Today, and thank you for standing by. Welcome to Riot Platforms Business Update. After the speaker's presentation, there will be a question-and-answer session. At this time, all participants are in a listen-only mode. To ask a question during the session, you will need to press star 11 on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star 11 again. Please be advised that today's call is being recorded, and the presentation is also available on the Riot website. I would now like to hand the conference over to Josh Cain, Vice President of Investor Relations at Riot Platforms. Please go ahead.
Good morning. Thank you for joining us, and welcome to Riot Platforms Business Update call. At this time, all participants are in listen-only mode. We will hold a question-and-answer session following management's prepared remarks. Please note this call may contain forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. All statements, other than statements of historical fact, are statements that could be deemed forward-looking statements. Such statements may include, but are not limited to, statements that predict future events or trends, forecasts or performance, and may contain words such as believe, expect, intend, project, plan, or words or phrases with similar meaning. Forward-looking statements are based on management's current expectations, forecasts, and assumptions, and are subject to risks and uncertainties that may cause actual results to differ significantly.
Listeners are cautioned not to place undue reliance on these forward-looking statements, which speak only as of the date on which they were made. For further information, we encourage you to review the risk factors discussed in the company's periodic filings filed with the SEC. Joining me on the call today are our CEO, Jason Les, our Chief Data Center Officer, Jonathan Gibbs, and Jason Chung, EVP, Head of Corporate Development and Strategy, and our next incoming CFO. We'll now begin with remarks from Jason Les.
Thank you. On today's call, we are announcing transformational milestones in Riot's evolution as a leading digital infrastructure platform. I am pleased to share two achievements at Rockdale, firmly establishing it as a premier data center development site alongside Corsicana, giving Riot 1.7 gigawatts of immediately actionable power capacity across both sites. First, we have completed the fee-simple acquisition of 200 acres of land at Rockdale previously leased by Riot. By converting our interest from a ground lease to full ownership, we have unlocked the ability to develop data centers on-site. With 700 megawatts of fully approved power already interconnected and contracted, full land ownership allows us to position Rockdale as a premier asset for data center development alongside Corsicana.
Second, we are announcing our first data center lease, a 25-megawatt agreement with AMD, one of the most significant hardware manufacturers in the world and a key leader in the AI ecosystem. This partnership includes multiple options for expansion and represents the beginning of what we expect to be a deep, long-term relationship. Our data center team has been designing and customizing this space to meet AMD's specific needs, and we will share more details on the full scope of this lease later during the call. Taken together, these milestones validate Rockdale as a premier site for data center development, establish Riot as a credible developer and operator for premier technology organizations, and create highly visible contracted cash flows with an investment-grade tenant.
More importantly, these transactions demonstrate Riot's disciplined approach to capital allocation, as we are strategically deploying capital to build high-quality assets, securing creditworthy tenants, and utilizing efficient financing strategies to recycle capital into our next wave of development. This is a major advancement for Riot and vastly improves our position to capitalize on today's lucrative data center market. I'll now walk through the details of the Rockdale acquisition and our new lease with AMD, following which we will open the line for questions. Turning to slide five, I would like to outline the key terms and strategic rationale of the Rockdale land acquisition. Our decision to acquire this site was driven by clear market feedback. In our discussions with potential tenants and lenders, it became evident that a ground lease structure creates unnecessary complexity and risk.
To maximize our ability to develop data centers at Rockdale and fully leverage the 700 megawatts of fully approved power on-site, we determined that fee-simple ownership was essential. We closed this acquisition in December for a total purchase price of $96 million, funded entirely by the sale of approximately 1,080 Bitcoin from our balance sheet. This allowed us to secure this critical asset using existing liquidity without the need for any equity financing. The acquisition encompasses 200 contiguous acres that are primed for immediate development, and we expect, based on an assumed PUE of 1.5, that Rockdale could eventually support approximately 436 megawatts of critical IT load capacity, with potential to expand further. Importantly, Rockdale has 700 megawatts of fully approved power, with an interconnection that is direct to Riot with no intermediaries between us and the utility.
This is a direct, non-interruptible, evergreen connection and is supported by a 700-megawatt substation already on-site and operating. In addition, we have secured sufficient water supply with ample capacity to fully support both our existing Bitcoin mining operations and the cooling requirements of future high-density data centers. Moving to slide six, I want to underscore why this acquisition is such a significant value driver for Riot. This was not just a land purchase. It was a strategic imperative to unlock the full potential of the Rockdale site and a critical step in our capital allocation strategy. First and most importantly, this transaction enables us to fully utilize Rockdale's power capacity for data center leasing. As we engaged with market participants, it became clear that top-tier tenants and institutional financing partners require absolute certainty regarding land control.
