Hut 8 Mining Corp

TSX:HUT

Good day. Thank you for standing by. Welcome to the Hut 8 First Quarter 2025 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your first speaker today, Sue Ennis, Head of Investor Relations. Please go ahead.

Good morning, and welcome to Hut 8's First Quarter 2025 Financial Results Conference Call. Joining us today are our CEO, Asher Genoot; and our CFO, Sean Glennan. Following the presentation, we will open the line for questions. This event is being recorded, and a transcript will be made available on our website. In addition to the press release issued earlier today, our full quarterly report on Form 10-Q is available at www.hut8.com, on our EDGAR profile at www.sec.gov and on our SEDAR+ profile at www.sedarplus.ca.

All figures discussed today are in U.S. dollars. Certain statements made during this call may constitute forward-looking statements within the meaning of applicable securities laws. These statements reflect current expectations and are subject to risks and uncertainties that could cause actual results to differ materially. Certain key risks are detailed in our Form 10-Q for the quarter ended March 31, 2025, our Form 10-K for the year ended December 31, 2024, and our other continuous disclosure documents. Except as required by law, we assume no obligation to update or revise any forward-looking statements.

During the call, management may reference non-GAAP measures such as adjusted EBITDA. We believe these metrics alongside GAAP results provide valuable insight into our performance. Reconciliations of GAAP and non-GAAP results are included in the tables accompanying today's press release, also available on our website.

With that, I'll turn the call over to our CEO, Asher Genoot.

Thanks, Sue, and good morning, everyone. Since our 2024 earnings call, we've made substantial progress against our 2025 roadmap. Last year was about restructuring the legacy Hut 8 business and setting the foundation for sustained long-term value creation. This year is about building on that foundation, investing in growth and advancing our evolution as an integrated energy infrastructure platform.

Against that backdrop, the first quarter marked a deliberate and necessary phase of investments designed to accelerate our development flywheel, unlock more capital-efficient growth and better position our platform to deliver sustained long-term value. With that, let's get started.

The significance of the first quarter is clearest in the context of our broader 2025 strategy. Today, I'll begin by outlining that strategic context, then review our results and what we believe to be a forward-looking drivers of value creation catalyzed by our work. Sean will follow with a detailed financial review.

In 2024, we fundamentally transformed the legacy Hut 8 business, optimizing operations, fortifying our capital strategy and developing a high-velocity utility scale power origination pipeline. In 2025, we are channeling that foundation and momentum into a new phase of growth and expansion.

Driving us forward is our development flywheel, a framework that aligns 4 drivers of value creation: origination, investment, monetization and optimization under a single power-first strategy.

The premise is simple. The more effectively we reinforce each driver, the faster we compound value. Our objective is to deliver returns faster and more efficiently than infrastructure development models constrained by traditional commercialization dynamics and capital cycles.

Origination and investment are the foundational drivers of our flywheel. And together, they shape the focus of the first quarter. Origination defines the scope and scale of our platform. In the first quarter, our focus was on maintaining the scale and velocity of our power origination pipeline.

As of March 31, 2025, it spanned approximately 10,800 megawatts with approximately 2,600 megawatts under exclusivity. Investment follows origination and transforms our pipeline into tangible revenue-generating assets.

Historically, our platform model required us to allocate capital across 3 businesses: Power, Digital Infrastructure and Compute, each with distinct risk profiles, return horizons and capital intensity. In practice, this often forced difficult trade-offs between power acquisition, data center build-out and mining expansion.

Breadth came at the expense of depth, limiting the velocity of our flywheel. This challenge isn't unique to Hut 8. Across the sector, operators face a structural choice, focused exclusively on mining or diversify into the broader digital infrastructure space. Few who pursue the latter do so with true strategic coherence.

We believe our power-first approach to digital infrastructure development provides a structural advantage in navigating that complexity. Our business is built around long-term access to high-quality energy assets and our approach to digital infrastructure development is application agnostic.

Developing data centers for Bitcoin mining ASIC compute allows us to scale our power layer aggressively while preserving the flexibility to potentially transition assets to other high-value use cases in the future.

Within this architecture, ASIC compute for Bitcoin mining introduced structural tension. Each dollar allocated to fleet expansion had to be weighed not only against expected returns, but also against its impact on our broader platforms, capital structure and strategic positioning, especially given the distinct investor expectations and capital demands tied to mining.

To strengthen the compounding effect of our flywheel, we sought to decouple investments in ASIC compute from our capital allocation framework with the aim of creating a dedicated vehicle that could scale independently without diverting capital from our core power and digital infrastructure businesses. That imperative led to the creation of American Bitcoin.

Now let's examine this transaction across 3 dimensions: the actions of the past that made it possible, its present date impact on our first quarter results, and the future value we believe it will unlock. First, I'll review the strategic arc that led to the launch of American Bitcoin.

From the outset, our objective was threefold: to create a structure that preserved our ability to monetize power assets through Bitcoin mining, maintain long-term capital exposure to potential Bitcoin upside for our shareholders and resolve the capital allocation constraints embedded in our integrated platform model.

Since I stepped into the role of CEO last year, we have methodically laid the groundwork for a purpose-built standalone entity capable of fulfilling that mandate. We began preparing for the carve-out even before capital partners like Coatue came on board, driven by our conviction that restructuring the platform was critical to unlocking long-term scalability and capital efficiency.

We restructured the business, instilling technology-driven operating rigor and improving unit economics. We built a utility scale power origination pipeline, unlocking access to near-term power at scale. We forged deep relationships across the mining value chain, including a partnership with BITMAIN to commercialize a next-generation ASIC miner.

We designed proprietary direct-to-chip liquid cooled ASIC compute infrastructure, which we believe will enable market-leading ASIC computing density and efficiency. And this quarter, we executed a fleet upgrade, increasing our deployed hashrate to 9.3 exahash at an average fleet efficiency of approximately 20 joules per terahash as of March 31, 2025.

Together, these initiatives form the foundation of what we believe is a durable structural competitive advantage for American Bitcoin. The business is purpose-built to accumulate Bitcoin at scale with exceptional speed, capital efficiency and operational leverage. As Sean will detail shortly, the resulting structure also embeds a dedicated anchor tenant into our broader platform.

