31, 2025. The primary objective of this ATM is to enhance balance sheet flexibility and provide us with the necessary liquidity to secure additional wafer allocations from TSMC, which typically require large advanced payments. This flexibility will enable us to significantly scale SEALMINER production to meet growing demand and position us to capture significant market-share in the multi-billion dollar ASIC market. Thank you, everyone. That concludes the prepared remarks sections of our earnings call. Operator, please open the call for questions.

[Operator Instructions] Our first question coming from the line of Mike Colonnese with H.C. Wainwright. Your line is now open.

Hi, good morning, guys. Great update today and thank you for taking my - going to be focused on the mining rigs here. So there have been several media reports out there that suggest some U.S.-based miners are experiencing delays in receiving their ASICs from your biggest competitor amid the trade tensions between the U.S. and China here. It'd be great to hear your thoughts on this and if this has changed the demand for your SEALMINERS.

So I'll take a quick try at that. We are just watching the news the same as you for in terms of what's happening with other folks miners. We have not had any direct experience of that ourselves. So I don't know, Jihan, if you want to elaborate on that.

Well, I think that this issue, according to the recent news, has been solved quite good by the lobbyist and by other industrial association's efforts to communicate with this White House. I think we consider this issue as well because our mining rig, we will need to explore the market over the United States. Hopefully that this issue can be resolved quite smoothly.

Got it. Thanks for the color there. And staying on the proprietary mining rigs here, so how should we think about manufacturing capacity you have available to produce your next batch of A2s in the next-gen A3 rig? I know you're shipping 35 exahash of the A2s this year. Just trying to think through total capacity to produce your next batch of rigs here.

We do expect the capacity to continue to go up. TSMC is really a trusted and very good partner for us on this, but we don't really pre-announce any of those. And - so as we get more allocation from TSMC, we will make those announcements in the future. We are optimistic and our relationship with TSMC remains strong.

Got it. Helpful. And just one more to slip in. How do we think about the revenue recognition in ASIC sales? I know you received a 20% downpayment for the full 7 exahash, but how do the payments progress through the stages of the delivery?

Sure. So I'll take that. So revenue is typically going to be recognized upon delivery of the A6. So the down payments right now are being held on just - on the balance sheet and will be recognized once they're fully delivered to the customer.

Great. Thank you for taking my questions.

Thank you. Our next question coming from the line of Nick Giles with B. Riley. Your line is now open.

Yes. Good morning, everyone. This is [Fedor Chabollin] asking question on behalf of Nick Giles. I wanted to touch on CapEx outlook for 2025. You - Jeff, you mentioned some number, but could you outline the allocation just kind of if you may specify the sites involved for vis-a-vis part and intendent and uses of these funds? So what is the breakdown between self-mining CapEx and say HPC or AI spending? Thank you.

Sure. So the CapEx numbers that we discussed on the call are specifically for assuming the sites are used for Bitcoin mining and that reflects the - just the infrastructure costs of those. So that infrastructure cost could obviously, change if we were to pivot it to, for example, HPC AI. It also does not include the cost of the steel miners that we would put in the - there for self-mining. So of the numbers that we referenced earlier, those would cover our expansions in Rockdale, Texas, the hydro expansion there. It's build out in [Baucon] Massillon, Ohio, and clearly the phase one of Clarington, Ohio, and would also cover the generation portion of the new facility in Alberta. And again, assumes all those will be going to Bitcoin mining, obviously, if one of them were to be pivoted to something else, that would change.

Great. Thank you for that. And then on Clarington side, what will be ready by the outlined date in 3Q 2025, so this year? And the same question from Masilon side, just infrastructure wise, what will be done by this?

So the Clarington site, the Phase 1 of 266 megawatts is expected to be available in 2024 - excuse me, 2025, same with Masilon, Ohio, the full 221 megawatts should be available there as well.

Yes, I mean - yes, but it's a - it's just going to be energized or substation in-place or what about the shelf data center?

