管理層發言
Ladies and gentlemen, good afternoon, and welcome to Applied Digital's Fiscal Third Quarter 2026 Conference Call. My name is Abby, and I will be your operator today. Before this call, Applied Digital issued its financial results for the fiscal third quarter ended February 28, 2026, in a press release, a copy of which has been furnished in a report on Form 8-K filed with the Securities and Exchange Commission, or SEC, and will be available in the Investor Relations section of the company's website. Joining us on today's call are Applied Digital's Chairman and CEO, Wes Cummins; and CFO, Saidal Mohmand. Following their remarks, we will open the call for questions. Before we begin, Matt Glover from Gateway Group will make a brief introductory statement. Mr. Glover, you may begin.
Thank you, Abby. Hello, everyone, and welcome to Applied Digital's Fiscal Third Quarter 2026 Conference Call. Before management begins formal remarks, we would like to remind everyone that some statements we are making today may be considered forward-looking statements under securities laws and involve a number of risks and uncertainties. As a result, we caution you that there are a number of factors, many of which are beyond our control, which could cause actual results and events to differ materially from those described in the forward-looking statements. For more detailed risks, uncertainties, and assumptions relating to our forward-looking statements, please see the disclosures in our earnings release and public filings made with the SEC. We disclaim any obligation or undertaking to update forward-looking statements to reflect circumstances or events that occur after the date the forward-looking statements are made, except as required by law. We will also discuss non-GAAP financial metrics and encourage you to read our disclosures and the reconciliation tables to the applicable GAAP measures in our earnings release carefully as you consider these metrics. We refer you to our filings with the SEC for detailed disclosures and descriptions of our business as well as uncertainties and other variable circumstances, including, but not limited to, risks and uncertainties identified under the caption Risk Factors in our annual report on Form 10-K and our quarterly reports on Form 10-Q. You may access Applied Digital's SEC filings for free by visiting the SEC website at www.sec.gov. I would like to remind everyone that this call is being recorded and will be available for replay through a link available in the Investor Relations section of Applied Digital's website. Now I'd like to turn the call over to Applied Digital's Chairman and CEO, Wes Cummins. Wes?
Thanks, Matt, and good afternoon, everyone. Thank you for joining our fiscal third quarter 2026 earnings conference call. This quarter, we continued to differentiate ourselves in the industry. Over 2 years ago, we were one of the first companies to recognize the surging demand for large-scale, high-power density AI data centers and broke ground on our first 100-megawatt facility. This early investment is now paying off in 2 important ways. First, we now operate one of the only 100-megawatt direct-to-chip liquid cooled data centers in the world online today. This, coupled with key learnings, gives us the experience and the ability to demonstrate to major hyperscalers and others that we can execute on time and deliver fully functional state-of-the-art facilities. Second, what investors are seeing today in our reported financials, including over $44 million in adjusted EBITDA for the quarter across our core businesses, is just the early stages of what we expect to achieve. In the HPC segment, this first 100-megawatt building represents only 1/10 of the total capacity we currently have under construction. While there are many variables and uncertainties involved in developing large-scale power infrastructure such as new power plant construction, transmission lines, and regulatory approvals, we currently estimate that we have contracted only a small fraction of our long-term power potential. Turning to execution. All buildings under construction at PF1 and PF2 are progressing on time and on budget. Building large-scale data centers through a North Dakota winter is no small task, but with years of experience and thousands of skilled professionals on site, along with trusted partners such as McGough, ABB, Adolfson and Peterson, and BASX, we're executing effectively. At Polaris Forge 1, the 400-megawatt CoreWeave campus, the first 100-megawatt building is now operating, and our 1,200 skilled craft professionals are progressing in parallel on 2 new 150-megawatt facilities. At Polaris Forge 2, the 200-megawatt investment-grade hyperscaler campus, both buildings are advancing well with foundations largely complete and work now shifting to precast direction as well as mechanical, electrical, and plumbing trades mobilizing for interior fit-out. During the quarter, we also broke ground on Delta Forge 1, a 300-megawatt critical IT load AI factory campus spanning more than 600 acres in a strategic Southern U.S. market with initial operations expected in mid-2027. We have some great videos reflecting our progress on X and LinkedIn pages. Last quarter, we shared we were actively marketing 3 potential sites. During the quarter, we made the decision to delay the South