By removing the ground lease, we have eliminated the primary barrier to entry, allowing us to market the full 700 megawatts to the world's most demanding tenants. Second, outright ownership fundamentally changes the risk profile of this asset. Under the previous structure, we faced lease expiry risk, where ownership of our improvements could eventually revert to the landlord. That risk is now eliminated. Riot owns this site in perpetuity, strengthening our long-term strategic optionality and ensuring that the value we create accretes to our shareholders. Third, we are looking beyond the current capacity. As part of this acquisition, we have secured easements that provide a pathway to expand power capacity beyond the existing 700 megawatts. In a power-constrained world, having the land and rights to expand capacity at an existing operational substation is a valuable asset.
This positions Rockdale not just as a 700-megawatt data center campus, but as a site with a pathway to grow even larger as grid conditions permit. Finally, there is a substantial and immediate financial benefit to Riot. By acquiring full ownership, we are eliminating approximately $130 million in future rental payments that would have been due over the remaining term and extension options. This immediately improves the operating cost profile of the site and enhances the returns on all future developments. I would now like to walk through the specific terms of our lease agreement with AMD, an investment-grade tenant with an S&P credit rating of A, and highlight why this is such a compelling transaction for Riot, both financially and strategically. We have entered into a 10-year lease to provide customized data center space at our now fully owned Rockdale site.
The lease includes three five-year extension options, giving this partnership a duration of up to 25 years. Securing a tenant of AMD's caliber, a global leader in the AI ecosystem, validates both our site and our development capabilities. The initial deployment under this lease is 25 megawatts of critical IT load, which will be housed in a retrofit of our existing Building G. This allows us to deliver capacity quickly and cost-effectively. We are delivering this in two phases to de-risk execution. The first 5-megawatt phase will be ready for occupancy later this month, with the remaining 20 megawatts following in May. The capital required for this initial deployment is just under $90 million. This translates to approximately $3.6 million per critical IT megawatt, a fraction of traditional new build cost.
By leveraging power distribution and building envelope already in place at Rockdale, we can deliver uniquely tailored capacity with highly attractive unlevered returns. From a financial perspective, the total contract value for this initial 25-megawatt deployment over the 10-year base term is $311 million. We expect this to generate average annual net operating income of approximately $25 million, representing highly visible contracted cash flows. Crucially, this 25 megawatts is just the first step. The lease includes an expansion option for AMD to take an additional 75 megawatts of critical IT load, which can be accommodated within the remainder of Building G and Building F using the same capital-efficient retrofit model. Beyond that, AMD holds a right of first refusal on an additional 100 megawatts of capacity.
If AMD were to fully exercise both the expansion option and the ROFR, their total footprint at Rockdale would reach 200 megawatts of critical IT load. In summary, this lease accomplishes several critical objectives. It establishes Riot as a credible operator for hyperscale tenants. It creates immediate, high-margin, long-term contracted cash flows, and it proves that our methodical approach to development can provide unique, flexible solutions for the most demanding tenants in the market. With that, I will hand the call over to Jonathan Gibbs, our Chief Data Center Officer, to explain how we designed this deployment to meet AMD's technical requirements while maintaining our capital efficiency advantage.
Thank you, Jason. I want to walk through the timeline and the execution milestones for our AMD deployment, which demonstrates both the speed of our data center team and the depth of this partnership. We signed the lease with AMD earlier this month, and we are moving immediately into delivery. The first phase, 5 megawatts of capacity, will be delivered and commenced later this month. The second phase, 20 megawatts of capacity, will be delivered in May 2026, bringing the total initial deployment to 25 megawatts of critical IT load. This phased approach allows us to accelerate delivery timelines and lease payments and demonstrates our ability to meet AMD's technical requirements and delivery timelines. From there, we are positioned to scale immediately.
Because we are utilizing the same retrofit model of Building F and G, we can execute the expansion options and ROFR capacity with the same speed and capital efficiency as this initial phase. We have the team, the plan, and the infrastructure ready to support AMD's growth as fast as they need it. The speed, scale, and quality of this deployment reflect the strength of the data center team we have assembled and the strategic advantage of our existing infrastructure at Rockdale. We are moving quickly. We are executing efficiently, and we are building a long-term partnership with one of the most important players in the AI ecosystem. I want to start by emphasizing why AMD chose Riot. While our power and assets were critical, the deciding factor was the strength and credibility of our data center team.
