Core Scientific Inc

NASDAQ:CORZ

Earnings Call Analysis

Earnings Call Transcript 2024-Q1

[Operator]

Good afternoon, ladies and gentlemen. Thank you for joining today's Core Scientific's First Quarter Fiscal Year 2024 Earnings Conference Call. My name is Tia, and I will be your moderator for today's call. [Operator Instructions]I would now like to pass the call over to your host, Steve Gitlin. Please proceed.

[Steven Gitlin, executive]

Good afternoon, ladies and gentlemen, and welcome to Core Scientific's First Quarter Fiscal Year 2024 Earnings Call. This is Steven Gitlin, Senior Vice President of Investor Relations for Core Scientific. At this time, all participants are in a listen-only mode. We will conduct a question-and-answer session after management's remarks. As a reminder, this conference is being recorded for replay purposes.Before we begin, please note that on this call certain information presented contains forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. Forward-looking statements include, without limitation, any statements other than historical or current facts that predict or indicate future events or trends, forecasts, performance, or achievements and may contain words such as believe, anticipate, expect, estimate, intend, project, plan, or words or phrases of similar meaning. Forward-looking statements are based on current expectations, forecasts and assumptions that involve risks and uncertainties that may cause actual results to differ materially.For further information on these risks and uncertainties, we encourage you to review the risk factors discussed in the company's annual report on Form 10-K filed with the Securities and Exchange Commission and the special note regarding forward-looking statements contained in the company's current report on Form 8-K filed today and the earnings release and slide presentation contained therein. Today's presentation is available on our website at corescientific.com in the Events and Presentations section. The content of this conference call contains information that is accurate only as of today, May 8, 2024. The company undertakes no obligation to update statements made today to reflect events or circumstances occurring after today.Joining me today from Core Scientific are Chief Executive Officer, Mr. Adam Sullivan; and Chief Financial Officer, Mrs. Denise Sterling.We will now begin with remarks from Adam Sullivan. Adam?

[Adam Sullivan, executive]