Dakota site as we evaluate its long-term viability and explore opportunities to reallocate the associated power agreements. As a result, we have brought 2 additional sites into the pipeline and are now actively marketing 4 development sites in total. These include Delta Forge 1 in the Southern U.S., an additional site in North Dakota, and 2 sites in unnamed states. Subject to receiving all necessary approvals for these sites and grid power capacity across these locations, the total grid power capacity across these locations is approximately 1 gigawatt, and the campuses are in various stages of negotiation with some in advanced stages. While there can be no assurances we will successfully match any specific site with a customer, and many variables must align to bring a new data center campus to fruition, we believe it is helpful to provide investors with visibility into our expanding development pipeline and future growth opportunities. Turning to our data center hosting business, where we host 2 sites for Bitcoin mining. This segment has our highest return on assets, and we had another strong quarter. Many of the sites in the U.S. are being converted to data centers and thus, anyone who has high-performance powered sites is sitting on very valuable assets, especially in lower-cost regions with a great climate like the Dakotas. Now turning to cloud. As discussed last quarter, after reviewing strategic options, the Board announced plans to separate Applied Digital Cloud and combine it with EKSO Bionic Holdings through our proposed business combination to form ChronoScale Corporation, a dedicated accelerated compute platform for GPU-optimized AI infrastructure. We believe this is an ideal time to pursue this transaction, particularly in light of the significant recent increases in demand and GPU rental rates we are observing in the market. This move positions the cloud business to raise capital independently, create differentiation, and drive accelerated growth with the long-term goal of spinning the business to our shareholders. With that, I'll turn the call over to our CFO, Saidal Mohmand, for a detailed review of financials.
Thank you, Wes, and good afternoon, everyone. This quarter, we realized a full quarter of lease revenue from our 100-megawatt data center in the HPC hosting business. Going forward, we expect revenues to ramp significantly over the next 12 months as our 2 150-megawatt buildings come online. We have also completed the majority of our equity and debt financing for our first 2 campuses. Note, this past March, we disclosed a $2.15 billion private offering of 6.75% senior secured notes due 2031 to support our 200 megawatts of critical IT load at our Polaris Forge 2 campus. We now have only one remaining tranche of debt to place for the final 150-megawatt building at our Polaris Forge 1 site. We have some very positive news for our debt and equity investors. On March 30, 2026, we executed amendments and related agreements with CoreWeave that included restructuring portions of the ELN-02 and ELN-03 leases through a special purpose vehicle, or SPV, subsidiary wholly owned by CoreWeave. This included delivering unconditional springing parent guarantees from CoreWeave, Inc. and securing a $50 million letter of credit. These enhancements were supported by CoreWeave's SPV receiving an investment-grade A3 rating, a meaningful improvement from its previous BB rating. We believe this improved credit support not only derisks the existing 250 megawatts lease capacity, but should also help lower our cost of capital when placing the remaining 150-megawatt tranche, although there can be no guarantees on timing or pricing. Longer term, we expect these enhancements will position us well to refinance that debt at more attractive rates in the future. We are actively working with top institutions to place that debt at the right time and at the lowest possible cost of capital. From here, we believe we have a straightforward financing model. We have access to $4.1 billion in preferred equity from Macquarie Asset Management following a mutually agreed upon executed lease with an investment-grade hyperscaler. We would then follow a similar approach for the debt financing. This structure allows Applied Digital shareholders to retain over 85% common equity ownership of future sites while significantly reducing reliance on the public capital markets. Now let's turn to the quarter. We reported total revenues of $126.6 million, a 139% increase from the comparative prior quarter. Our HPC hosting business generated $71 million of revenue, consisting of $44.1 million related to base rents, $18.9 million related to tenant fit-out services, and $8.1 million related to power pass-through arrangements and other ancillary revenue streams. This resulted in segment operating profit of $17.6 million. The Data Center segment, which operates our crypto data centers, had another strong quarter with $37.5 million in revenue, up 7% year-over-year. We are very pleased with this business, which continues to deliver the highest return on assets in the company, generating $13.9 million in operating profit in just 1 quarter, and that's on $119.6 million in reported assets. Given that the cloud business is merging with EKSO and that we will be a majority holder, we have consolidated cloud's revenues of $18.1 