AMD needed a partner who didn't just have power and land, but one who had the technical expertise to execute on a complex, high-performance requirement. The fact that a world-class technology leader like AMD trusted us with their critical infrastructure is the strongest possible validation of the credible team we have built. Slide 10 highlights the specific value proposition we delivered. First, this is a bespoke solution. Our engineering team worked hand in hand with AMD to design a facility that meets their specific, focused redundancy requirements. We did not simply offer a standard off-the-shelf data center product. We collaborated to build exactly what they need for their specific AI workloads, optimizing for both performance and cost. Second, by retrofitting the second Building G, we are leveraging the power distribution and building shell already in place.
This is the key driver behind our speed to market and what allows us to deliver the first five megawatts later this month, a timeline that greenfield developers simply cannot match. Third, this efficiency translates directly to attractive economics. Because we are repurposing infrastructure and optimizing the design, we have achieved a significantly lower cost of development. This allowed us to pass through those savings to AMD in the form of a highly competitive lease rate, while still generating the attractive returns Jason had outlined. This lease is proof that Riot has the technical capability to deliver for the most demanding tenants in the world. I will now turn the call over to Jason Chung.
Thank you, Jonathan. I want to address how we intend to finance these announcements and demonstrate why Riot is uniquely positioned from a financing perspective. Our approach to capital allocation is methodical, disciplined, and designed to maximize returns while minimizing dilution and maintaining financial flexibility. Riot is exceptionally well capitalized. As of the end of the Q3 of 2025, we held over $2 billion in total liquidity, consisting of approximately $400 million in cash and nearly 20,000 Bitcoin. In addition, our scaled Bitcoin mining operations continue to generate significant monthly cash flows, providing ongoing liquidity to support our business. Our capital allocation strategy follows a disciplined, three-phase approach that optimizes our cost of capital and enables efficient scaling. Phase one: initial development funding.
For the projects we have announced to date, including the Rockdale land acquisition, the AMD 25 IT megawatt buildout, and the previously announced initiation of the development of two core and shell buildings representing a total of 112 IT megawatts at Corsicana, funding for these capital expenditures has and will come through the sale of Bitcoin from Riot's balance sheet. We anticipate that these initiatives require approximately $400 million in total capital, and we've already begun executing on our Bitcoin sales strategy. In December, the Rockdale land acquisition was funded entirely through the sale of approximately 1,080 Bitcoin. For the AMD 25 IT megawatt buildout and Corsicana 112 IT megawatt core and shell development, these initiatives will be funded through the sale of approximately 3,134 Bitcoin based on current market prices.
This represents only 17% of the Bitcoin currently on Riot's balance sheet, allowing us to maintain substantial exposure to Bitcoin while funding these important growth initiatives. As our development projects progress, we will continue to strategically liquidate Bitcoin to fund the capital expenditures in a disciplined and measured manner. Phase two: tenant acquisition and project financing. As we secure high-quality leases with creditworthy tenants, we intend to utilize non-recourse project-level debt financing to fund the majority of data center CapEx costs. This approach will allow us to enhance returns on invested capital and preserve balance sheet liquidity by efficiently leveraging the contracted cash flows from our leases. By securing long-term leases with investment-grade tenants like AMD, we create assets that are highly attractive to institutional lenders and which can be financed at favorable rates.
We are targeting project financing for new leases at loan-to-cost ratios in the range of 80%, allowing us to externally finance a significant portion of total development CapEx while retaining full ownership of the assets and their ongoing cash flows. This financing approach serves multiple strategic objectives. First, it optimizes our cost of capital by replacing equity with lower-cost debt. Second, it allows us to recycle capital, deploying into our next wave of development projects and therefore scaling our business in a capital-efficient manner. And third, it allows us to maintain ownership of high-quality, cash-flowing assets that will generate ongoing, high-margin, and stable returns for years to come. Phase three: capital recycling and compounding growth. As we execute this strategy across multiple projects, we create a powerful compounding effect. For projects like AMD, where we will deploy equity first, we can refinance stabilized assets to recover capital.
For future projects, we will use project financing upfront to minimize equity deployment. In both cases, the result is the same. We preserve capital for redeployment while retaining ownership of cash-flowing assets that generate ongoing operating income to support additional development or be returned to shareholders. Going forward, our capital allocation priorities are clear. First, deploy capital into high-return data center development projects and secure creditworthy, long-term tenants. Second, utilize project finance to fund development CapEx in order to optimize our cost of capital and recover equity. Third, recycle capital into the next wave of development, and fourth, maintain balance sheet flexibility and strategic optionality. The ability to fund these critical near-term capital expenditures using our existing balance sheet represents a significant competitive advantage, allowing us to move quickly, capitalize on market opportunities, and maintain financial flexibility as we scale our data center platform.
