Bitdeer Technologies Group (NASDAQ:BTDR) Q3 2024 Earnings Call Transcript November 18, 2024 8:00 AM ET

Company Participants

Yujia Zhai - IR, Orange Group
Haris Basit - CSO
Jeff LaBerge - Head of Capital Markets & Strategic Initiatives
Jihan Wu - Chairman and CEO
Matt Kong - Chief Business Officer

Conference Call Participants

Greg Lewis - BTIG
Brett Knoblauch - Cantor Fitzgerald
Kevin Cassidy - Rosenblatt Securities
Darren Aftahi - Roth
Mike Colonnese - H.C. Wainwright & Company
John Todaro - Needham & Company
Nick Giles - B. Riley
Brian Kinstlinger - Alliance Global Partners
Mike Grondahl - Northland
Bill Papanastasiou - Stifel

Operator

Good day and thank you for standing by. Welcome to the Bitdeer Q3 2024 Earnings Call. At this time, all participants are in a listen-only mode. After the speakers’ presentation, there will be a question-and-answer session. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Yujia Zhai from the Orange Group. Please go ahead.

Yujia Zhai

Thank you, operator, and good morning, everyone. Welcome to Bitdeer's third quarter 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. Harris will begin today by providing a high-level overview of Bitdeer's third quarter results and then cover the company's strategy as well as provide 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.

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 from those expressed or implied in such forward-looking statements. 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. 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 Investor Relations website. Thank you.

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

Haris Basit

Thank you, Yujia, and good day, everyone. Welcome to our first earnings call since we became a public company last year. Bitdeer stands out in the industry with a unique value proposition, and we are excited to share the many developments happening at the company and walk you through the progress we've made since last quarter. Before diving in, I'd like to briefly highlight our Q3 financial results, and Jeff will provide more details in his section.

For Q3, total revenue was $62 million, gross profit was $2.8 million, and adjusted EBITDA was negative $8.5 million. This lower performance compared to Q3 last year was primarily driven by the impact of the 2024 halving, higher global network hash rate, lower hosting and cloud mining revenue, and higher R&D costs related to the one-off development expense of our SEAL02 chip. These negative impacts were partially offset by slightly higher average self-mining hash rates for the quarter and higher Bitcoin prices. While many of our peers have pursued a self-mining hash rate growth strategy, we pursued a more long-term strategy of first focusing resources on the development of our own ASIC technology. We believe this significantly differentiates our business from the rest of the sector in terms of our strategic positioning, revenue and cost structure.

In parallel, we strategically expanded our energy infrastructure pipeline to be online when our mining rigs are also available. This allows us to rapidly increase our self-mining hash rate using our own ASIC technology in the coming months and years. This strategic focus on technology has been a core part of our strategy and is driven by the DNA of our leadership team, which has a proven track record of founding and scaling successful tech companies.

As we look ahead, we anticipate the industry moving towards modernization, which will require Bitcoin miners to diversify and differentiate their business models in order to remain competitive. To this end, we are focused on building a vertically integrated business by developing industry-leading hardware and software solutions across our ASICs and AI cloud business lines, as well as ensuring geographical diversification across our global operations and supply chain. We are prioritizing strategic growth anchored in deep R&D to create a strong competitive moat around our business.

With regards to our ASIC business, we have made significant strides in our product roadmap this year and are ramping up to commercialize our advanced SEALMINER ASICs to disrupt what industry experts have estimated to be a $4 billion to $5 billion market annually over the next five years. We see strong demand for diversified ASIC solutions and suppliers. Commercializing our SEALMINER ASICs will enable us to enter this growing market, diversify our revenue streams, and significantly accelerate the growth of our self-mining business. Having our own ASIC mining rig is a big step towards full vertical integration and will provide us with distinct advantages, including a diversified revenue stream, lower cost structure, higher capital efficiency, and dramatically improved supply chain compared to the rest of the industry.

We announced our R&D technology roadmap for our SEALMINER products in early June and have already energized the first batch of SEALMINER A1 mining rigs powered by the SEAL01 chip. These chips are performing within our expectations. We began mass production in October and have plans to install 3.7 exahash in phases between December of this year and Q1 2025. This deployment will coincide with the energization of our Tydal, Norway Phase 1 Project as well as our hydro cooling lines at our Rockdale, Texas facility.

In October, we successfully incorporated our SEAL02 chip into our SEALMINER A2 mining rigs, including both air-cooled and hydro-cooled models. Testing of SEALMINER A2 mining rigs achieved 226 terahash per second for air-cooled and 446 terahash per second for hydro-cooled, with both having 16.5 joules per terahash efficiency. We commenced mass production of our SEALMINER A2 and the first production run is expected to deliver 18 exahash per second, which will be used for self-mining and selling to external customers. Notably, the commercial sale of SEALMINER A2 will mark a significant milestone as we begin delivering these machines to customers starting in Q1 2025. We will continue to provide progress reports on our hash rate production as a commitment to maintaining transparency for all stakeholders, including shareholders, suppliers, and especially clients who will need clear guidance to make their commercial decisions.

