Marathon Digital Holdings Inc

NASDAQ:MARA

Earnings Call Transcript 2024-Q1

[Operator]

Good day, ladies and gentlemen. Welcome to Marathon Digital Holdings' First Quarter 2024 Earnings Webcast and Conference Call.I would now like to turn the call over to your host, Robert Samuels, Vice President of Investor Relations. Please go ahead.

[Robert Samuels, executive]

Thank you, operator. Good afternoon, and welcome to Marathon Digital Holdings' first quarter 2024 earnings call. Thank you for joining us for our call today. With me on today's call are our Chairman and Chief Executive Officer, Fred Thiel; and our Chief Financial Officer, Salman Khan.Before we get started, I'd like to remind everyone that our prepared remarks may contain forward-looking statements, and that we may make additional forward-looking statements during the question-and-answer session. These forward-looking statements are subject to risks and uncertainties and actual results may differ materially. When used in this call, the words anticipate, could, enable, estimate, intend, expect, believe, potential, will, should, project and similar expressions as they relate to Marathon Digital Holdings are as such a forward-looking statement. Please refer to our earnings release for a full recitation of our forward-looking statements. Investors are cautioned that all forward-looking statements involve risks and uncertainties, which may cause actual results to differ materially from those anticipated by Marathon at this time. Some of these risks and uncertainties are more fully described in Marathon's public filings with the U.S. Securities and Exchange Commission, which can be viewed at www.sec.gov, and ir.mara.com.Finally, please note that on today's call, we will refer to certain financial measures that were not prepared in accordance with generally accepted accounting principles in the United States, including adjusted EBITDA and non-GAAP total margin. Marathon believes these non-GAAP financial measures are important indicators of its operating performance because they exclude certain items that are unrelated to and may not be indicative of its GAAP financial results. Please refer to our company's periodic reports on Form 10-K and 10-Q and to our website for a full reconciliation of these non-GAAP performance measures to the most comparable GAAP financial measures.As usual, we'll begin today's call with prepared remarks from Fred and Salman. After their comments, we will be going through some of the more popular questions from investors before transferring to a live Q&A with our covering analysts.And with that out of the way, I'm going to turn the call over to Fred to kick things off. Fred?

[Frederick Thiel, executive]

Thank you, Rob. The first quarter of 2024 was one of our most transformative to date. We officially transitioned Marathon from being an asset-light Bitcoin miner and laid a solid foundation on which we are building this organization into a globally diversified company that leverages digital asset compute to build a more sustainable and inclusive future.During the quarter, we significantly grew and improved the resilience of our portfolio of digital asset compute by acquiring and integrating our first fully owned and operated Bitcoin mining sites. At the end of last year, our portfolio consisted of 584 megawatts of capacity, only 3% of which we directly owned and operated. In less than 4 months, we initiated, negotiated, closed and integrated our first 2 acquisitions comprised of 3 sites. The first 2 being the sites in Granbury, Texas and Kearney, Nebraska, and the second being the site in Garden City, Texas, that is adjacent to a wind farm. We purchased these assets for approximately $458,000 and $437,000 per megawatt, respectively, which is approximately half the cost of what some of our competitors have paid to build new sites.In Q1, we effectively doubled the size of our portfolio to 1.1 gigawatts of capacity, and we gained far more direct influence over our operations by taking direct ownership and operational control of 54% of the portfolio. We did this while generating significant savings for our shareholders relative to building sites the way our competitors have, and we quickly gained enough capacity to meet our near-term growth targets of 50 exahash by the end of this year. As we cost-effectively laid this path to accelerate our scale and enhance our operational influence, we also launched our first products and services as an organization, each of which demonstrates our commitment to expanding and diversifying revenues by creating advanced technologies that transform digital infrastructure and diversify our revenue streams.In February, we launched Slipstream, which is a direct Bitcoin