Iris Energy Ltd

NASDAQ:IREN

Earnings Call Analysis

Accelerated Growth and Cost Efficiency in Bitcoin Mining

In the latest earnings call, IREN reported an adjusted EBITDA of $2.6 million and Bitcoin mining revenue of $54.3 million. As they approach 31 exahash and aim for 50 exahash, cash costs per Bitcoin are expected to drop to approximately $29,000, the lowest in the industry. With attractive power prices and a transition to a spot pricing model, overheads per exahash are projected to decrease by 70%. IREN anticipates generating substantial operating cash flows that may allow for investor distributions in 2025, contingent on Bitcoin prices reaching $125,000, potentially leading to $1.4 billion in adjusted EBITDA.

Navigating the Future of Bitcoin Mining

In the latest earnings call, the company has painted an optimistic picture of its prospects in the rapidly evolving Bitcoin mining sector. The company is at a significant inflection point, as it prepares to scale its operations from 21 exahash to an anticipated 31 exahash in the near term, with plans to reach 50 exahash by the first half of next year. This growth trajectory underscores the company's commitment to becoming one of the largest listed miners, particularly at a time when Bitcoin is approaching all-time highs.

Cost Efficiency and Improved Margins

One of the standout features of the discussion was the company's impressive cost control measures. The all-in cash cost of producing Bitcoin is projected at around $29,000, marking it as one of the lowest in the industry. As the company scales, it anticipates a 70% decrease in overheads on a per exahash basis by Q1 FY '25. Such operational leverage could significantly enhance profit margins moving forward, particularly as the firm transitions to a more cost-effective energy procurement structure at its Childress site.

Ambitious Earnings Prospects

The company provided notable financial guidance for the upcoming periods. With Bitcoin prices projected to reach up to $200,000, the company's adjusted EBITDA could soar from $435 million at 31 exahash to around $1.4 billion at 50 exahash. This dramatic potential for revenue growth illustrates the sensitivity of Bitcoin mining profitability to price fluctuations in the cryptocurrency market.

Transitioning to Renewable Energy

The management has reiterated its commitment to 100% renewable energy, with the shift in electricity procurement also serving as a strategic advantage. The transition to a spot pricing model from August 2024 is expected to streamline energy costs significantly. In fact, if this pricing strategy were applied during the last quarter, the average cost of mining Bitcoin could have been reduced from $35,400 to $26,700—a compelling indicator of potential cost savings.

Strategic Connections and Expansion Plans

The company is making substantial investments in its data center portfolio, including a new 1.4 gigawatt project in Sweetwater that aims to be one of the largest Bitcoin mining facilities globally. This site will enhance capacity by leveraging strategic land and power agreements, including an additional 800 acres already secured—a crucial move in a power-constrained market.

Potential for Investor Distribution

Interestingly, with the anticipated growth in operating cash flows, there is discussion of potential investor distributions as early as 2025. The firm is targeting alternative funding structures to finance its growth, with the potential for dividends arising from its position as a low-cost producer who capitalizes on advantageous market conditions.

Long-Term Growth and AI Integration

In addition to its Bitcoin operations, the company is actively pursuing growth in artificial intelligence (AI) through its cloud services. With the recent procurement of NVIDIA H200 GPUs, it is strategically positioned to capitalize on growth in this sector alongside Bitcoin mining. The dual focus on AI and Bitcoin mining showcases the firm's adaptive strategy in leveraging its data center capabilities.

Market Position and Shareholder Value

The management's emphasis on prudent capital allocation and growing shareholder value is evident throughout the call. With the company's strong balance sheet—totally debt-free and around $182.4 million in cash—the future looks bright for the firm's ability to expand operations without undue financial risk. The successful implementation of its plans has the potential to significantly enhance shareholder value, aligning with its long-term vision.

Earnings Call Transcript 2025-Q1

[Operator]

Good day, and thank you for standing by. Welcome to IREN Q1 FY '25 Results Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. [Operator Instructions]

I would now like to hand the conference over to your speaker today, Lincoln Tan, Director Investor Relations.

[Lincoln Tan, executive]

Thank you, operator. Good afternoon to those of you in North American, and good morning to those of you in Australia, and welcome to IREN's First Quarter FY '25 Results Presentation. My name is Lincoln Tan, Director of Investor Relations, and joining me on the call today are Daniel Roberts, Co-Founder and Co-CEO; Belinda Nucifora, CFO; and Kent Draper, Chief Commercial Officer.

Before we begin, please note that this call is being webcast live with an accompanying presentation. For those of you that have dialed in via phone, you can elect to ask a question via the moderator after our presentation.

I would like to remind you that certain statements that we make during this call may constitute forward-looking statements, and IREN cautions listeners that forward-looking information and statements are based on certain assumptions and risk factors that could cause actual results to differ materially from the expectations of the company. Listeners should not place undue reliance on forward-looking information or statements. Please refer to the disclaimer on Slide 2 within the accompanying presentation. Thank you.

And I will now turn the call over to Dan Roberts.