In addition, we taped out our third generation chip with an industry-leading target efficiency of 10 joules per tera hash and anticipate initial sample wafers in Q2 2025. Looking forward, we remain fully committed to executing a successful market entry into the multibillion-dollar ASIC market in 2025. We are already engaged in discussions with a number of potential customers and early demand is promising, indicating strong interest in our cutting-edge technology and the industry's desire for technology and supply chain diversification. We are excited to meet this demand and begin driving a differentiated and diversified revenue stream for our shareholders.

Last, but not least, I want to emphasize that we are planning to tape out a fourth generation chip in the second half of 2025, which we are targeting to achieve an unprecedented 5 joules per terahash energy efficiency. We believe this chip, along with the third generation chip, could position Bitdeer as the preeminent supplier of the most energy-efficient mining rigs on the market. We believe this will significantly strengthen our competitive position and unlock substantial value for our shareholders.

Now, moving on to our AI cloud business. Our NVIDIA DGX SuperPOD system in Singapore achieved around 98% utilization in September, establishing ourselves as a leading provider of advanced AI computing solutions in Asia. We possess strong capabilities in deploying AI infrastructure and offering cloud platform services such as bare metal, virtual machine, and serverless GPU. These support computing across multiple regions utilizing both our own and third party assets. We have also recently expanded our GPU cloud capacity into Canada. Our cloud services business model and technology allows us to collaborate with partners to scale quickly in response to customer demand. Demand has been steadily increasing for our GPU cloud services, and we are positioning Bitdeer to serve the rapidly expanding market of companies focused on LLM training and inference, catering to small and medium-sized companies as well as large corporations.

In terms of our energy assets, we are actively exploring opportunities to leverage our global power capacity of 2.5 gigawatts to capitalize on the significant boom in demand for power for HPC and AI data centers. In July, we engaged TLM Group, a leading consultant in HPC and AI data center development to perform a comprehensive suitability analysis on each of our sites. TLM Group completed their feasibility assessment of our US sites and confirmed the suitability of several of them for Tier 3 HPC and AI data centers. These sites have abundant power available in a short time frame, low latency fiber, and plentiful water resources. We have commenced discussions with development partners and potential end users for these sites. A shortage of reliable power for AI data centers is a critical challenge for the industry, and we are well positioned to leverage our substantial power capacity to meet this growing demand. We are actively collaborating with leading data center developers and advisors to secure long-term partnerships and strategic opportunities that can position Bitdeer to play a significant role in the rapidly evolving HPC and AI ecosystem.

Moving on to the build-out of our additional 1.6 gigawatt electrical capacity and infrastructure. Construction across our global sites continues to progress steadily. This quarter, we incorporated more detailed updates into our monthly production reports to provide greater visibility into the status of our infrastructure expansion. The 40 megawatt Phase 1 expansion at Tydal, Norway remains on schedule to be energized in December 2024. In Rockdale, Texas, the 100 megawatt hydrocooling conversion has been delayed by one month due to supply chain impacts from the US dock workers’ strike in early October. Base energization is now expected between January and March 2025. Meanwhile, the 500 megawatt project in Jigmeling, Bhutan is advancing well with the primary substation expected to be completed by Q1 2025. With these projects and other projects such as Phase 2 of Tydal, Norway and Phase 1 of Clarington, Ohio, we are poised to bring over 1.1 gigawatts of new power capacity online over the course of the next year.

In summary, we have many exciting milestones on the horizon. We remain on track to deliver our SEALMINER A2, A6, substantially grow our self-mining fleet at a competitive cost advantage, and leverage our industry-leading global power portfolio.

I'll now turn it over to Jeff LaBerge, our Head of Capital Markets and Strategic Initiatives, to go over our financial results for the quarter.

Jeff LaBerge

Thank you, Haris. Before I go over this year's third quarter financial results, I'd like to remind everyone that all figures I refer to today are in US dollars and that all comparisons are to Q3 of last year. Q3 consolidated revenue was $62 million versus $87.3 million. Self-mining revenue was $31.5 million, up 4.7%, primarily due to an increase in our average self-mining hashrate for the quarter, to 7.8 exahash from 6.1 exahash, and higher year-over-year Bitcoin prices. These positive factors were largely offset by the April 2024 halving and higher global network hash rate. Cloud hash rate revenue was $7.1 million versus $15.6 million. This decline was primarily due to long-term cloud hash rate contracts rolling off. This hash rate was subsequently reallocated to our self-mining operations. General hosting revenue was $9.6 million versus $22.2 million. Membership hosting revenue was $9.9 million versus $16 million. The decrease in hosting revenue was mainly caused by two aspects. First, we're converting 100 megawatts of hosting capacity at our Texas facility to hydro-cooling capacity, which is expected to be fully renovated and equipped with SEALMINER hydro-cooled mining rigs by the first quarter of 2025. Second, some hosting customers removed less efficient miners after the halving in April 2024. This extra capacity is currently being replenished by new hosted mining rigs.

