Q3 2024 Bitdeer Technologies Group Earnings Call
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 speaker's presentation, there will be a question and answer session. To ask a question during the session, you will need to press star 11 on your telephone. You will then hear an automated message advising your hand is raised.
To withdraw your question, please press star 1 and 1 again. Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Yujie Cai from the Orange Group. Please go ahead.
Speaker Change: Thank you, operator, and good morning, everyone. Welcome to BitGear's third quarter 2024 earnings conference call.
Speaker Change: Joining me today are Jihan Wu, Chairman and CEO, Matt Kong, Chief Business Officer, Harris Bassett, Chief Strategy Officer, and Jeff LaBerge, Head of Capital Markets and Strategic Initiatives.
Speaker Change: 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.
Speaker Change: Before management begins their formal remarks, we would like to remind everyone that during today's call, we may make certain forward-looking statements.
Speaker Change: 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.
Speaker Change: For a more complete discussion on forward-looking statements and the risks and uncertainties related to the Deere's business, please refer to its filings with the SEC.
Speaker Change: 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.
Speaker Change: 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 BidGear's investor relations website. Thank you. I will now turn the call over to Harris.
Thank you, Eugeo, and good day, everyone.
Harris Bassett: Welcome to our first earnings call since we became a public company last year.
Harris Bassett: Bittor 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.
Harris Bassett: Before diving in, I'd like to briefly highlight our Q3 financial results, and Jeff will provide more details in his section.
Harris Bassett: 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
Harris Bassett: lower hosting and cloud mining revenue, and higher R&D costs related to the one-off development expense of our Steel02 chip.
Harris Bassett: These negative impacts were partially offset by slightly higher average self-mining hashrate for the quarter and higher Bitcoin prices.
Harris Bassett: 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.
Harris Bassett: We believe this significantly differentiates our business from the rest of the sector in terms of our strategic positioning, revenue, and cost structure.
Harris Bassett: In parallel, we strategically expanded our energy infrastructure pipeline to be online when our mining rigs are also available.
Harris Bassett: This allows us to rapidly increase our self-mining hash rate using our own ASIC technology in the coming months and years.
Harris Bassett: 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.
Harris Bassett: As we look ahead, we anticipate the industry moving toward commoditization, which will require Bitcoin miners to diversify and differentiate their business models in order to remain competitive.
Harris Bassett: 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.
Harris Bassett: well as ensuring geographical diversification across our global operations and supply chain.
Harris Bassett: 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.
Harris Bassett: We have made significant strides in our product roadmap this year and are ramping up to commercialize our advanced sealed miner ASICs to disrupt what industry experts have estimated to be a $4 to $5 billion market annually over the next five years.
Harris Bassett: We see strong demand for diversified ASIC solutions and suppliers. Commercializing our steel miner ASICs will enable us to enter this growing market, diversify our revenue streams, and significantly accelerate the growth of our self-mining business.
Harris Bassett: 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.
Harris Bassett: We announced our R&D technology roadmap for our steel miner products in early June and have already energized the first batch of steel miner A1 mining rigs powered by the steel 01 chip.
These shifts are performing within our expectations.
Harris Bassett: We began mass production in October and have plans to install 3.7 ExaHash in phases between December of this year and Q1 2025.
Harris Bassett: This deployment will coincide with the energization of our Tidal Norway Phase I project, as well as our hydro-cooling lines at our Rockdale, Texas facility.
Harris Bassett: In October, we successfully incorporated our Steel-02 chip into our SteelMiner-A2 mining rigs, including both air-cooled and hydro-cooled models.
Harris Bassett: Testing of field miner 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.
Harris Bassett: We commenced mass production of our steel miner A2, and the first production run is expected to deliver 18 exa-hash per second, which will be used for self-mining and selling to external customers.
Harris Bassett: Notably, the commercial sale of SealMiner A2 will mark a significant milestone as we begin delivering these machines to customers starting in Q1, 2025.
Harris Bassett: We will continue to provide progress reports on our hash rate reduction as a commitment to maintaining transparency for all stakeholders.
Harris Bassett: including shareholders, suppliers, and especially clients who will need clear guidance to make their commercial decisions.
Harris Bassett: In addition, we taped out our third-generation chip with an industry-leading target efficiency of 10 joules per terahash and anticipate initial sample wafers in Q2 2025.