Thanks, Steve. I'll start today's call with a high-level summary of our positioning as we enter 2024 and some highlights of our exceptional first quarter performance. I will then hand the call over to Denise Sterling to review our first quarter financials. After Denise's remarks, I'll take some time to talk about the current industry environment and our strategy to drive continued growth and value creation for 2024 and beyond. We will then take your questions.Core Scientific is a market leader positioned for growth. Slide 3 summarizes the position of strength from which we operate today, highlighted by the following key points: First, we operate the largest owned bitcoin mining infrastructure in the industry in terms of operating megawatts, comprising approximately of 745 megawatts of operational power and contracts for a total of up to 1.2 gigawatts of power.Next, we own and control every structure, every transformer in every concrete pad in our 7 mining data centers. Finally, we have the experience, track record, and team to monetize our infrastructure for the highest value uses into secured additional infrastructure opportunistically. We started our business by identifying high-power sites with attractive power rate that could support emerging high-value compute applications.We focus on designing and building efficient low-cost proprietary infrastructure for bitcoin mining operations that offered attractive hosting opportunities for third parties. When the price of bitcoin increased, we use our expertise to mine for our own account. We invested in mining equipment and expand the geographic footprint of our infrastructure, increasing our revenue and the ROI of our original infrastructure investment.You can see our current infrastructure footprint on Slide 4. Our industry-leading infrastructure has allowed us to produce more bitcoin than any other public company for the last 3 years through our self-mining business shown on Slide 5. We now believe our infrastructure is well positioned to take advantage of the enormous demand for power and infrastructure required for high-performance compute, and we see that as the next major growth opportunity for our business.With the demand for ready high-power sites increasing rapidly, our infrastructure can be repurposed to provide access to HPC with healthy development, planning, regulation, construction, permitting, and supply chain time lines associated with greenfield HPC types.According to Bank of America Research, power demand from data centers is expected to double in the next 3 to 5 years. With this in mind, we would like to bring today's conversation around a simple central theme. Owning and controlling all our valuable high-power data center infrastructure gives us a significant advantage at a time when the demand for such infrastructure exceeds the available supply.Our high-power data center infrastructure places us in a uniquely valuable position where we can balance our portfolio between bitcoin mining and alternative compute hosting to maximize cash flow, minimize risk, and maintain significant exposure to bitcoin's upside potential.We can offer clients a shorter time to power as compared to them waiting potentially 3 to 5 years for new greenfield data center capacity to come online. We see this as a powerful mix that provides the potential for multiyear high visibility cash flows to buffer against the inherent volatility of bitcoin pricing.And because we own and control our infrastructure, we can optimize for the allocation of our infrastructure portfolio, investing in bitcoin mining position us well, and we now have the opportunity to maximize the value of these assets.Moving forward, we will continue to seek out low-cost abundant power for bitcoin mining as our entry point and will constantly evaluate the market for a way to pair that power with another higher value use case. With Bitcoin mining as our base business, our infrastructure becomes the platform upon which we will continue to grow and optimize.We have created a unique business opportunity for us and for you, our shareholders, to monetize our own infrastructure for both bitcoin mining and HPC hosting. I will discuss this further after Denise's comments.So now let's review our strong first quarter results summarized on Slide 6. We entered 2024 with strong momentum from 2023, continuing to set the pace for our industry by earning 2,825 bitcoin in the first quarter, more than any other listed miner. Our leading bitcoin production generated $150 million in revenue, plus $29 million from our hosting business for total revenue of $179 million, up 49% year-over-year.Nearly all our key financial metrics reflect strong performance in the quarter. Our gross margin was 43%. Operating margin was 31%. Net income was $211 million and adjusted EBITDA was $88 million, up 118% year-over-year. We exited the quarter with healthy liquidity, consisting of $98 million in cash and cash equivalents and $16 million in restricted cash.Shortly after the end of the quarter, we deployed capital to pay down $19 million in debt associated with outstanding Mechanic's Liens and fund a $1 million project at our Denton data center to add 72 megawatts of infrastructure.Throughout the first quarter, we continued to deliver strong hashrate utilization, which remains higher than the average for our peer group and for scaled miners illustrated on Slide 7. We also continue to refresh our self-mining fleet with new S21, completing the deployment of 2.5 exahash in April and improving our average miner efficiency to 25.78 joules per terahash. We are waiting to make countercyclical miner purchases to take advantage of improved pricing after the recent halving. We are already seeing that dynamic take shape with post halving pricing lower than pre halving.In March, we entered into a contract for high-performance compute hosting at our new Austin data center, which [indiscernible]. Importantly, we delivered a 16-megawatt data center to our client, CoreWeave more than 30 days ahead of schedule, helping them accelerate their time to power, which refers to how long it takes to establish operations and service their clients.Upgrading this data center was no small task and required a team effort to completely reconfigure 118,000 square feet of compute space, including pulling 18 miles of fiber and 500 miles of copper cable, removing 1,500 racks and installing 4,500 new PDU strips. Our team's performance was nothing short of spectacular.As we look to the remainder of 2024, we are confident. Our outstanding first quarter has positioned us well to continue building on our momentum and capitalizing on the significant growth opportunities we see ahead.Now I'll turn the call over to our CFO, Denise Sterling.

[Denise Sterling, executive]