million for the quarter. We also recorded a $59.7 million noncash write-down of the business due to the reclassification from held for sale. As a result, this segment reported a loss of $52.2 million. As the cloud business is pursuing a separate strategy from our core business and will be placed in a separately publicly traded company, we have excluded the segment from our non-GAAP results. Cost of revenues increased by $23.7 million for the quarter. This increase was primarily driven by $18 million in tenant fit-out services, an increase of $4.8 million in personnel expenses, an increase in $4.1 million of energy costs associated with our data center hosting business, and an increase of $2 million in D&A expense. These increases were partially offset by a decrease in $5.2 million in lease and lease-related expenses. SG&A expense increased $57 million to $79.7 million this quarter. The increase was primarily driven by $39.3 million in stock-based compensation due to increased headcount and performance rewards, $8.6 million in professional service expenses, mainly related to legal support for one-time transactions and business growth, $5.1 million in personnel expenses also related to the increase in headcount, and $8 million in other SG&A expenses. These increases were partially offset by a decrease of $3.9 million in lease and lease-related expenses. Net interest income was a positive $2.4 million this quarter. This was primarily driven by a $19.3 million increase in interest income from our money market accounts. Net loss attributable to common stockholders was $100.9 million or $0.36 per share. Adjusted net income was $33.2 million or a positive $0.09 per share. Depreciation for the quarter was approximately $18.5 million, and adjusted EBITDA for the quarter was $44.1 million. Turning to the balance sheet. We are exceptionally well positioned. We ended the quarter with $2.1 billion in cash and cash equivalents against $2.7 billion in debt with no significant maturities due in the next 2 years and approximately $1.6 billion in equity. Our goal is to maintain one of the strongest balance sheets in the industry throughout the majority of the construction phase, and we believe we are achieving those goals. Now I'll turn over the call to Wes for closing remarks.
Thanks, Saidal. We are seeing a clear acceleration in demand for high-performance AI data center capacity as hyperscalers are as aggressive as we've ever seen them. While some have questioned the slower pace of new lease signings industry-wide, I want to be clear. There is significant demand for credible, well-located data center sites almost anywhere in the world. Just 3 months ago, we referenced approximately $400 billion in annual capital expenditures from the largest U.S. hyperscalers. That figure has now been reported to have increased to nearly $700 billion. This represents one of the largest investment cycles in U.S. history compressed into an extremely short time frame. These enormous investments highlight the intense pressure on power and infrastructure. Leaders such as Elon Musk have publicly stated that even if we utilize all available excess power on the grid, it will still not be enough to meet the demand for new data centers. This concern is so significant that it has driven major strategic moves across the industry, including efforts to develop data centers in space. We believe these trends only increase the long-term value of high-quality, low-cost sites like those that we operate today. Recognizing this dynamic early, we are advancing our own power strategy through support of Base Electron, an independent power producer. Base Electron will work with Babcock and Wilcox to build a power plant that will supply initially roughly 1.2 gigawatts of natural gas-fired generation capacity to the grid in the Dakotas region. This power will be in front of the meter and developed in partnership with regional utilities. We are providing the support based on insights gained from discussions with some of the largest hyperscalers in the world. We believe that if we build it, they will continue to come to our region. The objective is to add reliable power to the Dakotas and help contain electricity costs for consumers, reduce the need for utilities to raise capital, and allow for the development of new large-scale sites in the region. Applied Digital is providing limited credit support through a guarantee on the project. As Base Electron successfully raises at least $50 million in financing or completes an IPO, Applied Digital's guarantee will be terminated. In exchange for the guarantee, Applied Digital shareholders will own approximately 10% of this new company. We believe we are once again ahead of the curve by supporting an IPP just as we were 2 years ago when we began building one of the first state-of-the-art liquid-cooled AI data centers. We expect to see more companies follow this model of developing dedicated power solutions in the coming years. We're not only investing in infrastructure and power, we're also investing in our communities through Applied Digital Cares, where we recently awarded our first round of grants supporting important local initiatives in education, health, wellness, innovation, and public safety, including upgrades for local fire departments. In closing, we recently celebrated our 5-year