Harris Bassett: Looking forward, we remain fully committed to executing a successful market entry into the multi-billion dollar ASIC market in 2025.
Harris Bassett: 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.
Harris Bassett: We are excited to meet this demand and begin driving a differentiated and diversified revenue stream for our shareholders.
Harris Bassett: 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.
Harris Bassett: 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.
Harris Bassett: We believe this will significantly strengthen our competitive position and unlock substantial value for our shareholders.
Now, moving on to our AI cloud business.
Harris Bassett: 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.
Harris Bassett: We possess strong capabilities in deploying AI infrastructure and offering cloud platform services.
Harris Bassett: such as bare metal, virtual machines, and serverless GPU. These support computing across multiple regions, utilizing both our own and third-party assets.
Harris Bassett: We have also recently expanded our GPU cloud capacity into Canada.
Harris Bassett: Our cloud services business model and technology allows us to collaborate with partners to scale quickly in response to customer demand.
Harris Bassett: Demand has been steadily increasing for our GPU cloud services, and we are positioning Bitgear 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.
Harris Bassett: We are actively exploring opportunities to leverage our global power capacity of 2.5 GW to capitalize on the significant boom in demand for power for HPC and AI data centers.
Harris Bassett: 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.
Harris Bassett: TLM Group completed their feasibility assessment of our U.S. sites and confirmed the suitability of several of them for Tier 3 HPC and AI data centers.
Harris Bassett: 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.
Harris Bassett: 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.
Harris Bassett: 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.
Harris Bassett: 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.
Harris Bassett: This quarter, we incorporated more detailed updates into our monthly production reports to provide greater visibility into the status of our infrastructure expansion.
Harris Bassett: The 40-megawatt Phase 1 expansion at Tidal Norway remains on schedule to be energized in December 2024.
Harris Bassett: In Rockdale, Texas, the 100-megawatt hydrocooling conversion has been delayed by one month due to supply chain impacts from the U.S. stockworkers' strike in early October.
Harris Bassett: Phase 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.
Harris Bassett: With these projects and other projects, such as Phase II of Tidal Norway and Phase I 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.
Harris Bassett: We remain on track to deliver our steel miner A2-A6, substantially grow our self-mining fleet at a competitive cost advantage, and leverage our industry-leading global power portfolio.
Speaker Change: 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, Harris. Before I go over Bitgear's third quarter financial results, I'd like to remind everyone that all figures I refer to today are in U.S. dollars and that all comparisons are to Q3 of last year.
Q3 consolidated revenue was $62 million versus $87.3 million.
Jeff Laberge: Self-mining revenue was $31.5 million, up 4.7%, primarily due to an increase in our average self-mining hash rate for the quarter to 7.8 exa-hash from 6.1 exa-hash and higher year-over-year Bitcoin prices.
Jeff Laberge: These positive factors were largely offset by the April 2024 halving.
and higher global network hash rate.
Jeff Laberge: Cloud hashrate revenue was $7.1 million versus $15.6 million. This decline was primarily due to long-term cloud hashrate 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.
Jeff Laberge: The decrease in HOCSI revenue was mainly caused by two aspects.
Jeff Laberge: 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 steel miner hydro-cooled mining rigs by the first quarter of 2025.
Jeff Laberge: Second, some hosting customers removed less efficient miners after the halving in April 2024.
Jeff Laberge: This extra capacity is currently being replenished by new hosted mining rigs.
Jeff Laberge: Total gross profit per quarter was $2.8 million versus $21.1 million and gross margin was 4.5% versus 24.2%.
Jeff Laberge: 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.
Jeff Laberge: Total operating expenses for the quarter were $42.9 million versus $27.3 million.
Jeff Laberge: The increase is primarily driven by higher R&D costs, including a $13.4 million one-off incremental development expense related to the CLO2 chip, higher stock-based compensation, and non-cash amortization expenses of intangible assets related to the acquisition of pre-chains.
Jeff Laberge: Other net gain loss for the quarter was a net loss of $14.7 million versus a net gain of $0.9 million.
Jeff Laberge: 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.
Jeff Laberge: 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 the halving, and higher R&D costs.
as described previously.
