Applied Digital Corp (APLD) 2023 Q1 法說會逐字稿

Good afternoon, and welcome to Applied Blockchain's first fiscal quarter 2023 conference call. My name is Paul and I will be your operator today. Before this call, Applied Blockchain issued its financial results for the first quarter of fiscal 2023 ended August 31, 2022, in a press release, a copy of which will be furnished in a report on Form 10-Q filed with the SEC and will be available in the Investor Relations section of the company's website.

Joining us on today's call, are Applied Blockchain's, Chairman and CEO, Wes Cummins and CFO, David Rench. Following their remarks, we will open the call for questions. Before we begin, Jeff Grampp from Gateway Group will make a brief introductory statement. Mr. Grampp, please proceed.

Thank you, operator. Good afternoon, everyone, and welcome to Applied Blockchain's fiscal first quarter 2023 conference call. Before management begins their formal remarks, we would like to remind everyone that some statements we're making today may be considered forward-looking statements under securities law 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 Securities and Exchange Commission. 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 reconciliation tables to applicable GAAP measures in our earnings release carefully, as you consider these metrics. We refer you to our filings with the Securities and Exchange Commission 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 IPO prospectus.

You may get Applied Blockchain Securities and Exchange Commission's filings for free by visiting the SEC website at sec.gov. I would also like to remind everyone that this call is being recorded and will be made available for replay via a link available in the Investor Relations section of Applied Blockchain's website. Now I will turn the call over to Applied Blockchain's Chairman and CEO, Wes Cummins. Wes?

Thanks, Jeff, and good afternoon, everyone. Thank you for joining us for our first-quarter fiscal 2023 conference call. Our results for the quarter were solid with revenue of $6.9 million coming in at the high end of our guidance range of $6.5 million to $6.9 million and adjusted EBITDA coming in at a loss of $1.9 million, which compares favorably with our guidance of adjusted EBITDA of $1.8 million to $2.2 million loss.

Our financial performance relative to our guidance reflects our steady execution and the consistency and predictability of our hosting only business model, which provides shareholders with revenue that is contracted out on a multiyear basis. We expect these dynamics to drive long-term high margin, sustainable cash flow generation for our company.

Also, I want to remind everyone that when evaluating our first fiscal quarter results, it is important to keep in mind that our 100 megawatt Jamestown, North Dakota data center operated at only 50% capacity during most of July and August due to the previously disclosed unexpected equipment failure at the substation powering the facility. Therefore, our first fiscal quarter results ended August 31, 2022 did not reflect a full quarter of contribution for our Jamestown facility. Since repairs were made in August, the facility has been operating as expected.

Outside of our currently operating Jamestown site, we remain laser focused on executing the build-out of our next-generation data center sites, a 200 megawatt facility in Garden City, Texas, and a 180 megawatt facility in Ellendale, North Dakota. These two facilities are expected to energize between now and mid-2023, which will bring our total hosting capacity to almost 500 megawatts.

All this capacity is already contracted out on multiyear agreements. And once fully online, we expect to be at a run-rate annualized adjusted EBITDA of about $100 million, which we think is quite compelling in the context of our current enterprise value.

I would like to highlight what I believe to be the largest risk factors to the timing of energization, which is regulatory approvals and weather. We're working through the regulatory approval process in Texas and North Dakota. Weather in North Dakota has been good to date, but we need to pour all concrete prior to extremely cold weather arriving. Our crews are working diligently to make this happen.

I want to commend the entire Applied team and especially the operations personnel for executing so well amidst an overall difficult operating environment with supply chain challenges and persistent inflation. Our team has stayed ahead of the curve by ordering certain long lead time items well in advance to avoid any meaningful delays related to construction.

Since our entry into the high-performance computing hosting industry, we have prided ourselves on the capabilities and experience of our team. So I'm pleased that we've been able to execute so well in an industry that has unfortunately been challenged to meet timelines.

I will now turn the call over to our CFO, David Rench, to walk you through our financials before providing my closing remarks. David?