Total gross profit for the quarter was $2.8 million versus $21.1 million, and gross margin was 4.5% versus 24.2%. The decrease in our gross margin was primarily driven by our lower hosting and cloud hash rate revenues and higher average power prices, primarily at our Rockdale, Texas facility. Total operating expenses for the quarter were $42.9 million versus $27.3 million. The increase was primarily driven by higher R&D costs, including a $13.4 million one-off incremental development expense related to the SEAL02 chip, higher stock-based compensation, and non-cash amortization expenses of intangible assets related to the acquisition of FreeChain.

Other net gain/loss for the quarter was a net loss of $14.7 million versus a net gain of $0.9 million. The net loss was due to a $14 million non-cash derivative gain on the Tether warrants offset by a $29 million non-cash derivative loss on the convertible bond issued in August. IFRS net loss was $50.1 million versus $1.8 million. Adjusted profit was negative $26.2 million versus positive $10.5 million. Adjusted EBITDA was negative $8.5 million versus positive $28 million. This quarter's losses were primarily due to the year-over-year revenue declines in our cloud hash rate and hosting businesses, lower gross profit margins in our self-mining businesses as a result of halving and higher R&D as described previously.

Net cash used for operating activities was $90.7 million. Net cash generated from investing activities was $10.2 million, including $30.1 million of capital expenditure for infrastructure construction and mining machines, and the proceeds from disposal of cryptocurrencies of $39.9 million received from our principal business. Net cash generated from finance activities for the quarter was $168.1 million, primarily as a result of $166.3 million raised from our convertible notes issued in August.

Looking forward, we expect CapEx spend to accelerate for the continued build-out of our global power and data center infrastructure. Between Q4 2024 and 2025, we anticipate infrastructure CapEx to be in the range of approximately $250 million to $275 million. 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, we ended the quarter in a strong financial position with $291.3 million in cash and cash equivalents, $39.7 million in cryptocurrencies, and $92.7 million in borrowing, excluding derivatives.

Thank you, everyone. That concludes our prepared remarks section of our earnings call. Operator, please open the call for questions.

Question-and-Answer Session

Operator

[Operator Instructions] We will now take the first question from the line of Greg Lewis from BTIG. Please go ahead.

Greg Lewis

Yes, thank you and good morning and good afternoon everybody. And thanks for taking my question. Guys, thanks for the color around the TLM. I was hoping you could kind of talk a little bit more about that, the process, and kind of how that positions you to potentially look to secure some HPC partners over the next couple quarters?

Jihan Wu

Hey, Greg, thank you for that question. So yeah, the TLM engagement has more or less concluded as we said on the call and we have very positive reports about our US sites, especially three of them, the two Ohio sites and Rockdale. We really can't say a lot about what we're doing now. We do have a lot of ongoing discussions, even today and over the past few weeks. And so we're not going to be ready to announce something, I think, for a little bit and we're just talking with folks that will make good partners for us in proceeding in an area here where it will be new for us. So, we're really trying to find the right partnerships. And I don't know, Jeff, did you want to add to that?

Jeff LaBerge

Yeah, I would say, we've had a number of conversations, especially with potential partners. We are really looking, I think, focusing more on the structure of this partnership and how this should look from both a financing perspective and an operating perspective as well. We think there will be interest in these sites given the details that were uncovered during the TLM report. And ultimately, I think the financing and the actual structuring of it will be a main driver of returns overall in the long term. So we've been focusing a lot on that aspect.

Jihan Wu

We can say that there's been significant interest expressed from multiple parties.

Greg Lewis

Okay, great. And then just, I did want to talk about the rollout of the SEALMINER. I mean, there's a lot going on there, especially when you talk about the efficiency of the rigs, realizing that you have a lot of existing infrastructure right now that's mining for Bitcoin and if we assume that kind of remains to be the case, any kind of way you could let us think about the rollout of those more efficient rigs and kind of thinking about maybe the scale-up of global, of your hash rate as kind of these new rigs are rolled out and deployed?

Jihan Wu

Matt, go ahead.