More importantly, our methodical approach to capital allocation (lease, finance, build, and recycle) positions Riot as one of the most efficient capital allocators in the data center industry and ensures that every dollar we deploy generates maximum value for our shareholders. I will now turn the call back over to Jason Les.
Thank you, Jason. I want to summarize the strategic benefits these transactions deliver to Riot across three key dimensions: our data center team's credibility, the structural advantages of the AMD lease, and the validation of Rockdale as a premier development site. First, the credibility of our data center team demonstrated with AMD and the speed at which we delivered provides clear validation of our operational capabilities and execution discipline. This transaction is the foundation that establishes Riot as the partner of choice for leading data center tenants. The speed, flexibility, and capital efficiency we demonstrated with AMD, namely customizing infrastructure to their specific needs, providing a rapid path to quickly energize 25 IT megawatts, and providing a clear path to 200 megawatts of expansion, makes Riot uniquely attractive to demanding tenants.
We have the energized power, the balance sheet to move quickly, and the operational expertise to deliver mission-critical infrastructure on aggressive timelines. This positions us as a preferred partner for hyperscalers who need speed, certainty, and scale. Second, the AMD lease significantly de-risked our data center strategy. We have secured a 10-year agreement with a top-tier investment-grade tenant rated A by S&P. This lease is highly financeable, meaning we can leverage contracted cash flows to support non-recourse project debt as we scale our buildout. By securing a blue-chip tenant such as AMD, delivering capacity in phases, and generating cash flows within weeks, we are proving our execution capabilities while minimizing capital at risk. AMD's credit quality, the long-term nature of the contract, and the high degree of visibility of the cash flows make this an exceptionally attractive asset from a capital markets perspective.
Third, these transactions establish Rockdale as a premier site for data center development. By efficiently converting existing Bitcoin mining infrastructure into a custom data center solution for AMD, we are unlocking the significant value of our underlying assets. More importantly, we are transforming the economics of Rockdale and our entire power portfolio, moving towards long-term contracted cash flows with strong, creditworthy counterparties. This will fundamentally improve the profitability and stability of our business, driving higher quality earnings and positioning Riot for a valuation re-rating as we continue our ongoing evolution into a leading digital infrastructure platform. Collectively, the transactions announced today highlight Riot's focus on efficient capital allocation and financial discipline. We acquired a premier asset using balance sheet liquidity, eliminating future lease obligations, and are developing capacity at high investment yields, all at a fraction of greenfield cost.
This methodical approach (secure creditworthy tenants, finance efficiently, develop with discipline, and recycle capital) is how we will scale our platform and maximize value creation for our shareholders. Taken together, these milestones represent a major inflection point for Riot. We have proven our ability to execute, secured a marquee tenant, validated our development strategy, and demonstrated a disciplined, repeatable approach to capital allocation that will drive compounding value creation as we scale. I'd like to conclude by outlining how the AMD lease fits into our broader pipeline and the significant growth opportunity ahead. The timeline on this slide illustrates the rapid progress Riot has made in developing our data center business. Less than 12 months ago, we announced a strategic evolution to explore AI HPC data center development. Last June, we hired Jonathan Gibbs as Chief Data Center Officer.
Today, we are announcing the Rockdale acquisition and our first data center lease with AMD, which will begin generating lease income this month. By May, we will have delivered the full 25 IT megawatts initial deployment. But this is just the beginning. AMD holds an expansion option for an additional 75 IT megawatts and a right of first refusal on another 100 IT megawatts, bringing their total potential footprint to 200 IT megawatts at Rockdale. Looking further ahead, our fully approved power portfolio has the potential to support 1.2 gigawatts of critical IT load, which, upon full buildout, could generate substantial annualized NOI in the range of $1.6-$2.1 billion from long-term contracted leases with creditworthy tenants, cash flows which command premium valuation multiples in the market. Before we open the line for questions, I want to step back and put these announcements in context.
Today's announcements reflect four critical advantages Riot brought to bear. First, we have the assets, specifically 1.7 gigawatts of fully approved, immediately actionable power at two of the most attractive data center development sites in the United States. Second, we have the balance sheet, including more than $2 billion in liquidity to fund development. Third, we have the team. The AMD lease is a direct result of the quality and speed of execution Jonathan and our team have delivered. And fourth, we have a disciplined, methodical approach to capital allocation that sets us apart. We identify high-quality sites with energized power, immediately monetize with Bitcoin mining where economic, attract financeable investment-grade tenants, develop with capital efficiency, and recycle capital to finance the next wave of growth. This approach makes Riot not just a developer, but one of the most efficient capital allocators in the data center industry.
