Who in the world is mining Bitcoin in Australia right now? I mean, I don't like a company going bankrupt, but whenever something like this happens, you can go into Pacer, find the court filings, and then you have all this juicy open source information about who is involved in these things. Some of the names that have surfaced from this. There's some really weird stuff in here. Like for instance, the Chicago Blackhawks hockey team. This is real. For 1.26 million, at first blush might be like, well, of course, just, you know, check the boxes, dude. But you are the trade, bro. The trade. That's not how it works. Now, this is very much a he said, she said story. Riddled with drama.
What's going on y'all? Welcome back to Blockspace Live presented by CleanSpark for the final day of week one of earnings week here at Blockspace. We've got CleanSpark and MARA's earnings that dropped yesterday after market closed to tear through. We will lead with CleanSpark right after our usual Friday hash rate index update. And then after CleanSpark, we've got a slew of interviews to chew through today. We've got Luxor's Khan Farahani on to talk about their most recent lookback series for the month of July. Following that, we've got Brandon Bailey of Nakamoto and Diametrics on to discuss Q2 earnings for the Bitcoin miners turned AI pivots that we cover and also unpacking some of the fun metrics that you can find on Diametrics. And after that, our keynote interview today, Hut 8 CEO Asher Genoot on to talk about their earnings. And we're specifically interested in if Batch Zero will throw a wrench into their expansion plans at Beacon Point in Texas. And then we will close the show going through MARA's earnings and taking a look at the fact that they could be getting close to their first AI tenant based on the analyst questions and Q&A at the end of the call.
That's right. Blockspace goes live weekdays at 1 p.m. Eastern. We are compute's daily show featuring quick hits on AI data centers, compute markets, and emerging technology. If you missed the live stream, you can catch us on podcast form shortly after this wraps up anywhere podcasts are found. Leave us a review if you haven't already. And if you can't get enough of Blockspace, you can find the rest of our content beyond just talking head live stream stuff. We do a lot of written stuff. You can find it on our website at blockspace.media. That's not blockspace.com, that's blockspace.media. This show is brought to you by CleanSpark, NASDAQ listed ticker CLSK. More on them later on the show. It's Friday, Colin. So, let's kick it off with a hash rate index. Little hash rate index update coming at y'all from Luxor. And Charlie, you'll be shocked to hear that it's a little more or less of the same. I mean, we're going to get into some of the numbers. July was actually a pretty interesting month on the Bitcoin mining side, specifically when we look at Luxor's derivatives and who won out on USD versus BTC hedging for hash price that is. But if we look right now, not too much good to glean, Charlie. I mean, Bitcoin making another run at 65K. We love to see that. 65K gang, undefeated for the last two months. 64K gang, we're at $32.60 per petahash per day for spot price. Lot better than being in the 20s, but still obviously not what miners want to be seeing right now. Network hash rate kind of making a little bit of a recovery here at 930 exahashes on the 7-day average. We're entering the hottest time of the year. Don't expect much growth here as we round out the summer. Could be wrong though. The most recent difficulty adjustment was minus 0.74%, marginal, and also it's looking like we will get a bump marginally again, almost offsetting this last difficulty adjustment at plus 0.7%. So hash rate kind of just rangebound, man. I mean, not really too much to say about that. Fees aren't really doing anything, you know. Transaction rate per day is going up a little bit, but really not too much to say on that front. Overall, still slim pickings out there for the Bitcoin miners who are staying true to the game and not chasing the shiny new thing. I'm not stuck in here with you. You're stuck in here with me. Look, the only interesting thing happening I think in blocks right now, we'll get into this with Con here in just a moment, is that the end of this difficulty adjustment, which is tomorrow on Saturday, that is when the BIP 110 fork is going to happen. So that's when the mandatory signaling period for knots nodes happens, which will result in a chain split if they are able to produce a block. So we'll have to see what happens then. I've obviously got opinions which I've been unrelenting with on Twitter. But that's the end of that story, Colin. It's day five of earnings week one. So that means we have to kick off with our final earnings update of the week.
Yes, that'll wake you up if you haven't already. Earnings, last night we had watch parties for both CleanSpark and MARA. I think we'll tackle CleanSpark first with you, Colin. Take it away.
Yeah, we'll tackle CleanSpark here. We've got a recap here on the Blockspace website. Go ahead and check it out. I won't spend too much time on what's written here. Got a bunch of notes to chew through. And I would like to start with numbers, but I'll be brief because I think the meat of the call honestly came from the Q&A session with analysts and there was a decent amount of alpha to be gleaned from that, Charlie. So going over numbers really quickly, revenue came in, and this is fiscal year 2026, Q3 came in at 138 million versus expectations for 154.85 million. So pretty decent miss, 11%. I think to me, you know, I don't know, analysts haven't been modeling their hash price out very well as we've actually seen this time and time again with some of the Bitcoin miners. Which is strange because a lot of them usually are like they know how to use this metric at this point. And usually they're pretty spot-on, but I guess I don't know, maybe CleanSpark might have had some more downtime. Maybe there was curtailment they weren't pricing in. But nonetheless, a little bit of a revenue miss. Although these numbers, I just want to caveat it going into this for MARA and for CleanSpark, like the analysts honestly don't really care about these numbers that much because they're really looking at the AI segment.