transaction submission service that is designed to streamline confirmations of large or non-standard Bitcoin transactions. In essence, it provides sophisticated users with a simple, transparent, and trusted means of adding complex Bitcoin transactions to the blockchain, provided they adhere to Bitcoin's protocol and it provides Marathon with an opportunity to increase revenue by capturing more transaction fees, which is particularly advantageous in a post-halving environment. Although the service is only a few months old, it has already proven valuable. Last month, just before the halving, we earned an additional 4.25 Bitcoin from Slipstream alone. MARA Pool, which powers Slipstream, captured 1 block with 10 Bitcoin in transaction fees and another with 16, generating nearly 10% greater fees than FPPS pools at the time. While these numbers are small relative to our overall production, Slipstream serves as a tangible example of the strategic importance of operating our own mining pool of being at scale and of deploying our proprietary technology stack. As a matter of fact, if you were to look at the mempool.space for the past week and look at mining pools, you would see that Marathon's pool performed 40% above average.As we demonstrated in March, Slipstream was just the beginning. During the first quarter, we also brought our first products to market, all of which contribute to 1 of our key competitive advantages, our vertically integrated tech stack, and provide us with an opportunity to diversify our revenue streams. These products include our industry-leading MARA firmware, controller board and 2-phase immersion technology which we call 2PIC. We showcased each of these products at the Empower Conference in March, and the reception was overwhelmingly positive. We already have paying customers for our firmware and 2PIC has a pipeline that is in the tens of millions of dollars and growing. We're currently integrating 2PIC at some of our own sites, which is part of our strategy to optimize our performance by increasing uptime, improving efficiency and reducing maintenance downtime and cost. For the market, our goal is to begin shipping these systems in volume to third parties by the end of this year.While the quarter was highlighted by the significant expansion of our portfolio of digital assets compute and the launch of our first products and services geared towards transforming digital infrastructure, our Bitcoin production was negatively impacted throughout the first quarter by unexpected equipment failures, predominantly transformers at our third-party hosted sites. Utility company transmission line maintenance and the higher-than-anticipated weather-related curtailments across multiple sites. However, I'm pleased to say that we have mitigated a number of these issues and are currently operating at a record high level of 27 exahash. While there will always be the potential for exogenous events that impact operations, we believe we can mitigate their impact over time as we continue to scale, standardize equipment, diversify globally and build redundancies and leverage economies of scale that stem from owning and operating multiple sites around the world. In the meantime, we can always control how we respond to challenges, and just as we always have, our team quickly began working on solutions that were within our control.We were able to immediately take advantage of the newly available capacity from our recent strategic acquisitions. Once it became clear that the transformer issues at Ellendale were more complicated than originally believed, our team quickly began reallocating miners from this site to Garden City, where there was available capacity. We successfully moved 9,500 idle miners from Ellendale and simultaneously began energizing new machines to expand our capacity. Our ability to adapt combined with the uptime improving across our sites, these challenges were resolved, enabled us to achieve an all-time high operating hash rate of 27 exahashes this month.The transformer issues that negatively impacted our ability to convert megawatts into terahash, and therefore, Bitcoin have been challenging, but with our hash rate continuing to grow to record levels, we have clearly been able to work through the challenges and to execute. Unfortunately, the resilience we've built into our diversified portfolio, our ability to adapt to change and changing circumstances and our HODL strategy allowed us to still capitalize on Bitcoin's positive momentum. Despite the operational challenges, we still produced record financial results during the first quarter.To unpack our financial results, I'm now going to turn the call over to Salman, who will cover the results in more detail. Salman?