Thanks, Lincoln. Good afternoon, everyone, and welcome to o

[Daniel Roberts, executive]

Thanks, Lincoln. Good afternoon, everyone, and welcome to our quarterly earnings call. Thank you for dialing in. It's a an interesting time in the market, and we're pleased to have the opportunity to present both our results from the prior quarter, but also importantly, step you through how we're thinking about the short term as well as the medium and long term. So it's an exciting time for us, exciting time for the industry. So let's dive into it.

So jumping straight through to Slide 3, make sure everyone's read the disclaimer on Slide 2. So I think it's fair to say that the Bitcoin mining sector is reaching an inflection point. It's a very dynamic time in the market, and we fielded a number of questions around why miners are underperforming the price of Bitcoin over the recent months. And what we're seeing is institutional interest returning to the space, but the question around why isn't miners necessarily or why aren't they necessarily moving with Bitcoin, given it's meant to be some sort of proxy exposure, what's different this cycle?

So from our perspective, we are seeing that the sector has matured. There are more ways for investors to express interest in Bitcoin. Miners that built their business around a HODL strategy now have more competition than ever for that business model. You're seeing institutional-grade self-custody solutions. You're seeing the emergence of the ETFs. You're seeing companies like MicroStrategy accumulate Bitcoin through capital market products.

In contrast, we're positioned as a low-cost commodity producer, focused on cash-on-cash returns for our investors and producing Bitcoin at low cash cost, circa $29,000 per Bitcoin as we will see in the coming slides. So mining business models are under scrutiny, this is an environment in which we expect a low-cost, strong cash flow business to thrive.

So as you can see there on the right-hand side, business fundamentals, they're coming into focus, production costs, operational focus, capital allocation and the outlook for growth. We believe we're well positioned for market leadership in every aspect of growth and ultimately, we view Bitcoin mining as Gold Mining 2.0.

Could I just ask that other speakers on the line from our side go on mute, please? There's a bit of background noise. Thank you.

So if we could then move to Slide 4. It's also an inflection point for our business. 31 exahash is weeks away. In fact, I believe that the last miner is going to be delivered in the next 7 days. Installation and commissioning will take a little bit longer than that, but the 31 exahash is now weeks away, making us one of the largest listed miners.

Just as Bitcoin is hitting an all-time high. However, those that know us, we're not stopping there. We are also pleased to announce that we are accelerating our expansion to 50 exahash into the first half of next year. Previously, it was the second half. So we've been able to bring all that forward and continuing to invest to generate more Bitcoin at an expected cash cost of $29,000 per coin.

The fundamentals of our business model have allowed us to continue to accelerate growth at low cost. Our strategy, focusing on large-scale projects, continuing to provide a lower execution risk and a rapid platform to expand. Years of planning and procuring long lead items to mitigate supply chain risks.

And finally, our management capability to deliver. If we have set a target in the market, we hit it, and we will continue to do that. This is a business model now, and this is a business in IREN where we believe we've hit that inflection point, we've hit the economies of scale now. While we will talk more on unit economics over the coming slides, at 31 exahash, we expect to be producing Bitcoin at an all-in cash cost of $29,000 per coin with those economics and those unit economics only to improve as we scale, as power costs and operating leverage really start to kick in.

In relation to AI and HPC, we've got almost 2,000 GPUs now operating, a combination of H100s, H200s, both NVIDIA generations. We're focused on contracting out this capacity and looking for opportunities to grow in a measured way based on robust contractual arrangements and demand that we're seeing in the market.

In respect to AI colocation and other monetization opportunities, they're also being progressed in parallel. We have high-quality assets that we believe will remain valuable in a power-constrained market with increasing competition from both Bitcoin mining and AI for this land and power.

We will continue not to provide guidance on specific terms or timing given the uncertainties, given the nature that we are dealing with counterparties. It is not all within our control. However, we continue to progress negotiations with some very large counterparties and hyperscalers. We will update the market as appropriate and as we have news.

In terms of what is within our control, as we know, the majority of the new generation Blackwells from NVIDIA, those GPUs will require liquid cooling. So notwithstanding successful testing of our GPU clusters in free cooling environments, including at Childress, we will also be installing our liquid cooling infrastructure at both our Prince George and Childress data centers in the coming months to ensure we are ready to support this next generation of GPUs. This creates opportunities for the business, both across our cloud platform as well as our colocation pathway.

In terms of corporate and funding initiatives. So in terms of funding, it's clear to say market conditions have improved materially over recent months. More funding options are emerging beyond the ATM. We are focused, very focused on alternative funding instruments as part of funding our upcoming growth plans, for example, convertibles, and we'll continue to progress that over the coming months.

In terms of becoming a U.S. domestic issuer, we intend to transition to this status in 2025, reporting on a U.S. GAAP basis. We have also been working on potential inclusion in major U.S. indices, for example, the Russell 2000 in the near term, given the majority of our assets are in the U.S.

Finally, our increased scale at 31 exahash and not long after that, 50 exahash, combined with our non-HODL approach, i.e., we don't hold Bitcoin on our balance sheet, we mine at an expected all-in cash cost of $29,000, it's anticipated to generate very substantial operating cash flows in the near term. As such, we believe this will support potential investor distributions in 2025. We're working through specific details around what that might look like. However, we wish to again signal our commitment and our focus on prudent capital management and generating strong cash-on-cash shareholder returns.