Now let's turn to the present. The impact of the initiatives we executed in the first quarter is reflected in our results. I'll share the highlights now, and Sean will walk through the numbers in detail. Note that our results for the comparison period have been restated under our new reporting structure.

This quarter, our performance reflects both our strategic investments during the period and the broader macroeconomic forces that shaped the Bitcoin mining sector over the past year. Revenue for the quarter was $21.8 million, down from $51.7 million in the prior year.

This decline was driven primarily by 2 factors. First was planned downtime associated with our fleet upgrade, which not only involved the installation of higher efficiency machines, but also targeted infrastructure upgrades at our sites to support higher rack level power density. Second was top line pressure from the April 2024 halving and resulting increase in network difficulty.

These dynamics contributed to a net loss of $134.3 million for the quarter versus net income of $250.7 million in the prior year and adjusted EBITDA of negative $117.7 million, down from $297 million. The majority of this variance stems from $112.4 million non-cash loss on digital assets under the new FASB fair value accounting rules.

During the first quarter, the price of Bitcoin declined from approximately $93,000 as of December 31, 2024, to $82,500 as of March 31, 2025, triggering a mark-to-market adjustment that impacted reported results. As expected, this phase of investment also introduced transitional cost pressure.

Our energy cost per megawatt hour rose to $51.71 from $40.06 in the prior year, driven primarily by fixed transmission and distribution charges at our sites, where we incur monthly costs regardless of consumption.

These fixed charges had an outsized impact during the quarter due to planned downtime from our fleet upgrade. With reduced consumption, these costs were spread over a smaller base, inflating our average cost. Normal operating periods, when consumption scales with uptime, these costs are amortized over a larger base, reducing per unit cost. While this dynamic led to a meaningful increase, we remain confident in our ability to maintain a track record of highly competitive energy costs.

With the upgrade now complete and our new ASIC fleet fully deployed, we expect a step change improvement in mining economics across American Bitcoin's footprint beginning in the second quarter.

Finally, while our results reflect the near-term impact of planned investment and external headwinds, they also highlight the strength of our balance sheet and discipline behind our capital strategy.

We ended the quarter with 10,264 Bitcoin held in reserve, representing $847.2 million in market value as of March 31, 2025, and remain capitalized to support continued growth. That strength reflects a consistent focus on capital efficiency and thoughtful structuring.

A clear example is the Bitcoin-based agreement we structured with BITMAIN to support our fleet upgrade. Under the terms, we pledged approximately $100 million in Bitcoin at a strike price of approximately $104,000, slightly above the spot price of approximately $101,000 at the time.

If 3 months after the shipment, Bitcoin is trading above that price, we reclaim the Bitcoin and pay cash. If not, we simply walk away with no further obligation. In effect, we secured new hardware while retaining a no-cost call option on Bitcoin upside.

Stepping back, this quarter represents the first in building on the foundation we established in 2024. The investments we made were deliberate, and we believe they will begin compounding meaningfully in the quarters ahead. We executed our fleet upgrade, launched American Bitcoin and advanced key infrastructure projects, including the continued development of our Vega data center and initial site work at our River Bend site.

In total, we invested $63.3 million in property and equipment we believe will drive sustained margin expansion, capital productivity and platform yield. Our fleet upgrade delivered a 79% increase in deployed hashrate and a 37% improvement in fleet efficiency quarter-over-quarter.

Despite a roughly 4-week delay due to customs clearance, strong coordination between our commercial and operation teams enabled us to accelerate deployment once the hardware cleared ports and complete the upgrade efficiently.

During the quarter, we also energized a test rack module at Salt Creek in preparation for the energization of our next-generation ASIC compute architecture at Vega. Vega is a key strategic asset due to its technical architecture, which we believe reflects where the broader industry is headed.

Our proprietary rack-based direct-to-chip liquid cooling system for ASIC compute is designed to meet the demands of denser, more power-intensive compute. The proprietary system, including pump skids, fluid distribution networks, server racks, switchboards and smart power distribution units was designed in-house to optimize thermal efficiency, minor stability and operational reliability.

This system design is expected to enable materially higher compute density, greater thermal control and improved uptime in high ambient environments like Texas, where air-cooled ASICs are prone to thermal throttling and failure.

Early results show clear performance advantages from liquid-to-chip cooling over traditional air systems, particularly during peak temperature periods. As we energize and scale the site, we expect these gains to translate into higher realized hashrate and lower failure rates.

To support this next-generation infrastructure, we took targeted steps to build durable technical leverage and drive long-term returns. We established site level operating infrastructure at Vega, onboarded dedicated management for the site and expanded our team with professionals from leading hyperscale operators like Microsoft to support enterprise-grade reliability and uptime.

In parallel, we developed new software tools within reactor and operator built specifically to optimize energy consumption at Vega and automate ASIC level operations across our platform.

Put simply, the first quarter was a deliberate and necessary phase of investment. We believe the returns on this work will become increasingly visible in the coming quarters.

With that said, let's look to the future. The launch of American Bitcoin marks a pivotal shift in our platform trajectory, accelerating our transition to power and digital infrastructure. The streamlined capital allocation framework made possible by the carve-out of our Bitcoin mining business reinforces our ability to scale lower cost of capital businesses such as high-performance computing. Because American Bitcoin remains strategically integrated in our platform as a dedicated anchor client, we also retain our ability to monetize power assets rapidly through mining.

The carve-out not only streamlines our capital allocation framework, but also drives cash flow predictability, positioning us to compound value through more predictable contracted sources of revenue. At the same time, our retained ownership in American Bitcoin preserves exposure to Bitcoin without the balance sheet burden of further parent level investments in mining hardware.

We believe this structure offers a distinct advantage across both the mining and broader digital infrastructure sectors, enabling us to deliver a more robust, diversified value proposition to our shareholders.

In summary, the first quarter was a deliberate and necessary phase of investment designed to unlock the potential of our development flywheel and reposition our platform for faster, more efficient value creation.