Thank you, Wes, and good afternoon, everyone. Before I begin my remarks, I would like to note that like last quarter's call, since we did not have operations in the year ago comparable period, we will not be providing any year-over-year comparison.

Revenues in the fiscal first quarter were $6.9 million, which, as Wes mentioned, was at the high end of our guidance range of $6.5 million to $6.9 million. Hosting revenues were generated entirely from our first hosting facility in Jamestown, North Dakota, which was brought online in phases between the third and fourth fiscal quarters of fiscal year 2022.

Note that the Jamestown facility operated only at about 50% capacity during a meaningful part of the quarter due to an unexpected equipment failure of the substation powering the facility. Repairs were made in August and the facility has been operating as expected since the event.

Cost of revenues in the fiscal first quarter were $6.1 million, consisting of $4.9 million of energy costs to generate our hosting revenues; $800,000 of depreciation and amortization expense; and $400,000 of personnel expenses for the employees directly working at our Jamestown hosting facility.

Operating expenses for the fiscal first quarter of 2023 were $5 million, which included $4.1 million in selling, general and administrative costs, $600,000 of stock-based compensation and $300,000 of depreciation and amortization expenses.

Adjusted net loss from continuing operations for the fiscal first quarter of 2023 was $3.4 million or a loss of $0.04 per basic and diluted share based on a weighted average basic and fully diluted share count during the quarter of 93.1 million. Net loss attributable to Applied Blockchain for the fiscal first quarter of 2023 was $4.5 million or a loss of $0.05 per basic and diluted share based on a weighted average share count during the quarter of 93.1 million.

Adjusted gross profit, a non-GAAP measure, for the fiscal first quarter of 2023 was $1.7 million, which equates to 25% of revenue. Adjusted EBITDA, a non-GAAP measure, for the fiscal first quarter of 2023 was a loss of $1.9 million, which compares favorably to our guidance range for an adjusted EBITDA loss of $1.8 million to $2.2 million.

Lastly, on our balance sheet, we ended the fiscal first quarter of 2023 with $40.8 million in cash and equivalents and $14.7 million in debt. During the first fiscal quarter of 2023, we received $4.6 million in net customer deposits and $15.3 million in net deferred revenue, which collectively amounted to almost $20 million in cash inflow due to the structure of our commercial arrangements with our customers that incorporate upfront deposits and prepayments.

In certain contracts, the prepayments are amortized back to the customer over the first year of their contract with no impact to revenue recognition. But the timing of cash flow with upfront cash to us is a major benefit for the company in that it helps our CapEx funding needs as we build our data centers. We also ended the quarter with approximately 92.8 million shares outstanding. And as a reminder, we fully converted all our outstanding preferred stock to common stock concurrent with our IPO in April 2022.

Now turning to guidance. We have decided not to give an explicit range for guidance as we have in the past for the next two quarters, due to the uncertainty around timing of energization of our new facilities. This timing can have a dramatic impact on revenue and EBITDA for a quarter. We do, however, want to be transparent with our expectations.

For our fiscal second quarter of 2023, ended November, we expect our Jamestown site to generate approximately $12 million of revenue, assuming it remains at full capacity during the quarter, which is our expectation.

Any other capacity that is energized will generate revenue at a rate of about 1.2 times the amount of Jamestown on a per megawatt basis. We will announce when megawatts are energized. We do expect to resume quarterly guidance in the future, possibly as soon as Q4 of this fiscal year.

That completes my financial summary. Now I will turn the call over to Wes for closing remarks.

Thank you, David. Before we open it to Q&A, I want to cover a couple of the key goals and initiatives for our company. First, we are focused on standing up and energizing the remaining two of our first three data centers, which would give us hosting capacity of almost 500 megawatts, all of which is contracted out on multiyear agreements. Once operational, this capacity should put us at about a $100 million annualized EBITDA run rate.

Achievement of this goal over the next several quarters would be a major milestone for the company that only a year ago had no revenue. To achieve this, we simply need to execute the same business model we've been executing during calendar 2022. We already have the customers lined up and are constructing the facilities as we speak.