So moving along to Slide 6, a little bit more about Bitcoin mining. So where we're at? We have been delivering consistent growth over the last couple of years. If we set a target, we have met it. We're now at 21 exahash above the 20 exahash previously guided and 31 exahash is just weeks away, as I mentioned earlier. We're not stopping there. We are continuing 50 exahash in the first 6 months of next year.

So we are committed to hitting these targets. We will maintain our disciplined approach to delivery, our focus on costs, our focus on safety. It's supported by our single site expansion at Childress, where we have over 400 people on site who will continue the cadence of building out 50 megawatts of data centers every month.

No M&A is required. It's all organic growth at cost. The land in Childress cost us a few million dollars. That is it. Everything has been self-developed internally by our team. Proven people, proven delivery processes and a strategy to focus on large-scale projects continues to provide lower execution risk, more certainty and a platform that allows us to rapidly expand at low cost.

On the mining hardware side, you will recall that we previously secured miner purchase options for S21 Pro miners at a fixed price of $18.90. These options were struck when Bitcoin was $65,000. It's taken the last 6 years to get our platform up to this stage. The vision to accumulate organic development of large-scale sites, building a team and know-how, raising capital, we've now hit that inflection point just as we are seeing the real world really start to lag the digital world. There is significant megawatts, significant miners, significant capital and management expertise required from this point for the network, i.e., the hash rate at a global level to scale in parallel with the Bitcoin price.

Against this backdrop, there's now increasing competition for a number of those inputs, transformers, land, power with the AI traditional data center world now also competing for exactly the same core inputs. To play this out with a simple thought exercise, what happens if Bitcoin moves up to 150,000? 50%. All of a sudden, you need another 7,000 megawatts of additional capacity, another $9 billion in CapEx, transformers, long lead items, cables, networking, concrete, steel, all these things that take time that are fraught with risk. All of it to bring into the real world, real projects, real data centers, the ASICs, you cannot plug an ASIC into a high-voltage transmission line. It just doesn't work. So we've set ourselves up to really become a leading force in Bitcoin mining.

On Slide 7, to really drill into this $29,000 cash cost per Bitcoin, we have driven it down. It is happening. Best-in-class efficiency at 15 joules per terahash from 22 in the recent September quarter. Expansion at low power cost. Childress, our average power cost is a tick over $0.03 since moving to spot price with our automated curtailment algorithm a few months ago. That is becoming a larger and larger proportion of our portfolio, driving down our average power cost accordingly. And then overheads, non-power costs, but all-in cost of our business, we're really starting to see the benefits of operating leverage come through. We'll talk more to this on coming slides, but it's effectively spreading our corporate cost base across a much larger number of exahash and much larger number of expected Bitcoin mined.

So finally, in terms of operating efficiency, we believe in building high-quality data centers that optimize operating conditions for our miners. It's really important to us that we maintain the highest levels of uptime and look after that fleet to maximize profitability.

Moving on to Slide 8. Illustrative mining economics. Some of you would have seen this on Twitter previously, but 1 exahash of operating capacity costs IREN approximately $30 million to deliver against a market backdrop where capital markets are valuing that at almost 4x that valuation, $120 million. If you have the ability to deliver at $30 million to have something valued at $120 million, then in my simple mind, that makes a lot of sense to continue pursuing.

We see investing in additional capacity is highly accretive. It's a really simple framework for thinking about this. The unit economics to cross-check that in terms of cash flow, $13 million of net illustrative adjusted EBITDA on a $30 million CapEx, it's around a 2-year payback, again, making a lot of sense.

So as we're looking at all-in cost to produce, you'll notice on the right-hand side, a step-up in renewable energy certificate costs as we continue expanding at Childress. This highlights our commitment to 100% renewable energy. We're seeing the sustainability focus really come to the fore. There are lots of different claims out in the market around supporting renewables.

At Childress, notwithstanding 80% of the network in that area is renewable, you need the certificates. If you do not have certificates, you are not operating with renewable energy. At the moment, a wind farm or a solar farm evacuates power into the grid, the commodity is bifurcated into 2 components. One is the electron that trades alongside black and brown power. The other is the certificate. If you do not buy those certificates, you are not renewable because someone else will own those certificates and we'll be double counting it. So this has absolutely been a critical part of our strategy. We've bought it for every electron since day 1, and we will continue to do that.

A key part of how we are also thinking about growth is accretion on a per share basis. How do we drive value for every shareholder in our business? One example of how we're thinking about this is in the bottom right-hand table. We've shown this analysis in terms of adjusted EBITDA per share, which highlights very strong accretion metrics as we continue to scale from 31 to 50 exahash.

We have also set out some illustrative share price levels at which equity is raised, whether that be through ordinary equity, convertibles or other alternative funding structures. We've seen a number of convertibles launched over recent months with creative strike prices, and we'll continue to look at that as all, but one of our pathways to raise additional capital.

So yes, growing from $400 million to $700 million of illustrative EBITDA in the top right-hand corner is super exciting, but more exciting for us is the accretive nature of this growth and our continued commitment to drive cash-on-cash shareholder returns.

On that note, I'd like to pass over to Kent to talk about AI and HPC.