Thank you, Adam. As Adam stated, our first quarter performance was strong across all financial metrics, driven by favorable fundamentals and outstanding execution. Total revenue for the fiscal first quarter of 2024 was $179.3 million and consisted of $150 million in digital asset mining revenue and $29.3 million in hosting revenue.Segment economics are highlighted on Slide 10. Digital asset mining revenue of $150 million for the fiscal first quarter of 2024 exceeded mining cost of revenue of $81.6 million by $68.4 million, representing a gross margin of 46%. Digital asset mining revenue of $98 million for the same period in the prior year, exceeded mining cost of revenue of $72.7 million by $25.4 million, resulting in a gross margin of 26%.Gross profit increased by $43 million, demonstrating a significant improvement quarter-over-quarter. The quarter-over-quarter increase in digital asset mining revenue of $51.9 million was driven primarily by a 134% increase in the price of bitcoin and a 20% increase in our self-mining hashrate driven by the deployment of an additional 18,000 new generation mining units.The increase in digital asset mining cost of revenue of $8.9 million for the fiscal first quarter of 2024 was primarily driven by an increase in depreciation expense resulting from the deployment of our new self-mining unit and an increase in payroll and benefits costs associated with merit and market adjustments made during the quarter.Power costs were relatively flat quarter-over-quarter as the increase in power consumption associated with the deployment of the additional self-mining units was offset by a 3.8% decrease in our power cost per kilowatt hour, which declined to $0.043 from $0.044 per kilowatt hour for the same period in the prior year.As a reminder, digital asset mining cost of revenue consists primarily of direct production cost of mining operations. These direct production costs consist of electricity and data center operating costs, which includes salaries, stock-based compensation, and depreciation of property, plant, and equipment.Hosting revenue of $29.3 million exceeded hosting cost of revenue of $20.1 million for the fiscal first quarter of 2024 by $9.3 million, resulting in a 32% gross margin. Hosting revenue of $22.6 million for the same period in the prior year exceeded hosting cost of revenue of $16.2 million by $6.4 million, representing a 28% gross margin.Hosting gross profit increased by $2.8 million or 44% quarter-over-quarter, driven by the onboarding of proceed-sharing clients beginning in the fiscal second quarter of 2023. Hosting costs consist primarily of direct electricity costs and data center operating costs.Operating expenses for the fiscal first quarter of 2024 totaled $16.9 million as compared to $24.2 million for the same period in the prior year. The decrease of $7.3 million was primarily attributable to lower stock-based compensation of $13.3 million due to [indiscernible] during the quarter, partially offset by a $3.4 million increase in personnel and related expenses and a $1.7 million increase in advisory fees related to the reorganization incurred during the fiscal first quarter.Net income for the fiscal first quarter of 2024 was $210.7 million as compared to a net loss of $388,000 for the same period in the prior year. Net income increased by $211.1 million, driven primarily by a decrease of $143 million in reorganization items, which included $143.8 million associated with the extinguishment of pre-emergence obligations in excess of the amount settled post emergence, lower Chapter 11 financing cost of $11.1 million, partially offset by a $12.8 million increase in claimant related bankruptcy professional fees and a $60.1 million mark-to-market adjustment on our warrants and contingent value rates.Non-GAAP adjusted EBITDA for the fiscal first quarter of 2024 was $88 million as compared to non-GAAP adjusted EBITDA of $40.3 million for the same period in the prior year. This $47.7 million increase was driven by a $58.6 million increase in total revenue and a $1.1 million decrease in impairment of digital assets, partially offset by a $4.4 million increase in total