Keel Infrastructure Corp. (BITF) Earnings Call Transcript & Summary

Good morning, everyone, and welcome to Bitfarms' Second Quarter 2024 Financial Results Conference Call. [Operator Instructions] It is now my pleasure to turn the floor over to your host, Tracy Krumme, Senior Vice President, Head of IR and Corporate Communications.

Good morning, everyone, and welcome to Bitfarms' Second Quarter 2024 Conference Call. With me on the call today is Ben Gagnon, Chief Executive Officer; and Jeff Lucas, Chief Financial Officer. Before we begin, please note this call is being webcast with an accompanying presentation. Today's press release and our presentation can be accessed at our website bitfarms.com under the Investors' section. Turning to Slide 2. I'd like to remind everyone that certain forward-looking statements will be made during the call and that future results could differ from those implied in this statement. The forward-looking information is based on certain assumptions and is subject to risks and uncertainties, and I invite you to consult Bitfarms MD&A for a complete list. Please note that references will be made to certain measures not recognized under IFRS and therefore, may not be comparable to similar measures presented by other companies. We invite listeners to refer to today's press release and our MD&A for definitions of the aforementioned non-IFRS measures and their reconciliations to IFRS measures. Please note that all financial references are denominated in U.S. dollars unless otherwise noted. Importantly, I would like to highlight that we are unable to comment on the ongoing legal process with Riot Platforms, outside of what has already been disclosed. I would also like to add that we will be attending the following upcoming equity conferences: H.C. Wainwright Global Investment Conference taking place in New York City from September 9th to 11th, and the AIM Summit in Dubai, taking place from October 21st to the 22nd. If anyone would like to meet with us on those dates, please contact me or a sales representatives from the firm. Now turning to Slide 3, it is my pleasure to turn the call over to Ben Gagnon, Chief Executive Officer.