[Kent Draper, executive]

Thanks, Dan. With respect to our AI cloud service, commissioning of our recently acquired NVIDIA H200 GPUs at our Prince George data center location is now complete. The transition from Bitcoin mining to GPUs was able to be completed in a matter of weeks, underscoring the multipurpose nature of our data centers and ability to rapidly scale to support customer growth ambitions.

Now that, that capacity is installed, we're focused on contracting out the new capacity and currently have customer trials underway. To that point, we've also been increasing our activity with respect to marketing and go-to-market initiatives.

As an example, in conjunction with the recent SuperCompute Conference in Atlanta, we hosted roundtable events in San Francisco with commercial leaders in the AI sector and have more activity planned for the future.

Our current fleet of NVIDIA H100 and H200 GPUs supports approximately $32 million of potential annualized hardware profits using less than 2.5 megawatts of data center capacity. To put that into perspective, that's less than 0.5% of our total year-end data center capacity.

This is one of the elements that we particularly like about the cloud services offering. It's highly complementary to our existing operations and doesn't cannibalize our data center capacity for other use cases such as AI colocation.

Our strategy is to continue focusing on measured growth of our capacity in response to customer demand, backed by good contracts and solid utilization of that capacity. In relation to colocation and other monetization opportunities, we believe that we have high-quality assets, which will become more and more valuable in an increasingly power constrained market. And as Dan referred to earlier, there's increasing competition for sites to support both Bitcoin mining and AI.

We expect competition for large scale grid connected power to intensify over the coming months, given some of the dynamics in the market, including continued growth and adoption of AI, regulatory complexities with behind the meter solutions. So for example, the recent FERC rejection of the AWS-Talen interconnect at the Susquehanna nuclear plant in Pennsylvania, and also the time line for new nuclear generation and other generation sources to come online.

As a result, we're seeing West Texas sites increasingly coming into focus, and there have been several hyperscaler projects announced in the region in recent months. For example, the Crusoe data center project in Abilene.

With respect to our discussions underway, we continue to advance negotiations with interested parties on a range of structures for our sites, including powered land leases and built-to-suit contracts. As Dan mentioned, we aren't providing guidance on terms or timing at this stage, but as we've said previously, we're actively negotiating potential transactions with hyperscalers and will update the market on material developments as appropriate.

Monetizing our data center platform into new verticals remains a key focus for management. We're committed to maximizing shareholder value, and any transaction must reflect the strategic value of our assets.

With respect to liquid cooling, our team has extensive experience designing and operating data centers, optimized for power-dense compute, and a demonstrated execution capability. We're leveraging that experience to best position our infrastructure for opportunities in the rapidly evolving AI sector.

For instance, as many people would be aware, certain variants of the Blackwell generation of NVIDIA GPUs require liquid cooling. Notwithstanding the success and the successful testing of our GPU clusters in our current free air cooling environment, we will be installing liquid cooling infrastructure at both our Prince George and Childress sites to ensure that we're ready to support the next generation of GPUs. We expect this to create additional opportunities for us across both the cloud services and colocation pathways.

As we touched on in the previous slide, I'm on Slide 12 now. We see significant strategic value in our land and power portfolio, and parties are turning their attention to power and capacity requirements for 2026 and beyond. Today, we're excited to announce the location of our 1.4 gigawatt Sweetwater data center project, which is immediately adjacent to the utility's high voltage substation and approximately 60 miles from the regional hub, Abilene in West Texas.

During the quarter, we've continued to push forward with development activities, including securing an additional 800 acres of land adjacent to our existing freehold land, bringing our total land holding at the site to over 1,300 acres. We've also accelerated the time line for upgrades and energization of the utility substation from October 2026 to April 2026.

Furthermore, we've commenced procurement to support the 1.4 gigawatt substation energization by April of 2026, including procurement of transformers, circuit breakers and ancillary electrical equipment. At 1,400 megawatts, this single site is larger than many entire data center markets in the U.S. and has the potential to be the largest Bitcoin mining facility or AI factory in the world, capable of supporting over 90 exahash of Bitcoin mining or over 800,000 GPUs.

On to Slide 13. In relation to our power portfolio, we've spent years developing our current 2.3 gigawatts of contracted capacity. Our approach with respect to the market is only to announce sites where we have executed interconnection agreements. This means they're not potential future megawatts. These are genuine megawatts that are contracted today through binding interconnection agreements.

We've observed firsthand within our own portfolio, some of the timeline delays in securing grid approvals. And as Dan mentioned, there are a number of risks associated with development and securing capacity, which underscores the strategic value of having these megawatts contracted today, and especially at large individual sites.

Without signed interconnection agreements, the quantum of megawatts, timing and cost is highly uncertain. That is why on our calls, we only highlight the 2.3 gigawatts of signed interconnections. However, we continue to develop our data center portfolio and hope to provide updates in due course on our pipeline of over 1 gigawatt of data center sites, if we are successful in signing additional interconnection agreements at those sites.

Importantly, the interconnection agreements that we have executed provide for uninterruptible power. So that means no mandatory curtailment. Any curtailment that we do is entirely voluntary in order to reduce our power costs at those sites.

Additionally, we've continued to invest internally in our development capability, including growing our commercial teams at our North American headquarters. And we'll continue to look to accelerate our development activities globally and add to our organic pipeline, which means less reliance on M&A transactions to support our future growth.