hashrate operating expenses, a $4.1 million increase in cash cost of revenue, a $3 million increase in realized losses on energy derivatives and a $0.5 million decrease in gain from sales of digital assets.Our power contracts vary in pricing terms. As mentioned previously, our fleet-wide power cost averaged $0.043 per kilowatt hour in the first quarter. We continue to expect average power cost in 2024 to be between $0.045 and $0.047 per kilowatt hour.At the end of the first quarter, our self-mining to hosted mining mix was 77% to 23%, respectively. We plan to increase the efficiency of our self-mining fleet through ongoing miner refresh and additional hashrate to achieve our 2024 goal. As we expand our self-mining fleet, we expect our hosting mining mix to decline over time.Our fleet-wide average energy efficiency was 26.85 joules per terahash as of March 31, 2024, and 25.78 joules per terahash as of April 30, 2024. The improvement was due to the completion of our S21 deployment in April. As of March 31, 2024, we operated approximately 173,000 miners in our self-mining fleet. The model mix shown on Slide 11, was 10% S19, 64% S19 Pro and S19j Pro, 24% S19j XP and 2% S21.Now I'd like to discuss the strength of our balance sheet. Cash and equivalents at the end of the first quarter was $98 million, up from $50 million at the end of 2023 and does not include an additional $16 million in restricted cash.Slide 12 compares our first quarter capital structure to year-end 2023. At the end of 2023, total debt was just under $1 billion. As of March 31, 2024, total debt was $608 million, a decrease of $390 million. The reduction in debt for the quarter was driven mainly by the equitization of legacy debt and the settlement of prior plans.As a reminder, a share price of $6.81 puts our tranche warrants in the money and their full exercise would provide us with $670 million in cash, allowing us to pay down debt and to build our cash balance. A share price of $7.79 triggers the mandatory conversion of the convertible notes, which would clear $260 million of our balance sheet.Now I'll turn to our CapEx plans. In the first quarter, we made all payments due this year on miners we ordered and have deployed in 2024. We anticipate purchasing additional miners in 2024 to complete our planned refresh and to achieve our 21.8 exahash self-mining hashrate goal. The precise amount and timing of this purchase will ultimately depend on us finalizing the details of our HPC strategy.In April, we announced the start of the expansion project at our Denton, Texas data center where we are increasing our operational power by 72 megawatts to 197 megawatts, an expansion of more than 50% by the end of our fiscal second quarter. The capital expenditures associated with this expansion were included in our 2024 CapEx plan and were paid in April of 2024. We will incur an incremental $4.5 million in CapEx associated with our new Austin HPC data center, which was not previously included in our 2024 CapEx plan.Now I'll turn to a review of the mining economics summarized on Slide 13. Our direct cash cost to mine a bitcoin in the first quarter was $18,915. This consists of power cost of $15,977 and a cash-based facilities operations cost of $2,938. Allocated based on the 77% of our fleet dedicated to self-mining and divided by total bitcoin self-mine in the first quarter of $2,825.Another way to look at this is by calculating the cash-based cash costs of these same items, which represents the same cost expressed as a cost per terahash per day. Our total cash-based, cash cost in the first quarter was $0.326 per terahash.In summary, we expect operating cash flow to be sufficient to support operating expenses, debt service and CapEx associated with our organic growth plans in 2024.And now a few housekeeping items. We continue to model a statutory effective tax rate of approximately 23% for 2024. We also have more than $300 million in net operating loss carryforwards, which will reduce our future cash taxes. Our share count as of March 31, 2024, is approximately 182 million shares.And now I'll turn the call back to Adam to discuss our expectations for 2024. Adam?