Thank you, Tracy, and thank you, everyone, for joining us today. I am so excited to have stepped up into the Chief Executive Officer role. This has been my personal ambition for the past several years, and it is truly an honor. As this is my first conference call as CEO, I would like to give a quick background on myself for those of you who may not know me. 14 years ago, I discovered Bitcoin. And 9 years ago, I began working full time in the mining industry. I started from the very bottom, investing every dollar I had, to build and operate a mining facility in Mainland China of my own design. Since then, I've been through multiple bull and bear markets, three halving epochs and have seen and been involved with every aspect of this industry from the mining floor to the C-suite of one of the largest publicly traded mining companies globally. I have been with Bitfarms for 5 years, and for the last 3 years, served the company as Chief Mining Officer, where I oversaw mining operations and strategy and worked intimately with every department in the company. I am a proof of work CEO, and the transition into this new role has been smooth and well received among our team and external stakeholders. Turning to Slide 4. Over the past 30 days since my appointment, I've set aside time to speak with every key employee of Bitfarms, in addition to all of our key partners and many of our investors. From these conversations and from my first months in the new role, I would like to share the following key points: We are excellent builders and operators. Our team is highly skilled and passionate but more importantly, they believe in the Bitfarms vision. From site teams through senior management, we are happy to stay up late and get up early in the morning to do the hard work that is necessary as a miner. Bitfarms is a meritocracy, and a lot of our staff have risen through the ranks due to their proof of work. Everyone here is incredibly excited to be a part of building the new Bitfarms. And there is so much energy in the air right now that we could probably mine a block with it. This energy and excitement is also shared by our external stakeholders. Over the last few years, I have had the pleasure of representing the company with sell-side analysts, institutional investors and in public speaking forums during which time, I've gone to know many of you and many of you have gone to know me. I am a known and trusted entity. Our energy portfolio and strategic approach to growth sets us apart. We manage what I believe is the best and largest internationally diversified portfolio of energy contracts in the Bitcoin data center business. We have been able to organically grow our footprint globally, while adhering to the decentralization ethos that is core to Bitcoin and the profit maximalism that is core to Bitfarms. Our exposure to different geographies, different sources of energy and providers, different climates, different government authorities, significantly de-risk our portfolio, and we lead nearly every market we operate in, at scale. Lastly, the opportunities ahead of us to continue to grow and create value for shareholders are second to none. We believe that we represent one of the best opportunities for investors to gain high-quality exposure to Bitcoin's upside through our fleet upgrade program as we continue to scale up our highly efficient operations as we continue to gain market share. That being said, there are always ways to improve and to grow more efficiently and more effectively. The more we are growing at a tremendous rate, I have identified areas where our organizational structure can be revamped to better match the scale at which we are operating as well as our ambitions to grow into the future. In the coming weeks, we will be reorganizing some of our teams to provide an even stronger foundation that supports an accelerated growth trajectory. Additionally, sometimes in the pursuit of growth, it is easy to miss out on smaller optimization opportunities. One question I asked every employee was what is the low-hanging fruit that you see? Numerous team members throughout the organization have suggested sometimes simple but powerful ideas to improve and optimize performance. In addition to our focus on growth, we are also implementing systems that will drive continuous optimizations throughout the organization with a focus on cost effectiveness. It is important to highlight that our 2024 growth plan and growth targets are not changing with my appointment to CEO. As Chief Mining Officer, I was the key architect behind our fleet upgrade and growth plan this year, and we remain committed to reaching 21 exahash and 21 watts per terahash by year-end. That being said, we are not stopping at 21 exahash. I am also laser-focused on growth into 2025 and beyond. With our new strategic plan, the Board has determined to end the strategic alternatives review process. The company is certain that the best path forward to maximize value for all shareholders is to move forward with our standalone plan. While it's too early to provide specifics today, I'd like to comment on some of the key initiatives I will be focused on moving forward. First, I am committed to continued diversification of our portfolio. This means both geographic diversification and diversifying beyond Bitcoin mining. Our greatest asset is our portfolio of competitively priced energy assets. As portfolio managers, we are constantly reevaluating how to maximize the value of these assets. This doesn't mean pivoting away from Bitcoin, but expanding into synergistic business lines that will increase our profitability and make us better Bitcoin miners. Some examples include HPC and AI, heat recapturing and recycling, energy generation and of course, energy trading. To reiterate, these activities will not detract from our Bitcoin mining operations, but rather be integrated into our portfolio in order to make us more efficient and more profitable. Second, over the past 2 years, we have focused on developing our international portfolio. In 2025 and beyond, we will be largely focusing on increasing our U.S. exposure. We anticipate that our recently announced deal in Sharon, Pennsylvania, will just be the first of many new sites in the U.S. Third, we will be pursuing more miner purchases with creative structures that give strategic advantage to Bitfarms. By way of example, last year, we were the first mining company to negotiate and announce a miner purchase option with Bitmain. They gave us the right but not the obligation to purchase a significant amount of miners at locked-in competitive prices. The structure was so advantageous that nearly every one of our peers followed suit with their own option in the weeks and months after we announced. As we look to the new, highly efficient miner models currently being announced, investors can expect Bitfarms to continue leading the industry and utilizing and developing new, highly accretive structures that maximize flexibility and value creation for our shareholders. Turning to Slide 5. We are on track to deliver record hashrate growth and efficiency improvements in 2024, and we continue to execute on this growth plan in quarter 2. Here, you will see a snapshot of where we are and where we plan to be by year-end '24 and year-end '25. In Q2, we increased our 2025 power capacity by 220 megawatts with agreements in Paraguay and the U.S. We grew our hashrate 70% from Q1 and hashrate growth will continue to accelerate in the second half of 2024 and into 2025. I'll let Jeff speak to the financial results in more detail, but I would like to highlight that we have a strong balance sheet with $195 million total liquidity at the end of Q2, over 1,000 Bitcoins at the end of July and a 2024 growth plan that is fully funded. Turning to Slide 6. Let's talk a little bit more about our new U.S. site, Sharon. We are so excited about this site for a number of reasons. First, Sharon will be our first mega site in the U.S. with over 120 megawatts of total capacity. This single location will increase our U.S. footprint sevenfold from 20 megawatts to 140 megawatts and kick-start our aggressive U.S. growth plan. Second, Sharon is located in Western Pennsylvania, which is close to major metropolitan areas like Cleveland, Pittsburgh, Philadelphia and New York and is in close proximity to major fiber lines. Pennsylvania is a conservative, business friendly jurisdiction with a notably pro-Bitcoin and pro-energy democratic governor. Third, the PJM grid is the largest wholesale electricity market in the U.S., offering abundant access to competitively priced and flexible power that is attractive for multiple uses, including Bitcoin mining, energy trading and even HPC and AI. Fourth, for Bitcoin mining specifically, the site supports over 8 exahash with the latest generation miners and there are significant curtailment, demand response and energy trading opportunities available to effectively hedge our