Yeah, these numbers aren't moving this stock at all.
No, not at all. Net loss came in at 239.8 million, largely driven by change in fair value of Bitcoin holdings. CleanSpark has the third largest Bitcoin treasury of any of the companies in this cohort, following Riot at second and MARA at first. Earnings per share came in at negative 89 cents per share and that fell short of FactSet estimate for almost 50 cents, 48 cents per share, negative 48 cents per share. Just evident. Cash stood at 22 million. Bitcoin holdings fair value stood at 814.9 million. Total assets 2.7 billion. Total debt, all long-term, no current, 1.8 billion. So those are the numbers, but again I think that they belie the actual thing that mattered on the call, which was the management commentary regarding Sandersville, also their expansions into Texas that are under exclusivity for the tenant at Sandersville. So, just to recap that deal for those who may or may not remember, the Sandersville deal for CleanSpark came in at, let me see, I...
It's a $6.6 billion deal. Yeah.
And I believe the contract terms are for 15 years. I just wanted to make sure I was being specific about it. It's getting buried in my notes here, but they did not announce who the tenant is for this site. There's been a lot of speculation that it could be Meta. There are a lot of unconfirmed reports that suggest as much, but Harry Sudock and other management like Matthew Schultz made very clear that they are an investment grade counterparty and that they are a global tech company with high investment grade ratings. So, there's really only a few companies this could be. And Meta has had some job openings for the Atlanta area. That might just be noise. It's hard to tell, right? It's a huge company. They probably have job openings all around major metro areas in the US, but they did not disclose who the tenant is at this point in time. Again, speculation that it could be Meta. We don't know for sure yet.
I will comment that like the term investment grade tenant has done a little bit of heavy lifting for some of these announcements. Because I've heard, if I can't remember who it is, but some folks have said that have used that term and they're referencing perhaps like an Anthropic or OpenAI who's not public and so they're kind of assigning by the transitive property the other folk on the other side of the deal who is providing said non-public company with perhaps some investment grade credit. So...
Yeah, a good example of that being Hut 8 and TeraWulf.
And Cipher with Fluidstack.
Google being on the end of that compute that's going to be hosted at those facilities.
Yeah. So, quick fact check for myself here. $6.6 billion deal over 20 years for 175 megawatts of critical IT load at the Sandersville site. On the call they revealed that they expect the cost of construction to be between 10 and 12 million dollars a megawatt. It's a total greenfield. By the way, this is important because the Bitcoin mine is...
Yeah, the Bitcoin mine's still going to be sitting adjacent to the HPC site so that they can continue to make use of that electricity. Substation's already in place, which is great news. If you don't have to worry about that, that's a huge cost. Also, lead time probably crazy. So, that's important. They also, the CFO mentioned that the equity portion has already been funded for Sandersville. Really important here, too, with regard to the investment grade counterparty, Charlie. There's a backstop, but they said that there's no equity given up and no credit wrappers for that backstop. If you remember when TeraWulf got the backstop from Google, which was the first of its kind for the whole cohort, they ended up giving up, I think roughly like 14% of the company in warrants to Google. And CleanSpark here saying that there's no equity component to that backstop. Wondering what they did to sweeten the deal to make sure that they could get that. They didn't really talk about that on the call, but I think that's absolutely worth noting. Few more updates. All data center long lead items are ordered and prepared for the Q4 2027 first data hall. The 122 acre greenfield parcel lets them build without disturbing the mining footprint like we said. And noticeably, engineering firms and general contractors are the tenant's preferred partners and that the firm also manufactures MEP and cooling components with a track record with the specific tenant. So they already have the contractors and the engineering acumen via this counterparty. Again, investment grade, high investment grade, large tech company, don't know who it is. Another big reveal from the call, CleanSpark, according to CFO...