[Adam Sullivan, executive]

Thanks, Denise. Before we turn to Q&A, I'll spend some time walking through our strategic priorities for the rest of the year and the macro environment factors driving these priorities.Now that we are 3 weeks from the latest halving, we have seen a normalization of the record high transaction fees immediately after the halving. Cash price has declined to around $0.05, and we are seeing a small decline in global network hashrate. Barring any dramatic and sustained increase in hash price over the next 3 to 6 months, we expect to see inefficient hashrate drop off the network as some machines are turned off and as operators seek new homes for their miners. We expect some difficulty decreases throughout this process, and we expect year-end hashrate to be higher than current levels.Speaking more broadly, in 2024 and over the next few years, we anticipate increased competition for blocks as scale miners continue to invest CapEx to increase their hashrates. We also expect to see increases in U.S. power prices over the coming years. The question for Core Scientific now is how can we best grow our business and continue to create economic value for our shareholders at a time when the value of our owned infrastructure is increasing?For Core Scientific, the answer comes in 3 parts. First, by continuing to build out our owned infrastructure, particularly through the completion of our Texas projects; second, by expanding our hashrate through fleet refresh and emerging miner options. And finally, as I discussed, we are focused on leveraging our owned infrastructure to capture the significant opportunity in HPC hosting.I'll describe each of these in more detail, starting with our partially built infrastructure at our 2 Texas sites. At these sites, we had 372 megawatts that require an investment of about $200,000 per megawatt on average to complete. These 372 megawatts can support more than 20 exahash of mining capacity over the next 3 years when complete. And as mining technology yields higher efficiencies and hashrate per megawatt, that total hashrate will increase.We can also dedicate a portion of this new infrastructure to HPC hosting depending on customer needs and opportunities. We are currently on track to energize the 72 megawatts in Denton by the end of our second quarter. We expect to purchase the remaining miners to achieve our refresh and hashrate expansion goals later this year.Second, we are taking advantage of mining market economics and new miner suppliers to expand our hashrate cost effectively, both through refreshing our fleet and expanding our rack space. For perspective, based on our current 745 megawatts of infrastructure, if we were to refresh all of our prior generation S19, S19 Pro and S19j Pro miners with S21, we would be able to increase our existing hashrate by more than 10 exahash without adding any new infrastructure.We are already seeing improved mining equipment economics in the post-halving environment. We are also working with multiple technology companies to develop and deploy new lower-cost miner technology with higher energy efficiency that will offer greater procurement options.And third is our emerging alternative compute business launched with our successful deployment of 16 megawatts of data center capacity for CoreWeave. As discussed earlier, buyers of advanced GPUs for workloads such as AI cloud and high-performance computing having limited supply of infrastructure options and often face significant and costly delays in the availability of new data center capacity.Core Scientific has more than 500 megawatts out of our total 1.2 gigawatts of contracted power that can be utilized for alternative compute workloads based on geographic proximity to major cities and fiber lines. Further, we have a successful track record of efficiently managing large-scale data centers, and we have a team from the data center industry leading our operations.We are in regular discussions with customers in the space and expect to build out this part of our business further over the course of the year. We think it's important to help frame the economics of this potentially significant business opportunity as follows: Based on industry data, the cost to build a new Tier 1 HPC data center ranges between $7 million and $12 million per megawatt. Based on our current assumptions, we project the cost to convert one of our high-power bitcoin mining data centers into a Tier 1 HPC data center at between $5 million and $8 million per megawatt. Even saving $1 million per megawatt represents a $100 million in construction saving for a 100-megawatt data center.We are pursuing clients that are able to prepay for construction CapEx as an offset against a portion of their monthly hosting payments. We aim to become a market leader in providing digital infrastructure for high-performance computing. The cash generating power of that business will enable us to keep some of our bitcoin production on our balance sheet in anticipation of future increase to the extent that we have cleared certain debts that prevent us from holding bitcoin today.Based on industry data, we target Tier 1 HPC hosting revenue on the order of $1.4 million to $1.6 million per megawatt per year with gross margin of 75% to 80%. Power costs and utilities are direct pass-through to clients. The complete conversion of 500 megawatts of bitcoin mining infrastructure to HPC hosting would likely take 3 to 4 years, but we expect to begin generating revenue earlier as capacity comes online incrementally during that process.HPC hosting provides stable revenue and gross profit. This is important for Core Scientific because it will provide stability and a greater degree of revenue predictability, which can also help moderate the variability in our bitcoin mining results against the more dynamic mining backdrop.As we consider these 3 points, we see a transformational opportunity to balance our portfolio in business between highly efficient bitcoin mining at scale and alternative compute hosting. Our bitcoin mining business generates profitable cash flow and preserves our exposure to the upside potential in bitcoin price. It also built the platform for an alternative compute business that could provide significant multiyear steady cash flows with strong financial returns.The potential to optimize our asset portfolio across these 2 attractive and high-value compute areas is only available to us because we own and control all our infrastructure. We cannot be better positioned to capture the opportunity in these 2 growing markets. We will provide more details about our emerging alternative compute hosting business when we reach any definitive agreements.Our Board, our leadership team and I are more excited than ever about Core Scientific and our growth plans. We truly believe that by executing on our balanced strategy of bitcoin mining scale and alternative compute hosting, we can enhance value for all our stakeholders, both in the near and long term and deliver compelling financial results that will unlock tremendous value in our company. Thank you all for your engagement and attention, and thank you to our customers, industry partners, and all our teammates for your ongoing efforts and support.We will now take your questions.

[Operator]

[Operator Instructions] The first question comes from the line of Joe Flynn with Compass Point.

[Joseph Flynn, analyst]

On the HPC front, with the 500 megawatts of potential infrastructure capacity, I was curious like what kind of customers are you currently having conversations with, whether they be hyperscalers, data center operators, start-ups? Just any color you could provide there would be helpful.

[Adam Sullivan, executive]

Of course, Joe. I would say our target base right now is mainly around our goal to have prepaid revenue as part of this contract, so them -- having the client pay for the CapEx. That definitely narrows the scope of potential clients, but that definitely puts it in the range of large tech companies that are looking at the development of their AI segments. So that's really our focus right now is mainly around large tech companies with a focus on AI, where the demands are for application-specific infrastructure.

[Joseph Flynn, analyst]

And just drilling down to the economics you guys mentioned, it looks like the existing CoreWeave contracts was roughly $100 per megawatt hour. And the numbers you just gave closer to $150 to $170. Just kind of curious, is that the $150 per megawatt hour level, is that with these existing agreements where the CapEx will be prepaid? And really just any other color you could provide on the margin profile, that would be great.