energy costs and bring down the total cost of power. Further, as PJM is rapidly adding renewable capacity and significantly contributing to the decarbonization of the grid, these grid stabilization programs make the site both economically and environmentally sustainable. Given these significant advantages, we are actively engaged in assessing additional new opportunities to expand our presence within the PJM region. In addition, based on numerous conversations with potential partners, we believe this site is very well suited for HPC and AI. 1, the PJM market has a very reliable power and the grid is much less prone to the weather-related stresses that you'll see in Texas; 2, accessible fiber lines support connectivity and redundancy in close proximity to the 4 major metropolitan areas I mentioned previously; 3, the site is located in a deregulated market and is not tied to any one power provider, providing unparalleled flexibility; 4, we have not yet started construction on the site, and so we maintain 100% flexibility in terms of the build-out plan. We have a clean slate and would not have to incur retrofitting expenses. This also allows for an expedited deployment schedule capable of meeting AI customers' aggressive time lines. We have, in fact, received so much interest over the past few weeks for U.S. sites in this 100-megawatt sweet spot that I'd like to spend a minute here discussing the HPC and AI opportunity. Turning to slide 7. The HPC and AI opportunity is a very exciting one that has been monopolizing the headwinds for the past few months and rightfully so. Bloomberg and UBS cite a total addressable market for the AI cloud GPU services of $28 billion in 2022, growing to $420 billion in 2027, with related AI infrastructure growing eightfold from $26 billion to $195 billion over the same time period. We own and operate a portfolio of high-quality energy assets that are currently monetized for Bitcoin mining. But when we take a step back and look at how we get the most value and utilization out of this portfolio, we believe that HPC and AI has real potential. Recent HPC and AI deals are boasting revenues from approximately $140 to $210 per megawatt hour. These are potentially very attractive and stable, high-margin revenue streams not correlated to Bitcoin prices. Comparatively, Bitcoin mining with T21 miners yesterday on August 7, yielded approximately $80 per megawatt hour. Properly timed, we still believe that investments in Bitcoin mining provide a better return on invested capital compared to HPC and AI due to their materially lower CapEx requirements and upside exposure to Bitcoin prices. That said, we believe that the most attractive opportunity is a combination of the 2 with a potential integration in Q4 2025 or Q1 2026. This would potentially provide us increased diversification and exposure to varied revenue streams in what this has historically been the top of the Bitcoin bull market cycle and lines with HPC customers' timelines. We are still in the early stages of evaluating the opportunities here, but we believe that our North American sites have the potential to be very well suited for these activities. To help us evaluate and develop this vertical, we are currently recruiting for HPC and AI talent, ensuring we have the expertise to capitalize on this huge opportunity. Additionally, our very active corporate development team who is constantly assessing new energy assets are now evaluating all opportunities through multiple lenses, including the HPC and AI lens. The key thing to drive home here is that HPC and AI will not replace Bitcoin mining for us, but rather seek to complement our current operations in order to create the most upside and value for our shareholders in line with historical market cycles. Moving to Slide 8. I would now like to switch gears and tell you about our progress to 21 exahash and 21 watts per terahash year-to-date. While we did experience temporary delays in hitting our midyear target 12 exahash, we did hit our efficiency target, 25 watts per terahash, representing a 19% improvement quarter-over-quarter and a 29% improvement year-to-date. The 12 exahash milestone was delayed due to some temporary equipment delays as well as a batch of nearly 3,000 miners, representing approximately 700 petahash that underperformed in even low temperatures. The delayed equipment has since been received and installed and these issues have been addressed with Bitmain and are not expected to be present in future batches of miners, including our August deliveries. Bitmain is also rapidly replacing these 3,000 miners with new units at their expense. These new miners are expected to arrive and be installed in 3 weeks. Our facility upgrades have also progressed rapidly and nearly all of our sites in Canada have now been upgraded, resulting in up to a 52% improvement in energy efficiency per site and a 29% improvement in energy efficiency across the company. With 7 of 11 data center upgrades now complete, the only remaining facilities to be upgraded are Villa Rica, Magog, Washington and Argentina. PDUs are currently being shipped to Villa Rica, Magog, and new T21 miners are scheduled to be sent in the coming days, with upgrades at both sites scheduled to be completed in September. Works are also progressing in Washington, which is both a data center upgrade and an expansion. Final works are scheduled to be completed in November. Lastly, in Argentina, we are currently working on a revised data center upgrade to marginally expand the total capacity of the site from 54 to 62 megawatts. With this expansion to 62 megawatts, we now expect this upgrade to be finalized in December or January. The first batch of PDUs and miners are being shipped to Argentina this month, and we are scheduled to begin seeing improvements in hash rate and efficiency as early as October. On miner deployments, we have now deployed approximately 48% or 42,000 of the 88,000 miners that we ordered for 2024. These miners were mostly deployed in facility upgrades. And this replacement of our older, less efficient hardware is largely responsible for our rapid improvement in energy efficiency year-to-date. Roughly half of the remaining miners will be deployed in the 4 remaining data center upgrades just mentioned, further improving our energy efficiency down to our target of 21 watts per terahash. The other half will be deployed in our new constructions and will be responsible for most of the remaining hash rate growth to 21 exahash. In terms of construction progress, we've made significant strides to date in 2024. I am pleased to report that our 70-megawatt site at Paso Pe is now fully online and is our largest site by both megawatts and hash rate. Our 12-megawatt expansion Baie-Comeau is well underway and is on track to be energized in September. In Yguazu, we started the year with 100 megawatts contracted and have since doubled the contracted capacity to 200 megawatts. This site will represent the largest site in our portfolio in 2025. In terms of construction progress, we have now completed all of the necessary purchase orders and broken ground in 7 of the warehouses. We expect a 100 megawatts to come online in December, contributing approximately 5 exahash with 20 watts per terahash efficiency and an additional 100 megawatts to come online in the first half of 2025. Turning to slide 9. I would like to share with you some beautiful aerial photos of Yguazu that show the tremendous progress we have made. 4 months ago, this was just a soy field, and in 5 months, it is expected to be between 1/2 and ¾ of a percent of the entire Bitcoin network, powered entirely by renewable energy. From breaking ground to energization, the construction schedule is only 9 months, far faster than what is possible in the U.S. The rapid scale at which we are developing the site is unparalleled and is an incredible testament to our amazing team and capabilities. Turning to Slide 10. I'd like to take a quick moment to welcome Fanny Philip to our Board of Directors. Fanny is a recognized expert in the blockchain technology field and an accomplished financial executive with an extensive background in audit, public company reporting and M&A. Her skill set will be invaluable as we continue to drive organic and inorganic growth. Fanny represents our fifth Director, 4 of which are now independent. Turning to Slide 11. I will close out with a summary of our impressive growth stats and trajectory for 2024 and 2025. In 2024, we will be tripling our hash rate, increasing our operating capacity by 83% and improving our efficiency by over 40%. These are industry-leading benchmarks and numbers that I am incredibly proud of. We will continue to build on this growth in 2025. We've already added an additional 220 megawatts to our energy portfolio, supporting over 35 exahash and stay tuned because there is more to come. And with that, I will turn to Slide 12 and turn the call over to CFO, Jeff Lucas.