Matt Kong

Yeah, I think now we have a lot of visibility in terms of doing self-mining and all the sale today of the main machines for the market. I think we are technically well positioned to do so. We have lots of powers secured, and also we have very good many machines now starting into the mass production. So going forward we have disclosed that we will have 18 exahash of hash power and this hash power will be allocated, I think, maybe for our self-mining and also part of it will go into the market to show our technology advancement. So going forward, I think we will have more along with the completion of our data center. So we will deploy more our A2 or even A3 mining machines to our data center to increase our self-hash rate.

Greg Lewis

Okay, great. And then just one more for me. Over the last couple quarters, kind of with the Bitcoin price where it was, it seemed like as a lot of the hosting contracts were rolling off, those kind of, like, that power infrastructure came back to Bitdeer and then you were able to kind of -- not kind of you were able to deploy more rigs on that infrastructure. Just given the real surge in Bitcoin over the last, I guess, couple months, has that kind of changed the dynamic around the Bitdeer’s ability to kind of reclaim some of that existing hosting infrastructure and convert it to self-mining?

Matt Kong

Yeah, okay, let me answer this question. After AI starts to join the competition for power in the data center area and the Bitcoin increased a lot in price of course. So the value we see in the power supply has increased a lot. So it is in our interest to have those hosting capacity back. But at the same time, we need to make sure that those clients who engaged with us earlier for hosting service have a trust and have kind of a sense of reliability on us. So we will not push too hard. So now we will just use those chances and opportunities when customers create it themselves and to reclaim back those hosting capacity.

Jeff LaBerge

And, Greg, I would add just to add to that, most of the general hosting customers were just contracts that were designed to roll off at this point right now and just were not renewed. So, like you said, that capacity will come back to us and we'll have it available to deploy for self-mining.

Greg Lewis

Okay, perfect. Super helpful. Thanks, guys.

Haris Basit

Thank you, Greg. Thank you.

Operator

Thank you. We will now take the next question from the line of Brett Knoblauch from Cantor Fitzgerald. Please go ahead.

Brett Knoblauch

Hi guys, thanks for taking my question. I wanted to dig in a bit into your CapEx commentary. I think $250 million to $275 million over the next, call it, a year. Can you maybe just outline what that is going into? Which sites and then what those sites are also going to be used for? Is it more hosting capacity? Is it going to be self-mining?

Jeff LaBerge

Yeah, I'll go ahead and take that, Fred. So yeah, the number that we quoted in the earnings release, that $250 million to $275 million, has primarily to do with our existing sites that we've announced under construction right now. So that would be the Tydal, Norway site, Texas hydro conversion, Jigmeling, Bhutan, and then it also is considering Massillon, Ohio for assuming that does go to the Bitcoin mining. And it also has part of the Clarington Phase 1 upgrades as well. So those are the numbers that are incorporated into that. And from a mining standpoint, again, those numbers assume that the sites will be used for mining. And again, in most cases, moving forward, we will emphasize self-mining over hosting capacity, but that's not to say we would not be open to additional hosting conversations depending on, again, market cycles, market conditions.

Brett Knoblauch

And when you say, kind of, for those sites, for Bitcoin mining, is that including all the electrical infrastructure as well as the powered shell? So the only thing in addition to that would be just machines being plugged in or is this just on the electrical infrastructure built-out of those sites and getting those sites energized from that perspective?

Jeff LaBerge

That would include the electrical infrastructure as well as the build-out. So essentially everything except the rigs.

Brett Knoblauch

Awesome, that's really helpful. And then just on the ASIC business, I saw kind of inventory had a big jump quarter-over-quarter on the balance sheet. I'm not sure if that was maybe due to the acquisition or just some working capital investments on your end in anticipation of those machines being mass produced. But could you just talk about the working capital dynamics needed for you guys to mass produce these chips? How much additional capital or inventory do you need or how should we think about the timing of that?

Jihan Wu

Yeah, okay. Working capital. Let me answer about that. TSMC has a very conservative approach into the crypto industry. So we need to pay that upfront when the wafer fabrication started. And we need another two months after the wafer comes back to us to mass produce the mining rigs. So for the ASIC mining rig business, roughly we need six months of working capital period for the cost of the mining rigs.

Brett Knoblauch

If you had to decide what that amount could potentially be for maybe SEAL01 and SEAL02, do you have a ballpark estimate for that? Is it too -- or too soon to tell?

Jeff LaBerge

The SEAL01, we are not making any more of those wafers. SEAL02, it's the 18 exahashes all that we've announced. And I think you recall what the capital requirements for that were. So, yeah, CapEx requirements for that would be on the order between $100 million, I could give a range, maybe $125 million to $175 million that we'd like to have.

Brett Knoblauch

Awesome. That's extremely helpful. And then Just to clarify one additional point, on the, I guess, pushback of hydro cooling conversion at Rockdale, I think you guys originally expected 3.4 exahash of your own rigs to be put there by year end. Is the reason why that's in pushback due to the delay there? Or am I thinking about that right?