Vitarelli. Vitarelli. Vitarelli. Excuse me, my southern tongue has trouble with Italian names, man. I just, we don't have many Italians down south. He said that they will utilize high loan-to-cost project-level debt, quote, as you've seen some of the recent deals financing in excess of 90% loan-to-cost is very common and that's our target. Obviously, the higher the loan-to-cost, the greater the internal rate of return is because you're bringing less equity to the project. This is a notable departure from our interview with Patrick Flury and we'll actually touch on this theme one more time when we talk about Salman Khan's comments on how they're funding the Long Ridge acquisition for MARA. We're seeing three different financing strategies. One with a lower loan-to-cost, more equity given by the parent company, TeraWulf. Here, higher loan-to-cost, less equity committed, CleanSpark. And then in MARA's case, lot of Bitcoin. They're actually taking on Bitcoin-backed loans for some of these acquisitions that they're seeking, specifically the Long Ridge one. Last few things here, Charlie, before we move on. The rest of the Q&A segment rested on Texas and there were questions as to whether or not the Batch Zero process will affect Celia and Brazoria, CleanSpark's two sites that they have acquired in Texas and which this tenant at Sandersville has exclusivity on once they decide whether or not they want to exercise that option. The Celia site, 285 megawatts, is base load designation in Batch Zero. They acquired this in 2025, or sorry, that base load designation was made in 2025 and so it just needs final determination for the interconnect through Batch Zero. So that one seems like it should be pretty secured. Brazoria phase one is the same, 300 megawatts, classified base load, shouldn't have any problems with Batch Zero. The only one up in question is Brazoria phase 2. That's 300 megawatts. That's the second half of Brazoria. That still needs a study. And you know, it also bears to mention that I assume they're probably pushing it through Batch Zero, but like considering the construction timeline on these, it might not be an issue. They're probably not going to break ground on that site for some time. The other really interesting point from the earnings call was what CleanSpark highlighted for future expansion. They highlight an 86 megawatt site in Washington, Georgia that is energized today, but they've also submitted a line study at MEAG's request to expand it by up to 500 megawatts and has secured options on several hundred adjacent acres for that expansion. So, if you're looking at where they might try to strike next, probably going to be down south in Georgia. Schultz also said the value of non-Texas assets for 2027 and 2028 went up significantly, quote unquote, because of the Abbott audit. They also confirmed 100 megawatts in Wyoming, 60 megawatts in Jackson, Tennessee, and a small footprint in Ripley, Tennessee. But it seems like the one that they're really going after right now is that Washington County or Washington, Georgia site, 86 megawatts looking to expand to 500. So, if you're looking at where they might strike next for potential AI expansion, my bet would be on Georgia. Overall though, some clarity on exactly what the financing will look, or not exactly, but the shape of what the financing will look like for Sandersville. Some assurances that lead times or prepaid items and long lead time items are secured for the first data hall and also most of the Texas portfolio won't necessarily be disrupted by Batch Zero other than just the timeline getting pushed out. So yep. So, I think that probably wraps up the first of the two earnings that we'll cover today. We'll cover MARA in the back half of this stream. But we've got Con in the wings. We're going to talk about this month's hash rate index lookback. But before that, a word from our sponsor, CleanSpark.
We are CleanSpark, America's Bitcoin miner, a publicly traded company with the largest operating hash rate, powered entirely by self-operated infrastructure across four states. This is our proof of work, and we are setting the standard for what's next. Learn more about the intersection of energy and Bitcoin at cleanspark.com.
All right, we got Khan in the audience for our monthly hash rate index lookback. Khan, welcome back to the show.
Thank you so much, Colin and Charlie. Good to see you both.
Good to see you, too, man. Okay, so it's July. Summer's, you know, over halfway done, but doesn't mean that things weren't interesting in July. I'll throw it to you. I've got the lookback pulled up here. We can hop around to it. Give me the high level TL;DR before we go into specific questions, Con.
Definitely. July was quite interesting. So let's start with the headline on hash price. First thing we saw was monthly average dollar denominated hash price rose to $31.21 per petahash per second per day, up 2.8% month over month, but it only recovered a sliver of June's fall. We saw that in June hash price slid 17% or $6.23 and July only regained around 84 cents of it. Now at $31.21, July ranks as the second lowest monthly average hash price on record behind June which was $30.37 and just under March which was $31.27. So these three lowest readings, these are the three lowest readings in Luxor's Bitcoin hash price index and they all fall inside the year of 2026. Now for comparison in 2025, monthly averages ranged from around $38 to $59 with a mean of $50.68. So this gives you an idea of what it feels like to be a miner these days.
Yeah, it looks pretty choppy and brutal, man. I just don't see... Do you see light at the end of the tunnel? Is there anything in the lookback that makes you think, okay, things could be improving from here? Because to me, the light would have been the 4CP season for the miners who could still remain online, but it hasn't made that much of a difference. So much hash rate's kind of already come offline. Is there anything that gives you hope?
Yes, there is some good news and I'm going to share some of that right now. But before I get into that, another interesting aspect here is on difficulty. So what we saw with difficulty trends is there was a double drop which of course contributed to this slight recovery in hash price. During July's epoch, difficulty fell twice. The first was on the 11th of July. It was a minus 5% decrease and then a second one on the 25th which was a 0.74% decrease. Overall a 5.71% net decline in difficulty over the month. Now this trend was driven by 4CP curtailment and also notably the July 11 difficulty drop was the fourth largest difficulty decrease of 2026. Now I mentioned that the trend was driven by 4CP curtailment. How do we know this or how do we get to this conclusion? The reason is because Luxor's energy desk saw that ERCOT dispatch events clustered heavily across this 10-day period at the very beginning of the month spanning from the 1st to the 10th of July. We saw that there were more than a dozen curtailment events during these 10 days and then the following difficulty epoch spanning from the 11th to the 25th saw less than half of that and produced a much lower difficulty adjustment, right from 5% down to 0.74. So the main message here is that curtailment intensity and the size of the difficulty adjustment tracked each other pretty closely. Another interesting insight which I think is highly relevant is at this stage difficulty has now closed below the one zeta equivalent which comes out to 139.70 trillion for 10 consecutive difficulty adjustments spanning from the 20th of March to the 25th of July. So right now in early August it's been over 130 days where we've firmly been below the one zeta hash milestone. And then a third distinction between the previous month of June and this month of July was that the difficulty drops that we saw in June coincided with two different factors. One was record low hash price which caused marginal mining machines to shut down and then there was also seasonal curtailment via 4CP peak avoidance behavior for flexible loads. July saw a hash price rise by 2.8%, which removed some of that economic layer. So that's why we're coming to this conclusion that the trend was driven by 4CP curtailment. Of course we can't be absolute about it but we believe that it was a significant contributing factor.