Thanks, Ben, and thanks, everyone, for joining us today. I want to underscore the great advantage of having Ben as our CEO. As a long-standing veteran of Bitfarms and the architect of our growth and profitability improvement programs, we are under his leadership, positioned to develop and act quickly in our initiatives. In such a fast-evolving environment, it is essential to keeping Bitfarms on the cutting edge of our industry. Now let's begin with an overview of our second quarter financials. Second quarter revenue of $42 million was down 16% quarter-over-quarter and up 17% year-over-year. The change was due primarily to the decrease in block awards following the April halving. During the quarter, we earned 614 Bitcoin, 35% fewer quarter-over-quarter, primarily the result of the halving and a 10% increase in average network difficulty. Mining revenue was $40 million compared to $49 million in the prior quarter. Gross mining profit was $21 million or 51% of mining revenue, down from $31 million or 64% last quarter and up from 48% in the prior year quarter. General and administrative expense, excluding noncash stock-based compensation and the sales tax refund was $13 million in comparison to $10 million in the first quarter. The $3 million increase pertained primarily to unusual costs associated with the strategic alternatives review process, our sponsor Riot Platforms hostile takeover bid in the shareholder rights plan as well as fees associated with the employment competition dispute with the former CEO. For the second quarter, our operating loss was $24 million, largely unchanged from the first quarter. The operating loss includes $57 million of depreciation expense on older miners compared to $39 million depreciation in the first quarter. Under our upgrade program, our existing miners are being depreciated on an accelerated basis over the remainder of their expected operating life as they replace more efficient miners. As such, a higher level of depreciation was expected in the first and second quarters of this year. Depreciation expense is projected to normalize from the third and fourth quarters as the miner replacement program was largely completed by the end of June. In the second quarter, financial expense includes a $1 million noncash expense for the revaluation of financial liabilities for warrants issued in earlier financings, compared to a $9 million noncash gain on the revaluation of its financial liability in the first quarter. Under IFRS, we are required to recognize the liability for these warrants even though they cannot and will never be settled for cash. Net loss for the second quarter was $27 million, or a loss of $0.07 per share compared to a net loss of $6 million or a loss of $0.02 per share in the first quarter. Now let's turn our attention to operating performance and per Bitcoin metrics. Our corporate cost of electricity for the quarter was $0.43 per kilowatt hour. That's an increase of $0.41 per kilowatt hour in the first quarter. Quarter-over-quarter, we benefited from the Canadian revenue agency ruling that allows us to recover the 50% VAT on our Canadian electricity purchases, which we calculated and reduced our overall electricity costs by about $0.04 per kilowatt hour. This savings was offset by higher electricity costs in Argentina as we shifted from lower summer to more expensive winter rates beginning in May and also a 3.2% increase in Canadian electricity rates effective April 1. Importantly, with our improvement in electrical efficiency from an average of 35 watts per terahash in the first quarter to 28 watts per terahash in the second quarter, our electricity cost per terahash decreased by 17% from [ $0.06 ] per terahash per day to $0.03 per terahash per day. Our direct mining cost per Bitcoin in the second quarter was $30,600. Our total cash cost to mine Bitcoin was $47,300 and our revenue per bitcoin was $65,800, result in cash profit for Bitcoin of $18,500. Turning now to Slide 14. For the second quarter, our adjusted EBITDA was $12 million or 28% of revenue compared to $23 million or 46% of revenue in the first quarter. The lower adjusted EBITDA largely reflected the impact of the halving along with higher G&A expense associated with the expansion of our operating activities. As we've noted in previous quarterly earnings calls, our adjusted EBITDA is very straightforward, being purely a measure of the cash profitability of our mining operations and the profit contribution of our Volta Électrique subsidiary. As an IFRS filer, we do not mark-to-market our Bitcoin Holdings, and we do not include this or any other balance sheet account valuation changes in our adjusted EBITDA. Stated simply, our adjusted EBITDA of $12 million in this quarter equates to cash profit per Bitcoin of $18,500, multiplied by the 614 Bitcoin we mined during the quarter, plus approximately $300,000 of profit from our Volta Électrique subsidiary. Turning now to Slide 15. At June 30, we had total liquidity of $195 million, consisting of cash of $139 million and Bitcoin valued at $57 million. As you've noted previously, our program to achieve our year-end 2024 targets of 21 exahash and 21 watts per terahash is totally funded. At July 31, we had 1,016 Bitcoin, up from 905 bitcoin at the end of June. Our higher Bitcoin treasury balance reflects our solid cash position and strong cash flow from operations. Further, our synthetic HODL continued to grow, increasing from 208 equivalent Bitcoin at the end of June to 333 bitcoin currently. As a reminder, under our synthetic HODL strategy, we utilized exahash bitcoin generated each month to fund our growth at a low cost of capital while maintaining upside potential by applying a portion of the proceeds towards the purchase of long-dated Bitcoin call options. In regards to our ATM facility, which we initiated in March and used solely to fund our growth initiatives and fleet upgrade, we raised $136 million for the second quarter. Since June 30, we have raised an additional $68 million under the facility. Moving on to Slide 16. I'll turn the call back over to Ben for a quick summary.