Jihan Wu

Yes, it's mostly due to the delay in the mining farm constructions and the labor strike in the port.

Brett Knoblauch

All right, thank you guys, really appreciate it.

Haris Basit

Thank you, Brett.

Operator

Thank you. We will now take the next question from the line of Kevin Cassidy from Rosenblatt Securities. Please go ahead.

Kevin Cassidy

Yes, thanks for taking my question. Just for some clarification, the TLM had said you're suitable for Tier 3 data centers. Can you describe what the difference is between a Tier 3, say, and Tier 2, Tier 1 data center? Like, I guess what kind of customers would you be targeting?

Jeff LaBerge

So, we're looking, and the instructions we gave to TLM was really that is it suitable for a leading AI type of data center and for that application. So, we never really set specifically Tier 3, but I mean that's incorporated into that sort of description. And so there are requirements here that are far in excess of the requirements for a Bitcoin mining site, certainly, and even things like, are you next to a railroad line where hazardous materials are carried, what's the 100-year floodplain, and things like that. So many additional requirements that are really about protecting the very large investment that goes into a modern AI or high-performance computing data center. So we asked them to find every bit of air on the on the sites and so they actually went through a pretty extensive list and there are a number of small things that we have addressed or can be addressed with a relatively small amount of money or time but they didn't find anything significant that could prevent or really delay those sites from use as AI or HPC. I do want to stress though that no decision has been made whether or not we will use those sites for AI or Bitcoin mining.

Kevin Cassidy

I see, I see. And then maybe as you're making that decision, like, do you have an estimate of profitability of the HPC AI data center versus mining?

Jeff LaBerge

Can you repeat that, Kevin? You mean the probability of Bitcoin [Technical Difficulty] together?

Kevin Cassidy

Yeah, how much more would it be more profitable to co-host the HPC data center versus mining when you're looking at the decisions whether to convert any of these facilities.

Jeff LaBerge

I think there's a number of ways of looking at that and it really depends on how you model that Bitcoin mining site and what you assume hash price is going to be over the next few years. And so you can actually make the model come out either way. So one is more profitable or the other one is more profitable. So -- but there's a benefit in diversification of our revenue sources as well. And having a -- so we are thinking about it not just from the pure ROI, but also what impact it might have to our shareholders and the effect of having a reliable, fairly high margin revenue stream over the next 15 years and how that will affect our share price and our shareholders I think is also on our mind. So there's a balance of this to make diversification of our revenue stream part of this equation, not just pure ROI, especially since that's difficult to calculate for Bitcoin going forward.

Kevin Cassidy

Okay, great. Thank you.

Operator

Thank you. We will now take the next question from the line of Darren Aftahi from Roth. Please go ahead.

Darren Aftahi

Good morning. Thanks for taking my question. Two, if I may. First, as it pertains to your ASIC business and building the machines, the incoming administration, does that change anything in terms of just your position in the market, potential tariffs, just any kind of thoughts on that topic? Thanks.

Jihan Wu

Could you repeat that question again?

Jeff LaBerge

If the incoming administration changes, anything we're doing with regards to our ASIC manufacturing?

Jihan Wu

I don't think so. Yeah, I think we're going to proceed the way we have been and I don't see any significant impact.

Darren Aftahi

Great. And then on your site, as you potentially think about moving forward with Tier 3 data center, like two things, how critical is it that you need to kind of hire human capital for that expertise in-house? And then two, as you approach a model, is it something that you want to kind of own and operate or you want to do something where it sort of outsources somebody else that's a cost plus model? And then, you kind of call the day, just trying to understand how you're thinking about the economics and the ownership of it. Thanks.

Jeff LaBerge

I think we're open to a number of those possibilities and it depends a lot on who we end up partnering with on this and also who the end customer is. So I know that I'm not really answering your question because at this point we are open to those possibilities. We're having discussions with different groups that would result in different model for each of these. At some point, I do think we are going to have to add human capital to the company. But exactly what those human capital will need to bring depends a lot on which model we choose, of course. I don't really have a very precise answer for you right now because we're still open to that.

Darren Aftahi

Squeeze one more in. Just given the regulatory environment in different jurisdictions, like, what is your views on just which of these markets in the US between Texas and Ohio might be a more likely site for, not necessarily from the characteristics of the site, but more in terms of like what we get through regulatory approval faster between Texas and Ohio? Thanks.

Haris Basit

Yeah, look, I think we think both are great environments to do business in. Our Clarington, Ohio, for example, we were in a meeting several weeks ago with the -- basically the entire council there for the whole county there. So they're very pro-business there. Everybody we've talked to there has been very excited about this opportunity. So from a -- I guess just overall regulatory standpoint, both markets are very attractive. And, I don't see the regulatory side really being a major factor in the decision. I think it's going to be more the commercial factors.