And this actually gets into something I think you've talked, you've brought up in the lookback, which is we peaked last year at like 1.1 zetahash and we're down to the moving average is the mid 900s or low 900s now of exahash, which means there's an estimated 227 exahash idle. Pretty interesting here. Any color you can give to this? Economically unviable? Is this just rotting? What are the implications of this for the foreseeable future?
Definitely. Yeah. So that's an interesting observation that's been driven by our latest mining economics projections. And basically what we're seeing is that since that peak of the one zeta hash milestone back in around September, October of 2025, difficulty peaked around 155.97 trillion, I believe we're down around 19% from there since then. Throughout this time, going back to what Colin just talked about in terms of good news, what this means is for any miner that's online and still hashing, that's relief, right? The Bitcoin network is paying more BTC per unit of SHA-256 hash rate at the moment for those that have survived and stayed online. Another interesting insight that we can draw from our projections is that at a price range right now of between 59 to 66K in terms of Bitcoin price action, we would expect USD hash price to range in the 28 to $33 range. Now, that's above the level that triggered the marginal shutdowns in June, but it's not far enough to bring back mid-generation mining machines back online. So we're talking about the 25 to 38 joule per terahash tier. So Charlie, a lot of that idle capacity is coming from this mid-gen mining machine tier. Even more interestingly, at $40,000 Bitcoin price, our sensitivity tables would put equilibrium hash price around the $25 mark per petahash per second per day. And what we think is that the price point that really, really matters is around $33,000 for Bitcoin because at that point, assuming an average industrial power cost of $48 per megawatt hour, we would see that the break-even efficiency would drop under our estimate of the network efficiency, which is around 20 joules per terahash. So the message there is that the mid-gen machine would stop covering its power bill at that point. So the differences in idle capacity versus active hash rate and these sensitivity estimates gives you an idea of where we're sitting right now and how Bitcoin price action might affect hash rate.
I've got kind of two questions to follow up on that, Con. The first one involves the machines that are still active because right now break-even efficiency at July's hash price is like 27 joules per terahash assuming what, like $48, $50 per megawatt hour, and the 25 to 38 joule per terahash tier has sat below break-even for eight months against an estimated network average cost that we just mentioned. Do you think we've reached a kind of ceiling for hash rate for right now in the sense that like we're kind of getting to that point where equilibrium is setting in? The machines that are still online are the only ones that can take it. The operators that can't hash profitably under this hash price are gone. And the second part of that question, are there any insights from the ASIC trading desk as to like where that 227 exahash is going to go? Is that just scrap at this point? Is there any room for that to migrate to lower cost jurisdictions? So just two parts there. You can tackle them in whatever order you want.
Yes, really good question. First is on the idle capacity. This idle capacity could be idle for reasons beyond economics as well. Some machines fail. We know this. Eventually all machines will fail and some might be undergoing maintenance or as you mentioned, Colin, they may be moving elsewhere around the world chasing the cheapest electrons. So it's a confluence of these factors. And then the second part to your question is a lot of our estimations for these mining economics projections are driven by insights we get from our hardware team. And specifically for our latest run, we saw that there was a major downward revision in terms of monthly production estimates for incoming hash rate supplied by ASIC manufacturers. And the reason is quite clear. Excuse me. The reason is quite clear. Mining economics has deteriorated quite a bit. So machines are incoming at a slower rate which of course affects the level of hash rate supply coming in as well.
So, I want to talk about hash rate contracts and how hash rate traded. Always a really interesting thing to have you look at, especially in the middle of the summer. Yeah. I guess first, how did hash rate trade and then we'll get into how the hedgers performed.