Turning to Slide 17. Before opening up the call for questions, I want to drive home a few key points. We continue to dramatically alter our operating profile via our ongoing fleet upgrades and our geographic expansion. This transformation will only accelerate as I work with the team to continually diversify our assets. We are taking a close look at all of our megawatts and evaluating several opportunities to expand beyond Bitcoin mining, including HPC and AI. We have industry-leading Bitcoin mined for exahash and industry-leading efficiency. We will continue to distinguish ourselves by improving our operational efficiency and growing our profitability in this highly competitive industry. I am very confident in our growth prospects and look forward to updating you as we, 1, continue to execute on our 2024 growth plan; and 2, continue to build out our team and expertise to ensure we are well positioned to capture additional market share, both within Bitcoin mining as well as within synergistic and additive business lines. With that, I will hand the call back to the operator for Q&A.

[Operator Instructions] Your first question is coming from [ Mike Colonies from H.C. Wainwright ].

So then for me, my first question and follow-up really around market trends here. You always have great insights. So I wanted to get your views on hash prices here, which just hit all-time lows last week at around $0.035 a terahash. Do you think we've reached the bottom here? And as a follow-up to that, how does the current market environment influence your decision and timing to execute on some of these non-mining-related opportunities over the near term, if at all?

Yes, it's a great question. Obviously, the pullback in Bitcoin price has an effect on hash price. And something that we've been talking about for a very, very long period is hash cost. Simply put, that's a very simple measure. We're combining the watts per terahash efficiency of the miners or our fleet, combined with the electricity cost that's powering it. And when you look at our $0.043 average electricity costs that we had in Q2 and you combine it with the 25 lots per terahash we entered the quarter, that gives you a direct cash cost around $0.027. So even with those pullbacks, we still remain cash flow positive in our operations on a direct mining basis. When we look towards how the market is responding, what we said for a long time is we think that the market starts responding in between $0.04 and $0.05, and that's exactly what we've seen happen. If you keep track of how the blocks are progressing and what that implies for difficulty changes at the end of the cycle, what you can see is that on Friday, prior to the major pullback, we were expecting a difficulty adjustment around negative 1%, and now we're trending significantly upwards. I think it's somewhere between 3% and 4%. This is happening halfway through the difficulty adjustment cycle, which means that to have that impact, it requires that much more hash rate in order to have a meaningful reduction. So yes, now we're between negative 3.8% and negative 4.7%. So the market is responding by either turning off miners or underclocking those miners. We think that with our fleet upgrade and our competitively priced electricity, we've positioned ourselves very, very well and have shielded ourselves from a lot of these events and short-term Bitcoin price is going to do what it's going to do, but long term, we believe that our growth and hash price is going to be well justified. And we're -- potentially, if the historical cycles play out, we're going to still see significantly higher high prices in the months to come.

And just as a follow-up there, does this influence or accelerate your decision to start to look at executing some of these non-mining-related opportunities in the pipeline, be it HPC AI?