[Salman Khan, executive]

Thank you, Fred, and welcome, Rob, to team MARA. As Fred mentioned, our Bitcoin production was negatively impacted in the first quarter by transformer and other issues. Despite these challenges, the positive momentum in Bitcoin price, our strategy to increase our Bitcoin holdings over time and our team's consistent ability to execute help us drive record financial results for the quarter.Let's dig into the details. We reported net income of $337 million, or $1.26 per diluted share in the quarter. This was a 184% increase from net income of $119 million or $0.72 per diluted share in Q1 of last year. The increase in net income was primarily driven by favorable Bitcoin price, higher production, and fair market value of digital assets on our balance sheet.Revenues increased 223% to a record $165 million from $51 million in the first quarter of 2023. With the average price of Bitcoin mine being 126% higher this quarter than the year ago period, the increase in revenue was primarily driven by a $77 million increase in the average price of Bitcoin mined. We produced an average of 30.9 Bitcoin each day during the quarter compared to 24.4 Bitcoin each day in the prior year period. These improvements in production also contributed approximately $10 million to our top line. As a result of closing the recent acquisitions, we are temporarily providing hosting services to existing hosted customers during the transition period. The revenue we generated from hosting revenues was $21 million, which was not present a year ago. We expect this revenue to taper off in the coming quarters as we work with current tenants off the sites on their transition plans and use that space for our own growth.Our hosting and energy costs were $71 million compared to $33 million last year. In both time periods, the increase was primarily due to growth in our mining fleet and related costs.In the first quarter of 2024, global hash rate or network difficulty levels continued to increase. Average global hash rate has increased by more than approximately 86% since Q1 of 2023. That combined with operational issues at our third-party hosted sites in Q1 of 2024 resulted in a higher cost per coin.Our non-GAAP cost of revenue without depreciation has improved 16% from 53.7 per petahash per day in Q1 of 2023 to 45.2 per petahash per day in Q1 of 2024. As we realize synergies from our recently announced acquisitions, reenergized production at our third-party hosted sites, and spread the fixed costs over larger 50 exahash capacity, we expect our marginal cost per coin to improve. Total cost of revenues, which includes depreciation and amortization was $168 million compared to $151 million [Technical Difficulty] of last year. Depreciation and amortization was $78 million, a $60 million increase from the same quarter in the prior year. The change was predominantly the result of growing our energized hash rate from 11.5 exahash to 27.8 exahash and recent acquisition of 2 sites in Texas and Nebraska. Our non-GAAP total margin, excluding depreciation and amortization was $75 million this quarter compared to $18 million in the same quarter last year. The change was predominantly related to higher Bitcoin prices, increased production and increased operational efficiency.General and administrative expenses, excluding stock-based compensation were $21 million compared with $11 million in the prior year period. This increase in expenses was primarily due to the increasing scale of the business. Headcount grew from 30 employees at the end of Q1 last year to 60 employees at the end of Q1 this year, which subsequently has increased to over 100 employees. As we thoughtfully grow our company, we continue to opportunistically add talent across the organization.Part of our treasury management strategy involved selling enough of the Bitcoin we produce to cover operating expenses and then HODL-ing the remainder on our balance sheet. By employing this strategy, we have essentially been dollar cost averaging into Bitcoin at a discount for the past several years. With our cost to produce a coin significantly lower than the Bitcoin price, the dollar average strategy generally has been even more accretive for our stockholders. With the new accounting rules that require the measurement of crypto assets at fair value and with the Bitcoins price appreciating nearly 150% from the end of Q1 last year to the end of Q1 this year, this strategy has proven very effective for us.In Q1 of this year, we increased our Bitcoin holdings 15% from 15,126 to 17,320 Bitcoins. Because of the significant amount of Bitcoin we hold on our balance sheet, we recognized a gain on digital assets of $489 million during the first quarter of 2024. Primarily due to this favorable value adjustment, our adjusted EBITDA increased 266% to $529 million from $145 million in the prior year period. A reminder that the fair value accounting reflects the changes in market value of Bitcoin that we hold on our balance sheet. In the current environment of Bitcoin price appreciation in recent quarters, we have continued to recognize gains through our income statement, reflecting the impact HODL strategy has had on our financial statements. For example, every 10,000 change in Bitcoin price period-over-period can impact our adjusted EBITDA by approximately $170 million to $200 million.Digging more into our Bitcoin holdings and cash position, unrestricted cash and cash equivalents totaled $324 million, up from $125 million a year ago. Also, at March 31, we held approximately 17,320 Bitcoin with a fair value of $1.2 billion on the balance sheet. Combined, our balance of cash and Bitcoin was approximately $1.6 billion as of March 31, 2024. We sold 730 Bitcoin during Q1, realizing cash proceeds of $45 million. These proceeds were utilized to fund operating expenses, including cost of revenues for energy hosting and other cash operating expenses and general and administrative expenses. During the quarter, we raised $489 million from at-the-market equity sales, which we primarily intended to use for miners, acquisition of infrastructure and for other general corporate purposes.We expect our future Bitcoin holdings will generally increase, but will fluctuate depending on operating and market conditions. We intend to add to our Bitcoin holdings primarily through our production activities, and we will also continue to sell Bitcoin as a means of generating cash to fund monthly operating costs and general corporate purposes.As Fred mentioned, we have systematically deployed capital for accretive growth and strategic investments, which our teams have been able to integrate successfully in Q1 of this year. These acquisitions have, not only been almost half the construction cost, but also have allowed us flexibility to move our miners from asset-light sites with operational issues as discussed above. If we were to construct such sites on our own, it would have taken us more than a year to energize and likely cost twice the amount on a per megawatt basis, making these acquisitions highly accretive for our stockholders. The acquisition of sites in Granbury, Texas and Kearney, Nebraska have been accounted for as a business combination for FASB rules. And we have recognized the fair value of the assets acquired and liabilities assumed on our balance sheet. As a result, we also recognized goodwill of $30.9 million and intangible assets of $22 million, primarily related to customer relationships. We also accrued early termination expenses of $22 million as we terminated certain hosted customers subsequently.One other -- another item worth highlighting is that, Marathon was added to S&P Small Cap 600 Index. Just a reminder, the S&P Small Cap 600 tracks 600 U.S. companies with a market cap between $1 billion and $6.7 billion that have positive earnings over the most recent quarter and 4 quarters combined. We are proud to be the first Bitcoin miner added to the S&P Small Cap Index. This inclusion is a testament to the consistently improving financial performance of our organization as we have scaled and implemented effective strategies to capitalize on market conditions.And that completes my update. I'll now turn it back over to Fred, who will talk more about our future plans. Fred?

[Frederick Thiel, executive]