Jeff LaBerge

We have good relationships with the local communities and governments in all of those areas.

Darren Aftahi

Thank you.

Operator

We will now take the next question from the line of Mike Colonnese from H.C. Wainwright & Company. Please go ahead.

Mike Colonnese

Hi, good morning, good afternoon everyone. Thank you for taking my questions today. First one for me, I appreciate all the details on the ASIC market opportunity being a multi-billion dollar one over the coming years. But I'm curious, what is your goal from a market share perspective as we look out over the next few years? And how do you plan to win share from large established players like Bitmain and MicroBT? If you could just speak to your ability to scale production capacity in that business over time.

Jihan Wu

I think we will secure the market share by our technology and our stability of the mining rigs. And we also will provide a very transparent data on how much mining rigs we are going to produce in the foreseeable future and what is the collective behavior of a customer combined together. So every customers will make the decisions based on rational number and projections into the future. And we cannot foresee market shares, I think, because it's quite a little bit unpredictable because what is the capacity of Samsung and TSMC will happen? We are not quite sure. Right now, the semiconductor of an advanced node is fully loaded. So the capacity allocation is quite uncertain right now. But we have this kind of confidence that we will meet our design in the chips and the overall stability and cost overall in the whole [manual] (ph).

Mike Colonnese

Got it. Thanks for the color there. And it sounds like the hosting business is still going through a bit of a transition period here over the last couple quarters. And it looks like you hosted about 9.5 exahash or so, give or take in October. Where can we expect that number to settle in as additional contracts really roll-off of the general hosting business?

Jeff LaBerge

So, hard to say. I would say we would not anticipate it growing significantly from where it is now. As we said, now that we are manufacturing our own ASICs, we will certainly favor self-mining over hosting. So, we do have some of that capacity that may fill back in, but I think it's safe to say the majority of our excess capacity in the future will be allocated to self-mining.

Mike Colonnese

Got it. Appreciate all the color, guys. Thank you.

Jeff LaBerge

Great. Thanks, Mike.

Operator

Thank you. We will now take the next question from the line of John Todaro from Needham & Company. Please go ahead.

John Todaro

Hey, great. Thanks for taking my question. I have two for you guys, one on the HPC opportunity and then second on the mining rig side. First, when you talk about development partners, could you just elaborate a bit more? Would this mean that someone actually comes in and shares the equity slice on those sites? And as part of that, there would be split economics? That type of development partner or more is it just on the construction side? And then after that, I'll ask my second question.

Haris Basit

So, yeah, it could mean that. And some of the folks we're speaking with would be in that category. But as I said before, we haven't actually chosen a business model or partner on this yet. So there's no way for me to exactly say which model will be there, but we are looking at models that include some kind of an equity split and that's part of the conversation.

Jeff LaBerge

Yeah, so John, that could look a couple ways. Obviously, we could bring in an infrastructure fund to just sort of financially back the project and then build that team through either outsourcing as well as internal team members, or we could go a partnership route where we would essentially partner with somebody who has those development capacity already. And so like Harris said, we've not settled down on a model yet. We're still exploring those.

John Todaro

Yeah, I understand. That's helpful, guys. And then second one, as we think about the mining rig side and just I guess taking a step back, how do you think about the ASIC cycle as it relates to the Bitcoin cycle and where would you kind of -- or when would you expect prices for rigs to expand, margin to expand as it relates to a Bitcoin cycle?

Jihan Wu

Well, usually when the Bitcoin price goes high, I can feel that the semiconductor is also in kind of a very bull cycle as well. So when we compare the supply of the mining rigs, most of the cases are likely to be highly limited. And the mining rig price will drive up by the hash price. We have seen quite a stable hash price around $40, $45 for quite a long time. But recently you have seen the hash price increase a little bit. Right now it's around $40 -- $55 per petahash per day, increase a little bit. But there are lots of older fleets hasn't been energized because of the low energy efficiency. But when the Bitcoin price goes up, maybe lots of generation fleets will energize. So we will need to see the Bitcoin price increase, maybe another 20% or 30% to see the hash price goes up more significantly. But the good news is that even if the Bitcoin price goes down, because of the older generation of the mining that are running now will shut down as well. So it's really unlikely to see the hash price goes down if the Bitcoin price goes down.

John Todaro

Got it. Understood. Thank you guys for that. Really appreciate it.

Operator

Thank you. We will now take the next question from the line of Lucas Pipes from B. Riley. Please go ahead.

Nick Giles

Thank you so much, operator. Good morning, everyone. This is Nick Giles on for Lucas. Guys, I was wondering just on TLM's assessment, are they still assessing sites outside of the US? And could you remind us on timing there, and could that assessment impact your expansion in Norway and how you go about that? Thank you very much.