The next key objective for the company relates to our recent proposed name change to Applied Digital, which shareholders will vote on at our Annual General Meeting on November 10 of this year. Our name change will more accurately reflect our services and broader business offerings that serve customers that require large computing power applications.

While these large computing power applications can include Bitcoin miners, which compromise our current customer base, we see a much larger market opportunity pertaining to high-performance computing applications relating to image processing, graphics rendering, artificial intelligence and machine learning.

While we continue to see robust demand from Bitcoin miners, we believe there is value in diversifying our revenue base to adjacent industries that can still leverage our expertise and capabilities in building and operating next-generation data centers. We are having active discussions with numerous potential customers that are in various stages with power needs ranging from only a few megawatts to potentially hundreds of megawatts, all within the broader HPC landscape.

Ultimately, we are committed to providing white-glove service and best-in-class digital infrastructure to any customers that have high-performance computing needs and require access to reliable power.

I spend a significant amount of time contemplating the highest and best use case of our company's assets. And when I think about what our assets are and what we can provide, I keep returning to our low cost power contracts in our cold locations in North Dakota. These sites can be leveraged to build ultra-low-cost, high-efficiency digital infrastructure for high-performance computing applications. The first application is Bitcoin mining, but it won't stop there.

We have found what I believe will be our second large application in the machine learning area. And I expect we will have a pilot program running with the customer as earlier as this month. If all goes well, we could be in a position to scale this application early next year. This will be the first of what we believe will be many applications that will allow us to grow a non-crypto HPC business to complement our base business.

To close, it's clearly an extremely difficult time for our industry. We're in a strong position to not only survive the turmoil, but possibly take advantage of it. I'm more excited for the future of our company than I've ever been, and would like to thank all of our team members for the dedication and hard work. We're now happy to take your questions. Operator?

(Operator Instructions) Lucas Pipes, B. Riley Securities.

Thank you very much operator, and good afternoon, everyone. Thanks for taking my questions. I first wanted to touch on -- a bit on the near-term outlook, and I appreciate there are -- some lack of full visibility with the guidance. But can you touch a bit on the possible regulatory delays in Texas? What exactly is holding things up on the regulatory front there? Is it ERCOT related? Is it state related? I would appreciate additional color on that.

And then, you mentioned the pouring of concrete in North Dakota with the weather turning cold. How many days of concrete pouring do you have left and when does it get too cold to continue work on concrete? Thank you very much for your color on these two topics.

Yes. Hey, thanks, Lucas. I appreciate the question. So on the first, we're one of the first companies, I think, to go through the large flexible load program in Texas. So there's multiple things going on in Texas and multiple parties that we're dealing with. We've been -- we started on this back actually in February. We've made a lot of progress. We've had a few roadblocks come up more recently kind of in the September timeframe, but I think we're in a good position to get past those.

So you have some -- we have kind of three parties in this, and it's ERCOT, it's Oncor, and it's an entity called WETT, W-E-T-T, which is where we interconnect in West Texas, and we're working with all of these parties. In fact, we have in-person meetings with two of them this month.

I'm feeling pretty optimistic about getting close to the end there. But I don't know if you've seen in certain publications, I think maybe it was in CoinDesk and I've seen it in other areas, but there's something in the neighborhood of almost 40 gigawatts of applications in Texas.

And so that has kind of delayed, I think, a lot of people because, realistically, I think maybe 1 or 2 gigawatts of that is going to come online. But you have to separate yourself from those other projects that are just people kind of throwing something against the wall to see if they can sell it to maybe a company like ours.

So I think we're quite a ways through that. If I were to handicap this, I think we could get through and get things turned on anywhere from early November to an absolute worst case at the first part of February.

If it does get delayed out, what I would say about that is we'll be in a position instead of staged turnarounds to really energize the full site when we get approvals to do so. So it's still a moving target. I don't want to give a huge amount of details, I think I gave quite a bit right there. But we are in discussions and having meaningful discussions with all of the regulatory bodies that we need to.