Thanks, Salman. Over the past year, we've evolved from outsourcing our operations to third-party hosting providers to owning and operating sites, developing and marketing technology products, and using waste and stranded power to generate energy and recycle heat. During Q1, we laid the foundation for Marathon to become a globally diversified company that leverages digital asset compute to build a more sustainable and inclusive future. And as we transition into the second quarter, we are reorganizing the internal structure of the business to better align with our growth opportunities, sharpen our strategic focus, bolster accountability and maintain our speed and agility as we scale.Starting in the second quarter, Marathon will be restructured into a matrix organization consisting of 3 business units, utility scale mining, technological innovations, and energy harvesting and supporting organizations such as Marcom, Finance, et cetera. In my opening remarks, I spoke at length about the first 2 of these verticals. Utility scale mining is our large scale global digital asset mining operations. This vertical is what Marathon is primarily known for today, large-scale digital asset compute that's geared towards enhancing grid stability and the security of the world's preeminent blockchain ledger. Our technology division is focused on creating advanced technologies that transform digital infrastructure. This group has already launched its first series of products, and going forward, its mission will be to deliver innovative hardware, software and services that maximize energy efficiency, improve performance and unlock new opportunities for those who operate at the edge and for those who are building the foundation of the digital future.In March, at the Empower Conference, we demonstrated how the combination of MARA firmware and 2PIC immersion enabled us to operate state-of-the-art digital asset miners at significantly higher performance levels with near-minimal impact to energy efficiency. Using this solution enables digital asset miners to significantly increase the productivity of their digital mining assets, earn a higher return on the compute investment dollars and reduce the total cost of operating sites.Our third vertical, which I did not discuss in my opening remarks is energy harvesting. The mission of energy harvesting is to support the energy transformation by converting environmentally harmful waste energy, stranded clean energy or otherwise underutilized sources of energy into economic value. For now, it's perhaps easiest to think of this as the next evolution of Bitcoin mining, involving leveraging digital asset compute to convert waste into energy, sequester methane, and generate and reuse heat for industrial and commercial purposes. This last part is quite important, as nearly 50% of total energy use is used for heat and a portion of that is low-grade heat, i.e., sub-50 degrees Celsius that can be provided by heat harvested from our digital asset compute systems. This includes applications such as pre-heating materials in industrial processes, heating buildings and water for communities and food production applications such as greenhouses and shrimp farms. We created these business units in response to the opportunities we saw in the market and how we believe this industry will evolve.As you all know, the industry just underwent a halving event, during which the block reward was reduced by 50%. As block rewards will continue to reduce by 50% every 4 years, it is critical to find new ways to leverage expertise and create value for others in order to maintain a position as a leader in the industry. And we believe that we can do so successfully and we can, not only mitigate the impact of halving events, but also excel. While utility scale mining is by far the most significant portion of our business today, we believe each of these businesses will contribute meaningfully to our organization over the coming years. Long term, our goal is for 50% of our revenues to come from non-utility scale mining. It will take time to reach that scale, but these new businesses are already showing progress. We already have a pipeline valued at tens of millions of dollars for our technology products. Our pilot project in Utah, where we're capturing methane from landfill to power Bitcoin miners is our first energy harvesting project, and we expect to launch several others in the coming quarters.The demand for these verticals isn't just domestic, it's global, and we're chasing after it. In April, we announced that we appointed a new Managing Director to lead our expansion in Europe, Middle East, Africa and parts of Asia. Since his appointment, our team has generated promising opportunities across the region from Finland in the north to Kenya in the south. While Marathon is already operating on 3 continents, our plan is to continue scaling our international business over coming years such that approximately 50% of our revenues come from overseas by 2028.In the more immediate term, we're doubling down on our utility scale mining business. We originally targeted to grow our operations to 35 or 37 exahash by the end of 2024 and 50 exahash by the end of '25. However, with the expansion capacity gain from our recent acquisitions and the 45 exahash of additional compute available to us between current orders and options, we're now targeting 50 exahash by the end of this year. We already have 12 exahash of new miners arriving in early Q3 to take advantage of our acquired and expanded capacity with more coming to fill waiting capacity over the balance of this year. This represents a doubling of our compute capacity and further increases our energy efficiency in under 1 year and will strongly extend our position as the leader in our industry. While we will always be raising capital to continue growing our business, we currently have ample liquidity to fully fund the balance of our 50 exahash growth target this year.Q1 was a transformative quarter for Marathon, during which we battled against operational challenges to produce record financial results. We launched and began building the pipeline of our first product and we doubled the size of our portfolio of digital asset compute. With 11 sites diversified across 3 continents and the road to 50 exahash clear ahead of us, we look forward to evolving into and excelling as a globally diversified company that leverages digital asset compute to build a more sustainable and inclusive future for all.And with that, I'll turn it back to Rob for Q&A. Rob?

[Robert Samuels, executive]

[Operator Instructions] We'll start by answering some of the most popular questions submitted by investors through our Q&A platform.So our first question comes from [ Alexander Ren ], who asks, post halving, in what ways does MARA in a better position to stay profitable and grow compared to other mining companies?Fred, do you want to take that one?

[Frederick Thiel, executive]

Sure. So a number of things. By acquiring sites and vertically integrating, we lower our cost structure. So essentially taking out the middleman who was the third-party hosting operator and taking out their margin, which, in some cases, can lower our cost by as much as 20%. Additionally, the diversification of our revenue streams, technologies such as Slipstream, which enhance our ability to earn transaction fees over and above what normal FPPS pools earn, provide us with abilities to, in a lower Bitcoin block subsidy world such as post-halving generate, generally, greater revenues on a per block basis. So we believe we're very well positioned with our diversified revenue streams that we'll be developing over the coming years to be positioned to take advantage of whatever the market brings to us in the near- and long-term.

[Robert Samuels, executive]

Our next question, and we have a couple of questions related to M&A, but [ Parminder S ] asks, is MARA looking for a merger or planning to buy out other miners in the short-term?Fred, do you want to take that one as well?

[Frederick Thiel, executive]

Well, we've already shown that we're acquiring hosting companies and the sites for capacity. Now, to date, these have been mostly sites where we have miners installed currently. So it's a very easy integration. The challenge with integrating and acquiring miners that have their own fleets of miners is the age of miners that they have in their fleets. Most companies today are still transitioning to the latest state-of-the-art S21 class of machines and may be running S19j Pros or even XPs. The challenge is, if we were to acquire a miner today and attribute any value to their fleet of existing machines, especially if their machines were S19j Pro or older, that wouldn't be a very smart idea.So essentially, as we acquire miners, you'll tend to find that pricing will be very focused on the infrastructure value as opposed to the miners that are installed, unless, of course, they have state-of-the-art miners installed, at which point you can value them at their replacement cost. But I think typically, most publicly traded miners have valuations which exceed essentially the cost to rebuild that infrastructure, and it comes down to a question of time to revenues. And what is it worth for you to acquire a miner that has capacity running and miners installed versus building it yourself and how much extra do you want to pay for that time to market.