anniversary. In that short time, we have successfully navigated multiple business lines, built billion-dollar state-of-the-art data centers in remote locations, and secured approximately $16 billion in contracted lease revenue. Given the significant demand we are seeing, our focus is on scaling the platform, where new leases will continue to be a natural outcome as we expand across campuses in a disciplined, repeatable way. Our long-term vision is to build a dominant data center region in the Dakotas with multiple hyperscalers while also expanding into strategic locations across the United States. Every new campus we secure is intended to create one of the most valuable annuity streams available, a 15- to 30-year revenue stream backed by some of the strongest credits in the world. Once the site is secured, we will focus on growing that site. From a financial perspective, we know that today our cost of capital is higher than it should be, but we plan to refinance that down over time as we shift from project finance loans into ABS or equivalent market at lower rates. We believe that should be the key tipping point where shareholders' return on investment will significantly ramp and the majority of our shareholder value will be unlocked. We believe our first 2 hyperscaler partnerships are just the beginning. We remain confident in our ability to exceed our long-term goal of $1 billion of NOI within 5 years. To drive accountability, we've implemented new internal targets for our leadership team at both $1 billion and $2 billion of NOI levels. With that, operator, we're happy to open the call for questions.
分析師問答
And our first question comes from the line of Mike Grondahl with Northland Securities.
Two questions. First, Saidal, could you provide more insight into the restructured leases at PF1? If you had to estimate, what kind of cost savings do you think could arise from refinancing? Secondly, Wes, can you elaborate on how the demand environment has changed over the last 90 days and what discussions you've had with various hyperscalers?
Saidal, why don't you go first?
Yes. Regarding the lease restructuring, there has been a notable improvement in the pricing of the outstanding bonds. This change is driven by a couple of factors. First, the offtake for CoreWeave is considered a high investment-grade offtake, providing a beneficial perspective. Additionally, there is a lockbox arrangement that ensures lease payments, which are crucial for operating the GPUs, are prioritized. We are contractually obligated to receive those payments through their financing facilities, offering us a financial advantage. Furthermore, we maintain a parent guarantee from CoreWeave, enhancing our lease position with a minimum credit improvement. There are also other protective measures in place, like letters of credit. Overall, the situation has improved significantly. CoreWeave has managed to lower their financing costs, and while we can't guarantee specifics, we anticipate that our borrowing costs will align more closely with those of an investment-grade tenant as we progress. Current trading levels of our bonds appear to be quite positive.
Mike, regarding demand, we consistently observe fluctuations quarter-to-quarter, where some players are very aggressive while others take a step back, and these patterns can last anywhere from six to twelve months. However, we find that every hyperscaler we target remains actively engaged in the market, although this varies by location, making it difficult to provide a comprehensive market perspective. What I can share is based on our observations for the areas we are focusing on, where we see multiple hyperscalers expressing interest at each location. In terms of contracting capacity, we have Polaris Forge 1 and 2, and now Delta Forge 1, with two customers currently at those campuses. My primary consideration is to diversify our customer base. Instead of signing additional contracts with existing customers, my goal is to acquire new customers at those sites. Although it might be straightforward for me to secure more business from a current client, my priority is to diversify. This focus is very specific to Applied Digital. Additionally, we aim to increase our total contracted revenue to 70% that is investment grade. Presently, we have $16 billion in total contracted revenue, consisting of $11 billion from CoreWeave and $5 billion from an investment-grade hyperscaler. You can calculate how we plan to achieve this goal based on our current split and the campuses we are actively marketing, where negotiations are in advanced stages. We feel optimistic about the assets we possess, and it's crucial to differentiate our assets from others in the market. All our marketing efforts are based on grid power, which remains our top priority, surpassing any off-site generation options. We are confident in our assets and are focused on securing the right tenant and contract. While investors may want to know how quickly we can finalize and announce these deals, we do not impose deadlines on ourselves; our priority is to ensure we partner with the right customer and establish the right contract, and I believe we can achieve this with the campuses we are currently marketing, as they are strong assets.
And our next question comes from the line of Darren Aftahi with ROTH Capital.