Jeff Laberge: Net cash generated from investing activities was $10.2 million, including $30.1 million of capital expenditure for infrastructure construction and mining machines.
Jeff Laberge: and the proceeds from disposal of cryptocurrencies of $39.9 million received from our principal business.
Jeff Laberge: 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.
Jeff Laberge: Looking forward, we expect CapEx spend to accelerate for the continued buildup of our global power and data center infrastructure.
Jeff Laberge: Between Q4 2024 and 2025, we anticipate infrastructure CapEx to be in the range of approximately $250 to $275 million.
Jeff Laberge: It should be noted that this infrastructure spend assumes the sites are developed for Bitcoin mining and does not include CapEx for steel miners used for self-mining.
Jeff Laberge: 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.
Speaker Change: Thank you, everyone. That concludes the prepared remarks section of our earnings call. Operator, please open the call for questions.
Thank you.
Speaker Change: Thank you. As a reminder to ask a question please press star 1 1 on your telephone and wait for your name to be announced.
Um
Speaker Change: To withdraw your question please press star 1 and 1 again.
We will now take the first question.
Thank you.
Speaker Change: from the line of Greg Lewis from BTIG. Please go ahead.
Yes, thank you and good morning and good afternoon, everybody.
Speaker Change: 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.
Hey, Greg. Thank you for that question.
Speaker Change: So yeah, the TLM engagement has more or less concluded, as we said on the call, and we have very positive reports about our U.S. sites, especially three of them, the two Ohio sites and Rockdale.
Speaker Change: 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 are.
Speaker Change: We'll make good partners for us and proceeding in an area here where it will be new for us So we're really trying to find the right partnerships
Speaker Change: 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.
Jeff Laberge: 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 offering perspective as well. You know, we think there will be interest in these sites.
given the details that were uncovered during the TLM.
Jeff Laberge: 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. We can say that there's been significant interest expressed from multiple parties.
Speaker Change: Okay, great. And then just I did want to talk about the rollout of the steel miner. I mean, I mean, there's a lot going on there, you know, especially when you talk about the efficiency of.
You know the rigs
Speaker Change: 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
Speaker Change: You know maybe the scale up of global of your hash rate as kind of these new rigs are rolled out and deployed
Matt, go ahead.
Remarkable. I think we are technically
We are all permissioned to do so.
Speaker Change: We have lots of our powers secured and also we have
Very good many machines now
Speaker Change: So going forward, we have disclosed that we will have 18 exa-hash 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.
Speaker Change: 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 the efficiency.
to increase our capacity.
Speaker Change: Okay, great. And then just one, just one more for me.
Speaker Change: You know, over the last kind of couple quarters, kind of with the Bitcoin price where it was, it seemed like a lot of the hosting contracts were rolling off, those kind of like...
Speaker Change: that power infrastructure came back to Bitdeer, and then you were able to kind of, you know, not kind of, you were able to deploy more rigs on that infrastructure, you know, just given the, you know, the real surge in Bitcoin over the last, I guess, you know, couple months, has that kind of changed the dynamic around Bitdeer's ability to kind of reclaim some of that existing hosting
infrastructure and convert it to self-mining.
Yeah, okay, maybe I'll have to add this question.
Speaker Change: After AI starts to join the competition for power in the data center area,
Speaker Change: And the Bitcoin increased a lot in price, of course. So the value we see in the power supply has increased a lot.
Speaker Change: 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.
Speaker Change: have kind of a sense of reliability to us, so we will not push too hard. So now we will just use those chances and opportunities when customers create themselves and to reclaim back those hosting capacities.
Thank you.
Speaker Change: Greg, I would just add 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.
Okay, perfect.
Super helpful. Thanks, guys.
Thank you, Greg. Thank you. Thank you.
Speaker Change: From the line of Brett LaBlauche from Cantor Fitzgerald, please go ahead.
Bye.
Speaker Change: 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 –
Speaker Change: 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?
Speaker Change: Yeah, I'll go ahead and take that, Brett. So yeah, the number that we quoted in the earnings release, that $250 to $275 million
Speaker Change: has primarily to do with our existing sites that we've announced under construction right now.
Speaker Change: So that would be the Tito Norway site, Texas hydro conversion.