[Adam Sullivan, executive]

Yes, of course. So our target for -- what we laid out is really our target for conversions of sites. And so when we talk about things like the existing CoreWeave deal or any conversion of existing space where we leased and then subleased to a potential client, our total revenue and our margin profile would be a bit different and a bit lower. Our focus going forward is really on the conversion of sites. And so what we've laid out are based on discussions with potential clients as well as industry data that's helping guide us really to the answer. And so what we're looking at on the margin side is really that 75% to 80% is really what we're targeting today. And now it's on the back of about $1.4 million to $1.6 million in revenue per megawatt.

[Operator]

The next question comes from the line of Lucas Pipes with B. Riley Securities.

[Lucas Pipes, analyst]

Adam, I also wanted to ask about the HPC opportunity. And you mentioned kind of 3 to 5 years for greenfield. And if I understood you right, you mentioned 3 to 4 years for your conversions. Is that correct? And maybe more importantly, kind of what is the process for developing greenfield? I'd like to understand kind of the competition. So if someone comes in looking at a greenfield, how long does it take power? How long does it take to construction? And how do you compete against that?

[Adam Sullivan, executive]

Of course, and thanks, Lucas. I want to start with the second part of the question. What we're seeing from traditional operators today, traditional data center operators they have long-dated contracts, 10 years or greater. And so for on the existing infrastructure side, they have a very hard time competing with the part of the industry that we're focused on today. And then going forward, they've sold forward, I would say, at least 3 to 5 years of capacity at which they've locked up. And so converting any of that in the short term is very difficult for them.Now if you look forward right now, you're seeing some of the large tech companies securing power of 2028, 2029, 2030, that's just to secure the power aspect. You tack on top of that, a lot of supply chain constraints for equipment luckily that we already own. So in the traditional data center industry, it's at minimum 3 to 5 years for them to really start attacking this industry.From our perspective, what we're looking at, we said 3 to 4 years to fully develop the 500 megawatts. We mentioned we're going to have incremental capacity come online throughout that time period, and that's mainly driven by the fact that we have a lot of the long lead items already owned inside of our business today. And a lot of those constraints are around the electrical infrastructure that you could see.

[Lucas Pipes, analyst]

To follow-up on this. In your presentation, you show that valuation arbitrage between some of the leading bitcoin miners and data center companies. In light of everything discussed, why haven't we seen M&A yet? What's your take on that?

[Adam Sullivan, executive]

Yes, Luca, if I might just follow-up your question, you're referring to M&A in our industry itself?

[Lucas Pipes, analyst]

Yes. Well, specifically what I'm referring to on Slide 8 is a forward EV-to-EBITDA multiple for data center companies 20 time, you cite 914 times multiple for the highest multiple bitcoin miners, certainly consistent with some of the data I've looked at. Why haven't we seen M&A from the data center side to the bitcoin mining side to date? I would appreciate your thoughts on that.

[Adam Sullivan, executive]

Yes. No, it's a great question. I think Morgan Stanley put out a very good report related to the opportunity that bitcoin miners actually have today given the fact that just on electrical equipment alone, it's at least 36 months lead time for traditional data center. So just having access to the power is a significant advantage and it's actually a much higher value to traditional data centers and really the valuations that we're seeing bitcoin mining infrastructure traded today. I wouldn't rule that out, traditional data centers are definitely trying to find ways to bring power online more quickly.What we're seeing across a number of reports is that data center capacity is going to double over the course of the next 6 years. So I think that's something that we're still in the early stages. I would imagine that companies throughout the industry are having those types of conversations. From our perspective, we're focused on executing this because we believe we can drive a significant amount of short-term and long-term value for our shareholders.

[Operator]

[Operator Instructions] The next question comes from the line of Kevin Dede with H.C. Wainwright.

[Kevin Dede, analyst]

So Adam, maybe you could offer a little operational insight. I know the hash price has trended down, right, maybe a little bit lower than you expected or had modeled. I'm wondering, and I know you said that you expect miners to come off and you look for the next difficulty adjustment this week and 2 weeks beyond. Is there anything that you're doing sort of in-house to maximize the performance of the fleet?