When we look at the HPC and AI opportunity, we see a lot of potential here to achieve a high-value revenue stream. But just to be clear that the time lines for integrating and building out that infrastructure is at least 12 months away. Personally, I think that is actually a very advantageous time line because when you look at the value that you would get out of a T21 for instance, with current hash price at $0.04, it's roughly $80 a megawatt hour. But if hash price were to go up to $0.08, it would be $155 a megawatt hour. If we went up to $0.12, it would be $230 a megawatt hour, and if it went up to $0.16, it would be $310 a megawatt hour. So there's still a lot of potential for Bitcoin prices and Bitcoin mining economics and a Bitcoin bull run. And we don't necessarily want to rush and pay a premium to diversify away from that, but continuing with the same time lines of roughly 12-ish plus months really means that we might be able to integrate these diversified revenue streams towards what we would expect to be the top of the Bitcoin bull market cycle according to historical trends. And that time line is something that I think is very, very compelling.

Your next question is coming from Mike Grondahl from Northland Securities.

On the HPC AI strategy, what pieces do you still have to put in place there to execute on that? You talked about recruiting the team. Do you -- have you had discussions with hyperscalers? And kind of give us a sense of the time line and some of the pieces you still need to put together.

We are great builders and operators of electrical infrastructure. So when it comes to building out the facility and building out the substations and doing the interconnection work and doing the maintenance, we're fantastic at that. The area that where we need to bolster our bench here is the area of HPC and AI specifically because the data centers required for HPC are fundamentally different than the Bitcoin data centers that we are specialized at building and operating. So we do need to bring in more talent in order to help us properly evaluate and develop these opportunities, which we're actively recruiting for now, but this is more specifically regarding the data center engineering for HPC as well as the data center management. It's going to require different systems, different technology, different equipment, different software. And those are things that we either need to bring in-house or partner with a third party. We have had numerous discussions with both hyperscalers and third parties to work with on different projects. But we're still in the early days of evaluating that. And I think for us, that time line that most of the HBC in AI and hyperscalers are targeting kind of a Q4 '25, maybe Q1 2026, is ample time line for us to bring the proper expertise into their forms, integrate it into our team, evaluate the various partners to be working with and identify and start working towards the right deployment at the right locations for a target maybe at the end of next year or beginning of 2026.

And then just one more, with the 120 megawatts in Pennsylvania, do you have a goal, something aspirational, like you would like to get another 100, another 300 of committed megawatts in '25? In the U.S., how do we think about what you want to do next in the U.S?

Yes, it's a great question. When you look at our portfolio of energy assets, what you'll see is that right now, the U.S. is very underrepresented. We only have currently active and hashing the 20 megawatts that we have in Washington. But we do have, I think, the best developed international portfolio. What we'd like to do next year is rebalance that portfolio with a much greater emphasis on the United States, but we're not going to pursue growth opportunities that are too expensive, they don't make sense for us. We're very, very disciplined and selective in choosing growth opportunities that we think are long-term economically sustainable. And that means that we're not targeting a specific megawatt, what we're targeting is a specific profile, and that profile is a long-term competitive energy profile that is not just suitable for Bitcoin, but is suitable for multiple applications. And so I can't provide a specific megawatt number that we're targeting, but what we are looking to do is constantly integrate competitive energy prices that are going to help us manage our energy price and drive those costs lower across the entire portfolio.

Your next question is coming from Lucas Pipes from B. Riley.

This is [ Fedor Shabalin ] asking questions on behalf of Lucas Pipes. And my first one is on your Sharon site in Pennsylvania. So last week, PJM held an auction and prices in 2025, '26, were settled roughly 10x higher compared to 2023, 2024. So, the question is where are you right now in terms of power supply for this site and how are you going to negotiate it? Is it going to be PPA or something else? And what pricing range do you anticipate per megawatt hour here?

When we're looking at the Sharon site, one of the benefits there is that we do have, as we said in the call, we have multiple different power providers that we can work with at that site. We haven't locked in prices or a power provider yet, but what we do know is that because of the reduction of thermal generating assets in PJM and because of the increase in renewable assets, what that means is that there's a lot more opportunity around the energy trading specifically. The capacity auctions that you're referencing are referencing what is the value that these very reliable base loads are able to secure. And the reason why that has increased so much is because of the massive reduction in thermal assets that is taking place in replacement with those renewables. And so, we haven't locked in a fixed price, and it would be hard to guarantee exactly what that price will be because the prices are subject to a lot of change, but what we do know is that it increases the amount of opportunity around trading the energy.

And my follow-up is on HPC side. Just a little bit more detail here. But beyond what you mentioned in the release, you touched on cadence and things in that segment, but other players already building HPC capacity and some AI cloud. And what is your perspective on how you're going to use some of your capacity for this purpose, just potentially? Is it going to be HPC or AI cloud, you're going to purchase the GPUs for your own use or going to be just kind of mix of it? And where it will be possibly deployed additional to Sharon? And what's your plan? And the important one is what is you plan to finance this capital-intensive segment, how we going to -- use ATM or debt?