With the launch of American Bitcoin, we streamlined our mandate, refined our capital allocation framework and cemented our focus on power and digital infrastructure.

While projects in these sectors take time to commercialize, we believe demand remains strong as we continue to make progress with creditworthy partners. We're executing alongside proven leaders in data center design, construction and operations, and we're increasingly confident in our ability to play a differentiated role in this evolving market.

That conviction is already taking shape across key projects on the ground. Development at Vega continues to advance on track for energization in the second quarter. During the quarter, we secured 592 acres of land in Louisiana for our River Bend data center campus, which is being developed with the aim of supporting a utility scale data center campus for high-performance computing.

Initial site work is now underway, including clearing and grubbing across 75 acres for the switchyard, substation, laydown yards and utility corridors. In parallel, we've advanced 2 additional AI data center projects, which if secured, would add over 230 megawatts of IT load capacity to our platform.

We're still in the early innings of this transition, but we believe the path forward is clear and the logic behind our platform has never been stronger. We look forward to updating you next quarter as we continue to execute on our 2025 roadmap and as we build an enduring generational business at the intersection of energy and technology.

With that, I'll turn it over to Sean.

Thanks, Asher, and good morning, everyone. I'll start by reiterating that the first quarter was a focused phase of investment designed to accelerate Hut 8's evolution as an integrated energy infrastructure platform and shift our revenue base towards more predictable and financeable lower cost of capital segments.

That strategic objective shapes the work we did this quarter, most notably the structuring and launch of American Bitcoin. My remarks today will follow the logic of that transaction.

I'll begin with the long-term economics and structural value unlocked by the commercial framework between Hut 8 and American Bitcoin. From there, I'll review our first quarter results by segment, including the cost dynamics introduced by this period of investment. Finally, I'll close with an update on our capital planning and investment priorities for the remainder of the year.

Let's begin with the long-term value creation embedded in our agreements with American Bitcoin. Going forward, all Bitcoin mining operations previously reported under Hut 8's Compute segment will operate through American Bitcoin and will be consolidated under Hut 8's existing reporting structure.

While our objective is to execute a go-public transaction for American Bitcoin and our ownership position is expected to be diluted over time, American Bitcoin's economics will, for now, be consolidated under Hut 8's Compute segment given our control position.

To enable this transition and enhance long-term revenue visibility for Hut 8, we've implemented a commercial framework across 3 agreements that convert the cyclical economics of our legacy Bitcoin mining operations into stable contracted revenue streams within our Power and Digital Infrastructure segments.

An additional benefit from having American Bitcoin as an anchor tenant is that it reinforces our development flywheel as we now have a dedicated utility scale offtaker to accelerate the monetization cycle of our robust powered land pipeline. American Bitcoin also benefits from a dedicated partner providing access to a pipeline of new sites to support its growth.

Let's discuss each aforementioned agreement in turn. First is the colocation agreement, which gives Hut 8 the exclusive right to host American Bitcoin's ASIC miners at its sites. This agreement generates recurring fiat-based revenue for Hut 8's Digital Infrastructure segment under our ASIC colocation business line. From an economic perspective, we have structured the colocation agreement to achieve a payback roughly equivalent to the depreciation cycle of the miners that they host.

Second is the managed services agreement, which you can think of as an asset management and O&M agreement. Structurally, this comprises a fixed price component along with the pass-through of certain operational costs. This creates a second recurring fiat-based revenue stream within Hut 8's Power segment under managed services, while allowing American Bitcoin to leverage Hut 8's proven mining operations platform.

And third is the shared services agreement through which Hut 8 will manage American Bitcoin's core business and back-office functions. This agreement is designed to minimize the burden of building and maintaining business functions in-house allowing American Bitcoin to scale with a lean SG&A-light cost structure. We believe this will make American Bitcoin one of the most efficient miners in the industry.

Now let's turn to our segment results. We begin with Power. Segment revenue declined from $9.9 million to $4.4 million year-over-year, driven primarily by an $8.2 million reduction in managed services revenue following the termination of our managed services agreement with Ionic Digital. This was offset partially by a $2.7 million increase in power generation revenue from our natural gas power plants in Ontario, which we own and operate through a joint venture with Macquarie.

In our Digital Infrastructure segment, revenue declined from $5.8 million to $1.3 million, driven primarily by the termination of our ASIC colocation agreement with Ionic Digital. While this termination impacted top line during the quarter, it released critical rack capacity at our sites, which were reallocated to support our fleet upgrade ahead of the launch of American Bitcoin.

Prior to the launch of American Bitcoin, we expected our Digital Infrastructure segment to reflect revenue from our ASIC colocation agreement at our Vega site with BITMAIN starting in the second quarter. However, with the launch of American Bitcoin, we now expect the following to happen. If we exercise our purchase option with BITMAIN for the 15 exahash of machines that are heading to Vega, that ASIC colocation revenue would instead be recognized through our colocation agreement with American Bitcoin.

Finally, we turn to our Compute segment, which bore the greatest impact from both our investments and sector-wide headwinds. Segment revenue declined from $32.1 million to $16.1 million year-over-year, reflecting 2 primary drivers: planned downtime related to our fleet upgrade and top line pressure from the April 2024 halving, which drove higher network difficulty and lower hashprice. These dynamics were both anticipated and necessary. The downtime enabled us to complete infrastructure upgrades and our fleet upgrade.

We ended the quarter with 9.3 exahash of deployed hashrate at an efficiency of approximately 20 joules per terahash, representing a 79% increase in hashrate and a 37% improvement in efficiency quarter-over-quarter. These gains lay the foundation for a step change in top line performance beginning in the second quarter.

As Asher noted earlier, the launch of American Bitcoin introduced a set of onetime investments and setup costs required to execute the carve-out and establish the foundation for American Bitcoin's ambitious growth roadmap.

General and administrative expenses were $21.1 million versus $20 million in the prior year. The net increase of $1.1 million reflects approximately $5.8 million in strategic investments, including a $3.4 million increase in salary and benefits related to headcount expansion, a $1.3 million increase in transaction costs tied directly to the carve-out and a $1.1 million increase in legal and tax advisory fees. These increases were partially offset by reductions in non-core and non-cash expenses, including restructuring charges, acquisition-related costs and stock-based compensation.