Now I'll pass over to Belinda, our CFO, to walk through the financial results.

[Belinda Nucifora, executive]

Thank you, Kent. Good morning to those in Sydney and good afternoon to those in North America. Thank you for joining us for our Q1 FY '25 earnings update. As a start, I'm going to highlight our operating leverage as we scale our Childress operations.

Earlier in the presentation, Dan highlighted the illustrative mining economics on our path to 50 exahash. As shown on the slide, as we scale from today's installed capacity of 21 exahash to 50 exahash, we see potential for significant improvement in our unit economics. The key drivers to this being attractive power prices at Childress, with the transition to spot pricing made on 1 August, 2024, a reduction in the energy costs for Bitcoin mined across the portfolio due to the increased mining contribution from Childress at a lower cents per kilowatt hour price, coupled with improved fleet efficiency to be industry leading at 15 joules per terahash.

Furthermore, as we scale, overheads are spread over a larger revenue base. As such, overheads on a per exahash basis are expected to decrease by approximately 70% from Q1 FY '25 as we scale to 50 exahash. This contributes to a significant reduction in our all-in cash cost per Bitcoin of approximately $28,000, being the lowest in the industry.

As touched on earlier by Dan in the presentation, to finance this growth will be focused on alternative funding instruments. In addition, at scale, the achievement of positive operating cash flows may support a potential for investor distribution in calendar year 2025.

Moving on to the next slide. Further to the operating leverage we just discussed, this sets out the illustrative economics at various Bitcoin prices. The first column of the table shows illustrative economics at 31 exahash and 50 exahash with a Bitcoin price of $90,000 and a current network hashrate of 732 exahash. This results in an estimated adjusted EBITDA of $435 million and $714 million respectively. The sensitivity analysis shown in the table sets out the illustrative economics associated with Bitcoin prices starting at $125,000 and scaling by 25,000 intervals to $200,000 at an implied network hashrate of 1,000 exahash.

This results in adjusted EBITDA of $1.4 billion at a Bitcoin price of $200,000. The estimated increase in overheads as we scale from 31 exahash to 50 exahash is approximately $23 million, primarily attributable to variable Childress site-related costs, including property insurance and property county taxes.

Also mentioned earlier, we're 100% committed to renewable energy and our investment in renewable certificate costs increase as we scale from $9 million at 31 exahash to $16 million at 50 exahash. Our overheads reflect a business contributing to deliver significant growth. We have onboarded technical and go-to-market resources for our AI business and have built internal site development expertise to scale our broader portfolio. This makes us less reliant on competitive M&A processes to scale our platform, and we see this as a strategic advantage.

Moving on to the next slide, I'll touch on the quarterly results. During the quarter, we achieved an adjusted EBITDA of $2.6 million with Bitcoin mining revenue of $54.3 million and AI cloud services revenue of $3.2 million. The average operating hashrate for the quarter increased from 9 exahash to 12.1 exahash and we mined 821 Bitcoin at an average realized price of $66,000.

Net electricity costs for the quarter increased by $4.6 million to $28.7 million. This was primarily due to the increased Childress megawatt usage as we scaled as well as the Childress energy procurement in July 2024, which was under a fixed cost, fixed quantity contract. From 1 August, 2024, Childress procurement transitioned to a spot pricing strategy. If that spot pricing strategy had been in place for the entire quarter, the cost of Bitcoin mine would have been $26,700 versus a $35,400.

As discussed in the previous slide, other costs of $21.4 million reflects a business today that is delivering significant growth and projected continued expansion over the coming years. The primary driver of the quarter-on-quarter increase of $0.4 million is due to Childress construction and operational insurance costs.

Now moving to our cash flows. Closing cash at bank at 30 September is reported as $98.6 million with strong receipts from Bitcoin mining activities of $49.6 million and AI cloud services of $3.7 million. The increase in electricity payments made during the quarter includes the one-off liquidation payment of $7.2 million related to the transition to spot pricing strategy from 1 August, 2024.

We increased our investing activities and spent a total of $387 million, which relates to expansion at Childress data center, the purchase of BITMAIN S21 Pros and T21 Miners as part of the pathway to the 31 exahash and the purchase of NVIDIA H200 GPUs. We raised $84 million of net ATM proceeds during the quarter and post that $142 million from net ATM proceeds. Cash increased to $182.4 million at 31 October, 2024.

We'll now turn to the balance sheet. At 30 September, 2024, we had total assets of $1.3 billion, no debt and a strong balance sheet providing flexibility to fund future growth. Total equity increased to $1.1 billion with gross proceeds of $76 million from 9.1 million shares sold under the ATM. As touched on earlier in the presentation by Dan, to finance future growth, we'll be focused on alternative funding instruments and in addition, at scale, the achievement of positive operating cash flows may support a potential for investor distribution in calendar year 2025.

I'll now turn over for Q&A.

[Operator]

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

[Lucas Pipes, analyst]

My first question is on Sweetwater and how you think about the strategic process versus the organic opportunity at the site. Thank you very much for your perspective on that.