Jeff LaBerge

So, yeah, TLM is not actually evaluating the Norway sites. We will likely engage another local company to do that. What we do know obviously is that the Norway sites do have -- we're going to have, once our expansion is done, just right around 300 total megawatts. Norway's actually very well connected from a latency standpoint. Very number of submerged sea cables that run all throughout the country there. So, Norway's done a very good job of connecting to the rest of the EU. Obviously, water's there, power's there, and all hydropower. So we would anticipate those being very attractive sites from an HPC standpoint. But like I said, we'll engage a local firm to do a more detailed analysis.

Nick Giles

Got it. Thanks. That's helpful. You have a number of expansion projects within your current footprint, obviously, but I was wondering if you could speak to any opportunities outside of that. Are you actively looking at any potential new sites, or have you put things on pause just given the amount of projects ongoing?

Jeff LaBerge

We're always looking. We get new sites and new opportunities sent to us all the time. We're taking a look at all of them. Obviously, we have a significant pipeline right now to develop very attractive assets. So Any new acquisition would need to be very strategic for us. Pricing would come into play as well. So we're always open and looking for new opportunities, but we don't feel we're in the position where we need to force anything new, given the extensive pipeline that we already have.

Nick Giles

Makes sense. Last one for me. How should we think about the split of the 18 exahash just between self-mining and external sales?

Jihan Wu

Right now, the priority is to do self-mining because we see the incoming years we will have lots of our mining facilities will accomplish construction. So later those facilities idle is not quite optimal. But at the same time, we need to plan for long term of our ASIC selling business. That's big. And we do see a kind of responsibility to distribute our money mix to the whole ecosystem, not to concentrate too much of the hash rate to make sure that the Bitcoin network is decentralized as well. And if the end-of-the-year customers will order a really big chunk from us, we will need to sell the mining rigs from right now so the customers will have a test use of our mining rigs. So right now we got an 18 exahash that is planned to be manufactured and we have a kind of 30,000 units of air cooled or air cooled equivalent of hydrocooling manually being capacity locked through our website in the last week. That's around one-third of the total hash rate we planned to manufacture at this point of time. So the majority will go to our self-mining and then we will allocate some of it to our customers. So, potentially they will have the confidence and the kind of mechanism to order more from us in the future.

Jeff LaBerge

And Nick, what I would just add to that is, again, one of the advantages of our business model being both a manufacturer and a sell finder ourselves is that it allows us to be very market-based in our decisions. So if the conditions for selling mining rigs appear to be more attractive, we would be able to obviously capitalize on that. If we were, saw this to be less attractive, we could obviously use those rigs internally. So it allows us to essentially have the highest and best use for every rig that we ever manufacture.

Nick Giles

Absolutely. Well, I appreciate all the color and continued best of luck.

Jeff LaBerge

Thank you.

Operator

Thank you. We will now take the next question from the line of Brian Kinstlinger from Alliance Global Partners. Please go ahead.

Brian Kinstlinger

Great. Thanks so much for taking my questions. The first question I have in terms of the A2 going into mass production, could you share what you think monthly production capabilities are? Maybe how we should think about those will also impact efficiency maybe over the next 12 months?

Jihan Wu

Okay, we have been transparent as much as possible. I mean, how much hash rate we are going to manufacture is public, 18 exahash at this point in time. If we have more plans, we got more signals or securement from the supply chain, we will update this number more. And we don't see any problems right now to mass production the SEALMINER A2. In terms of the high efficiency of our own mining operations, I think it's quite noticeable that our mining rig that is running on mining farm ourselves is quite older generations. It was like a four-year circles machine. That's because we concentrated our financial resources into building the mining facilities and develop our own mining chip, our own mining priority mining ASICs. And we did not buy quite much from other mining rig providers, except for from a few from [indiscernible]. But with our SEAL02 and SEAL03 expected, the mining hash rate power efficiency will increase a lot. Right now it's above the 35 to 40 joule per hash right now average on our mining operations, but it will be below 20, 16.5, or maybe even like around 11 or 12 in the next coming years. So the power efficiency of our hash rate will increase dramatically in the coming year.

Brian Kinstlinger

Great. My follow-up question is on the financials, quickly. With the SEALMINER03 tape-out, I think I saw on the press release in October, should we expect another one-time non-recurring R&D expense similar to what we saw in the third quarter?

Jihan Wu

Yes, doing chip design, [ethical design] (ph) is a large chip. We will discuss the very accurate R&D expenses number.

Brian Kinstlinger

Okay, thank you.

Operator

Thank you. We will now take the next question from the line of Mike Grondahl from Northland. Please go ahead.

Mike Grondahl

Hey guys, thanks. Two on the HPC side. How many people does Bitdeer have working on the HPC initiative? And as you've talked to financing partners, at a high level, what do some of those terms look like? And then maybe one on the ASIC side, you mentioned in the release that you're talking to a number of potential customers. How close are we to some orders there from those customers?