When you look at the different compute markets, very clearly, the capital requirements associated for HPC and AI are significantly different. When we look at what is our rough cost to build out 1 megawatt of compute power with T21, it's roughly $1 million a megawatt. Just to build out the infrastructure for HPC and AI, most estimates are putting it at $10 million a megawatt or more. When you include the value of the compute in that, it can easily go up to $30 million to $40 million a megawatt. We don't have any interest in buying the actual HPC and AI compute. What we have an interest in is developing, owning and operating the infrastructure that provides the power to the HBC and AI compute. And there are so many hyper-scalers and so many companies right now trying to deploy GPUs that we don't have to worry about how those -- we don't have to worry about the actual value of the compute hardware and who's going to pay for it because there are so many parties who want that space and who are valuing the infrastructure. Our core competencies and strengths is on building and operating energy infrastructure. And so that's where we're going to focus our efforts. When it comes to the financing question, I mean, this is still very much open to very different structures. But what I can say is that based on the numerous conversations we've had with different parties so far, there is no shortage of different financing opportunities and financing structures available to help finance the construction of these different infrastructures and these different opportunities.

Your next question is coming from Martin Toner from ATB Capital.

You mentioned not having broken ground in Sharon and that, that creates some flexibility for you. Can you talk to how long you're willing to wait before you need to start committing to Bitcoin-specific infrastructure development in Sharon and the possibility of that putting pressure on your 2024, 2025 exahash target?

Sure. So, there's 2 different components to Sharon. 1, there's an immediately available 12 megawatts, which we're going to be -- we're still finalizing the electrical plans for that and the site plan, but we'll be dropping in containers for that site to quickly deploy 12 megawatts by the end of the year. For the remaining capacity for that 120 megawatts, we have to build out the substation. We're expecting to break ground on that at the beginning of next year. And when we look at when we actually need to start building out and finalizing what we're doing there with the data center and the allocation of those megawatts, we really need to have that decision. We'll be working towards that decision within Q4 with hopefully, a final decision sometime at the end of Q4 or at the beginning of Q1 if we want to meet those time lines at the end of 2025.

Yes, putting words in your mouth, some of the initial build-out is the same regardless of HPC or Bitcoin mining?

Yes. For the first 12 megawatts, since there's already location, since there's already capacity there, it's a very, very quick build out. We are going to drop to containers, both for the convenience, the speed and also the flexibility that if we do want to convert those 12 megawatts later, they can be repurposed at a different location.

And can you talk about whether or not you have the building capability in Pennsylvania and the necessary materials like access to materials, to the substation, transformers, switches, et cetera, to be able to deliver that HPC compute in your stated time line, which starts at 12 months?

For HPC & AI compute specifically, no, we do not have on hand the necessary equipment for that. And the reason for that is because we haven't determined the final plan there. When we have determined the final plan for what we're going to do with the Sharon megawatt, that's when we'll go out and start procuring that equipment. So no, currently, we do not have the necessary equipment for an HPC and AI build in Sharon specifically.

Do you see that as being a potential bottleneck? Or are you guys confident that you have the relationships and access to necessary materials and equipment.

We have, at this point, over 12 months, 12 to 16 months if we wanted to get that energized by the end of the year, and I think the most realistic time line for HPC AI build is really Q4 or Q1, which extends that time line out to possibly 19 months. That gives us, I think, sufficient time line to go out and procure those equipment given our position in our existing industrial relationships with suppliers.

Just wondering on the -- I think we're all very, very curious on the pipeline of development opportunities. Wondering, what are the chances that there's another Sharon out there? Is it possible that it could be in a different U.S. ISO? And is it possible it could be somewhere in some other countries?

Yes, it's a great question. We do have a global view towards new opportunities and growth opportunities, again, with a focus on securing the right energy profile and securing something that we believe is going to be economically sustainable for many years to come. And certainly, there are opportunities all around the world. But what we're going to be focusing on over the 2025 and beyond, and even currently right now in our development pipeline, what we're currently focusing on is increasing our U.S. exposure. We think that there are tremendous opportunities in the United States, and we do really like the PJM region. It's got numerous benefits in terms of its curtailment programs, demand response programs, opportunities for hedging and energy trading. We think there are more opportunities like Sharon out there, and we would love to pursue more opportunities in PJM.

Is it fair to say the market has tightened since early June for power development opportunities?

There's always power available. Sites are constantly turning over. What I don't know if necessarily the market has tightened, but certainly, the demand and the interested parties for more power is increasing rapidly. So, I don't know if market conditions have necessarily tightened – [Technical Difficulty]

Your next question is coming from Joe Flynn from Compass Point Research & Trading.

It looks like you've built up a pretty strong balance sheet here, but maybe you could walk us through the remaining CapEx spend to get to the exahash targets in 2024 and 2025? And ultimately, how we should think about usage of the ATM going forward?