[Daniel Roberts, executive]

Thanks, Lucas. Nice to see you. Look, the short answer is there's no decision point right now beyond procuring long lead items around the electrical infrastructure, so transformers, substation components, et cetera. So we're in a position to prepare dual pathways, both for Bitcoin mining as well as AI-related opportunities on the colocation or otherwise.

So we're pursuing it in all parallel. And the reality is we're facing a number of different potential pathways. And the opportunity is to compare each pathway to each other as an opportunity cost. And it's everything from building out Bitcoin mining. And if we can continue to build at $30 million and have the market value that exahash at $120 million, then that sounds pretty good. All the way through to build-to-suit options where we might build an AI data center for a counterparty under a build-to-suit model.

And again, internally deliver a component of that. There's structures like power and land, leases in there as well. So at this stage, it's working through all the options in parallel and just playing them off against each other.

[Lucas Pipes, analyst]

I appreciate that. That's helpful. And my follow-up is on the capital intensity on the Bitcoin mining business. Should we kind of think of that $30 million per exahash in a linear fashion up to 50? And if so, could you just remind us what the capital requirement is from where you are today to that 50 exahash? How much capital is needed?

[Daniel Roberts, executive]

Yes, for sure. So we haven't guided a specific number, partly because there's a number of moving parts around do we buy additional GPUs, because there's a spike in demand and the customer would like that. We're buying long-lead items for Sweetwater. But if we look in really high-level terms, we previously announced that we were fully funded to the 30 exahash. Since then, we've raised about $200 million-ish under the ATM. That gives a $400 million funding requirement to go from 30 to 50.

Yes, we've had higher operating cash flows than expected because of the Bitcoin price action. But again, that's somewhat offset by other CapEx. So we haven't guided specific numbers, but in general terms, you can probably triangulate roughly what we need.

[Lucas Pipes, analyst]

Dan and team, I really appreciate all the color and wish you best of luck.

[Operator]

Our next question comes from Joseph Vafi with Canaccord Genuity.

[Joseph Vafi, analyst]

Nice to see all the progress on so many fronts. It's great. Just kind of starting with a theoretical one, we've got a much higher Bitcoin spot price than we did, and you're now mentioning potential. It sounds like maybe a dividend or something like that next year. Just wondering at what Bitcoin level and kind of what level of exahash do you think at least the Bitcoin business can keep growing kind of more on a self-funded basis moving forward versus balance sheet-related financing? And then I'll have a quick follow-up.

[Daniel Roberts, executive]

Yes, sure. So we think about operating cash flows a little bit separate to investing cash flows where there are 2 separate decision points in our mind. So the first is operating cash profits. So how much are you generating from your operations as a going concern, and what do we do with that cash flow? The second is what is the decision around reinvestment of those cash flows?

Now, I think it's fair to say that for quite some time, 100% of all the capital we raise is going into revenue generating CapEx and using our operations to continue to support building our corporate overheads, our operating base, et cetera. And as you can see in the presentation, we're now hitting that inflection point where operating cash flows are going to be potentially substantial.

So the decision to reinvest those cash flows as distinct from distributing those cash flows to investors is something that we'll work through. But when you look at the market today and you see a number of different companies accumulating Bitcoin on their balance sheet, paying market price or close thereof to it, and we've got the opportunity to generate Bitcoin and effectively acquire Bitcoin at a cash cost of $29,000 and distribute that Bitcoin/cash out to investors who can then self-custody.

We're not that big of believers in third-party custody. We've been around Bitcoin since 2013, ridden Mt. Gox, FTX, et cetera. So I think the opportunity to generate a $29,000 cash cost Bitcoin for investors and effectively distribute that coin out either through the physical coin, we'll have to look into that, or as cash flow is pretty powerful.

And I think it's important to keep that separate from investing in cash flows and how we might fund that to continue to drive that cost potentially down even lower, but drive it in aggregate.

[Joseph Vafi, analyst]

That's great color. And potentially exciting there in actually distributing the underlying Bitcoin instead of cash. And then just congrats on being able to announce the Sweetwater site. That's nice progress. Just -- I know you started to talk about some long lead time items there. Are you going to try to get enough long lead time items for the full 1.4 gigawatts? Or do you think that there's a step progress there? And then any update on fiber and if you're going to have to invest on fiber interconnect for that site?

[Kent Draper, executive]

Dan, I'm happy to jump in and take that one. So with respect to the full capacity there on the data center side, it will take some time to build it out. As Dan referred to earlier, we're currently at Childress able to build out about 50 megawatts a month in terms of data center capacity. So to build out a site the size of the Sweetwater data center site, it will take some period of time.

But what we're doing with the procurement of long lead items is enabling the full capacity on the high voltage substation. So this is the substation that allows you to connect into the high voltage utility grid. And then we will build out the remaining site substations and data center capacity on a progressive basis.

But the long lead procurement is set up in a way to enable us to energize the entire bulk high voltage substation to the full capacity on that April 2026 time line.

[Joseph Vafi, analyst]

Great. And then any update on fiber at Sweetwater?