Yeah, good question. So how future hash rate traded throughout the month was we saw that the forward market lowered its difficulty and hash rate expectations again in July, deepening rather than reversing the reduction it made in June. So every forward contract from August to December fell by an average of around 6%. And the implied network hash rate for end of year December 2026 dropped from 1.05 zeta to now 985 exahashes. So Colin, touching back on what you just mentioned, the forward curve is expecting hash rate to remain below the one zeta level throughout the end of the year. And then second, Charlie, when it comes to hash rate hedging performance, what we've seen is instead of looking at single month contract performance, I think it's more interesting to look at rolling hedge performance over different time frames. Two main messages here. The first is that over the past 12 months, the trailing 12 months, it's generally made sense to hedge against dollar denominated hash price and to remain exposed to spot Bitcoin denominated hash price. This is driven by two different factors. First is weak Bitcoin price action which obviously affects dollar denominated hash price and then the second is the decline in difficulty that we've seen which relieves Bitcoin denominated hash price. So generally dollar denominated hash price hedgers have won and those remaining exposed to spot Bitcoin hash price have also won throughout this period. Now, if we zoom out further to the halving since April of 2024, the picture changes a little bit. The main message being there that regardless of your hedge horizon or your contract denomination, it's generally made sense to hedge against hash price. So, as a group, miners that have engaged in consistent hedging have come out on top versus spot mining alone. So let's get into how the USD and BTC denominated sellers performed. It looks like...
USD won what, two out of five horizons. So...
That's correct. Kind of a... Yeah. So what... Yeah, dive into this for me. So for this month in specific, dollar denominated hedges won throughout April and May and any hedge struck outside of that time period was out of the money. And conversely on the Bitcoin denominated side, buyers of Bitcoin denominated hash price won across the board. For the fourth consecutive month, anyone who's selling forward in terms of Bitcoin denominated hash price has been out of the money. And the reason is because difficulty is declining. So why would you hedge against difficulty or transaction fees as well? And there's some interesting insights on transaction fees that we can touch on as well. Of course, there's a very relevant exploit which shows up through these onchain signals.
I assume you're talking about the Coldcard exploit and specifically increased fees from sweeping those funds.
Exactly. Yes. The data clearly showed up towards the end of the month. So throughout the month, fees accounted for 0.69% of total block rewards. This is still below 1% for the 13th consecutive month. So we've seen this trend since July of 2025. Bone dry fees. In dollar denominated terms, average fee revenue per block came out to just under $1,400 per block. 8% decline month over month and the estimated network-wide fee revenue for the full month of July was around $6.2 million. Now, the fee distribution was heavily skewed towards a single day and we believe that the cause was the Coldcard exploit rather than organic blockspace demand. We saw that on the 31st of July, fee collection peaked, which was 67% above the monthly average. It also saw the highest fee revenue per block coupled with the lowest transaction count of the month. So the 31st of July saw an average of 3,900 transactions per block against the monthly average of 4,700. Why is this the case? It's because sweeping thousands of compromised addresses into a handful of consolidation addresses produced a small number of input-heavy transactions. So blockspace was being filled with these handful of very large heavy transactions rather than many ordinary ones. And miners were able to capture some of this fee competition. But yeah, safe to say that the data showed quite clearly that something happened towards the end of the month and we attribute it towards this Coldcard exploit.
An unfortunate note to end on but salient insight. Khan, thank you for joining. Really appreciate it. Be looking forward to next month. See if hash price improves and we'll see if that ceiling on hash rate holds. Have a good weekend, man.
Thanks so much. Take care.
All right, I'm gonna pause. Can I do a quick live mic check? Colin, is my mic super quiet still? It's a little better. Okay, let me try this. Boosting a bunch of manual stuff. Is this better? Yeah, that's better. Okay. Anyway, moving on. We've got Brandon Bailey, the guy behind Diametrics in the wings, and we'll bring him on up here to talk about the latest in data centers and data center insights. Before that, Brandon, a word from our sponsor, Luxor.
This episode of Blockspace Live is brought to you by Luxor's Commander, Bitcoin miner management software for enterprise operations. Luxor's Commander gives you real-time fleet monitoring, bulk remote commands across your fleet, and Intelligent Miner. That's an automated profitability engine that runs every 5 minutes and adjusts your fleet's power settings to live energy and hash rate markets. In fact, our backtest showed 10% improved profitability with Intelligent Mining versus old-fashioned binary mining. Commander Pro is roughly half the cost of competition, $100 per megawatt or a 25 basis point pool fee adder, but you can also try it free for 60 days. So, if you'd like to learn more, go to luxor.tech/commander.
All right, let's get data whisperer Brandon Bailey up here for a little Q2 recap and also a little showoff of Diametrics. Brandon, welcome to the show, man. Welcome back.
Thanks, guys. It's good to see you.
Good to see you, too, man. So, we're about halfway through, a little over halfway through in terms of the companies we cover for earnings. And I just wanted to start this segment off with what have been some of your key takeaways so far from what you've seen specifically on, it could be financial metrics or it could be on the operational side and the qualitative side with company updates.