[Robert Samuels, executive]

Our next question comes from [ Roshan U ], who asks, what's the status of your open ATM? And do you think this is the reason that MARA has not been tracking the price of Bitcoin?Salman, I think this one is for you.

[Salman Khan, executive]

[ Roshan ], in Q1, there was a general rotation out of miners into the recently launched ETFs. We believe the top performing miners will see price appreciation as the investors see the results. Similar to other industries like oil and gold, ETFs have their own position in the value chain, while oil and gold companies create value for their stockholders and generally track the commodity price with premium on top of the commodity price.In terms of the ATM, as of March 31, 2024, we had not accessed the new ATM. And subsequently, we raised about $200 million. As a reminder, the ATM gives us low cost of capital to invest in accretive acquisitions or grow the company as we continue to use Bitcoin to fund our operating costs and G&A.

[Robert Samuels, executive]

I think the next one is for you as well. [ Orlando C ] asks, what's your current breakeven for Bitcoin?You want to take that one?

[Salman Khan, executive]

Great question, Orlando. There seems to be a lot of diversity in the industry in terms of how the companies report their costs and where is it parked and all that stuff. So it's a difficult question to answer in terms of apples-to-apples comparison. So with that in mind, typically, you could look at multiple ways. One is, marginal cost of revenue without depreciation, which is cost of revenue minus depreciation non-cash items. And that's one way of doing it. The other way of doing it is to look at cost per hash, which we believe is a better representation in this industry. So cost per petahash, as we talked about previously in our prepared remarks, we have improved about 16%. And as we fully integrate these recently announced acquisitions, as Fred had mentioned, some of those acquisitions can be -- can save up to 20% on our costs. All these costs will continue to show up in the future quarters as these synergies kick in and we integrate these acquisitions completely.As a reminder, our mining fleet remains the industry's one of the most efficient, which will even get better and more efficient as we progress with time and we hit the 50 exahash towards the end of the year. And that larger scale of exahash, it will allow us to spread the fixed cost across the larger pool, which, on a unit basis, will make us even more efficient.Moreover, just a quick reminder, our G&A per hash or coin remains one of the lowest in the sector, and that shows how efficiently the company has operated historically.

[Robert Samuels, executive]

Our next question comes from [ Tarik P ], who asks, what are your plans for increasing efficiency and reducing the recent downtime at several bitcoin mining facilities? Fred?

[Frederick Thiel, executive]

Yes, great question, [ Tarik ]. So if you think about what happened in Q1, as I mentioned in my prior remarks, we had transformer issues at third-party hosted sites and we had a utility line maintenance that was done that shut down one of our sites for a short period of time, and then we had weather events. All of these are exogenous factors.If you think about when you own and operate sites, you look to do things like standardize equipment across sites so you have standard operating procedures, you also have spare parts for everything you're doing. One of the challenges with the Ellendale site were transformer issues, where some of the transformers that were installed could not operate at their stated spec. And so, as those transformers started to fail, they had to be replaced and/or modified. And had we built that site, we most probably would not have used Chinese transformers but rather transformers of higher quality that would operate within spec. So again, over time, as more and more of the sites that we operate at come under our own control, I think we'll see uptime continue to tick up and we'll see performance improve.When you think about just operating efficiencies, again, stripping out the third-party hosting provider dramatically impacts your cost because you're stripping out some of these built-in margins. And we also begin deploying our immersion technology. The 2PIC immersion technology does a number of things. For one thing, it increases the -- essentially, the lifespan of miners because these miners that traditionally would be air cooled in many sites essentially have to be serviced every 30 days, they have to be cleaned, filters have to be changed, fans have to be replaced, and the mechanical load at the sites is higher. Using 2-phase immersion, you don't have a lot of those issues. We have immersion systems that have run for months without having to be serviced at all. And so, that reduces the overhead at the site, which reduces your operating cost, makes you more efficient, increases uptime.And the other advantage with 2PIC immersion is that, you become fairly immune from heat waves because, at the end of the day, the liquid in this tank is boiling, which means that it's likely to be hotter in the tank than the ambient temperature would be. So that makes heat much less of an issue, especially in places like Texas and in the UAE, for example. And if you look at our site in UAE, that site operates with 99.8% uptime. And this is in an environment where the ambient temperature is over 100 degrees and humidity is well into the 90% range. So I think that's a great example of how our sites operate more efficiently and more effectively when they're under our control.

[Robert Samuels, executive]

In the interest of time, I think we'll have to wrap up this section of the Q&A. And again, we really appreciate the questions and interest.I'm now going to turn the call back to our operator to open the line to questions from our covering analysts. Alicia, back to you.

[Operator]

[Operator Instructions] Our first question comes from the line of Lucas Pipes with B. Riley Securities.