[Kent Draper, executive]

Yes. So in terms of fiber, there are multiple Tier 1 carriers in the area, including right up to the existing utility substation. So access to fiber is not an issue. In terms of latency, because these are Tier 1 providers, the latency to all major hyperscaler hubs is very good as well. So sub 20 milliseconds latency and sub 10 milliseconds in a number of cases. So the fiber position at that site is suitable for almost all AI applications, as well as Bitcoin mining, obviously.

[Joseph Vafi, analyst]

Sure. That's all positive news.

[Operator]

Our next question comes from Darren Aftahi with ROTH.

[Darren Aftahi, analyst]

Two, if I may. First, I'm just curious, strategically, can you walk through just the mindset of accelerating the 50 exahash in the context of maybe looking at opportunity costs relative to other opportunities?

And then my second question on your balance sheet, the $275 million of prepayment on the cash flow, what is that exactly related to in terms of payments? I assume it's all for the acceleration to 50 exahash. And I guess like what portion of that CapEx is already paid for?

[Daniel Roberts, executive]

I'll pass over to Belinda for the balance sheet question in a minute. But to address the first question around opportunity cost, I mean, we continue to have conversations with very large counterparties about the prospects of doing a deal at Childress.

But the reality is you've got to make a decision at some point, and the decision is right this second, the preference is to build out Bitcoin mining. The opportunity cost is comparing each pathway to another. Yes, we've received multiple offers for capacity there. But in our opinion at this stage, the prospects and the risk return proposition of building all that out at Bitcoin makes sense.

Equally, there's still some optionality there. We haven't exercised all those minor options. We've got some time to do that. So if those contractual negotiations with other parties went a slightly different direction or improved, then, I mean, we reserve the right to change our mind and pivot. But by definition, it needs to be better than the alternative which we can control, which is building out Bitcoin mining at $30 million and having the market seemingly value that at $120 million.

[Belinda Nucifora, executive]

Thank you, Darren. Yes. In relation to the prepayments, so we've got mining hardware prepayments of $122 million as well as on the GPUs $8.1 million. We also -- in our Childress new contract, we have to keep some collateral on board, so we have some prepayments in relation to the electricity contract in Childress.

[Operator]

Our next question comes from Reggie Smith with JPMorgan.

[Reginald Smith, analyst]

You had a really interesting quote in your press release, and you actually mentioned it again during the call. Some of the effect of making sure that a deal reflected the value of your assets, strategic value of the assets, I think, was the exact quote. I'm not asking for a number, but I'm curious how you guys go about appraising the value of that, like how you think about it. We've tried to take some stabs at it, but I'm just curious how you think about that. And I have a follow-up.

[Daniel Roberts, executive]

Yes. Thanks, Reggie. To be honest, it's a bit of art and science, both hand-in-hand. The science element is putting the numbers down and modeling out the various pathways and scenarios. You've got Bitcoin mining and then you've got various AI colocation structures that all result essentially in an NPV valuation to the business today. So that's the science.

Notwithstanding, you've got to put assumptions into those spreadsheets, but those assumptions are informed largely by reality. When it comes to Bitcoin mining, it's CapEx, it's returns, Bitcoin price, network hash rate. When it comes to AI transactions, it's plugging in contractual terms that are under negotiation. So it's quite easy in terms of working out at any point in time what the preference might be. Notwithstanding there's assumptions that go into that, including discount rates, cost of capital, market valuation of different types of revenue streams. So all of that goes into the bucket.

The art element is saying, well, where are we at in the market? What are the conversations we're having with market participants all the way from hyperscalers to data centers, to the banks, to real estate companies, and triangulating all that information to try and form some sort of directional view on where the market is at. And I think it's fair to say that the market is still in transition. You've gone from the old world of effectively available capital, available data center capacity to increase the number of GPUs and meet this AI demand.

All of a sudden, you seem to have hit this point in the market where hang on, we are short potentially substantial amounts of power to meet this. I think it's absolutely fair to say there's a transition period where every counterparty is going on their own journey around how scarce is power, how real is the current demand on the AI side? And what does this mean in terms of capacity to pay an interest in transacting in a specific valuation point.

And as time goes on, what we're seeing, again, this is anecdotes, is people are now starting to realize the 1 and 0 difference between having an interconnection agreement and talking about this made up pipeline of megawatts. I mean, we have been in development for decades, renewable energy projects through to other infrastructure to now developing power for data centers, as we've done in IREN. It is a 1 and 0 business. You could have whatever made up pipeline of megawatts you want. But unless you get to the finish line and you get that signed connection agreement, it is worth 0, absolutely nothing.

In sporting terms, the analogy is going through a home and away season, undefeated, not losing the game, being the hot price favorite, you then get to the final. You have a few injuries, the weather changes, something else goes wrong and you don't win the final or you've lost. It's a 0. The season counts for nothing and you go on to the next season. It is exactly the same in this. And I think the market is now starting to realize the value of that.

[Reginald Smith, analyst]

That's good. It actually leads right into my next question. It was going to be about kind of the urgency that you were sensing. And it sounds like you're starting to get that urgency. One of the things that I've noticed is that a lot of these -- not a lot, but several of these big tech companies have announced long-range projects that deliver in 2029, 2030, haven't really seen outside of CoreWeave a deal that's like for more immediate capacity. It sounds like there's more urgency. Maybe can you talk a little bit about how conversations have changed and whether there is more urgency on the front end, post end of kind of capacity and opportunities, how that's changing?