Yeah, I would characterize this most recent earnings season as being another one of sort of strong execution across many of the miners. A lot of what we heard on various earnings calls is that demand for power capacity continues to be strong and robust. So, what's constructive about that is the fact that we're seeing some pricing power with respect to some of the companies as it relates to their lease rates. They feel confident that their ability to sign new deals at higher rates is continuing to trend in the right direction and I think that that's a function of just scarcity, right? I think something we might end up touching on is, you know, the Batch Zero process, other regulatory sort of hostility as it relates to data centers that is making it even harder to ultimately get capacity online, which is making the near-term power capacity that's already available even more valuable. Another thing that I think we heard on earnings transcripts or from various CEOs was that they're starting to try to differentiate their sort of capabilities from how they pursued the debt financing terms and then also with respect to key terms in the leases they sign. So that's pursuing triple net leases versus modified gross leases. Some of the credit backstops or other various, I would say, covenants or clauses that they're including in some of their leases or even how they go about the project financing debt. They're really starting to focus on how they go about the execution across these items for differentiation between their peers. And I think that that's going to be a really important point as we go forward as most of these companies now have signed a lease. I think we're past the point of can these companies actually sign a lease and now it's coming down to who can actually get the best economic terms or sign the best deals. Another big point that we heard was around the behind-the-meter power generation. I think that is another big sort of key component that's going to start to come into play. And I think another element that kind of is a little bit related to the Batch Zero process, again, as it becomes more difficult to get net new capacity permitted and approved, looking for ways that you can kind of go behind the meter or find alternative routes to getting power capacity online is going to be another big opportunity for a number of these companies.
Let's touch on Batch Zero here, Brandon, because I think that was probably the defining factor across a bunch of these earnings calls and so much so that some stocks with exposure to Texas and those who were more exposed to Texas than others actually sold off pretty significantly following the earnings calls. First question, is that justified to you? Second question, kind of building on the first, do you think that that should be nullified by the idea that suddenly electrons outside of Texas are now much more valuable? But then there's also increased competition from everyone for those electrons.
I think that's right. I mean, in my personal opinion, I think there's been a little bit of a market overreaction to the news. I actually think for groups like a Hut 8, a Cipher, many, many of these companies, it's actually more of a, it could even be a bullish thing for them because I think that Greg Abbott's comments and wanting to delay the process is really about distilling the signal from the noise, like who are the real players here versus who are sort of the pretenders or the people that are just trying to be opportunistic with respect to this opportunity. And so I think, you know, companies like Cipher, Hut, you know, Wolf, others that we're talking about here have, you know, taken all the appropriate steps when it comes to, you know, their applications with respect to the Batch Zero process. So I think that this maybe even gives them an opportunity to demonstrate, you know, why they are a competent sort of, you know, player and should be selected or be one of the companies that should be granted, you know, the permission to kind of, you know, continue to operate in the space. I think it also creates opportunities for these companies on the M&A front as there are likely to be some sites that, you know, either don't make it where they might be able to pursue interesting sort of purchase option style contracts for sites. So it might actually even present an opportunity for them to boost their pipeline capacity at fairly competitive or cheap rates. And then the last thing, again, it goes back to this just being another item that is ultimately going to make it more difficult and more challenging to get power capacity online, right, which increases the value of already approved and permitted capacity that is available.
And permits, that's becoming like the constraint. And with Diametrics, you've been tracking this moratorium area map. I have to bring it up because it's the thing everybody's thinking about. It's very difficult to track because it's all inconsistent, but so what's your assessment? How would you characterize moratoriums across the United States right now? And we can get into some more narrow questions.
For sure. I think you're definitely seeing a growing number of moratoriums happen. The trend has mostly been, I would say, the counties or towns that have been the loudest with respect to wanting moratoriums against data centers are typically the more educated, affluent sort of urban towns or counties that, you know, ultimately it's sort of like a NIMBY effect. So I found that most interesting. I think that a lot of people assume that it's mostly Democratic sort of counties. That is a correlation, but it's not as strong as ultimately the level of affluence and educational attainment in a general area. And so what you typically see is that more rural areas aren't necessarily against data centers, or at least they haven't been. It's mostly, you know, those urban areas that I've been mentioning. And so again, overall, you continue to kind of see the noise around data centers continuing to just trend up and to the right. One of the things that I think is somewhat unfortunate for counties is the fact that these data centers have the opportunity to bring significant tax revenue, especially in smaller counties. So, you know, I kind of understand some of the public outcries. But I think that counties should look at the bigger picture of what allowing a data center development in your county can do for the local budget. That could be new schools, that could be new parks and recreation, it could be, you know, getting a new fire station, whatever proffer conditions that ultimately add value to a community. There's a real opportunity there if they can work collaboratively with data center developers where I think everybody could really win.
I wonder how much of that is not properly communicated. I'm inclined to think that that's not necessarily the case for most of these companies. I mean, like if we look down the roster, right, you can look at, you know, Hut 8's done this. They've done a lot of community engagement. Meta has done a decent amount of community engagement. I believe TeraWulf as well has done quite a lot of community engagement. So I wonder how much of it is, you know, are they not being educated enough on like what exactly the money will, like how much money it will generate in local property taxes and other injections of capital and what that will mean for infrastructure in these areas, or I wonder if it's a woodenheadedness of just, this feels like a deal to the devil with a lot of people who, like, don't fully understand what these things are. I mean, when you go out into some of these rural towns, man, it's like everyone grew up there. Their great-grandparents, like, settled there and they've known everyone in the 600-person town their entire lives. And there's almost like an element to where they don't want anything to change. And I'm not really saying that derisively, but like when you think about just like putting a massive data center, they might see it as like, you know, counter to the town's, you know, values or whatever. I'm kind of rambling here just to ask like if you think this is something of an...
intractable problem for certain areas, or if you think that it just needs—that there just needs to be more education and outreach.