Importantly, our Bitcoin mining expertise provides a unique advantage. It allows us to immediately monetize power capacity while we develop data center infrastructure, generating positive cash flow and de-risking our development timeline. This optionality is something traditional data center developers simply do not have. The transactions we announced today demonstrate a disciplined, repeatable capital allocation strategy that will enable us to scale efficiently, maximize returns, and create substantial shareholder value as we build one of the premier digital infrastructure platforms in the industry. With that, we will now open the call up for questions.
Thank you. As a reminder, to ask a question, please press star 11 on your telephone and wait for your name to be announced. To withdraw your question, please press star 11 again. Please stand by while we compile the Q&A roster. And our first question comes from John Todaro of Needham. Your line is open.
Hey, great. Thanks for taking my question. Congrats on the lease and also unlocking Rockdale. That's huge. Represents a lot of immediate power. Question for me, can you just talk a little bit more about maybe the timeline of the AMD deal in terms of how it came about and how you guys approached the opportunity? And then if I can just get in a follow-up, walk through a little bit more of the economics of the design and then, in particular, why you guys found this project attractive and also maybe on their decision to go with Rockdale in comparison because we know, of course, that Canada is out there as well.
All right. Thank you for the question, John. I will tackle the first part, and then you can remind me of anything I missed in the follow-up. How the deal came about. This relationship with AMD originated from the sustained interest that we've seen from the Rockdale campus, and that's been driven by the scale of the opportunity there, our power position, and our ability to execute on a very quick timeline. So we've developed a process here, and we approach this opportunity with the same customer-first process that we are using across our pipeline. We first confirm the technical fit, the feasibility of what the customer is looking for. We validate the schedule and delivery certainty. We align with our partner on a commercial structure, and then we move into due diligence and finally documentation.
From a timeline perspective, this has been a process over multiple months that had several clear phases. We had early engagement and technical requirement discovery. We had iterative solutioning and pricing. We had a diligence process on our site and the infrastructure, and then finally documentation and execution planning. I want to keep here is that AMD is a strong validation of our go-to-market strategy. This deal validates the process that we've been talking about. When we see a high-quality demand signal, we can translate that into a structured process, and that process protects the certainty of delivery while still preserving our strategic flexibility for the entire campus here. Importantly, what I want to highlight is we are seeing similar momentum across our pipeline.
The outcome here with AMD reflects both the attractiveness of our assets and our repeatable approach to converting interest into actual contracted demand without compromising execution rigor. And then I think one of your follow-up questions was why Rockdale over Corsicana. AMD needed a custom solution here. They needed it on a timeline. We were able to work with them collaboratively to leverage infrastructure we already had that could be repurposed for their specific requirements. And that allowed us to move collaboratively with them over the process and start delivering revenue this month, obviously just a few weeks after the deal signing.
That's great, Jason. Thank you for that. Appreciate it.
Thank you. And our next question comes from Paul Golding of Macquarie. Your line is open.
Great. Thanks for taking my questions, and congrats on both deals. Big developments here. My first question is around the bigger picture vision for Rockdale. You have 700 gross megawatts at the site, 200 megawatts critical potentially with AMD if all options and the right of first refusal are exercised. Would you look to do something akin to Corsicana and proactively build out Tier 3 cores and shells? And just thinking about that relative to the efficient retrofit model you're doing for this initial 25-megawatt critical deployment. And then as a follow-up, just anything that you could speak to in the executed deal or right of first refusal in terms of timing or whether there are any limitations regarding negotiating with other potential tenants on the balance of the site, whether customers or competitors of AMD's? Thanks so much.
So let me first kind of just give you our vision of the layout of our Rockdale campus here. The lease with AMD is just our step in building our data center campus at Rockdale. As we think about the site today, we expect that outside of just the initial 25-megawatt deployment with AMD and their 75-megawatt expansion option that we've announced today, the bulk of the capacity that we have will be delivered in a more traditional Tier 3 format with an eye, as always, towards maximizing the value of our sites. So think the basis of design that we've talked about, deploying that over the remainder of the capacity. And we'll have some updates on that on our next update call.
But for the avoidance of doubt here, AMD's expansion option for the 75 megawatts of critical IT load capacity is allocated to AMD, and that would not be immediately available to any other customers. With Rockdale, we have 700 megawatts of total capacity. So that gives us a significant amount of runway for future expansion. And we believe that capacity is that capacity that would interest a lot of hyperscalers and major tenants. And the ROFO with AMD does not impact the viability of those discussions, whether or not AMD exercises its ROFO right or not. I think we've talked about this before, but our campus design, we have the ability, if we wanted, to segment that up and really make customers in distinct areas from each other.