In the state of North Dakota, so we're on the same schedule really that we were last year. And last year, we were able -- we completed pouring concrete, but we could go until about November 15. I think October 31 is always -- what I've been told from natives there, is kind of a good time to try to be finished by. And so we're pouring now. And generally, this goes really fast once we start, but I did want to flag it as that's one of the gating factors.

If we don't get that poured, if we get only ten of the pads poured, then we won't be able to pour again probably until March or April. And so then you would slide out maybe until June. I don't think that's going to be the case for us, but that's definitely one of the issues to deal with in North Dakota. It's just that -- we love that cold weather, but it is a problem for construction in the winter.

That's very helpful, really appreciate it. A follow-up question on the Texas side, are you continuing development construction while you're waiving this regulatory approval?

Yes, the regulatory approval is -- there's -- we've been through a lot of it. I feel really confident in it. I don't think there's any issue that we wouldn't get the approval. It's just a matter of timing. It's a process to work through. We're largely through the process and so we have -- I mean, we're, I would say, 90% complete on the site. So we'll be ready to energize, basically, the minute that we're approved.

Got it. And then my -- very helpful, thank you. And then my second question is on call it counterparty risk, it's obviously topic in the industry, given the distress out there. Can you speak a bit about who your counterparties are today, maybe your three or four largest counterparties and how you go about evaluating counterparty risk in the current environment? Thank you very much for that.

No, I think that's a really relevant question. So our largest customer in -- that's operating right now is Bitmain and that's in the North Dakota site. There are some other companies that are investors in the company. So I mean, that's been disclosed, so I don't mind saying F2Pool, which is one of the largest Bitcoin mining pools in the world, is one of our customers in North Dakota. Then GMR is one of our customers, of course, in North Dakota. I feel confident in all of those counterparties. I think the other large one, obviously that we've announced is Marathon.

And Lucas, you have your -- can have your own opinion about this, and I'm certain that you do and all the other analysts on the call do as well. But I think Marathon is one of the best counterparties, if not the best of the publicly traded miners in the industry. They have a relatively clean balance sheet. They have really the one piece of debt outstanding that I don't think is due for over four years now. And I think they have access to better capital access as far as capital markets or anything than almost any company in the industry.

Where a lot of the others with -- plaguing a lot of the public miners is really especially equipment financing, which has a lot of very fast amortizations and, at this time last year, didn't look like it would be a problem given the Bitcoin price. But it's turned into a massive problem, I think, all over the industry.

So that's the primary exposures that we have and I feel really good about them. But it's something that we monitor, I mean, almost on a daily basis because there's a lot of companies that are struggling in the industry.

Very helpful, Wes. Really appreciate your color and best of luck.

Thanks.

Chris Brendler, D.A. Davidson.

Thanks. Good afternoon. Just I guess a little bit of a follow-up on some of Lucas' questions, just on the overall competitive environment. Just given some of the recent news, the hosting business, there may be some -- even more demand amid constrained supplies. So I'm just curious, I think you mentioned there's an opportunity to take advantage of the challenging conditions. [It should be] a little more color on the demand side and where you sit in terms of parsing out your additional capacity.

So, thanks for the question. We don't have any additional capacity right now really to offer. So it's not -- we're not actively seeking other customers for capacity. We're still in the market for additional sites. We're looking for good power locations, the same stuff we always do -- good power location, good site.

And of course, we're looking at there's some really well-publicized sites that are out there that -- there's our one competitor filed for bankruptcy. I think there's other -- there's many, many other distressed properties in the market and we're looking at all of those. I haven't found anything that we're moving to take advantage of just yet. But there's a lot of that that's out there.

I will say again here, I've talked about this a lot and I'll reiterate, we spend a lot of time focused on the power side, and that's the biggest part of our business, right? We're -- purchasing power, that's our biggest cost. Then in the customer side, we're contracting, we're offering the lease with basically the spread in the middle.

I think last year a lot of companies went out and signed variable rate contracts or didn't sign contracts or whatever it might be. And so I'm seeing some assets that just -- they're hosting sites in remote locations that don't have very good power contracts, and so I don't know that those are worth very much.