Congrats on all your progress. Two things, if I may. So Delta Forge 1, your commentary about potentially being operational mid-2027. I guess what does that say or infer about when a lease effectively needs to be signed? And then on your last call, you talked a fair amount about being in exclusivity with a hyperscaler, 3 sites, 900 megawatts, if my memory serves me correct. Are you still in exclusivity with that potential tenant? And is there any update on that project in general?
Yes. Regarding Delta Forge 1, I expect to see a lease signed in the near term to meet our goal. We've been working on this for a few months and have made significant progress. I'm confident about finalizing the lease in time to meet the RFS date. Additionally, we had to pause development at the South Dakota campus because we didn't receive the expected tax exemption from the legislature in this session. However, we are actively working on two other sites and still have three sites in exclusivity with a hyperscaler. We're optimistic about these assets and aim to sign those leases this year, ideally in the near term, but we will not rush into signing a subpar lease just to make an announcement. Overall, we are pleased with the progress on these sites.
And our next question comes from the line of George Sutton with Craig-Hallum.
So for those of us that are non-fixed income guys, I wondered if you could just walk through what it generally means if you go from BB to single A, if you were to go into the refinance market, what kind of spread differential is there?
Yes, that’s a great question. Currently, single A ratings are considered investment grade, with spreads typically below 300 basis points, ranging from the low 2s to mid-2s, depending on the structure and placement of the lease. Historically, you can think of it as being in the mid-2s. For BBs, the spreads for single Bs to BBs usually range from 350 to 450 basis points, again depending on the offtake and the specifics of the contract structure.
Okay. That's quite significant. Wes, I'm interested to know about the sites you're currently working on. Some of these have six-month moratoriums imposed by local counties. My understanding is, and please correct me if I'm mistaken, that as time progresses, the ultimate value you can obtain from these contracts and properties tends to increase. In other words, while we are all anticipating the near-term deals, if these extend a bit longer, the value you can capture could ultimately be greater. Is that accurate?
George, we've observed a consistent trend so far, and I believe that's accurate. Regarding the moratoriums, we are actively addressing those and feel positive about progressing through the educational aspect. Specifically, in North Dakota, where we have an operational site, we've accumulated concrete evidence from the Polaris Forge 1 campus in Ellendale, showcasing both its economic advantages and its positive impact on ratepayers on the grid. Since the site has been in operation, we've saved ratepayers approximately $31 million due to the effective use of the infrastructure and strategic site selections. Our collaboration with the community has received positive feedback, making it easier for us to advance our efforts in that state, including managing the moratoriums and zoning issues. We are optimistic about our continued expansion in North Dakota. Looking at the broader picture, I believe our sites possess premium utility power available by 2027, and there's significant demand for such assets. This year, our primary objective is to achieve a total contract value with 70% classified as investment-grade and 30% in other categories, which implies substantial growth in total contracted value is necessary to meet that goal. We aim to market four new campuses, with a vision to reach a total of five or six campuses, all of which we expect to grow over time, some significantly. This presents a clear pathway to over 5 gigawatts of critical IT load across our campuses as we progress. From my perspective, once we secure a customer at a particular campus, it becomes easier to either expand that customer’s capacity or onboard additional customers at that location. Therefore, when I assess the future, I consider both the potential for new sites as well as the expansion opportunities within our existing campuses. If we can establish a clear trajectory towards 5 or 6 gigawatts across the contracted campuses and those we're aiming to secure this year, it will provide a strong growth pathway for the company that is relatively stable and likely easier to manage than continually launching new campuses.
And our next question comes from the line of Nick Giles with B. Riley.
Nice job, guys. So Wes, I think you mentioned you're marketing 4 sites, one of which is Delta Forge 1, and 2 unnamed sites. And I think other than maybe Garden City way back when this is new geographic exposure for you. So what drew you to the south? And what kind of contrast would you draw between it in your Dakota sites? And was this really a result of customer indications or more applied led?