So we have a lot of flexibility to maximize value here, whether it's one customer taking all the capacity or we are dividing it up. The lens that we're looking at this is always maximizing the value of the full portfolio that we have. So I think another part of your question was we are absolutely continuing to have discussions with a variety of customers to lease out our capacity at Rockdale, the full remaining capacity there, and of course, our capacity at Corsicana as well.
Great. Thanks so much, Jason.
Thank you. And our next question comes from Greg Lewis of BTIG. Your line is open.
Yeah. Hi. Thank you. And good morning, everybody. And thanks for taking my questions. And obviously, congratulations, guys. You've been working hard on this for over a year, so it's good to see. I did kind of have a question around it. It was interesting to us when you kind of talked about the press release. You led with the fact that you acquired the land. How important was that to own the land and kind of building this relationship with AMD?
Owning the land was critical. From our engagement with the market, we've got feedback from both potential customers, from financing partners, that a ground lease that had a risk variation, even though we had multiple renewal terms, was a real impediment to getting the deal done. So unlocking, purchasing the Rockdale site, and getting out of that lease was really a transformational step here, and it's why we're so proud of it today, to fully unlocking that 700 megawatts of capacity that we have at Rockdale. So yeah, Greg, it was very important for Rockdale being a viable site to buy the land, eliminate any lease risk, and put it 100% over in Riot's ownership.
Okay. Great. And then just the other one for me is, and I think it's a follow-up to Paul's question around Corsicana. I mean, clearly, that's I feel like over the last year, that was viewed as the better side of the two. And so just a couple of questions as we think about Corsicana and the potential to develop both at the same time. Curious how you're thinking about that. And then kind of curious the fact that we have AMD as an anchor. Does that at all change how we were thinking about the go-to-market, i.e., customer that we're looking to sign in at Corsicana? Thanks.
Yeah. Greg, this does not change at all how we're thinking about Corsicana. In fact, we view this announcement as just enhancing the attractiveness of Corsicana. We are continuing to engage in commercial discussions for both Corsicana and Rockdale simultaneously, and we're very confident in our ability to develop both sites at the same time. The fact that we are now a proven developer and that we have signed a lease with an investment-grade tenant, the fact that we are going to demonstrate execution, delivering capacity this month, I believe only enhances our ability to. So this improves our ability to execute across our whole portfolio.
All right. Super helpful. Thank you for the time. And congrats again.
Thank you.
Thank you. And our next question comes from Darren Aftahi of Roth. Your line is open.
Hey, good morning, and I'll offer my congratulations as well. Just to, if I may, you guys in the release talked about the aggregate build cost. I'm curious if you were to layer on the initial CapEx that you spent on the Rockdale site, what's the aggregate all-in cost per megawatt? And then as a follow-up, I know Oncor has plans to build a substation in Rockdale, I think, this year. Just any kind of updates there? And then what kind of process would you need to go through with ERCOT in order to potentially expand Rockdale's power capacity? Thanks.
Absolutely. So we laid out the key details for making this development possible: the retrofit cost of the building. I want to highlight we took mining capacity at Rockdale in order to support this space. That was equipment that has been deployed for almost five years, so fully depreciated miners, immersion tanks that have been in there for a while. But of course, we're using the other electric infrastructure at the site. So even if you are accounting for the existing infrastructure cost in place that have already been spent, this is still a very attractive deal from a yield perspective. But there's a multitude of factors here, right? We were taking off revenue-producing capacity, even though it was already depreciated and older equipment at this point. And we were moving rapidly and retrofitting infrastructure.
That's the cost that we've highlighted in the business update. So I think you look at all elements combined, it's still a very attractive deal. And that's based on leveraging existing infrastructure and the retrofit cost that we needed to just change the building that we already had.
Great. And then any update on the Oncor substation?
Oh, yeah. I'm sorry. Second part of your question. So at Rockdale, the first thing I want to emphasize is that we own the land and we control our destiny at that site. And we already have a 700-megawatt connection, which is fully approved, and we intend to use that for data center use. But to your question, as we allude to, we believe expansion to have its power capacity expanded beyond that 700 megawatts. That whole surrounding area is a hotbed for manufacturing. There is a Samsung fabrication plant in Taylor, Texas, nearby. There is a Tesla facility nearby. So there's been a significant amount of transmission infrastructure being built up in that area.