Sure. Joe I am always glad to answer that question here. So, let's first talk about where we are here, regarding the 21 exahash and 21 watts per terahash goals and targets that we have for the year-end here. So, we've already pointed out that we are fully funded for that, and let me give us some hard context to that. We pointed out that we had almost $200 million of liquidity and that our cash flow from operations over the next several months is projected to be around anywhere between $8 million to $12 million a month or roughly $50 million to $55 million here. In terms of what we have left to spend here for the year-end targets here, we've figured out we have roughly at this point in time, we need to spend about another, we see another $50 million that we have remaining here for the infrastructure build-out. That's primarily really with Yguazu in finishing that completion. And then secondly here, for the miners, we have remaining payments of roughly about $67 million from miners. We figure another $20 million to $25 million just for imports, duties and insurance, things of that sort. So overall, we're envisioning additional CapEx requirement for this year, roughly around $140 million, which, again, we're very well positioned here forward. Now what we are not reflecting here yet is what our plans are for next year in terms of our CapEx, and that's still unfolding, particularly I think move ahead with Sharon and other opportunities here. So, while we feel we're pretty well positioned at this point in time, that's something that we'll be addressing more in the future here. In reference to our ATM here, we've actually -- given the fact that our shares are now below $2, we've actually stepped back very, very dramatically in what we're doing here in the ATM because we're going to be very selective going forward here. We again have the benefit of a strong cash flow from operations to fund ourselves going forward here. And secondly, I think given the fact that we are now well positioned to be fully funded for this year and have actually a kitty for next year as well here, we are going to continue to be very judicious more so, I think, with the ATM over the next several months.

And maybe if you could just comment on the ending of the strategic alternatives process? Just really any color there would be helpful.

Sure. Happy to. The special committee, obviously in consultation with its financial and legal and strategic advisers did conduct a thorough review of all the different strategic alternatives in an effort to maximize shareholder value. Really following the completion of this process based on the new strategic plan that we're moving forward with, the special committee has unanimously determined that and really is certain that given all of the compelling opportunities ahead for value creation, the best path forward to maximize value for all shareholders is to just move forward with our standalone plan. Obviously, the Board and management team remain open to any and all opportunities that may deliver value to shareholders, but right now, by far, the best opportunity is moving forward with standalone.

Your next question is coming from Brett Knoblauch from Cantor Fitzgerald.

I guess, on Sharon and new site acquisitions, you mentioned that Sharon being close to a bunch of metropolitan areas is a good characteristic for the AI HPC. As you're looking in the PJM region, are you considering this characteristic for future site acquisitions?

Yes. It's one of many things that we're looking at now with the different energy assets with the corporate development team. We're not only looking for sites that are good for Bitcoin mining, we're looking at sites that will help us, as we said, diversify and expand beyond Bitcoin mining itself. So, this means we're looking for sites that have more than one good application. Sometimes a site may only could be good for 1 of the 2 applications. Maybe it's only good for Bitcoin mining. Maybe it's only good for HPC and AI, ideally, and it's much harder to find those opportunities, but ideally, you find an energy profile, that's good for both.

And then just on the customer front, I know this is a little bit always down the road, but do you have your idea on an ideal colocation customer on the AI and HPC front? I know there's talk of hyper-scalers, enterprises, AI start-ups who are a bit of a more risky play because you don't know the longevity of the hosting contract. So just your thoughts there.

Yes. It's a good question. One of the things that we like about Bitcoin Mining is that we have no customers, right? When we plug in Bitcoin and mining and compute the customers the network itself, and it becomes very, very easy for us to operate our business. When we look at HPC and AI, we don't want to get into the business of developing a sales force and building out a software platform and marketing out to individual customers. Our primary focus here is, again, on developing, building and operating the high-quality electrical infrastructure. And so our ideal customer would be a hyper-scaler probably or somebody with a very, very strong credit profile and strong balance sheet that we can work with to take all the capacity and only work with one customer and leave them to handle how they're going to sell off the compute or monetize the compute on their end.

Your next question is coming from Martin Toner from ATB Capital.

My last question, was the result of the PJM reserve auction that you were more confident about what your cost of mining would be in that region?

We're still working on locking in our power provider there in Sharon. So, the auction that just took place really is primarily for people who are providing a baseload capacity to the grid. That doesn't necessarily directly impact what we do here. What it means is that those thermal generating assets and those more stable baseline assets are increasingly more valuable. And that's due to the increase of amount of renewables on the grid and the increase in value associated with grid stabilization programs. So, it's not entirely clear the exact mass and the exact uptimes for these different sites, given the auction that took place, but what it is certain of is that these energy trading and grid stabilization opportunities have tremendous and increasing value in the light of a changing energy profile in PJM.

That concludes our Q&A session. I will now hand the conference back to Ben Gagnon for closing remarks.

Thanks. I just want to thank everyone really briefly for joining us on today's call and reiterate, I'm very excited about our growth opportunities and look forward to updating you on all future developments. Thank you very much.

Thank you, everyone. This concludes today's event. You may disconnect at this time, and have a wonderful day. Thank you for your participation.