So, Nick, our primary focus is always on the availability of power when selecting our sites. Following that, we consider fiber availability and several other factors. One important factor is the market saturation. We look at how crowded the market is because higher density can lead to an easier acquisition of customers due to the existing infrastructure. Currently, crowded markets pose challenges, particularly regarding the labor force. We seek locations where we can reliably secure the necessary labor to build our sites. For instance, we are not currently considering opportunities in West Texas; instead, we are targeting states that will help us attract a different labor force. We prioritize states that are pro-business, with supportive governors and legislatures eager to promote data centers and business expansion. While there are many considerations, our foremost concerns are power availability and access to the grid. We are still primarily focused on grid power. We've explored various projects where properties claim to be powered land, but typically they have only a nearby gas pipeline for gas extraction, necessitating off-grid power generation. Although we see some of these projects progressing, the overwhelming preference among hyperscalers we aim to partner with remains grid power. Therefore, we continue to develop sites that meet this criterion and will maintain this focus in our marketing efforts.
Got it. Makes sense. I appreciate that, Wes. And then it sounds like things are on track, but just would be nice to get an update on the next building at PF1. Can you just remind us when we would first see revenue recognition? I think the guide is sometime 2026.
The RFS date for PF1 is July 1. Just to remind you how these buildings are energized, there are 6 data halls in each building, and we don't energize all 6 at the same time. In July, some of the data halls will be energized, and we expect all of them to be energized by September. Later in the year, the first building at PF2 will come online, and it will follow a similar energization process. You'll notice some revenue increase in the August quarter from the new building, and then a full quarter of revenue in the November quarter, along with some from the Polaris Forge 2 building, with close to full quarters expected in the February quarter of fiscal '27. As we progress into '27, those buildings will continue to ramp up along with additional projects like Polaris Forge 1 and Delta Forge. This quarter has been a strong one for us in terms of revenue as it highlights the earnings potential of our developments. Although we are still somewhat subscale given the large workforce required for our construction, with nearly a gigawatt and 900 megawatts under construction, you will start to see the revenue flow-through, making it clearer for you to model the earnings potential of our platform.
And our next question comes from the line of Rob Brown with Lake Street.
Congratulations on all the progress. I wanted to follow up on the commentary regarding power availability and the base electron strategy. Can you provide some insight on when you anticipate running into constraints in the North Dakota market and how you see that developing?
We have the Polaris Forge 1, Polaris Forge 2, and another site in North Dakota, totaling three sites that we plan to develop by 2028. These sites will utilize much of the excess power available in North Dakota. Towards the end of 2028, we will begin to commission some of these new power generation assets through Base Electron. This initiative is intended to address the point at which we anticipate reaching the limits of available grid power, allowing Base Electron to contribute additional power to the grid. It's important to note, as we've mentioned before, that Base Electron's business model aims to increase grid power rather than solely providing on-site generation for the Applied Digital data centers. The strategic addition of power in North Dakota is designed to enhance the overall grid, making it more resilient, which benefits all stakeholders and ratepayers in the state rather than just supplying electricity for Applied Digital. This outlines our timeline for addressing the anticipated shortfall in available grid power and the necessary additions to support the growth of these campuses. As we’ve previously indicated, the Ellendale, Polaris Forge 1, and the new North Dakota campus have significant potential for electrical infrastructure expansion, and we want to ensure that we facilitate that growth.
And our next question comes from the line of John Todaro with Needham & Company.
Congrats on all the progress. First question, Wes, you made a couple of comments about not just signing any deal; you want the right terms. And also, it's taken maybe a little bit longer than you had hoped or expected. While appreciating that the demand is quite strong, has there just been any aspect, whether terms or rates that have changed that have maybe made conversations a little bit more difficult in getting the leases done? Is there any kind of like sticking point that is coming up?
Every lease is unique, John, with different parties involved and varying requirements. Sometimes, there are a lot of negotiations regarding utility guarantees and the overall agreement, which is something new for us. This complexity doesn't apply to every lease, but it does occur in certain cases. However, I don’t think the overall landscape has shifted; each campus, with its specific utility and counterparties, has its own unique aspects. I can tell you from my team's perspective that every lease seems to take longer than I would prefer. I wish we could quickly reach favorable terms and finalize them. Nevertheless, I feel confident about our progress and our ability to secure many campuses this year. We will ensure that we partner with the right tenants and establish the correct structure. While it's difficult to assess any market-wide changes, each site always presents its own specific details and nuances.