While the net impact on SG&A was modest, these investments and fees enabled the carve-out of our mining business, positioned American Bitcoin for capital-efficient scale and realigned our broader platform around contracted fiat-based revenue streams. Moreover, as a former banker, I can say that we managed to keep fees well below what I've seen often in my career.

I'll turn now to an update on our capital planning and investment priorities across the broader Hut 8 platform. Near-term CapEx will primarily fund the completion of our Vega site and advance the development of our River Bend campus. At Vega, we have deployed more than 70% of planned capital as of the end of the period. We remain on track to complete the final phase of investment and construction in the second quarter, targeting an all-in cost of approximately $400,000 per megawatt or approximately $80 million in total across 205 megawatts.

As American Bitcoin scales, we expect Vega to serve as a replicable blueprint for rapid, capital-efficient build-out of next-generation Bitcoin mining infrastructure. At River Bend, our strategy reflects the disciplined capital allocation philosophy. To date, we have deployed only a de minimis portion of the project's expected total cost or around 2%, sufficient to maintain execution and power readiness while avoiding significant speculative exposure.

Certain opportunities we are underwriting today follow a yield on cost model with pass-through economics, which mitigates cost risk. Under this model, we expect our capital exposure to remain a fraction of total system cost even at full build-out.

As we think about ultimately funding some of these larger-scale data center projects, we believe we are well capitalized to execute on our strategy and ended the quarter with a strong liquidity position. From inception to quarter end, we have raised $275.5 million in net proceeds from our ATM program, selling 9.8 million shares at a weighted average price of $28.23.

Our ability to move decisively on high-return growth opportunities is further underpinned by 10,264 Bitcoin held in reserve, all of which will remain on the parent company's balance sheet and were valued at $847.2 million at quarter end. With ABTC now focused on becoming a Bitcoin accumulation vehicle, the Bitcoin on Hut 8's balance sheet is available to be deployed into high-growth projects.

As always, we will continue to balance managing enterprise risk with shareholder dilution. Fortunately, few asset classes in recent history have attracted capital at the pace and scale seen in the data center sector. As such, there is a plethora of funding mechanisms at the project level.

On the equity side of the project level, there is a range of options from third-party equity, preferred equity and operational JVs. And on the debt side, the project finance market remains robust and healthy. In total, these options allow us to fine-tune capitalization to drive the highest risk-adjusted returns for our shareholders.

I'll close by revisiting the foundation Asher laid at the start of this call. Last year was about restructuring the legacy Hut 8 business and setting the foundation for sustained long-term value creation. This year is about building, investing in growth and advancing our evolution as an integrated energy infrastructure platform.

The first quarter reflected that shift. It was a deliberate and necessary phase of investment designed to accelerate our development flywheel, unlock more capital-efficient growth and better position our platform to deliver sustained long-term value. We believe the benefits of that work are already beginning to materialize and that their impacts will compound in the quarters ahead.

With that, operator, please open the line for Q&A.

[Operator Instructions] Our first question comes from the line of George Sutton of Craig-Hallum.

Quite a bit of information on the call. So I'm just curious if you could walk through the River Bend message that you had. I think you were suggesting that this would be a yield on cost transaction, but I'm not sure if you were specifying that specific project. But any update on sort of the work that you've started to do at River Bend would be helpful.

George, Asher here. Happy to jump in and go through that. We did not clarify the type of structure yield on cost. But across the different projects that we have, some of the projects are more triangulated around yield on cost model and other projects are a fixed rate per kilowatt month. So depending on the project, the structure is different, but we haven't shared what we're doing at River Bend yet.

In regards to River Bend, as we shared in the last earnings, we closed on the project. it's 592 acres of land in Louisiana at the River Bend campus. We've already started investing in initial site work. And so that includes kind of civil work, initial substation work. So that's about 75 acres of the land and being able to hand over that switchyard to Entergy so they can start building out or hand over that land to Entergy so they can start building out their switchyard while we're building out the substation in parallel with it.

And so for us, as we think about investment projects on data center builds in areas like River Bend where it's not a primary data center market, we will invest in some of the initial work, which is a nominal percentage of the overall budget of the data center to make sure it continues to be on track while we're working on the overall data center build, wherein other sites where they're in more primary data center markets, we may invest a bit more capital in some of the data center critical components. But we're thoughtful on making sure that the capital that we deploy that it's recoupable in all scenarios and with all kind of customer profiles.

And just one other thing on the colocation agreement with American Bitcoin. You mentioned the payback would be roughly equivalent to the depreciation of the miners. Can you just kind of walk through the logic of that agreement?

The way that we think about our economic interest with American Bitcoin and just our overall exposure and value creation from the Hut 8 perspective is threefold. The first is through the agreements that we have and the revenues we generate through that agreement. So that is the colocation agreement in which they are a colocation tenant. They place their ASICs in our data centers, and we operate -- we allow them to use the data centers in order to operate their ASICs.

The second is a managed services agreement in which we're the service operators to actually operate those data centers on behalf of them. And the third is a shared services agreement where we essentially split SG&A, think kind of supporting functions such as legal, HR, accounting and so forth, and we're able to share those talents and those resources with American Bitcoin to fuel that business, and we do that one at cost and that pass-through. And so that's kind of the first layer.

We look at the returns of colocation relative to some of the deals that we're looking at actually on the traditional data center side and account for, obviously, the risk of the market, the halving cycles and so forth. And we'll share more details on the exact economics between those 2 as those numbers get reflected in our Q2 results.

The second area of value creation for Hut 8 is obviously our equity ownership in American Bitcoin. We're large shareholders in our company. We believe that the company has a ton of upside and has extreme value that can be created both in being a Bitcoin mining leader now not having to be restrained and can really fuel growth, but also in being a Bitcoin accumulator. So we're very excited by our ownership and our equity stake within the entity.

And then lastly is the strategic alliance here. Something we've always shared over the course of the last couple of quarters is that Bitcoin mining, we've always seen as continuing to be strategic to our overall energy infrastructure platform.