Speaker Change: Jigmeling Bhutan, and then it also is considering Maslin, Ohio for assuming that does go to the Bitcoin mining And it also has part of the Clarington phase one Upgrades as well. So those are the numbers that that are incorporated into that
Speaker Change: 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.
Speaker Change: And when you say kind of for this site, you know, for Bitcoin mining, is that including all the electrical infrastructure as well as the powered shell?
Speaker Change: So, the only thing in addition to that would be just machines being plugged in, or is this?
Speaker Change: just on the electrical infrastructure built out of those sites and getting those sites energized from that perspective.
Speaker Change: That would include the electrical infrastructure as well as the build-out, so essentially everything except the rigs.
Awesome. That's really helpful.
Speaker Change: And then just on the ASIC business, I saw kind of inventory at 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. Could you just talk about the working capital dynamics needed for you guys to mass-produce these chips?
Speaker Change: You know, how much additional capital or inventory do you need, or how should we think about the timing of that?
Yeah, okay, working capital. Let me answer about that.
PSNC
Speaker Change: has a very conservative approach into the crypto industry, so we need to pay food up front.
Speaker Change: 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...
a working capital period for the cost of the maintenance.
Speaker Change: And if you had this size, what that amount could potentially be for maybe CL01 and CL02, do you have a ballpark estimate for that, or is it too soon to tell?
Speaker Change: So CL01, we are not making any more of those wafers.
Speaker Change: So CL02, it's the 18 extra hashes, all that we've announced, and I think you recall what the capital requirements for that were. So yeah, capital requirements for that would be on the order of between 100, I could give a range, maybe 125 to $175 million that we'd like to have.
Speaker Change: Awesome. That's extremely helpful. And then just to clarify one additional point, on the, I guess, pushback of hydro cooling conversion at Rockdale,
Speaker Change: I think you guys originally expected 3.4x a hash of your own rigs to be or the CL1 rigs to kind of be put there By year end is the reason why that's been pushed back due to the delay there or am I thinking about that right?
Speaker Change: Yes, yes, it's mostly due to the delay in the manifold constructions and the labors strike in the port.
All right. Thank you, guys. Really appreciate it.
Thank you. Thank you.
We will now take the next question.
Speaker Change: From the line of Kevin Cassidy from Rosenblatt Securities, please go ahead.
Speaker Change: 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 and Tier 2, Tier 1 data center? I guess what kind of customers would you be targeting?
Thank you.
Speaker Change: 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 said specifically Tier 3, but I mean that's incorporated into that.
Speaker Change: into that sort of description. And so there are requirements here that are far in excess of, you know, the requirements for a,
Speaker Change: Like a Bitcoin mining site certainly and you know, even things like are you next to a railroad line where hazardous materials are carried? You know, what's the?
Speaker Change: a hundred 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.
Speaker Change: So we asked them to find every bit of hair on the on the sites and so they actually went through a pretty extensive list and, you know, 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 would
Speaker Change: 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 for Bitcoin mining.
Speaker Change: I see, I see. And then maybe as you're making that decision.
Speaker Change: Do you have an estimate of profitability of the HPCAI data center versus mining?
Can you repeat that, Kevin?
Speaker Change: You mean the probability of it going one way or the other?
Speaker Change: Yeah, how much more would it be more profitable to co-host a HPC data center versus mining when you're looking at the decisions whether to convert any of these facilities?
Speaker Change: I think there's a number of ways of looking at that, and it really depends on how you model the Bitcoin mining site and what you assume hash price is going to be over the next few years.
Speaker Change: And so, you can actually make the model come out either way, so one is more profitable or the other one is more profitable.
Speaker Change: So but there's a benefit in diversification of our revenue sources as well and having
Speaker Change: So, you know, 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
Speaker Change: you know, a reliable, fairly high-margin revenue stream over the next, you know, 15 years and how that will affect
our share price and our shareholders, I think is.
Speaker Change: 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.
Okay, great. Thank you.
Thank you.
We will now take the next question.
Speaker Change: From the line of Darren, after he from Ross, please go ahead.
Speaker Change: 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.
Thank you.
Speaker Change: Could you repeat that question again? If the incoming administration changes anything we're doing with regards to our ASIC manufacturing?
I don't think so.
Yeah, I think...
Speaker Change: Yeah, I think we're going to proceed the way we have been, and I don't see any significant impact.