The fact that its capacity is directly around our existing site and the fact that we are proven good stewards of the grid and our ability to deploy and operate, we believe puts us in a great position to expand our existing interconnection agreement. We're in a position to capitalize on that. We also highlighted that we have the easements in place in order to bring in new supply lines of power to that site should that be approved. We will be going through that process. Of course, it's an unpredictable process. A lot of people are, of course, confused to get a power capacity. It's hard to give a timeline or much color around that. The location that we're in and our proven ability to execute and operate positions us very well in that process.
Appreciate it. Thanks, Jason.
Thank you. And our next question comes from Mike Grondahl of Northland. Your line is open.
Hey, thanks, guys. Great progress and very helpful update. Specifically here, this kind of looks like a different design for AMD than maybe what you guys envisioned several months ago. Can you talk about what some of the key differences are?
Yeah, absolutely. Let me turn that question over to Jonathan Gibbs.
Thank you, Jason. And thanks for the question, Mike. Simply, yeah, there are differences. The AMD is a purpose-built configuration, and it's different than our standard basis of design simply because we optimized it around AMD's specific high-density liquid-cooled compute requirements versus our standard design, which focuses on maximizing the ability to align investment-grade lease-up with design, construction, and the associated deployment of capital.
That's the standard. Maybe to help break this down at a high level, I can think of three differences: cooling and physical layout differences, redundancy, philosophy, and perhaps delivery-driven choices. From a cooling and physical layout, I would say that Tier 3 design is engineered to serve a broad set of IT loads efficiently. AMD, so we increased the proportion of liquid cooling and adjusted the supporting infrastructure within the white space to match how their equipment is deployed and serviced.
For the redundancy philosophy, which we talked about often here, traditional Tier 3 facility redundancy is applied in a very standardized way across major systems. AMD's configuration, while the redundancy is still very robust, it's applied more intentionally around specific components and paths that really matter most for their environment. In other words, it's not less redundancy. It's right-sized redundancy designed to produce the same outcome: high availability, reliable operations, things that our customers need, and in this case, certainly an unnecessary complexity and cost that don't add value. The third I mentioned, another one we speak about often, is delivery-driven choices, selections that support a compressed delivery schedule and reduce execution risk, all without compromising reliability. So for Riot, that includes proven equipment choices, constructability, and designs to support repeatable commissioning and operational turnover within a compressed schedule.
I think it'd be helpful possibly to also add what I mean by that. In my past nearly two decades of experience in this data center industry, it's told me that it's important to have this flexibility, but it needs to be within a disciplined framework. And what I mean by that is we can tailor the design to meet a customer's requirement, but we do it in a way that preserves what we're known for. We're known for predictable outcomes, safe delivery, schedule certainty, things like that. That said, it's probably important to note too as well that our integrated delivery model matches our customers' buying habits, right? So we have to have sales, delivery operations, all operate as one team from the start in order to be able to deliver this for our customers. At Riot, we treat design as part of execution, not just handoff.
We manage that risk proactively across the areas that are most commonly seen to disrupt projects. Labor availability, supply chain timing, utility coordination, all those types of things. With all that said, a significant strength of Riot's is our ability to be able to align our product delivery philosophy, which I just outlined, with Riot's path to help drive that execution and certainly maximize returns, of course. This sort of strategy is what gives Riot the opportunity where we feel the approach really matters and certainly why we're comfortable delivering a bespoke solution without taking on bespoke risk. I know there's a lot there, but I hope that certainly outlines the differences.
Yeah, definitely does. Congrats again, guys. Thank you.
Thank you. And our next question comes from Reggie Smith of JPMorgan. Your line is open.
Hey, guys. Congrats on the deal, and thanks for taking the question. I had one on the financing, and listening to you guys, I guess my key takeaway is that project-level financing debt, you're using Bitcoin as well to finance it. Importantly, it doesn't sound like equity is in your kind of on your roadmap at all. I guess my question is thinking about the mix of kind of using cash and Bitcoin. You guys want to say you're the first Bitcoin operator to talk about selling Bitcoin to build data centers. Maybe talk a little bit about, am I reading that right in the decision to use Bitcoin rather than cash on the balance sheet or some mixture of, and then have a follow-up on the CapEx side?
Yeah. Jason Chung, why don't you take that one?
Sure. Hey, Reggie. So the sale of Bitcoin is a fundamental part of our financing strategy. And to answer your question directly, you should expect to see further Bitcoin sales in the future. So from our perspective, we've always focused on having a strong balance sheet with multiple different financing levers, and we continue to have that today. We've got the Bitcoin, whether from our monthly production or on our balance sheet, we have a lot of cash on hand. We've got operating cash flows, access to debt facilities, and strong access to capital markets. But we will continue to look to Bitcoin sales, both from our monthly Bitcoin production and our balance sheet as our primary funding source for growth.