[Adam Sullivan, executive]

Thanks, Kevin. I think it comes down to really 2 items. And the first is comes down to operations. Prior to halving, we actually moved our machines based on their efficiency amongst our sites based on their power contracts, really to prepare for a time period that could be much worse than what we're seeing today in terms of the $0.05 hash price level. The second part is our in-house software development team has developed a significant amount of firmware around the ability to adjust machines on a minute-by-minute basis amongst different types of firmware settings.And really, what that does is it allows us to change our efficiency of our machine fleet and it allows us to do that based on power prices at each of our sites as well as prevailing hash price metrics. And so for us, that provides a significant advantage over our peers who have outsourced much of that capability set, whereas we've been able to integrate really all 3 parts of the software stack, the energy management, the fleet management and the firmware, all into a single software stack that allows us to provide a significant amount of control greater than our peer set today.

[Kevin Dede, analyst]

At what point would you consider developing fresh megawatts to address the HPC market versus conversion? And how would you balance that infrastructure spend vis-a-vis 10 or so exahash that you could gain in improving your fleet?

[Adam Sullivan, executive]

Yes. We're in a very unique position today where we have infrastructure that could support both bitcoin mining and HPC hosting. Obviously, bitcoin mining is the platform for which we're able to expand into new markets and find new sites. And on the contract side for the HPC business, our focus today is really on customers who are able to prepay for that CapEx. And so as we evaluate new opportunities, it's going to come down to a month-by-month, quarter-by-quarter basis in terms of how we allocate capital, but we'll continue to -- we'll be able to continue to grow our bitcoin mining exahash over the course of the next few years.Yes, as you mentioned, we have a 10 exahash opportunity just within our existing footprint today, if we were to upgrade our machines to the newest generation. And so what we're looking at is an exciting opportunity to bitcoin mining site and a very exciting opportunity on the HPC side.

[Operator]

The next question comes from the line of Greg Lewis with BTIG.

[Gregory Lewis, analyst]

Adam, I was kind of curious on your thoughts to Kevin's question where you kind of addressed your firmware and your stack. You mentioned other miners using third-party solutions. As I think about your firmware solution that you're using internally, is there an opportunity potentially to bring that out into the market and have other smaller miners be potential customers, i.e., is this a potential other revenue stream for Core?

[Adam Sullivan, executive]

Yes. Thanks for the question. This is something that we've evaluated in the past, potentially rolling out to a broader market. We view this as a significant competitive advantage over our peers. And what we've seen over the past few years in terms of the development of software is that we've continued to lead the path in terms of our development. And so from our perspective, we're going to continue to keep this as our proprietary in-house software so that we can maintain that competitive advantage over our peers as we continue to grow.I think one important note on that as well is that our peers when they're running third-party software, they're paying a pretty significant development fee to the ultimate owners of that firmware. That's a fee that we're able to not pay and actually generate greater gross profit than our peers on the same machine type at holding all the variables constant.

[Gregory Lewis, analyst]

And then I also wanted to follow-up on the HPC opportunity. I mean, you mentioned the ability for prepaying, I guess, prepaying is one thing, and that's obviously, it sounds like the ideal solution. But what about approaching it using your existing infrastructure with a joint venture? Is that something that could make sense? Or at this stage in the game, it's something that we're not really interested in bringing on a partner?

[Adam Sullivan, executive]

Yes, I would say the way we look at it right now and really what it says this most is utilizing the prepaid revenue structure allows us to own that infrastructure free and clear after that prepaid revenue drop. Owning this infrastructure, whether it's at the end of the contract or whether it's after that prepaid revenue is paid off, will give us significant advantage in terms of being able to refinance that infrastructure and potentially pull some capital out to fund future growth. And so we view this as we have the technical capabilities in-house. We have the infrastructure base. And so from our perspective, really, the capital is the only other part of that that's necessary to execute. And if we can get that from our client base, that will provide us a significant advantage going forward.

[Operator]

The next question comes from the line of Tyler DiMatteo with BTIG.

[Tyler DiMatteo, analyst]

I just wanted to follow-up really quickly on the prepayment from the HPC customers. I mean, do you have a threshold level in mind? And I guess like what's the lead time you're thinking about for that payment and then maybe rolling it down?