Understood. Appreciate that. And then maybe one for Saidal. Just trying to maybe reconfirm the cadence of the fit-out service revenue. Has all that been recognized now? Or should we expect some more in the coming quarter to contribute?
Yes. So for ELN-02, a majority of the fit-out revenue has been recognized. There will be a small amount remaining for the first building. And then towards the end, you'll see some ramp on ELN-03 or the second building in PF1 start to ramp up. Once again, timing, right, timing can be lumpy from quarter-to-quarter, and it's a low-margin line item that's nonrecurring.
Correct, around like 5% or so, right?
That's fair. Yes, correct.
And our next question comes from the line of Michael Donovan with Compass Point.
Congrats on the progress. Saidal, could you walk us through what still needs to happen between now and June 30 for the PF2 financing escrow tied to the $2.15 billion of 2031 notes to be released?
Yes, that's right. Great question. The ESA needs to be finalized between the utility and the involved parties, and that is progressing as planned. For example, there has been recent activity regarding the substation construction. The most significant challenge has been the construction of the substation. We signed a construction agreement for it last October, and the progress on the substation is looking really good.
Appreciate that, Saidal. And I guess for Wes, what was the strategic rationale for structuring Base Electron outside of Applied rather than owning the generation directly?
We put a lot of thought into this question. The power generation business is fundamentally different from the data center business, so we felt it wasn't wise for Applied Digital to take on the risk of building power generating assets to expand our data center capacity. Looking at the risk profiles, at Applied Digital, we sign long-term leases for our data centers, securing 15 years of lease payments regardless of whether the customer uses the facility. In contrast, while power generation supports the data center, if the data center isn't being utilized, our customers still owe lease payments, but there is no power being consumed. This creates distinct risk and return profiles for each business area. Thus, we established Base Electron as a separate company, allowing Applied Digital shareholders to benefit from its success without bearing the risks associated with potential failures. We anticipate it will eventually go public, giving investors the option to engage with power generation, data centers, or GPU cloud through ChronoScale's spinout. This approach provides more choices for investors instead of integrating these risks into Applied Digital. We believe it was better to keep it separate, allowing it to develop its own funding structure and secure the necessary capital to build these assets. This was the main reason for not simply folding it into Applied Digital as another unit.
And our next question comes from the line of Paul Meeks with Freedom Capital Markets.
Excuse me if this was asked and answered, but do we still have 100 megawatts at PF2 that is still uncontracted?
That's correct.
And going forward, you'll make an announcement, not to who the hyperscaler may or may not be. Those are always easy to figure out, but you will make an announcement when it is contracted.
Yes. And we do expect that to be contracted in the near term.
Next question is for both PF1 and PF2. Are you sticking with the site NOI margins that I think you last showed in the presentation last fall?
That is correct. Yes, that is correct. So high 80s to 90s is the range that we've been operating at on a cash basis.
Right. And last quick one. When you meet this 5-year NOI target, you're going to start with the project financing and then switch over time. But once we get there, 5 years out, what does your firm's capital structure look like?
So this is Saidal. So there are a couple of different ways. So one, as you complete the construction period and construction risk is removed from the overall financing, your cost of capital comes down. If you look at some of the private peers, leverage tends to be very high in excess of 10x NOI. We feel it's a prudent way to be in that 5 to 6x NOI leverage, which when you're against, call it, high investment grade and investment-grade credits for long-term leases with escalators, that's very prudent. So I think as you get to that 5-year mark, once our platform is fully humming. And as we surpass our NOI goals of $1 billion and $2 billion of NOI, that, call it, 5 to 6 turns of leverage is prudent. Now once again, the caveat being there's always going to be new potential opportunities that we're building out, and we are also opportunistic across the spectrum for financing with the view of always having paper that is constructive both for shareholders and obviously for other stakeholders in the company as well.
And ladies and gentlemen, that concludes our question-and-answer session. I will now turn the conference back over to Wes Cummins for closing remarks.
Thanks, everyone, for joining us today, and I want to make sure that I thank all of our employees who are working overtime to make all of this a reality for the company and its shareholders and look forward to speaking with you in July.
Ladies and gentlemen, that concludes today's call, and we thank you for your participation. You may now disconnect.