Great, and then on the
Speaker Change: your sites as you potentially think about you know moving forward with
Tier three
data center
Speaker Change: 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? It's a cost-plus model and then, you know, you kind of call it a day. Just trying to understand how you're thinking about the economics and the ownership of it. Thanks.
Speaker Change: 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.
Speaker Change: and also who the end customer is. So I know that I'm not really answering your question because.
Speaker Change: At this point we are open to those possibilities. We're having discussions with different groups that
would result in.
Speaker Change: a 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.
Speaker Change: I don't really have a very precise answer for you right now because we're still
open to that.
Speaker Change: Squeeze one more in, just given the regulatory environment and 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, you know, site for not necessarily from the, the.
Speaker Change: characteristics of this site but more in terms of like what would get through regulatory approval faster between Texas and Ohio. Thanks.
Speaker Change: Yeah, look, I think we think both are great environments to do business in. You know, our clearance in Ohio, for example, you know, we were in a meeting several weeks ago with the basically entire.
Bye.
Speaker Change: Council there for the for the for the whole county there. So they're very pro-business there Everybody we've talked to there has been very excited about the this opportunity. So from a
Speaker Change: 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 of the commercial factors. We have good relationships with the local communities and governments in all of those areas.
Thank you.
Thank you. We will now take the next question.
Speaker Change: From the line of Mike Colonese from HV Wainwright and Company, please go ahead.
Speaker Change: 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, and if you could just speak to your ability to scale production capacity in that business over time.
Bye-bye.
I think we will secure the market share by
our technology and our
Bye.
our stability of the mining rigs.
And we also will provide a very transparent.
Speaker Change: data on how much money we are going to produce in the foreseeable future and what is the collective behavior of a customer combined together so every customer will make a difference.
based on rational number and projections into the future.
Um
Speaker Change: And we cannot do a thing, the market shares, I think, because it's...
It's a quite a bit
Thank you.
predictable because
Speaker Change: But what is the capacity of Samsung or TSMC, what will happen? We are not quite sure. Right now the SIM connector...
of an avalanche load is fully loaded.
overall stability and
Bye.
Kost?
overall.
in their home anyway.
Speaker Change: Got it. Thanks for the color there. It sounds like the hosting business is still going through a bit of a transition period here over the last couple of quarters. It looks like you hosted about nine and a half X a hash 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?
Speaker Change: 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.
Got it. Appreciate it. All the better, guys. Thank you.
Great. Thanks, Mike.
Thank you. We will now take the next question.
Speaker Change: from the line of John Todaro from Needham and Company. Please go ahead.
Speaker Change: Eddie, 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.
Speaker Change: 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?
Speaker Change: 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.
Speaker Change: 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, you know, we haven't actually chosen a business model or a partner on this yet. So You know, there's no Way for me to exactly say which model will be there, but there you know we are looking at models that include some kind of an equity split and
That's part of the conversation.
Speaker Change: Yeah, so John, that could look a couple of ways. Obviously, we could bring in, you know, an infrastructure fund to just sort of back financially back the project and then, you know, build that team through either outsourcing as well as, you know, internal team members, or we could go a partnership route where we would, you know, essentially partner with.
Speaker Change: somebody who has those development capacity already. And so like Eric said, we've not settled on a model yet. We're still exploring those.
Speaker Change: Got it. Understood. That's helpful, guys. And then second one, you know, as we think about the.
Speaker Change: 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?
Bye-bye.
Well, um...
Uh, Unity wouldn't have Bitcoin.
Price goes high.
All right.
I can feel that
This is an actor
Speaker Change: It's also in kind of a very low cycle as well.
Speaker Change: So, when we go and buy the supply of the maniwakes,
Speaker Change: Most of the cases, it's quite likely to be highly limited.
And the money rate price
Speaker Change: You only will drive up by the hash price. We have a stable hash price around $40-$45 for quite a long time.
Speaker Change: But recently we have seen the hash price increase a little bit. Right now it's around $55 per peta.
Have a good day.
increase a little bit, but there are lots of out-of-flip.
I'm going to be
uh, the energized because of the low
Speaker Change: But when the Bitcoin price goes up, maybe lots of older generation fleets will energize. So we will need to see the Bitcoin price increase.