[Daniel Roberts, executive]

Yes. I mean 2 hours ago, we got an e-mail from $1 trillion hyperscaler that said, they weren't interested in Sweetwater, and now they are. I mean, that's one data point. And there's lots of little anecdotes along the way. I'm kind of reluctant to go into too much detail at the risk of building expectations because at the end of the day, a lot of this is outside of our control in terms of how counterparties think about value, their capacity to transact at terms that we will deem attractive.

So it's just a matter of working through all that, knowing that we control our own destiny, being a low-cost producer of Bitcoin and generating that strong cash-on-cash returns with this incredible optionality that we're super excited about around a large-scale AI HPC deal that might happen.

[Operator]

Our next question comes from Brett Knoblauch with Cantor Fitzgerald.

[Brett Knoblauch, analyst]

Congrats on the quarter. Maybe just to start, I think a big news event that happened in the quarter was the FERC announcement. Have you seen like a material increase in demand post that announcement given they kind of used Texas assets much more favorably?

[Daniel Roberts, executive]

Yes. I mean, I've got not a lot to add as a follow-up to answering the previous questions. I think, yes, like we are seeing more perceived demand coming through. But again, I'll use the sporting analogy. All these conversations are great, but unless you close a deal, it's 1s and 0s. And we're just not in the business of speculating and telling everyone every step of the way how good these negotiations are because unless you get a signature on a piece of paper that delivers a transaction that delivers value for shareholders, then it's worth 0. It's the same in development. It's the same in AI deals. And we can -- you can't really control the time lines of each. You've got to continue to progress things will go wrong, things will go right, but get the signature and then you've created a proper tangible value for your shareholders.

[Brett Knoblauch, analyst]

And then maybe just one more on AI cloud business. I know you guys guided for year-end, it being 10% of profits. Obviously, Bitcoin has gone up a lot since then. I guess, what's the plan beyond this year for that business? And how do you view that for CapEx on GPUs relative to maybe some of the Bitcoin mining CapEx given what we've seen happening with Bitcoin over the last couple of weeks?

[Daniel Roberts, executive]

Yes. It's a great question and something we think a lot about. I think in the short-term, we're probably less disposed to investing CapEx in the GPUs. And that's for 2 reasons. One is our current cost of capital and where we should be focusing use of funding. And I think when you look at the Bitcoin opportunity, that's very, very clear and tangible. GPUs, they are capital intensive. We have explored debt options. Do you want to lay in the business with debt at that -- this point in time in the absence of having a long-term contract?

Again, things can change quickly, so I'm going to hedge my bets here. But sitting here right here right now with our cost of capital where it is. And also, what we're seeing is a slight weakening of demand in that market. It may be attributable to this change in generations of the NVIDIA chips. We're going from the H100, H200 generation to the Blackwells being released next year, and that will be a reasonably material step-up in compute. And I think it's fair to say that we're seeing end customers be a little bit cautious around that and maybe preferring shorter-term contracts on the H100 and H200s in anticipation of those new generations come out. So we just think it's prudent to just take a measured approach.

If tonight, we get a contract, a call for a longer-term contract, and it's a big opportunity, absolutely, we're going to do it. But right here right now, it feels like it's a little bit more measured, and let's just wait and see how the next few months plays out.

[Brett Knoblauch, analyst]

Awesome. And maybe if I could just squeeze in one more. There's been a lot of news over the past couple of weeks surrounding BITMAIN, and I know you guys have a big purchase order in place. I guess, are you guys concerned that those shipments could potentially not be delivered or held off at customs? Or just any thoughts or comments on that?

[Daniel Roberts, executive]

Yes. I mean, for 6 years, we've had people throw concerns around BITMAIN add-ups, and I'll just continue to say exactly the same thing. We have never had anything but a fantastic experience with BITMAIN. The purchase process, the post market, sorry -- the aftermarket.

In terms of supply chain hiccups, we deal with it every day. It's the consequence of dealing with a real-world business and a million and one moving parts, but we've had no material issues. And as I mentioned on the call, I think the last of the 31 exahash in miners will be delivered in the next few, maybe 7 days.

[Operator]

Thank you. I would now like to turn the call back over to Dan Roberts for any closing remarks.

[Daniel Roberts, executive]

Thank you, operator. Well, that rounds out our Q1 FY '25 results call. We've enjoyed presenting our results, but also articulating our vision in terms of both funding side, looking at alternative structures beyond just the ATM to raise capital on potentially less dilutive terms. But importantly, the ability to invest that capital in accretive projects for our shareholders, looking at per share metrics, for example, adjusted EBITDA as outlined in our presentation and maintaining a prudent approach to capital allocation to generate those strong shareholder returns.

It's a consequence of long-term planning in this business to get it to this point, years of developing sites organically, years of planning and procurement of long lead items. And we feel that the mining sector as a whole as well as our business are at an inflection point. And over the next 3 to 6 months, we'll see that start to play out in even more tangible terms, and we're excited to position ourselves as a market leader in this area. So thank you, everyone, for dialing in, and have a good evening.

[Operator]

Thank you. This concludes the conference. Thank you for your participation. You may now disconnect.