So that would just cover the infrastructure - the power infrastructure. So again, we are under-construction in Masilon, Ohio, and that will be ready and that is targeted for the second-half of this year. Again, if we continue with down with Bitcoin mining, we should have that completed by the end-of-the year. And the Clarington substation, we are currently still in the development-stage right now, so we would expect construction on the substation this year.

Got it. Thank you for the color. I'll turn it over for now. Thanks. Best of luck.

Thank you. Our next question coming from the line of Darren Aftahi with ROTH Capital. Your line is now open.

Hi, good morning, good evening. Thanks for taking my questions. I'm just kind of curious with the full verticalization with the acquisition in Alberta. How are you guys kind of strategically thinking about Bitcoin mining for sites you currently have where your maybe - I guess, question is in the vein of full verticalization of behind the mirror versus not? And then my second question, how do we kind of think about your strategic decisions about self-mining versus selling those rigs to third-parties just given, Haris, your comments about the six time over subscription? Thanks.

Okay. I think I can take a first crack at that. So with regards to our existing sites, we haven't changed our position there. We're still very bullish on Bitcoin mining. And for the most part, you can assume that we will continue allocating Bitcoin mining resources to our existing sites. As we pivot any one of those sites, we will definitely announce that. But it hasn't really - the acquisition of the site in Canada or Alberta has not really changed our thoughts on the other sites. The allocation between selling the machines versus using them. Until we have satisfied most of our own spare capacity, we will have a strong preference for using the machines for self-mining. But as our capacity fills up and especially with the A3, we will lean further and further towards sales versus self-usage. So you can expect that transition to happen sometime in '26 probably where we'll have more emphasis on sales than on self-usage. But for now we're focusing very much on using the A2s, in particular, for our self-mining.

Helpful. Thank you.

Thank you. Our next question coming from the line of Kevin Cassidy with Rosenblatt Securities. Your line is now open.

Hi, thanks. Thank you for taking the question. And my question also is around the steel mining rigs. You're going to start shipping in March for the A2. How long will it take your customers to qualify it and say that they have accepted it in what's the next phase after that?

So it typically doesn't take very long for mining rig to be qualified, but maybe I'll ask either Matt or Jihan to - who are closer to some of these customers that are qualifying, if they have a timeframe in mind.

I think the time we committed has already included the time to do everything, including their qualification acquired. We committed the time, we're shipping out the money rig.

Okay, maybe as a follow-up - yes, I was just wondering as you go through the A3, if that's going to be the larger model for sales, is that a quicker qualification stage?

From verification of the chip to mass production, to really happens, usually it will take six months. I think this is quite fast speed in the industry to bring the chip into mass production.

Okay, great. And maybe just one other question. How many customers have you - will you be shipping to, the A2?

I think we have said in earlier earnings call that it was more than 60 customers.

Okay, great. Thank you.

Thank you. Our next question coming from the line of Brian Kinstlinger with Alliance Global Partners. Your line is now open.

Great. Thanks so much for taking my questions. As it relates to external sales for SEALMINER, how are you thinking about the short-term pricing strategy and medium-term given the efficiency? And then can you call out the - maybe total expected tape-out costs by quarters, if you know them roughly for 2025?

Okay. So our existing pricing for the A2s we announced earlier was $15 per terahash, and we think that's a very competitive price based on our - being a new entrant into the market and just wanting to be - get a large market presence initially. So I think as we get our A3s and our energy efficiency is leading of any other vendor, we have much more flexibility in our pricing and we expect that our pricing will reflect the energy efficiency of the machines. So you can expect to see that. And then what was the second part of your question?

It was just the tape-out cost. And I think you mentioned the first quarter is when you'll see some. Just maybe for modeling purposes, you can kind of discuss what you are…

Yes. So the tape-out costs for A2 and A3 are already paid for. So those have obviously, already been paid. The tape-out costs looking forward are primarily for the A4, and so we - that should be in line with A2 and A3 costs. We will incur that most likely in Q3. So there may be other small tape-out costs for other things, but I don't think that's going to be as meaningful as what we've already incurred for the A2 and A3.