[Adam Sullivan, executive]

Yes, of course. I think what we're looking at right now is really on the 100% payment terms. So for the total CapEx bill receiving 100% of that from our potential clients. I think some high-level guidelines here to think about in terms of how we're thinking about it is really not utilizing more than 50% in any given year towards the revenue that we could be generating. And I think that really brings us to a point where we're still able to experience significant free cash flow generation off of these megawatts even with the prepaid structure rolling off over the course of the contract.

[Operator]

The next question comes from the line of Lucas Pipes with B. Riley Securities.

[Lucas Pipes, analyst]

The $1.4 million to $1.6 million that you mentioned, Adam, is that including the pass-through on power? And I think you mentioned another pass-through, if you could remind me of that. And so for the EBITDA margin, I think you mentioned 75% to 80%, that would be still kind of straight on top of that revenue line. Just wanted to make sure I didn't miss anything there.

[Adam Sullivan, executive]

Thank you, Lucas. Yes. So the $1.4 billion to $1.6 billion, that is not inclusive of the pass-through power and utilities. And so when you think about that number, that's really the lease rate on a per megawatt basis for the entirety of the year. All of the other expenses that are incurred are pass-through to the client. And so when you think about the 75% to 80%, we like to think about that as the gross margin on that per megawatt number that's given to that $1.4 million to $1.6 number.

[Lucas Pipes, analyst]

So that, call it, 20%, 25%, those would be more or less maintenance costs of the site, security, fixing the plug, a leak here and there. So that's kind of the way to think about it?

[Adam Sullivan, executive]

Absolutely right, Lucas.

[Operator]

The next question comes from the line of Kevin Dede with H.C. Wainwright.

[Kevin Dede, analyst]

Denise, you mentioned 2 stock price thresholds for warrant conversion and the flood of cash that would offer you and perhaps clearing the balance sheet at that point. Could we take a step back and think about that process without the stock moving? I mean at what point would you consider retiring debt given cash generation and favorable mining economics?

[Adam Sullivan, executive]

Kevin, I'll take that question. So really, how we're thinking about today -- yes, of course, the $19 million that we paid off early in the second quarter was related to Mechanic's Liens to build out that 72 megawatts at our Denton facility. We have some additional debt service over the course of this year that will pay down some additional principles.As we evaluate the evolving market that we're in today, what we believe, based on our analysis, is that it's better to put capital towards growth than paying down debt at this point. And so we're going to continue to fund growth, and that may change over the course of the next month, the next quarter. But today, our focus is on continuing to fund the growth of this business. So we believe it's more accretive to our bottom line.

[Operator]

The next question comes from the line of Jack Chan with Imperial Capital.

[Jack Chan, analyst]

On the HPC, curious to know if the potential clients reached out to you directly or did you reach out to them? And how much of the 500 megawatts would be potentially taken by these clients you're in talks with?

[Adam Sullivan, executive]

Yes. Thank you for your question. I would have to say it's a bit of a mix of both. People recognize the platform that we've built. They know the locations of our sites, and they know the capabilities that we have on our internal team. You look at our operations team up and down, coming almost directly from the data center industry with decades of experience across each member. And so that is experienced at the traditional data center industry as well as the tech industry know and they know many of our team members very well. And so it was a bit of a combination of both inbound as well as some outbound calls to certain partners that we knew may be interested.Our goal right now is to repurpose about 500 megawatts to HPC. And really, what we're focused on is trying to accomplish this potentially depending on how negotiations and discussions go with potential clients over the course of the next 3 to 4 years. And so that's really what we're focused on today, executing on our growth plan, not only in the bitcoin mining side but also on the HPC side.

[Jack Chan, analyst]

And I think you said the timing of a potential deal could be in this year or did I mishear that?

[Adam Sullivan, executive]

Yes, that's something that we're going to be updating the market as these negotiations, conversations evolve. So as we get to more definitive decisions around this, we will be announcing that to the market.

[Operator]

There are no additional questions left at this time. I will hand the call back over to Steven Gitlin for closing remarks.

[Steven Gitlin, executive]

Thank you, Tia. With no further questions, we thank you for your attention and your interest in Core Scientific. An archived version of this call, all SEC filings, and relevant company and industry news can be found on our website, corescientific.com. We wish you a good day, and we look forward to speaking with you again following next quarter's results.

[Operator]

That concludes today's conference call. Thank you. You may now disconnect your lines.