We will be flexible, and we're always weighing the cost of selling Bitcoin versus some of these other financing levers that we have, always with a view towards maximizing shareholder value. But for the immediate future, utilizing the Bitcoin on our balance sheet for some of these key CapEx spend and key growth initiatives, we see as the most efficient use of our capital. So you should expect in the near term going forward.
Got it. That's good to know. And then just looking at you guys had an image in the slide deck. And it looks like from what I said, that those two buildings that are allocated for AMD appear to be a little wider than the other ones. I'm not sure if that's just the image or the graphic or whatever. But is there anything unique about those two buildings in particular that made them extra, I guess, ideal for the AMD deal? And then just remind us, if you could, what the general CapEx expectations are for your standard, I guess, kind of, what do you call that, prototype build, if you will. Thank you.
Yeah. Thanks for that question, Reggie. Let me first address your CapEx, answer your CapEx question, and then I'll turn it over to Jonathan to talk about the building designs and why those were chosen here or why they were viable here. We don't have a full expected schedule, or I should say an estimated cost for what a full build-out data center development would be. I think what we've headed towards is we expect to be within industry norms. Right now, that's around $10-$13 million per IT megawatt. Partially, those costs are driven by the customer. Like we talked about here, our costs are driven by the customer's power requirements. So we are a shell seller. We're going to be effective on delivery solutions that the customer wants. And there are things that that is costing more than anything.
But your question on the building and why those projects are viable is something that's much more important than our job.
Yeah, certainly. Certainly, the width, the height, clear height of these buildings are exceptionally important for our costs when we're looking at specific product requirements and the feasibility and our ability to be able to deliver product and mitigate the risk as we go through. But I would not discount the width of the other buildings in the site that they can certainly be applicable as well. A bit more in terms of the logistics of where those are, adjacency to substations, things like that, and our ability to be able to de-risk the delivery of the product for our customers.
I think we broke up like the last 15 seconds. I don't know if anybody thought that. Maybe if you could repeat that, the last 15 seconds of what you said or so.
Certainly. Yeah. No problem. Essentially, there was some wind at play. I wouldn't discount the existing buildings for future deals. The logistics and the ability for us to be able to de-risk delivery for this timeline certainly came into play.
Got it. Thank you. Congrats, guys. Glad to see it.
Thank you. And our last question comes from Stephen Byrd of Morgan Stanley. Your line is open.
Hey, guys. Congratulations on a great deal. Very exciting to see. One question for me, I guess just given the size and the economics of recent deals that we've been seeing in the market, could you walk us through why you think a 25-megawatt deal with these economics is the best deal that Riot could achieve with one of the premier sites in the country? Could you just sort of walk us through that?
Yeah. Thank you for the question, Stephen. So we think this is a fantastic deal for a number of reasons. First, we are in the business of providing solutions for our customers. So what we've done here is leverage existing infrastructure, and we're able to meet the needs of our customer in a very complex timeline delivered by custom solutions. So what we've been doing here is a customer-based approach that delivers in a very complex and very quick timeline. We're able to deliver this revenue from this lease starting this month, this quarter. Very quick timeline. I think you have to look at the deal for all the elements combined. So yes, on a $ per kilowatt monthly rate, it is a lower price. However, that is because we had significantly less required spend on CapEx in order to achieve this deal.
So you combine that, and we have demonstrated a very attractive yield on cost for this deal. And of course, the other element is the tenant here. This is an investment-grade tenant. We are demonstrating our ability to execute for an investment-grade tenant. And we think this opens the door for different financing tools that we can use in the future around this deal and will, as I'm emphasizing, lead to other deals in the future because of what we demonstrated here. So in summary, strong economics from a yield on cost basis because the CapEx costs were eliminated and, I'm sorry, the CapEx costs were less than other deals that we've been seeing out here lately. But also delivering a customer-focused solution to an extremely strong investment-grade tenant. So we're very proud of this deal.
This is just the first step that we're taking, and we look forward to executing across the rest of our pipeline here.
Makes sense. And congrats again. Thank you.
Thank you.
Thank you. This concludes our question and answer session. I would now like to turn it back to Jason Les for closing remarks.
Thank you, everyone, for tuning in today. We are very proud of this deal. I want to thank our full team that came together across the organization to make this possible. You can imagine the amount of effort it takes to deliver capacity on this quick of a timeline. I also want to thank all of our stakeholders and partners, now our customer out there for working with Riot to deliver this. And we look forward to sharing more updates on our data center strategy with the rest of the market as we progress. Thank you all.
This concludes today's conference call. Thank you for participating, and you may now disconnect.