Haris Basit

I think I'll take that first question. So unfortunately, I'm not going to be able to answer you directly here. We have three people that are working on looking at opportunities for partnerships and deals on the HPC AI side. We don't have anybody we've specifically hired with that background or expertise. We're looking at the business development side of it first. But we are engaged with consultants and other folks that are assisting us in that. And your second question…

Jeff LaBerge

Just about the terms. I mean, nothing we can really disclose just yet. Like I said, I guess it would fall into a couple different buckets as far as the types of terms we're looking at. Obviously, one would just be potentially an infrastructure fund that would essentially come in back the project and either invest at project level equity or end or debt. The other would be a development partner that would come in with equity as well as some of the development expense capital available as well. But nothing specifically as far as terms we could disclose at this time.

Haris Basit

Mike, did you have an ASIC question also?

Mike Grondahl

Yeah, the press release just says you're talking to a number of potential ASIC customers. Just kind of curious how close you are to maybe getting your first order there or when we begin to see some of that.

Jihan Wu

Yeah, actually in the last week, we started a procedure called capacity lock-in, which means that the interested customers of the SEALMINER A2, they will need to submit the interest of the amount. And actually we got 253,000 of units of interest and only 30,000 of those are satisfied. This kind of ordering, capital locking ordering is from 64 customers. And every unit, they will need to pay around $200, $199, specifically. And so far, they have -- let me check -- so far, they have paid around 12,000 units of capacity locking fees, $199. So I think the order, the interest, the demand, the enthusiasm, the supportive emotion from our customer basis is quite rare.

Mike Grondahl

Sure. Yeah, that's helpful data. Thank you.

Operator

Thank you. We will now take the last question from the line of Bill Papanastasiou from Stifel. Please go ahead.

Bill Papanastasiou

Yeah, good morning gentlemen. Thanks for taking my questions. For the first one, I was just hoping you could speak to the feasibility of sites outside the United States for AI HPC and whether there are -- whether those sites are a priority relative to the US sites. Can you draw any comparisons with respect to AI HPC demand and key considerations with respect to financing and counterparty risk in economics? Thank you.

Jeff LaBerge

Yeah, I guess I can address the outside of the US. So as I mentioned before, we believe the Norway sites will be very attractive given the renewable energy component as well as the fiber connectivity, water, obviously the climate there will be ideal for cooling as well. From the people we've spoken with, EU is likely 12, maybe 18 months behind where the US is as far as adoption. But in our view, EU is even more power constrained as far as the ability to bring new generation online than we are. So we think there will be a significant opportunity in Europe, given some of the data privacy laws that they have there, they will favor EU domicile data centers. So one advantage it gives us is that it gives us essentially a longer tail on this opportunity. Right now, the window for this HPC opportunity in the US is maybe two years, maybe three years, depending on who you believe. But we think that tail extends out longer in Europe and possibly in Asia as well with our Bhutan site. So that's sort of how we're looking at that right now.

Bill Papanastasiou

Appreciate that. And then…

Jeff LaBerge

Second part of your question, Bill?

Bill Papanastasiou

Yeah. Was just, if you could just draw some comparisons with respect to how financing counterparty risk economics and so forth might differ between outside versus US.

Haris Basit

I think it will be different. We don't have enough information right now. I think in Europe it will be less different than of course in Bhutan. In some cases, the tenant may end up being the same large hyperscaler as in the US, but it's too early for us right now to really intelligently answer that question about how it might differ. We do expect it to be significantly different but we are focused on the US right now.

Bill Papanastasiou

Appreciate that. And then just for a second question, I know you guys touched on this on the call, but can you just provide an updated capital allocation methodology given the progress on SEALMINERs and the rise in Bitcoin mining economics? When you look at your over 1 gigawatt portfolio of power capacity, how are you thinking that mix of Bitcoin mining and HPC could change over the years and whether there's been any update to that? Thank you.

Jeff LaBerge

Yeah, so I would say it hasn't changed significantly. I think we're taking a very long-term view on this. Obviously, the thesis behind implementing some of the HPC into our power portfolio kind of stems just from looking to diversify and as Jihan mentioned earlier, not decentralizing the Bitcoin network as well. So having that mix of revenue, that stable revenue, a site on top of the Bitcoin mining revenue makes a lot of sense. So from that standpoint, I don't think it changes a ton, but obviously, we certainly welcome better Bitcoin mining economics right now and think that will help us out in the short term.

Bill Papanastasiou

Appreciate the color. Thanks guys.

Operator

Thank you. That concludes our question-and-answer. Thank you for joining Bitdeer’s Q3 2024 earnings conference call. Have a great day.