And the main reason for that is when competitors of ours that are just data center operators cannot go and acquire a land because they're not sure if they can bring a customer to an area that is more rural, we have the ability to place a Bitcoin mining facility there as a consumer of last resort, and we can consume and monetize those assets.

And over the long term, if the power dynamics change, demand profile change, American Bitcoin can say, "You know what, why don't we move these ASICs to another facility that may have more competitive power since more data center demand is coming here?" And at Hut 8, we can convert that to data centers or to other high-intensity use cases as well.

So there's a kind of cash flow financial benefit through these agreements between the 2 parties. There is an equity interest in the value of that equity as we believe it will appreciate. And then the third is a strategic alignment between the 2 companies as well. And so that's why I believe that American Bitcoin is extremely well positioned, because without those 3 prongs, it's really hard for this type of structure to work. But because there's such a synergistic alignment between the 2 entities, really 1 plus 1 here is greater than 2.

Our next question comes from the line of Paul Golding of Macquarie.

Asher, I wanted to ask about the Vega direct-to-chip liquid cooling proprietary architecture for greater rack density. How is that being considered with hyperscalers in mind and in the context of potential use cases for GPUs for HPC and AI?

And as a follow-on to that, to what extent is this proprietary architecture or maybe not in current terms, but maybe how do you foresee this proprietary architecture potentially factoring into discussions with hyperscaler tenant prospects as you further build out Vega and that architecture?

Thanks, Paul. Appreciate the question there. When we look at the infrastructure that we're building just from the lens of Bitcoin, you could argue, "Okay, if I'm spending a little bit more on the infrastructure, is it actually worth it? Would I rather spend as least of amount as possible in initial CapEx and just go with a [ 4 server ] facility even if we see efficiency increase of those chips.

So the overall kind of decision, I think, is twofold. One, from a Bitcoin mining perspective, we think that these chips will allow us to have longer lifetime and longevity with the actual ASICs that we run and operate them at a much higher throughput regardless of environment and weather conditions. But you hit on an area that we're extremely passionate about, and it's a bigger story and the bigger reason why we're so focused and dedicated to innovating on this infrastructure stack.

When we have operated traditionally in the Bitcoin space, our mindset has always been around how are we the lowest cost operator in the industry above all us compared to everyone else. And that's been the focus and it's been kind of our ruthless priority since day 1 of starting U.S. Bitcoin, which became Hut 8.

As we look at the traditional data center market and as we're negotiating these customer contracts and conversations and site diligence, in some of these yield on cost models, at the end of the day, I mean, you're operating more as a regulated utility mindset than actually as kind of a merchant operator. And what I mean by that is you're in the business of deploying capital and receiving a yield on that capital, not necessarily in the business of driving down costs as much as you can to increase the yield that you receive.

And so I think for us, long term, as we think about what is Hut 8's competitive advantage as we build? We execute on a data center contract. We deliver that to the market. Great. Long term, what are we as a business and where is our competitive moat above and beyond just monetizing some of the tower assets that we have? Are we competing on cost of capital with some of these other long-term yield vehicles, permanent capital, private equity firms and such?

Or are we competing on innovation and are we competing on driving a product offering on the infrastructure stack that we believe can give customers what they need for the compute that they're running at a lower and more efficient cost. And that's where we've won in the past, and that's an area that we don't want to give up on. It is the philosophy that drives through our business.

And so as a result, when we think about the liquid cooling solution at Vega, I mean that site is being built from greenfield to server racks cooling systems, not including the actual ASICs themselves, but everything else for a little above $400,000 a megawatt far cheaper than the $10 million to $13 million per megawatt that people are building data centers at.

The big kind of structural difference is you don't have a generator backup, right? You don't have UPS systems. But even if you added a distributed redundant loop on the switchgears, transformers, I mean, you're talking an extra $50,000 to $100,000 on that piece.

And so as we look at where we believe the industry evolves long term, and I think we've gone through a similar chip efficiency cycle in the Bitcoin side of the business where we went from year-over-year efficiency increases on the chips that made older generation chips harder and harder to operate.

So the way you had to operate it was either by decreasing your initial CapEx, driving down energy costs, building efficiencies through software. And as a result, a lot of the machines that we're selling that are older generation machines as we did the fleet upgrade are being shipped to places like Africa or South America, where they have low CapEx, low energy rates because of hydro and they're able to operate.

And so when we fast forward 3 to 5 years, the value of building out these facilities that don't have these diesel backups, but can power, can cool down these systems is that you can get them up quicker, right? But in 3 to 5 years, what's the value add? The value add is when you have tens of billions of dollars being deployed into these GPUs today and then every 6 to 12 months, you have a new generation chip, does it still make sense with where the chips are depreciated and where their market value is to put them in a data center that was built across $10 million to $13 million a megawatt?

Or can you put that into a lower resilience and redundancy data center where it's raw compute that we're running. And so we've had early conversations around what is the meaning and value of Vega. And when we talk about the cost that we built, people don't believe us.

And so a big part of us really building and innovating and making -- say, let's do a U design, so it's a data center rack so we can actually interchange with the GPU is as the site comes online, people can torque and see what we've been able to build at that cost. And even if we add more functionality into the infrastructure, we're still able to be far cheaper than the market cost for a data center.

And so V2 f our design, which we've already been working on innovating and we'll be deploying on the next project, actually allows for room in the spacing where we can add in things like additional distributor redundant, even generators and some designs as well.

And so I think, Paul, we're very, very excited for this innovation and being able to kind of share this. We've spoken to end users on the hyperscale side about this. To be honest, for some of them, we're a little too early for them, and you really need senior management to get behind this. And right now, they're just focused on deploying as many GPUs as possible.

But with others, there's real interest and we've gone through real conversations. But also with other data center operators. As we've looked at commercializing data centers, some of them we say, "You know what, we can do on our own." Others, we say, "Let's JV together if you can help us decrease execution risk."

And as we've built some of those relationships, they've also been surprised by what we've been able to do at Vega and have said, "Look, we have our typical offering for five 9 redundancy Tier 3 data center, but we also would like to talk about a different offering that we can give to customers as well.