[Lucas Pipes, analyst]

Fred, my first question is on the pool performance you mentioned. I think you said 44% better. To what extent was this chance versus you being able to pick off kind of above average rewards?

[Frederick Thiel, executive]

Yes. So over the past week period, which that number comes from, it's about 40-point-something percent, not 44%. But there were certain transactions that had very high fees. And at the end of the day, you have to win the block to get paid the transactions. But what Slipstream enables us to do is essentially generate additional fees on top of the standard fees that the pool otherwise would earn. And so, even if you look over longer periods of time, whether it's October through May, October of last year through May, MARA Pool has performed amongst the top pools from an average fee structure. A way to look at it is, if you were to optimally fill a block with the highest value transactions, that would be an index of 0. And then how does a pool perform against that with blocks? And most of the FPPS pools operate at a minus 1% to 2%, sometimes 3% below the optimum, where MARA Pool together with Slipstream has been performing above that. And for the period really for the early part of this year-to-date, it's been almost 6%, 7% and sometimes higher than FPPS pools that it competes against.So this is an advantage that, over time, may decline as more people start trying to leverage the types of technologies that built into our pool. But this is the whole reason why having a vertically integrated tech stack is so important. It allows us to do things that you can't when you're just buying best-of-breed technologies and putting it together. An analogy some people use when describing what we're doing is it's kind of the difference between the PC industry and the Apple ecosystem. Apple is all vertically integrated from their own chips, through their own software, their own cloud services, et cetera. Whereas the rest of the industry uses microprocessors from 1 vendor, operating system from another vendor, graphics, et cetera, from other vendors. So we believe a very tightly integrated tech stack allows you to operate large-scale digital asset compute in a much more efficient and optimized manner than just buying third-party services and leveraging third-party pools.

[Lucas Pipes, analyst]

I appreciate that, Fred. And from a modeling perspective, would it be too early to guide on an additional revenue on that? Or would you...

[Frederick Thiel, executive]

Yes. I think you could definitely work with Rob to look at the data. And you can actually just see this in mempool.space. You can just run reports for different periods of time and you can see how our pool fees have averaged compared to others. But I think that over time, 2%, 3% would be a conservative estimate to use for modeling purposes. And then there will be times where it's higher and there may be times where we operate at the same as FPPS pools, but 2%, 3% on average should be fine, which over time becomes quite a large number.

[Lucas Pipes, analyst]

Fred, and turning quickly to M&A, I know this has come up before, but where are you spending your energies right now? And are there kind of more opportunities like the ones you've executed on recently in front of you? I would appreciate your thoughts on that.

[Frederick Thiel, executive]

Sure. So with the acquisitions we've already done, we have capacity to get to our 50 exahash goal by the end of this year. And we obviously have the miners inbound to fill that available capacity. So as we look at acquisitions, going forward, we believe that, as I've said before, there's a mix of kind of third-party hosting sites that may be attractive. There's a mix potentially of sites where somebody has tied-up power, but hasn't fully built out the infrastructure. And then there are true greenfield where you're out talking to energy companies, solar project developers, wind project developers and trying to negotiate power deals and then having to build infrastructure from the ground up. So we're looking across all 3 of those buckets pretty equally. We believe that this period post-halving now is going to be quite rich when it comes to M&A opportunities.And we believe that there will be some great opportunities to acquire more capacity. We don't think that there will be a shortage of compute. We look at pricing in the marketplace and what we're seeing is more efficient rigs coming to market at essentially the same prices as they are now. And we don't see an additional buying wave now post-halving of compute. So we think that, that market will likely remain fairly stable with relatively easy access to be able to buy more rigs. And we continue to think that capacity, in other words, sites are kind of the shortage and there's a lead time if you're going to build sites also. So we think that as we look towards 2025 expansion because, obviously, we're not going to stop at 50 exahash and we look at '26, '27 and towards the halving in '28, we think that companies like Marathon, where you have very large scale, utility scale mining, if you would, combined with energy harvesting and technology sales provides you with a really robust and resilient business where you can weather ups and downs and then also get the leverage of being able to continually operate a little bit better than everybody else. That's the goal, be the most efficient, most optimized operator out there.

[Lucas Pipes, analyst]

This is helpful. I want to turn it over. Fred, to you and the team, best of luck.

[Operator]

Our next question comes from the line of Joe Flynn with Compass Point Research.

[Joseph Flynn, analyst]

I had a question regarding the Generate Capital acquisitions. I was wondering how many megawatts of capacities were currently hosted by legacy customers. And if you could provide more color on, ultimately, the timelines of freeing up that capacity and transitioning it to self-mining, that would be helpful.

[Frederick Thiel, executive]

Yes. So about 390 megawatts of total capacity of which we already had somewhere around 180-ish megawatts that we were using. There is about 75 or 80 megawatts, which will be vacated by the end of June, and then the balance, we think shortly thereafter. So we think we'll have 100% utilization of that site for our miners in Q3.

[Joseph Flynn, analyst]

Great. And then a follow-up on kind of the power costs. The hash costs look to be impacted just due to the uptime. But just going forward, how should we think about maybe just power costs from a cost per kilowatt hour basis? And ultimately, how owning your own infrastructure can layer in to lower those costs and how much improvement can we see in kind of the second half as you install more efficient machines?