Maybe another 20 or 30 percent.
To see the hash price goes up more significantly
Speaker Change: But the good news is that even if the Bitcoin price goes down
Because of
the older generation of a Manifred.
that Ronnie now will shut down as well. So.
Speaker Change: It's really unlikely to see the hash price go down if the Bitcoin price goes down.
Bye.
Speaker Change: Got it. Understood. Thank you guys for that. Really appreciate it.
Thank you. We will now take the next question.
Speaker Change: From the line of Lucas Pipes from B Reilly, please go ahead.
Speaker Change: Thank you so much, operator. Good morning, everyone. This is Nick Giles on for Lucas.
Speaker Change: Guys, I was wondering just on TLM's assessment, are they still assessing sites outside of the U.S., 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.
Speaker Change: 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 –
Speaker Change: We were going to have, once our expansion is done, just right around 300 total megawatts. Norway is actually very well connected from a latency standpoint.
Speaker Change: A 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 HBC standpoint.
Speaker Change: But like I said, we will engage a local firm to do a more detailed analysis.
Got it. Thanks. That's helpful.
Speaker Change: 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?
We're always looking, you know, we.
Speaker Change: 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 a very attractive assets. Any new acquisition would need to be very strategic for us. Pricing would come into play as well. We're always open and looking for new opportunities, but we don't feel we're in a position where we need to force anything new given the extensive pipeline that we already have.
Makes sense. Lots of them for me.
Speaker Change: How should we think about the split of the 18x hash just between self-mining and external sales?
Speaker Change: Right now, the priority is to do self-money because we see
Speaker Change: The incoming years, we will have a lot of our many facilities will be open to the public.
Speaker Change: accomplish construction, so let those facilities idle is not quite optimal. But at the same time, we need to plan for long-term about basic selling business. That's big, and we do see a kind of responsibility to distribute our money.
Speaker Change: to the whole ecosystem, not to concentrate too much of the hashrate, to make sure that the Bitcoin network is decentralized as well. And if in the long term the customer...
We are there.
Speaker Change: really big chunk from us, we will need to sell the MiningWix from right now so the customers will have a test use of our MiningWix.
So right now, we've got an 18x hash.
Thank you, sir.
Planned it to be
manufactured and we have a kind of 30,000 units.
of air-cooled or air-cooled equivalent of
Hydrocooling
Manu Raikopin.
capacity locked through our website.
in the last week.
That's around one-third of the total hash rate.
Speaker Change: we planned to manufacture at this point in time. So the majority will go to our self money and 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.
Speaker Change: And Nick, what I would just add to that is, you know, again, one of the advantages of our business model being both a manufacturer and a self-miner 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.
Speaker Change: If we saw this to be less attractive, we could obviously use those rigs internally. So it'll allow us to essentially have the highest and best use for every rig that we ever manufacture.
Speaker Change: Absolutely. Well, I appreciate all the color and continue best of luck.
Thank you.
Thank you. We will now take the next question.
Speaker Change: From the line of Brian Kingslinger from Allianz Global Partners, please go ahead.
Speaker Change: 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, you know, those will also impact efficiency maybe over the next 12 months?
Thank you.
Okay.
we will we will have been
transparent as much as possible.
Speaker Change: I mean, how much hashrate we're going to manufacture, it's public, 18 extra hash, at this point in time. If we have more plans, we've got more signals or securement from the...
Speaker Change: supply chain, we will update this number more. And we don't see any problems right now to mass production of CO2. In terms of the power efficiency of our own
mining operations.
I think it's quite noticeable that our money rake
Speaker Change: that is running our mining farm ourselves is quite an older generation. It was like a four year ago 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 index, and we
did not buy quite much from other
Speaker Change: many providers, except a few from Markzware Manager, but with our CO2 and CO3 expected,
Speaker Change: the mining hashrate power efficiency will increase a lot. Right now it's...
It's above the 35 to 40 droplet hash right now
Speaker Change: But it will be below 2016.5 or maybe even like around 11 or 12 in the next coming years. So the power efficiency of our hashrate will increase dramatically in the coming years.
Great. My follow-up question is,
Speaker Change: on the financials quickly. With the CO-minor 3 tape out, I think I saw in 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?
Yes, doing chip design is a large chip.
We will discuss the very accurate on the expenses number.