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

I was hoping we could get a little bit more commentary around HPC. So my first question is around HPC customer conversations and also the JV -- potential JV conversations. Any milestones we should be able to look out for the next couple of quarters? And then I have a follow-up.

Thanks, John. River Bend is the campus that we've been speaking about. We obviously started moving dirt there and starting investing as we shared across River Bend, Vega and some of the other projects, we spent about $63.3 million in these CapEx investments. And so a lot of that was on the Vega project and the infrastructure that we're building. But nevertheless, in the substation switchgear equipment, civil work as well.

We're working on securing and announcing a customer contract and a definitive agreement. As we've shared with the market before, we have not and chosen as a company philosophically to not share LOIs. We believe that LOIs are not definitive agreements and thus should not kind of get the market excited for something that isn't finished nor are negotiating leverage in these conversations.

We've seen an increased amount of appetite. We've been focused on really driving the discussions to bring a customer to finality on these sites, but have increasing amounts of interest across Louisiana and other sites as well. And I think that goes to a testament of the team's ability to build these relationships with these counterparties and being an active dialogue around what we're building, what we have.

And so I think we feel comfortable that if we step away from current discussions, there are plenty of interest based on kind of where Louisiana is today. And then some of our other projects as well as we shared, we have another 230 megawatts of IT projects that we're in discussion with as well and at different stages. And in some of these projects the energy is sooner than the other projects. We've kind of gone through some of the definitive agreement negotiations and so forth.

And so I would say, John, I think we have less kind of small milestones and much more focused on kind of bigger announcements when these deals actually come into fruition. But we continue to remain excited, focused and ruthlessly executing on these initial projects and bringing them across the finish line because we do believe that they're great campuses for the customers that come here.

That's helpful. And that kind of led into my follow-up question on that additional 230 megawatts of critical IT capacity. And apologies if I missed this, but do we have like a location or earliest when that power would be available?

Both of those sites are near-term power similar to the Louisiana campus within the next 18 months, and that's why we focus on those to commercialize relative to our broader energy pipeline. We shared we have 2.8 gigawatts of power under exclusivity today. And the reason we're focusing on these is we believe that they're interesting markets from an overall power fiber story, land story, regional location.

Out of those, one is in a proven kind of Tier 1 data center market. And then the other one is kind of similar to Louisiana, where it's a little bit outside of the data center, but really well on kind of the worst located relative to other data center markets and kind of the fiber pass that runs across.

Our next question comes from the line of Patrick Moley of Piper Sandler.

So the power under diligence and power under exclusivity declined this quarter for the first time, I think, since you started breaking out. So could you talk about just the overall decline? Is that indicative of you just completing due diligence and taking it out of the pipeline? Or could it be competitive pressures where other folks are getting to these sites before you can get to them? Any color there on the decline would be great.

Thanks, Patrick. A bit of focus coming from my direction, 10.8 gigawatts is a lot of power, and there are a lot of sites. And so when we think about the process in which we develop data centers, we have an origination team that falls into our corporate development team, which we're rapidly scaling. And it's funny because, [ James Robles ], I think this kind of business we're in is similar to renewables 1.5 decades ago, and his advice was scale, scale, scale and capture, capture, capture. So we've been doing so.

But even with that scale and increasing our energy pipeline substantially in what we've disclosed quarter-over-quarter, it's really important for us to stay focused on the projects that we think have the highest likelihood and chance of converting. And I think where we are today, we have a much faster ability to vet projects with customers than we did a quarter or 2 quarters or 3 quarters ago. We're able to directly now text or shoot a quick e-mail to a portfolio of customers and say, "What do you think about this site? Is the site interesting, tell us why or why not?"

And so as a result, it allows us to stay a lot more focused on the sites that we believe can actually get to the finish line and are able to deploy capital. And the reality is, as we think about our development pipeline and the amount of CapEx dollars needed to actually bring those to fruition, we're also sensitive to what can we execute by ourselves? What can we execute with capital partners? And Sean has been looking at that as well into actually having special purpose vehicles to execute projects as we continue to scale.

But the decrease in the pipeline in our minds is primarily around our focus and our ability to say, if we don't think it's a perfect site and a really competitive site, we're willing to walk away rather than just show the market a larger and larger pipeline. And so that's been kind of the drive. It's more a filtering of projects that we don't want the team spending time on and redirecting their areas of focus.

And then as a follow-up, Sean mentioned in his prepared remarks that the Bitcoin on the balance sheet would be available to be deployed or invested in projects as you see fit. Could you maybe talk about just the overall HODL strategy? Is there a level of Bitcoin longer term that you'd like to hold on the balance sheet? Any color there would be helpful.

The big strategic mindset in forming American Bitcoin was that, that equity ownership would become Hut 8's ownership and exposure into Bitcoin and Bitcoin upside and optionality. As American Bitcoin builds up its own strategic reserve at Hut 8, we see now that, that balance sheet becomes investable capital.

Sean mentioned in his prepared remarks that we had raised around $275 million through our ATM at an average price of $28.23 net proceeds. And so for us, that's important because where we're trading now relative to the projects that we're working on to our balance sheet, we're not interested in raising capital and equity dollars, and we haven't over the last couple of months. And so as a result, as opportunities arise and if we need cash to fund those opportunities, we will look at that Bitcoin stack that we have, whether it be through some revolvers and being able to pull capital against the Bitcoin or selling the Bitcoin itself and investing in it.

And we'll also be creative in strategies that we deploy. As we shared, when we purchased the BITMAIN upgrade is about $100 million rather than just using cash and letting it out the door, we actually structured essentially a free call option on Bitcoin optionality and upside. But the whole idea behind American Bitcoin and Hut 8 is Hut 8 over time will drive down its volatility and become an energy infrastructure platform, and we hope with a lower vol, we'll be able to decrease our cost of capital.