[Frederick Thiel, executive]

Sure. So if you look at the sites in Texas where we can take advantage of a variety of programs to monetize energy, when we were using sites that were operated by a third-party, we didn't get the benefit of economic curtailment in power trading, which is something we now have the benefit of. And we have actually reaped some very material gains already in the first quarter from some of that activity. But that's very opportunistic. It depends on temperature spikes, things that happen in the market that allow you to really benefit from that.The big challenge -- the big difference rather with owning sites versus using third-party sites as you strip out this kind of upwards of $0.02 a kilowatt hour that the hosting provider was charging you, which was their way of covering their infrastructure costs, their operating costs, et cetera. We believe we can operate sites that we own for somewhere between $0.05 and $0.01 per kilowatt hour. It kind of depends on the where and what type of technology. The more immersion technology we have, the lower the cost to operate the site as an example.When it comes to power prices, there's some wide ranges from PTAs that are in the sub $0.03 range to things that run upwards of $0.04. So somewhere in that range across the different sites. But the key savings from the shift from asset-light to owned and operated is stripping out that margin of the third-party hosted provider. So you can imagine $0.01 to $0.015 a kilowatt hour reduction in your cost to mine is pretty significant.

[Operator]

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

[Kevin Dede, analyst]

Could you, Fred, maybe peel the onion back a little bit on power? I know you moved some rigs from Ellendale. Could you give us your thinking on when you might be able to move rigs back? And how you plan on building or filling out? I think you have available capacity at the old Generate Capital sites. Just give us some feeling for where you're going with power and how you'll be able to support 50 exahash this year?

[Frederick Thiel, executive]

So let me see if I understood your question properly. So from a capacity perspective, we'll be filling out a lot of the existing capacity over Q3 and a little bit into Q4 as well. As I said earlier to somebody's prior question, we should be seeing most of the hosted customers vacated, majority of the capacity really by end of this quarter and leading into a little bit into Q3. And we halve a whole slew of miners coming in to plug in that are all state-of-the-art S21 class machines.Within the case of Ellendale, power has -- we've been able to bring the site back up partially. Right now it's operating at about 60%, 65% of its max capacity, and each week, a little bit more comes online as transformer fixes are made by Applied on that site. And so far, we've been ahead of schedule there. So we -- miners ready to go to plug in there. We moved 9,500 miners from Ellendale to Garden City just as a way to -- we had the capacity, it was easy to make the move for the miners. And so, it made sense to do that. As Ellendale continues to come back online, we have machines ready to go to plug in to take up those empty slots. So no change in power cost, if that's what you were asking for in regards to power or if that was more related to energization. But essentially, the only site that we're really seeing a challenge at from an energization perspective is just what's left to do at Ellendale. But that's progressing on a weekly basis, as I mentioned, and that will be all back up over the balance of the next coming months.

[Kevin Dede, analyst]

Well, the subtext of the question, Fred, apologies for not making it clear was, I thought that the 2 sites that came with Generate Capital offered you flexibility and adding capacity, power capacity. I was just wondering how you were thinking about that. And whether or not you thought you might need it this year to reach 50?

[Frederick Thiel, executive]

No, we don't need to expand capacity at any sites. The capacity that we have that is energizable, if you would. So no need for additional expansion covers us to 50 exahash. So anything at the Generate sites that would be an increase in site capacity would be additive and think of that as more 2025 growth target.

[Kevin Dede, analyst]

One more if you let me, Fred. Could you add a little color please to your energy harvesting theme and your heat reuse theme? You mentioned a couple of things on calls, but not a ton of detail.

[Frederick Thiel, executive]