But we're still looking, we're still looking, but there is a lot of distressed assets that we're seeing and we'll continue to look. And I think we're in a pretty good position to take advantage of that. As you see the balance sheet at the end of the quarter, I think especially in our industry, we've kept ourselves fairly clean on the balance sheet side.

No, it's been a breath of fresh air for sure. Just -- I'm just a little confused. You talked about HPC, is that different capacity? And I guess my question was I think last time we talked, the opportunities and the economics in the Bitcoin hosting space were much larger, certainly bigger chunks than there is in the HPC space. So maybe is HPC sort of like a way to getting your toe in in some of the contracts you're pursuing now are not nearly as big as the Bitcoin ones?

So I would -- I view it this way, so Bitcoin has given us the opportunity to go and build these sites out. Like I said, we spend a lot of time focused on getting, what I think are really attractive sites with really attractive power contracts in good locations. And then I think a lot about how do we leverage that. So obviously, we've got this big beachhead, 500 megawatts that we're building in Bitcoin, we expect to continue to grow.

But if there's other applications that are -- I think about highest best use case of our assets at all times if there's other applications for our sites that are not just Bitcoin: one, I think diversification is attractive -- for me myself as the largest shareholder it is, but I think for all of our shareholders; and then two, when I look on a per megawatt basis, some of the HPC applications we're seeing, the revenue is five plus times per megawatt, the EBITDA is five plus times per megawatt. So that maybe is something that's more attractive from an economic perspective.

But I expect to see a blend as we go forward. But I think one of the key takeaways here is we do have a GPU-based machine learning application that we expect to be running a pilot as early as this month. And then we'll expand that if it goes well later this year and then maybe be in a position this is -- and I think this one application is an extremely large market, but be in a position to scale that in '23 and '24. And I wouldn't say the sky's the limit, but it's definitely a really attractive potential for us. And so we're spending some -- a significant amount of time on that, honestly.

But, when you think about these sites, low-cost power, cold location. If we can figure out the application to run ultra-low cost infrastructure that we're going to be running the same NVIDIA, or AMD, GPUs, we're going to be running the same Intel-based or AMD-based servers. So no one's going to have necessarily some kind of equipment advantage on us. It's really we're focused on the ultra-low cost digital infrastructure for HPC. Hopefully that's helpful, Chris.

No, it's super helpful. Thanks for the color. Just one more quick numbers question, the -- just your non-GAAP gross profit margin came in as expected, through -- also really nice to see given power costs at a low, but then your mining peers are struggling with. And I just was -- my question was, given the reduced capacity this quarter, does that -- was that impacted by that, should that go higher from here? Or is 24% a good run rate?

So it can fluctuate quarter to quarter, from -- our power month to month can change a little bit at the Jamestown site, but over a year it stays roughly the same.

The thing I would point out for last quarter, the one that we just reported, we were fully staffed. We didn't go and lay people off. And so there was some definite inefficiency in the gross margin line with the staff because we -- basically that same amount of staff will run $12 million of revenue versus the $6.9. So there's some leverage there.

Great. Thanks so much.

John Todaro, Needham & Company.

Great, thanks for taking my question. The regulatory issues in Texas, is that say the Texas market from your guys' viewpoint is going to be closed for new sites moving forward? I guess just how big are those concerns or is it something you can continue to push through?

I don't think that Texas -- I think there is a bit of a pause and then trying to work through and, from our perspective, the machine is moving. So I don't think in any way, Texas is closed to new sites moving forward. I mean, the governor here has been very supportive. Everyone has been very supportive, but also Texas has had some very widely publicized issues, really -- was that, February of '21, I think was the last big issue for Texas, on the grid.

So I think the regulatory body there, ERCOT just wanted to make sure that there's -- that the grid is stable and these are large loads being added. But I think the machine is moving. I don't think Texas is closed to new sites.

Okay, got you. And just to confirm, so that facility where already 15, 16 buildings are constructed, there's not really any possibility that permitting doesn't go through. It is really just comes down to that delay in energizing, right?

It's just time working through the process.