And on Hut 8 and on American Bitcoin side, we are building a business that gives people the best exposure to Bitcoin and Bitcoin accumulation. And as we've seen with companies like MicroStrategy in '21 with the Cantor SPAC, that business not only can monetize its volatility by being able to gain access to cheaper capital if people want to trade around it. But with the addition of a leading Bitcoin miner, we're able to not only buy Bitcoin at the market price, but mine it at cheaper than the market and at a discount.

Our next question comes from the line of Brett Knoblauch of Cantor Fitzgerald.

Maybe one for Sean. The Coinbase credit facility, I believe, matures next month. Is that something that you guys would look to repay or extend?

Yes. So it's a good question, and thanks for joining today, Brett. Yes, we have that maturity coming up in the middle of June. And I think that's something we're discussing with Coinbase at the moment. And I think there's some optionality to extend that potentially on better terms. If you look at some of the refinancings or financings done by CleanSpark and Riot in particular, those have been done at pretty attractive cost of capital.

So as we think about ultimately lowering our overall cost of capital of the company I think some of potentially extending that is something that is very attractive. And I think we'll look to do that on the best terms that we can, not necessarily just at the terms that we have at present.

And then maybe a question for Asher. Just on timing of the carve-out, I feel like the past couple of weeks or months, we've seen kind of market sentiment for these type of businesses or anything with, call it, meaningful Bitcoin ownership improve. Has that maybe wanted to accelerate your timing of the carve-out? Or has the timing changed?

So the carve-out happened when we announced the deal on April 1, right, at the end of the first quarter. We've been working on this structure and this transaction, I mean, for well over a year now when Coatue came in as an investor, as a part of our negotiation on our convert, we contemplated the carve-out of these assets and talked through the overall long-term strategy here.

Really, what we focused on was what is the catalyst that allows us to split these 2 businesses apart? Is it going to be when we announce a data center deal? What is going to be the thing that allows American Bitcoin to be successful?

And at the time, it wasn't called American Bitcoin. It was called [ ASICCO ], to be honest. And American Bitcoin was really ideated with Eric Trump and we came up with the name together. But when we got closer and closer with Eric and talked to what we were building and what we were doing at Hut 8, and this was one of the ideas that we had and we talked to American data centers and what they were doing, we felt like that was a strong catalyst, that would give American Bitcoin the highest chances of being successful, but not only successful market leader within Bitcoin mining, but also Bitcoin accumulation strategy.

And both of us have built great and dear relationships with Michael Saylor and I've talked to him about this over the last couple of quarters. And so it's funny, we told them that he kind of started rolling the snowball for both of us that kind of compounded together and are rolling even larger now. But overall, we announced the deal at the end of the first quarter to make it easier and simpler on our accounting team, but we had this structure and idea in mind for a while now and needed the right moment and the right ingredients in order to guarantee success.

And so for American Bitcoin, the plan is it is operating at its own company today. It's mining Bitcoin to its own balance sheet. It's continuing to mine and accumulate that Bitcoin, and we look to take that company public as well, and that is the path that we're running forward with.

Our next question comes from the line of Bill Papanastasiou of KBW.

I was just hoping you could describe the cadence of negotiations with prospect AI HPC counterparties. Have the talks been accelerating? Or are you seeing any form of hesitancy by counterparties to pull the trigger and deploy capital following recent market headwinds? Per my discussions, one operator in the space highlighted concern regarding GB200s overheating and customers waiting for the next generation of equipment. Curious to hear your thoughts.

Depending on the customer that people speak to, some customers are driven by their own demand signals and other customers are powering some of the large AI infrastructure platforms. And so some of the signals are changing less often because they're internally driven and others are driven by customers and they fill that need, then that demand signal quiets out until they have another need come up.

And so I think for us, the way we've approached these relationships, and I think it's come to fruition is rather than saying we have one site, let's figure out and monetize it, this is all we want to do, we treated these as more long-term partnerships. "Tell us what you're looking for, what you're trying to build. Do our sites fit within that framework? If not, no problem at all. Let's keep in touch, and we'll share with you sites as they come up and tell us if it's interesting or not."

And for the folks that do say the sites are interesting and we're going deep into diligence, as we negotiate how we structure these deals, how do we drive the most value, I think we're a lot more long-term greedy than we are short term. And what I mean by that is we're trying to find the best win-win solutions for us and the customer and if we bring on any partners either to build a flywheel that we can continue to scale and replicate with those relationships.

And so we've seen obviously ebbs and flows in the market and some of the larger market data. However, it hasn't impacted the conversations that we're having nor the demand profiles and kind of being calling the spade a spade. We're not working on 5 gigawatts of data centers today, right? The River Bend campus with 300 megawatts of utility. We have another 230 megawatts of IT. That demand will continue to exist regardless of the volatility of demand profiles for these customers that they hold when they think about their multi-gigawatt demand profiles. And so as long as our relationships continue to be strong and we continue to be good partners, we believe that we can bring these to fruition.

Appreciate that color. And then for the second question, perhaps you can share some more color on the go-public transaction for American Bitcoin Corp. Do you have a time line that you're targeting? And have tariff impacts played any part in reaching the 50 exahash target?

We'll share more time lines in the coming future and definitely by our next earnings call. We're sprinting forward on this aggressively, both in terms of capitalizing the company while it's private, but also looking at its ability and access to be able to be in the capital -- in the public capital markets. And so we'll share more on that, Bill, as we have news to share on those milestones.

And in regards to tariffs, we've looked at where these chips are being manufactured, right? We have them manufactured all over Southeast Asia. There's U.S. manufacturing now as well in North America, something we're spending a close eye on. And as we commit to the purchase price, we're obviously thinking about the all-in price of CapEx to actually deploy, not just the price we pay to one individual counterparty. And so we'll have that a part of our overall underwriting.

And when we think about the phases of growth, we are not going to grow at all costs. If at times, we're able to find the right entry point for ASICs to grow and we have the right sites with the right energy profiles, we will grow and we have confidence to grow at a fast and rapid rate like we've shown the market.

At other times where ASIC prices have not come down relative to adjustments in tariffs or hashprice, we will accumulate Bitcoin and purchase Bitcoin. And so that's the beauty about American Bitcoin. The mandate is to accumulate Bitcoin at the lowest cost possible, and we have different vectors in its ability to do so rather than a grow on all cost mindset regardless of return profiles. Our job is to accumulate Bitcoin at a discount to just being able to buy it in the open market.