Great. My second question - yes.

So the - sorry. For the A3, we tape-out two type of designs simultaneously. So for the A3 R&D may be extra spending expenditure on the design and R&D cost. I'm not sure whether this has been delivered to the analysts quite clearly or not. So actually, we spend two months on the R&D expenses.

Got it.

Yes. Just to emphasize that, that A3, we had two different competing designs and - so the tape-out costs for the A3 were almost double because of that.

Makes sense. And then my other question, I guess, a larger HPC AI question that I'm sure investors are thinking about. Since the DeepSeek moment, if you will, and obviously, suggesting anyway that they were able to perform with unusually low cost in power, and then we've seen some of the hyperscalers kind of have some uncertainty about their plans. Have you seen any changes in conversations about HPCAI opportunities or has there been zero change?

So we have after each of those events queried our connections and the people that we're discussing partnerships with, and we have not seen any pullback from the people that we're speaking with.

Great. Thank you so much.

Thank you. Our next question coming from the line of Mike Grondahl with Northland Capital Markets. Your line is now open.

Hi, guys, two questions. At a high level, how would you describe the demand environment for HPC AI, and specifically, how many megawatts of your power are you offering? And then two, post the 35 exahash for A2, do you have a rough season when we should expect the next batch or when you'll get some insight as to when you'll get another batch?

I'll take the first part of that. For how many megawatts we're looking at for HPC AI, it's more about whether - which sites we're looking at. And once a site is allocated to HPC AI, it's likely to be the entire site. One of the first sites we're really looking at is our Clarington, Ohio site. So you can - the size of that site, the initial phase is 266 megawatts. The total is 266 plus 304, so 570 megawatts. Those are still in discussion, so it's really site-by-site basis rather than partitioning of the megawatts. And I think that your next question was about the next allocation of A2s or A3s, is that correct?

Yes, roughly when do you - I mean, is it a spring, summer? I know you don't know exact timing, but I don't know how you're thinking about it at a high-level.

So we have said that we could have 40 exahash by the end of this year. And that we're optimistic that, that might be increased. So we expect that there may be more allocation that hits even this year in order to increase that. So - but as we said earlier, we really don't want to pre-announce anything until we have it securely in hand.

And Mike, just to clarify that, that 40 exahash is - includes the 35 and the 3.5 or 4 of the A1s that we already have allocated and that allocation is already given to us and those are under production right now.

Got it. Okay. Thank you.

Thank you. Our next question coming from the line of Greg Lewis with BTIG. Your line is now open.

Hi, thank you and good morning and good evening, everybody, and thanks for taking my question. I just was hoping to talk a little bit more about the decision to go vertically-integrated with the acquisition in Alberta. I guess as you think about building out your Bitcoin mining footprint and maybe your HPC footprint, should we kind of look at this as kind of a one-off or is this something where - this was the first one that kind of met all of our wants and needs and - but don't be surprised if we're back in the market acquiring more infrastructure to do this longer-term.

When we look at the Canadian site, what attracts us is that this site is not only about the 100 megawatt we are quite sure right now we can build. We already got the license almost everything ready. We just need to spend money and time and the team on the site to constructing it. But we are also very passionate about the potential future to build a gigawatt level of power plants and data center or mining facility infrastructures on that side because that side got lots of our gas productions that right now have low-ready pipe to sell to outside the world. So potentially we can expand that to really large operations. Our strategy will like a very large like a giga mining operation or a giga site to gain the economics of scale. So I think we will be quite careful to expand into new sites, but I think we will always be interested in looking for new sites to develop.

Okay, great. Super helpful. Thanks. Go ahead.

Greg, I would just add to that too. But what's unique about this site also is that not only do we have the ability to produce behind-the-meter for ourselves, but it also comes with the interconnection to the grid, which is increasingly more difficult to find. So that gives us the ability to sell power back to the grid, help balance the Alberta grid, and potentially reduce our power costs even further there. So we think it's a very unique asset and --

Okay.