Okay. Great. And then turning over just to the new facility, the second one in the Dakotas there. I've noticed it looks like this facility might cost a little bit additional per megawatt versus some of the past ones. Is it because those are being outfitted with HPC capability? Or is that facility completely contracted for Bitcoin mining. And just more broadly, could you give a little bit more context into the costs related to HPC versus the Bitcoin mining facility?

Is there something specific, John, that you're seeing on that? I don't think it's going to be any more per megawatt for the $180 million that is all Bitcoin. Maybe you're seeing some news items that talked about kind of $100 million. That wraps in a lot of other spending, just so you know. But the construction costs should come in in line with the other areas.

In fact, we're seeing some equipment show up on the secondary market that we recently purchased some switchgear transformers for a pretty significant discount versus what we have been. So I think we'll be able to run that in line if not lower.

Got you. Okay. Yes, in relation to some of those media articles.

Okay, great. And then I guess just to be clear, so far, nothing has been contracted on the HPP side. There's that pilot program that could be done this month, and that's the furthest along on that side?

Yes. So to take that a step further. It's -- we do a pilot program. What we really need to focus on is what can we do with this low-cost infrastructure with the -- for applications that can tolerate a somewhat higher latency environment. We'll still have very good fiber connectivity, but it will be a higher latency environment than other Tier 4 data centers have. And so we're working through that with our partner.

But it's very exciting, the work we're doing or will be doing with them, because it has the potential to bring down the cost to a lot of people of machine learning significantly. So it's early, but very excited about it. And so we'll hopefully have the pilot up as early as this month and then we'll have a pretty good idea. We will move that to a little bit larger scale later this year. And then in '23, really, hopefully scale that up. I'm pretty excited with the potential.

Got it. Understood. Thank you, guys, for taking my questions and congrats on the quarter.

Thanks, John.

Mike Grondahl, Northland Securities.

Hey, guys. Just a couple clarifying questions. In North Dakota as of today, how many more days do you have of concrete to pour or what do you think the end day is, if everything goes well from today?

We're shooting right now for the end of October. I mean, here's what you're -- you're dealing with construction outside, right. There could be rain days, there could be things like that, but we're shooting for, if all goes well, maybe the end of October.

Got it. So you've got a little bit of cushion depending on the weather. I got it.

Yes, a ton of the groundwork and a lot of that's been done. I mean, we're in a position to start to really -- once we start going, generally, our experience is the crews knock this out pretty quickly.

Sounds good. In Texas, if there were not these regulatory delays headaches, when do you -- when would you have been able to sort of energize Texas.

We would have been able to energize pretty close to now to -- between now and November 1. So if we can work through these -- what I will say though is those would be staged ramps, right? It would be like we did in Jamestown where you try to bring online about 12.5 megawatts per week, every week. So, like I said, if we can start sooner, that's what it will look like. And if we have to start later, we should have everything racked and ready to go by the time we get approval.

Fair. No, that's good. I mean, it sounds like ready to go except for this regulatory stuff.

Yeah.

And then you guys have a lot on your plate, but I know you're looking for number four, too. Is it too early to talk about the fourth site, the fourth project, the incremental 100 megawatts or what does that look like?

it is a little bit early. I mean, like I said, we're looking at everything. We want to make sure we find a very attractive site. Again, stable low-cost power, good location for us and maybe one of those is a site that someone else has that's already up. Like I said, I'm seeing a lot of assets that are being shown to us. So it might take that form instead.

Got it. Okay. Hey, thanks, guys.

Thanks.

George Sutton, Craig-Hallum.

Hey, great. This is Adam on for George. Thanks for taking my questions. Wes, I thought it was really encouraging, your commentary on the HPC pipeline of a few megawatts to 100 megawatts. I was hoping you could provide a little more detail there about the types of companies that are making up that pipeline in addition to how you're approaching the go to market and how you're building that pipeline?

Yes. So we -- I mean, this is really early days -- discussions anywhere from educational institutions to top-five hyperscalers in the world. So it's a really broad range, and it's to be seen how that comes together. But like I was talking about, Adam, is that if we can get these applications to work, and I'm very encouraged that we can, I don't think it will be hard for anyone to compete with us on the price that we can offer for the infrastructure in the locations that we're in.