Okay. Thank you.
Thank you.
Thank you. We will now take the next question.
Speaker Change: From the line of Mike Crandall from Northland, please go ahead.
Speaker Change: 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,
Speaker Change: At a high level, what do some of those terms look like?
and then maybe one on the ASIC side.
Speaker Change: You know you talk you mentioned in the release that you're talking to a number of potential customers
Speaker Change: How close are we to some orders there from those customers?
Speaker Change: 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 HPT-AI side. We don't have anybody we've specifically hired with that background or expertise.
Speaker Change: We're looking at the business development side of it first. We are engaged with consultants and other folks that are assisting us in that.
And your second question
Thank you.
Speaker Change: Just about about the terms. I mean, nothing we can really disclose just yet. You know, like I said, we're there's 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
Speaker Change: You know, potentially an infrastructure fund that would essentially come in and back the project and, you know, either invest at project level equity and 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.
Speaker Change: But nothing specifically as far as terms we could disclose at this time. Mike, did you have an ACIC question also?
Speaker Change: 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, you know, first order there or when we begin to see some of that.
Yeah, actually, in the last week, we...
Speaker Change: started a procedure called a capacity lock-in, which means that the interested customers of the CO minor, A2, they will need to submit the interest of a,
amount. And actually we got
$253,000.
of
30,000 of those is satisfied with this kind of ordering.
Our capital lock-in ordering is from 64 customers.
Speaker Change: And every unit, they will need to pay around $200, $199 specifically. And so far, they have...
Let me check so far they have
Speaker Change: They have paid around 12,000 units of capacity locking fees, $199. So I think the order, the interest, the demand, the supportive emotion from our customer basis is quite rare.
Sure. Yeah, that's helpful data. Thank you.
Bye.
Thank you. We will now take the last question.
Speaker Change: From the line of Bill Papanastosiu from Stifel, please go ahead.
Speaker Change: 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 of the United States for AIHPC and whether those sites are a priority relative to the U.S. sites. Can you draw any
Speaker Change: comparisons with respect to AIHPC demands and key considerations with respect to financing counterparty risk and economics. Thank you.
Speaker Change: Yeah, I guess I can address the outside of the U.S. 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, you know, water, you know, all of these things.
Speaker Change: up to the climate there will be ideal for gluing as well.
Speaker Change: From the people we've spoken with, EU is likely 12, maybe 18 months behind where the U.S. is as far as adoption. But in our view, EU is even more power-constrained as far as adoption.
Speaker Change: ability to bring new generation online than we are. So we think there'll be a significant opportunity in Europe, given some of the data privacy laws that they have there, that will favor EU domiciled data centers.
Speaker Change: So, you know, one advantage to give this is that, you know, it gives us essentially a longer tail on this opportunity, you know, right now the window for this HPC opportunity in the U.S.
Speaker Change: you know, 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.
Appreciate that.
What was the second part of your question, Bill?
Speaker Change: how financing, counterparty risk, economics and so forth might differ between outside versus U.S.
Speaker Change: 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, you know, the
Speaker Change: tenant may end up being the same large hyperscaler as in the U.S.
Speaker Change: 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 U.S. right now.
Speaker Change: Appreciate that. And then just for a second question, I know you guys have touched on this on the call, but can you just provide an updated capital allocation methodology, given the progress on the seal miners and the rise in Bitcoin mining economics?
when you look at your...
over 1 gigawatt portfolio of power capacity.
Speaker Change: How are you thinking that mix of Bitcoin mining and HPC could could change over the years and Whether there's been any update to that. Thank you
Speaker Change: Yeah. So I would say it hasn't changed significantly. I mean, I think we're taking a very long-term view on this. You know, obviously, the thesis behind implementing some of the HPC into our power portfolio kind of stems just from looking to diversify. And, you know, as Jian mentioned earlier, you know, not
Speaker Change: you know, decentralizing the Bitcoin network as well. So, you know, having that that mix of revenue, that stable revenue aside on top of the Bitcoin mining.
Speaker Change: 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'll help us out in the short term.
I appreciate the call. Thanks, guys.
Speaker Change: Thank you. That concludes our question and answer. Thank you for joining the DSQ3 2024 earnings conference call. Have a great day.
Thank you. Thank you.