Yes. And I can just chime in there that Alberta is a location that has really a surplus of engineering and construction talent related to the energy industry. And so it's a very easy place to build these kinds of assets because there's a lot of engineering and construction talent available locally.

Okay, great. And then just one quick one from me. You've mentioned all the customers that are going to be getting the SEALMINERs. Is there plans or are there thoughts around selling rigs to companies, which then you will then host? Or is it really just going to be third-party other companies that are going to go then take those rigs and mine them elsewhere?

Yes. We have no objection to that. But those are two separate businesses and we - there is some advantage in this vertical integration that we can, in many cases, sell mining rigs and possibly host them as well. I don't think it's going to be a major thrust that we're going to try to get a lot of that business, but I think that we're open to it.

Okay, great. Super helpful. Thank you very much.

Thank you. Our next question coming from the line of Brian Perrault with Needham & Company. Your line is now open.

Great. Thanks guys. For the purchase orders for new ASICs, could you remind us just how much folks are paying for the deposits, maybe what's come in so far? And then on HPC, would you guys be open to partnering with another peer in the space to develop the Ohio sites for HPC? Thank you.

Yes, so we collected a 20% deposit from all - everybody that's ordered. So that's already been collected. And again, we'll collect the balance before shipment. And on the HPC side, I mean, look, we're always open to any strategic opportunity that makes sense that would add value to our shareholders. So, we - I don't think we've foreclosed out any potential partnerships.

Thank you.

[Operator Instructions] Our next question coming from the line of [Rocky Wang] with Yong Rong Asset Management. Your line is now open.

Okay. Thank you. Good morning and good evening. So my first question is, could you tell how long is the wafer prepayment cycle to the TSMC? I mean, how many months does your $200 million kind of one payment correspond to the TSMC V3?

Rocky, can you repeat that one more time? You kind of broke up.

I mean, how long is your wafer premium payment cycle to the TSMC?

…is four months.

So your $200 million advanced payment is for four months wafer implant?

Yes, the TFS schedule, so we have the payment in advance in four months.

So okay. Thank you. So, my second question is, so I see you have $1 billion ATM plan. So just from the current year stock price, it seems that if you want to get all the financing from the ATM plan, it will dilute the stock price greatly. So, will your ability to obtain cash, a fact you are processing acquiring the wafer capacity from TSMC. So - and what are the main factors considered by TSMC when you obtain a wafer capacity from TSMC?

So I guess with respect to - the ATM is your question, are we planning to use the ATM to finance this, or?

I mean, you are recently and the - your stock price, will that affect your - obtain the money from the market, and will that be a fact your ability to acquire the wafer capacity from TSMC? I mean, if you don't have that much cash, you can't get the wafer capacity from TSMC.

As the way we finance the wafer, it's not only one way because sending shares at the market. We've got a multiple ways to finance it. We can swing comparable amount. We can borrow money. We already got $17 million of credit line from Singapore Bank. We can also sell our minor rigs as we showcased that we already got 30,000 units of minor rigs from market, and we can also deploy those minor rigs and mine, we can sell it on the market and we can go to the cash flows to finance future with ordering from fab. So we are not to - only. We have more than one way to finance the wafer ordering. And…

So what…

-- I think, essentially it's because if you look at the mining activity with our advanced design and a super effective mining rig production, it's a very high return on assets business. I think, of course, there can be challenges, for example, for the wafer price crashes and the stock price crashes. But if you look at our cash balance of - at the end of the year, we already financed lots of money and we already built up inventories and we already have lots of many capacitors ready. So I think it will be okay for us to venture into the future and to see how the situation goes. Definitely, we will need to finance the money through model with, but by the way, we not necessarily rely on the ATM.

All right. Thank you. That's all my questions.

Thank you. And there are no further questions at this time. This concludes today's conference call. Thank you all for participation and you may now disconnect.