I think there definitely needs to be more education and outreach. I think, you know, it's also a little bit depending on the county. I could understand how, you know, a massive data center development coming to your county could feel like a big boogeyman. You know, big business sort of, you know, having a new business development come to your town that kind of seems counter to the culture, right? Or you have these big giant organizations that, you know, you may envision don't really appreciate the culture of a town and having some hostility towards that. I could totally, you know, have sympathy for that understanding.
But I do also think that there are large social campaigns going around that are spreading misinformation. You know, there have been many studies that have been conducted that, you know, debunk effectively, you know, the power consumption as it relates to the data centers and what it means for local electricity rates, the water consumption and how much water is actually being used, the environmental considerations. But there's a lot of social media-led misinformation about these things. And not to be too tinfoil hat about it, but I also think that there's a lot to gain from other nations like China to try to incentivize that data center development is bad in the United States and is something that should be banned. Anything that can ultimately slow down this train and curtail data center development is advantageous for somebody like China.
There could even be smaller politics or games being played whereby, you know, the competition at the frontier is fierce. Imagine—I'm not saying I have any evidence of this, but just imagine this for a second. Anthropic is planning a mega data center, a mega campus. What if you are Meta? Could it be in your best interest to try to stall that data center from ever coming online? Even if you can delay development by six months by maybe pushing a campaign, it gives you an opportunity to catch up. I'm not saying those things are happening, but it's not that hard for me to believe that there are some political games that are being played across the country, maybe between different countries, and then certainly at a much larger scale as well.
So many competing incentives. Yeah.
And we've covered this—we've covered with Sam Lyman of the Bitcoin Policy Institute on the Chinese program to drive, to elevate these concerns. We've seen it with oil and gas for the past 20, 30 years. Okay, one last question. Again, got to quadruple down on moratoriums. That said, all this that you've said before, are the moratoriums or delays actually affecting data center projects? There's been headlines of X number of gigawatts, X number billion dollars of data centers like canceled or delayed. Is that true? Are those indicative of what's actually happening? Are there delays, or are these project cancellations, or are these projects actually ever really materially going to happen? What actually is the impact of the moratoriums in your view, as of right now?
I would say there's definitely been some impacts. We've had a couple of different data center projects absolutely be impacted by some of these moratoriums or just like public outreach that has caused either delays in the permit process, additional community hearings, requests for more time to study and evaluate these projects. So it's certainly having an impact on certain developers. But I would say, that being said, it's in very specific geographies. So I would say we've seen more of these impacts be felt in a market like PJM, so Ohio, states like Illinois, Michigan, and others. Certainly in Virginia, I know in Loudoun County there's a lot of pushback, Ashburn, places like that. So it's not all created equally. It very much is region dependent, but I would also say that a lot of the new planned capacity is mostly in Texas. You know, there's not a lot of new data center capacity planned in a location like Ashburn, Virginia or Loudoun County because of known hostility. And so I think that on a go-forward basis, it's going to be really, really important to watch sort of what happens with, you know, a place like Texas with the PUC process. Any other changes that ERCOT plans to make as it relates to how capacity gets studied, that will have a really material impact. But I would say for now it's been more sort of one-off projects here and there.
All right, Brandon, thank you so much for your time. Shout out—if you like what Brandon's talking about and you want more info, go to diametrics.ai. Diametrics spelled diametrics.ai. Brandon, thanks for coming on the show.
Thanks for having me. Happy Friday.
Happy Friday. All right, we have Asher, CEO of Hut 8, in the wings and we are bringing him on up here. We'll talk about the latest going on with Hut. Let's bring him on up.
Asher. Asher, welcome to the show.
How's it going, guys? How you doing?
Pretty good, man. Thank you for joining. Obviously, super busy couple weeks for you coming fresh off of Q2 earnings this week. We really appreciate you taking the time.
Of course. And I would love to start with a question that you've probably gotten a hundred times this week and was a feature of most of the calls this week for the cohort of Bitcoin miners turned AI companies, and that's explicitly: how does the new audit from Greg Abbott and the delay to batch zero affect Hut 8 specifically as it relates to the Beacon Point data center?
Overall, what I think Greg Abbott has done is extremely good for the ecosystem, and you actually saw even QTS came out and supported him. We need more people to give kind of local voters, constituents, community members comfort that data centers are not this big scary thing that's going to come and take everyone's water and increase energy bills and so forth. I think what we're seeing a lot across the country is that this misinformation has spread and really gone into people's heads, and for a data center developer it's harder to get there and say, hey, look, we don't do these things, because they think, okay, well, you are biased in saying that because you're trying to invest into this project and obviously you're going to say the data centers are fine and not going to consume water and so forth. So I actually think what Governor Abbott did was extremely smart because there's so much investment in data centers going into Texas and it's better to get ahead of this and say we're going to make sure that this investment is done in a responsible way that helps build communities rather than let the fear become so large that it's hard to reverse later. And so, like, we're extremely supportive of the initiative.