Yes. So, again, energy harvesting is taking essentially wasted or stranded energy. And that can come in multiple forms, right? It can come in the form of methane coming off of landfills and we've talked about the pilots that we have running in Utah in that regard. You can also capture methane flare gas from oilfields, for example, and there is methane coming out of a variety of agricultural processes. Obviously, dairy farming generates a lot of manure. You can take a lot of these things such as bio waste. Think of food processing plant, essentially, the waste that comes out of that, the stuff that comes out of ethanol, methanol manufacturing, which is all agricultural product base. You can take those base products and, through anaerobic digesters, generate electricity with them. You then use Bitcoin miners to consume that electricity. And then the heat generated by the miners can be fed back into the industrial process. So in the case of ethanol and methanol manufacturing, you need to pre-heat the biomass that you're using to make ethanol, alcohol, for example, manufacturing, spirits, beer, et cetera. You have to do similar things for the fermentation mass.And in the case of agriculture, you'll find there's a lot of bio waste that can be used and then you can generate heat and feed that back into things like greenhouses or shrimp farms, et cetera. So heat reuse is very important. You also have the simple use case of just heating buildings. You can take electricity that normally is going to power a boiler and you can replace that boiler with our 2PIC technology with miners in it and you can generate hot water just the same. Same amount of energy generates, same type of water at similar cost. And so, if for the building owner, it's cost-neutral to them or even a cost savings to them to deploy our solution versus traditional boilers or heating or even more importantly, if they're transitioning to electric heat from natural gas or oil, as part of the energy transition, then they can get the benefit of having potential cost savings, and we essentially get a situation where our cost to mine Bitcoin is free.And long term, the benefit of energy harvesting is that, the processing of the waste material and the sale of the heat back into an industrial process should hopefully mitigate 100% of your energy cost. And so, the mining of Bitcoin is done for free. Now granted, these are smaller scale sites, but some of these sites potentially are 5 to 20 megawatts in size. And you have the additional benefit that eventually, if transmission lines were available, you could just sell the electricity you're generating. So we believe that, as a way to diversify our business, lower our costs, decentralize the mining of Bitcoin, and this is an important factor because with the halving, you're going to see more and more concentration of Bitcoin mining amongst the big miners. It's very important that you start to develop this long tail of Bitcoin mining that I talked about where you essentially have hundreds of thousands if not millions of things out there that are leveraging energy to mine Bitcoin, all to secure the Bitcoin blockchain and process transactions because, over time, large utility scale miners will eventually come into a place where you can't get energy at utility scale at a lower cost than its cost to produce in those cases. And if Bitcoin price trends upward but not at a accelerated pace, then over the next 4 to 8 years, potentially, utility scale mining will become potentially a more challenging business. And so, the combination of technology, the combination of using alternative energy sources, all these things become really important to optimizing your operations.And our goal is to remain one of the lowest cost operators when we get to the next halving and be positioned to really continue to lead the industry well beyond the point where the block subsidies are the primary driver of revenues for Bitcoin miners.

[Operator]

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

[John Todaro, analyst]

I guess 2, if you could just kind of remind us through the cadence of getting 50 exahash by the end of '24?And then second question, do you anticipate G&A to kind of stay mostly flat from here? It sounded like there was still some hires you needed to do. But just wondering if we could drill into that a bit more.

[Frederick Thiel, executive]

So I'll let Salman answer the G&A question. But relative to the cadence of deployment, as I said before, capacity at the sites starts freeing up really at the end of this quarter and the early part of Q3. We have a large number of miners inbound to fill that capacity. And then the balance of capacity growth will happen over the latter part of Q3 and into Q4 as we continue to get Ellendale back online fully and fill out the little nooks and crannies of capacity that exist out there. We're at the same time also upgrading some of the sites that we acquired from air to immersion and doing things which will increase efficiencies and increase uptime. And this is all before we start talking about the benefits of 2PIC immersion and what it can do from an overclock perspective with minimal impacts to efficiency.So I think longer term, these types of technologies are going to become really important because if you can overclock S21s to the tune of 50% or 60% and you can do it with a de minimis loss of efficiency, then you can save a lot of CapEx, plus have the flexibility to overclock, underclock all depending on energy cost and Bitcoin price. It gives you a lot more flexibility and potentially saves a lot of capital. So we're very excited about kind of the impact of that on the longer term, especially as we grow well beyond 50 exahash at the end of this year.

[Salman Khan, executive]

On the G&A, if you think about it from a cash G&A standpoint, excluding stock compensation, which is a non-cash item, we expect it to be flattish on a quarter-over-quarter basis going forward from here. It's a quick reminder that, on a unit basis, we continue to improve from here as we deploy more exahash in the company and spread that across the larger exahash.

[Operator]

Our next question comes from the line of Joe Flynn with Compass Point Research.

[Joseph Flynn, analyst]

I had a question. I was wondering what the progress is on Auradine. And also, Fred, when you talk about like having the latest and greatest S21, are you referring to 17.5 joules per terahash or the most recent Pro version where it's 15 joules per terahash?

[Frederick Thiel, executive]

Yes. So on Auradine, I'm not going to comment specifically on all the company's achievements because I'll let the management team do that. But what I will say is, we are in the process of -- we'll be deploying Auradine in our 2-phase immersion tanks. We'll have essentially the ability to put almost 800 kilowatts of mining capacity in a 8-foot container, which is a huge power density factor and you can essentially take a -- what is today a -- if you look at our UAE sites, we have essentially each 40-foot container is about 1 megawatt. Well, we can retrofit that container to dual phase immersion and end up with 3.2 megawatts in that same form factor at fairly low cost for the upgrade. Same miners, just more of them and the ability to consume more power, more power is available. So we believe that, that's going to be a big differentiator going forward.The advantage with Auradine's miners for us is, a lot of the intelligence that's built into them regarding fleet management. And like any fabless semiconductor company, Auradine is building wafer capacity with their vendors and continue to grow supply. And we'll continue to buy product from them as the supply chain increases.In regards to the Bitmain question, yes, there is a mix of Pros and the 17 joules per terahash. The immersion version of the 21 does not come in a Pro version. But the immersion version operates more efficiently than the air version. So kind of mitigates the 2. Hopefully that answers your question.

[Operator]

Thank you. There are no further questions. I'm going to turn the call back over to Rob Samuels for closing remarks.

[Robert Samuels, executive]

Thank you all for your time today. If you have any questions that were not answered during today's call, please feel free to contact our Investor Relations team at [e-mail], and thank you. Enjoy the rest of your day.

[Operator]

This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.