Just this is some of the stuff that we've been going through. When you look at some of these data centers, if you're doing high-performance compute, a significant amount of the power is electricity and we're still working through those numbers. We haven't operated these. But, 40%-plus to run these have would be electricity. And then out of that, you split that half and half or 40%-60% to actually running the equipment versus running climate control or AC.

And so if we can do this with low cost power and then also in cold climates where we can use a lot of air-cooled, if not all air-cool then -- I think it would just be hard for anyone to compete with the cost basis that we can provide for HPC.

That's great. And in terms of this opportunity and your strategic thinking going forward with site construction, is there any change there on how you're going to approach building site four, five, six with this opportunity?

So there could be. There's time spent right now on the additional -- because it's going to cost -- in the buildings that we have, if you've looked at these, they're big steel structures. So in those -- the modifications that would be needed, if you decided to switch over to GPU applications and then also new-build sites, if we have it specifically for this.

The team's working on what that would look like right now, but I think it's pretty early to talk about what exactly that's going to look like. Let's get through the pilot that's -- scale the pilot up a little bit and make sure this works, and then we'll be focused and much more ready to talk about what new sites look like.

Great. Thanks, guys.

Rob Brown, Lake Street Capital Markets.

Yes, good afternoon. Just wanted to talk about the expected CapEx to complete the remaining two sites. What do you expect to -- you need to put into them at this point?

I'll let David -- do you have that?

Yeah. So we will use a combination of cash on the balance sheet, some community bank loans and the prepayments from customers that we're collecting, and we'll have the required capital to finish the CapEx.

And I'll add in there. So Texas is 90% complete at this point for capital. And then the experience -- I always like to point this out -- that we've had is, we have a construction loan that's barely -- we haven't started to draw that yet in Texas. And so that's about $15 million for that site. But the experience that we've had, Rob, is that in -- so in Jamestown, we had $7.5 million from the same bank for the construction. And then once the site was up and operational, there was another $15 million, we're able to expand that to $15 million.

So in Texas, given it's double the size, we're starting with $15 million and then once operational, we should be able to go to $30 million. So we haven't used any of that yet. So, as far as build goes, when you think about the cash on the balance sheet, plus $30 million coming in at the Texas site and eventually $27 million-ish to come in on the debt side for the Ellendale site, plus customer prepays, we think we're in a good position for that.

Okay. And just to be clear, you feel like you've got most of the Texas CapEx spend, but the North Dakota CapEx, you've got pretty much the whole facility CapEx to go on that one?

Yeah.

Okay.

You're -- there's actually -- it's a fair amount has been spent there, right? The procurement of the transformer switch gear, a fair amount of the building procurement has been done, so you probably have 20% spent there right now.

Okay. Good. And then I think in the press release you said the North Dakota power contract is now signed. Could you give us a sense of what's the length of that contract? How does it compare to some of the -- or your other contract there?

We don't ever talk about pricing, but it's very attractive pricing and it's a five year deal.

Okay. Thank you. I'll turn it over.

Kevin Dede, H.C. Wainwright.

Hi, guys. Nick with H.C. Wainwright here on behalf of Kevin Dede. Would you be able to quickly confirm expected energization timeline for Garden City and Ellendale?

Nick, we went through those a little bit earlier in the call, but we expect things to start to energize any time basically between now and mid next year. The big timing issues for us right now in Texas are regulatory approval. And it's -- I went through the three components there, but we have some in-person meetings this month, and I feel like the machine is moving well there.

And then in North Dakota, it's very much weather related. So we get all of our concrete for this month to have all the pads done. And we're targeting early '23 for North Dakota, but hopefully we can pull that one ahead a little bit. We started the same time in North Dakota last year as we started this one, and we energized -- the first energization was in late January.

Got it. Thank you, guys. I'll turn it over.

Yes.

Thank you. This concludes our question and answer session. I would now like to turn the call back over to Wes Cummins for any closing comments.

Thank you, operator. Thanks for everyone for your interest and participation and look forward to speaking to you next quarter.

This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.