And also we were one of the few that actually responded to the PUC in their questionnaire that related to a lot of these topics. And frankly that questionnaire was many of the same things that Governor Abbott and his office released. And so overall we're going to work through the process but think this is good and healthy for the ecosystem because he also didn't take an extreme approach of like New York, they said, all right, we're going to do a moratorium, no more data centers. He said, 'We're going to do an audit process to make sure that these investments are good for the communities and don't have a negative impact.' And I think that's great.
This is kind of off the cuff for me, but going back to the PUC survey, if I recall correctly, it was like 26 companies out of over 100 that actually ended up—or projects that actually ended up responding.
Um, 30. Yeah. Thank you, Charlie. And I think that underscores what Brandon was talking about earlier about the fact that what this is really about is chasing phantom load out of the equation. It's like if you're not serious about building, you need to just get out of the queue and then let the people who actually have their ducks in a row kind of tackle this.
So to dig in on that one just a little bit more though, Asher, as I understand it, the second phase of Beacon Point already has the agreement with ERCOT Texas. Correct.
Correct. So we already have an interconnection agreement in place. So, and it seems from what we've seen from Morgan Stanley that there's a high chance—and I'm not going to try to get you to say this with certainty, I understand that—but there's a high chance that Hut 8 will be considered base load for this batch zero and that it should smooth over the process. So, you're generally confident about Hut's prospects, I would assume, going in even to this audit and into the extension of the batch zero.
Yeah, before the extension happened, obviously today was when they were supposed to release those and we felt very confident and such. And I think at the end of the day, like, the criteria doesn't change, right? The amount of developments doesn't change. And I think all the questions that Governor Abbott had asked in his audit are things that we've already answered publicly through the PUC survey. And so we think that this is net positive and it's not helpful—obviously it is creating some volatility into the ecosystem, volatility into the Hut stock today, but overall, like, we are proud of the projects that we're building because, like, I'll give you an example and I spoke about this to our team the other day. Our first site that we ever built back in the Bitcoin days was a former DuPont sodium smelter that was a brownfield site that we retrofitted because it had a 50-megawatt substation into a Bitcoin data center. And when we were building that site, the subcontractors that were there were talking about how important that factory was for their parents and their grandparents. And it was like it was a center part of the community, right? It helped employ a lot of folks. It was—people thought fondly of the factory. And I think data centers should have that similar impact on communities they go into, whether it be by the folks they employ, whether it be by the technologies that they can teach people, right? We're looking at different things where we can work with local schools and teach them on the technologies, the infrastructure stack and so forth, or the different community projects that they invest into. And so overall, like, there's too much negative sentiment on data centers out there. And I think what I've heard through kind of some folks is some of that is driven by foreign adversaries, some of that is driven by different kind of stakeholder groups. But at the end of the day, what should come out is the truth versus what's real and what's not. And any process that helps verify what's real and what's not I think should be welcomed because it's hard to fight against things that are not real. It's much easier to just say, you know what, these things we don't do. We do not consume water because we have a closed-loop water system. We pay for energy infrastructure outputs and our supply. We're not driving price up. We're actually helping the grid by increasing infrastructure, right? We don't have big noise concerns because we've run all the noise and traffic studies. All of these things a responsible developer should do, and that's how we're going to allow us to go and compete against other countries in this AI race. But like, we've got to get ahead of this, otherwise we're going to just be dealing with all of this FUD that is going to be hard to actually develop responsibly.
Yeah, that's a good point I think with regards to making sure that the process is buttoned up so that we can address some of these concerns that oftentimes are overblown in the media and in the cultural conscience. On the note of expansion, Asher, Hut 8's got 5.4 gigawatts under due diligence, 11 sites across six states. Can you provide any insight into where y'all are at with the timeline for some of these? And specifically, can you give us an update on Batavia, Illinois, and what y'all are thinking in terms of prioritizing these sites going forward?
Sure. We have sites across the whole US today. We have, I think, over 10 different states that we're operating within and looking at different projects within. Those sites range from end of '27, '28, '29, and some in 2030, and some of the sites ramp up per year as well. And so it's a whole slew. You'll see that number continue to increase. We are spending heavy investment on growing that pipeline. That pipeline does not include all of the M&A opportunities that we're exploring right now as well. We have a whole team that explores those opportunities or behind-the-meter generation opportunities where we can bring generation and power where there's a lot of opportunities there as well. And Batavia is a relatively small site of ours, it's 50 megawatts right outside of Chicago. That site is going through a restudy process with ComEd now, and when it goes through that process there's ample demand there, but it's a relatively small project in our whole portfolio.