Our next question comes from the line of Mike Grondahl of Northland.

Two questions. The first one on the HPC side, the 2 other sites that you're marketing non-River Bend, is it possible that those leases could be announced before River Bend? Just trying to get a sense of how far you're along on those 2 other sites.

And secondly, as it involves American Bitcoin, you're going to own 80%. I think you've kind of covered the milestones there, the IPO and whatnot. But is there any other milestones or items you're really trying to accomplish in the first year of American Bitcoin being established? Just an opportunity there to answer that.

Thanks, Mike. I'll start with the first and then go into American Bitcoin. On the first, the other projects we're working on, one of them, I would say, is in a similar kind of place in terms of its evolution. Some areas were farther along, some areas were less far along compared to River Bend. So yes, there is a possibility.

The other site, I would say, is not as far along, however, is in a Tier 1 proven data center market. So the kind of the ability to move quicker with a lot of local contractors who work in that area, the talent is there. And so that could catch up quickly. So River Bend may not be the first campus that we announced, and I think that's accurate. And candidly, we would not be talking so much about River Bend if the permits didn't kind of get leaked out into the news and we felt like we had to address it. And so it is one of the campuses and may not be the first campus that we announced.

In regards to American Bitcoin, if we think about kind of the roadmap and Eric has been pretty open and transparent that anything he wants to be a part of, he wants it to be the leader and he wants it to be the biggest. And so that is very, very much a focus. When we think about key milestones, we have the kind of plans to take the entity public.

We have a private capital transaction that we're looking at funding before it goes public, so a raise there. We have the increase in exahash. So we're at around 10 exahash today. The Vega side is 15 exahash of optionality to purchase those machines from BITMAIN and average fleet efficiency of 13 joules per terahash there. So leading fleet efficiency with the newest generation chips, so the financing and purchase of those. And then we have 2 other projects behind that when we think about Phase 3 and Phase 4 to get to 50 exahash and committing to those builds and structuring those deals.

And so there's a lot to do there. We're really excited, and there's a lot of growth. And we're also looking at are there interesting convertible transactions in the private market that we can do that can convert into kind of a public convert similar to what we've seen kind of Michael Saylor strategy lead the industry with and how he's been able to monetize the volatility within that company and the exposure to Bitcoin accumulation.

Our next question comes from the line of Brian Dobson of Clear Street.

Just another follow-up question on American Bitcoin. So if you could step back from a broader perspective, how does that team see the Bitcoin mining business evolving? And what will a successful mining business look like going into the next halving? And do you think it may consider a Bitcoin treasury model?

The 3 layers we see in American Bitcoin are layer 1, which is the foundational layer that it started with, which is the Bitcoin mining business. The ability to accumulate Bitcoin cheaper than competitors in the Bitcoin accumulation strategy are able to buy, we think, is extremely compelling and extremely interesting. And so we'll continue to scale and build in a cost-efficient way, right?

The key thing there is, I think some companies have grown at all costs and the economics haven't really been as competitive. However, here, we're growing as long as that narrative, we believe, continues to hold as we model, which is by investing in Bitcoin mining, we'll be able to accumulate at a cheaper cost than we can buy for.

The second layer of the platform is the accumulation financing structure. So that's being able to raise capital to acquire Bitcoin as well in both equity and debt. Because when we think about it, how do we create a competitive advantage relative to someone just buying Bitcoin and ETF, we increase Bitcoin per share that they own. How do we do so more competitively than other companies that are out there are coming out there now is let's add other layers of the business that can accumulate Bitcoin and cheaper than we can just buy it for. So that's kind of layer 2 is the Bitcoin accumulation through different financing mechanisms.

And then layer 3 is we want American Bitcoin to be a Bitcoin household brand and that we can use that to create other Bitcoin businesses that are able to continue to increase Bitcoin per share for shareholders. So it's a 3-pronged approach that will be layered in as we continue to evolve and grow the business.

Our next question comes from the line of Stephen Glagola of JonesTrading.

Asher, in light of the competitive colocation market, I just was hoping you could revisit and maybe elaborate on the specific ways your data center team differentiates Hut 8 and enhances the company's credibility and negotiating position with hyperscalers and other enterprise customers.

Thanks, Stephen. I would start with sites are important. The quality of your site is extremely important because even if we have a good relationship with counterparties, this site doesn't work for where their footprints are, you can still have a great relationship with them, but that site might not be the site that works, right? And we've seen that happen. We continue to maintain relationships with those that we're not in active discussions with, and those are the ones that we share new sites with that enter our development pipeline if they would have interest or not at the early onset.

Second, once a site passes the interest level, then the question is, what is the deal in the economic terms that you're willing to agree to? How do you minimize perception of execution risk, especially for entities like ourselves that don't have a long track record of building these types of infrastructures for customers, both execution on operating and build-out, but also execution on financing.

And so it's bringing in counterparties, not necessarily just joint venture counterparties, but counterparties that we were able to say, "Hello, these are our financing counterparties. These are our engineering, A&E, MEP, GC counterparties. This is potential joint ventures that we can bring in if you want us to bring in." And I think that amount of humility for us has given us a lot of runway in building deeper trust.

And in conversations where we said, "Look, we understand that we have not built these before, but we have all these people around us. And we built infrastructure and energy infrastructure at scale. So we believe we have the ability to do so. And with the counterparties around us, we will execute. But if you want, we can bring in a joint venture partner to be able to make you feel more comfortable because we care about the long-term relationship here." And in some of those conversations, as those have evolved, they said, "You know what, you do it, it's all good. We trust you. Let's try to get a deal done here."

And so I think it's really having that long-term oriented mindset and being willing to say, my job here is not to just monetize the most amount of value I can out of this project for our shareholders but to create an opportunity where the customers feel comfortable and happy and win as well because that will lead us into getting the second, third and fourth project with them as well because just getting one project is not how we're going to build the business that we envision to build here.

This concludes the question-and-answer session. Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.