← Blockspace: AI & Bitcoin
Blockspace: AI & Bitcoin
MARA’s $600M Texas Acquisition, PUC/ERCOT’s New Data Center Rules, Hunterbrook’s Bloom Energy Short Report
01:06:06
Loading summary
Transcript102 lines
- [00:00]
A
What's going on, y'?
- [00:00]
B
All?
- [00:00]
A
Welcome back to Blockspace Live, presented by Clean Spark. We've got a fat docket today up top. First, Mara entering into an agreement to acquire a Matagorda County, Texas site that has the potential to scale up to a gigawatt in 2027 and 2 gigawatts in the near future. Following that, we hit on a landmark meeting today at the Public Utility Commission of Texas that will decide the future of of how large loads interconnect to the grid and specifically how much of the cost they will bear for not just transmission to their sites, but also for transmission maintenance for the grid in Texas. Following that, a bombshell short report from Hunter Brook Media for its capital arm. Hunter Brook Capital claims that Bloom Energy could be the Theranos of the AI cycle. That's their claim, not ours. But they do get into some numbers that say the math doesn't quite make sense for Bloom. And also their CEO may not be telling the whole truth about where it sources the Scandium for their energy cells. And we will cap today's stream with a letter from IRN chairman to say, hey, yes, we approved a fat stock package for our co founders and CEO. Our co CEOs, but what do you expect us to do? The stock's ripping.
- [01:26]
C
Yeah, I learned a new word reading prepping for this scandium. We'll hit on it in a second. Block Space goes live Monday, Tuesday, Wednesday, Thursday, Friday every weekday at 1pm Eastern. Featuring quick hits on AI data centers, Bitcoin mining, data centers, emerging tech and markets. If you like what you hear, you'll love reading our newsletter and dropped in your inbox every single day, covering the high points, the news we cover, and some other Block Space content. If you missed the live stream, it shortly turns into a podcast right after we wrap up. You can find it wherever podcasts are streamed. And this show is brought to you by CleanSpark. Nasdaq listed ticker CLSK on CleanSpark later in the show. So, Colin, we're going to kick it off with Mara holdings, formerly known as acquisition King of the Acquisition King. They're on the AI leaderboard now.
- [02:27]
B
They.
- [02:27]
A
They would we say that. I think they get on the leaderboard yet that revenue.
- [02:33]
C
We. We'll get into it. We'll get into it. But they said, they said AI so now you can start pricing them as an AI. So okay, so here's, here's the. Here's the headline as we at Block Space Report, How do I change? There we go. As we at Block space report. Mara is buying a powered land site in Matagorda County, Texas, just southwest of Houston from E Fuels Co. HIF USA, giving it access to approximately 2 gigawatts of power capacity on a 1,200 acre campus. It'll build with Starwood Digital Ventures for AIHPC and Flexible Compute. This deal closed July 2 and is a milestone based deal with the total maximum amount that could happen pending some triggers being $600 million. These milestones are regulatory approvals, site being under contract, site under contract, being acquired, approval to receive the power, and signing a data center lease with a third party tenant. Right now Mara stock is ripping at one point was up 18% on the day, now back down to 11 to 12% on the day. The investors like this news. Colin, I can dig into the deal a little bit more or I can toss it to you for a quick take.
- [04:16]
A
So at 600 million total consideration, if all those milestones are hit, this is $600,000 per megawatt. Just with the one gigawatt.
- [04:28]
C
Are you sure it's 600,000? My math. Did my math. Oh yeah, one gigawatt at 600.
- [04:32]
A
At one gigawatt it's 600,000. But my question is, what's actually active at this site currently? And that is conveniently left out, obviously.
- [04:43]
C
Right.
- [04:43]
A
And so the biggest question I think going forward with this and this is coming in concert with the PUC meeting that we're going to cover in the next segment, will Mara be able to reliably secure up to a gigawatt by 2027? I imagine they would not be touting this if there weren't a clear path to that. But if I had shares of Merrill, my immediate next question would be, okay, well, what's actually active at the site? How are you going to scale when everyone in Texas and their grandmother is looking for every juice of power that they can find? And also, are you in talks with any tenants for this potential site already or are you shooting first and asking questions later? So I have answers to those. Obviously the Mara press release left us with very little by way of answering those, but that to me seems to be or those seem to be the most outstanding questions with regards to this acquisition.
- [05:41]
C
Yeah, we don't know that I'm aware very much about this specific site. What's already there, what's the development landscape? Is this simply just a greenfield power land deal? You know, don't know. As per Mara's press release shown here, this can scale up to 2 GW total. By April of 2028. But, but, but initially up to a 1 GW of grid capacity by October 2027 is this kind of a hybrid thing because it's a deal with an E fuels company, HIF usa, the seller retains a bit of skin in the game. They will retain a minority ownership interest upon the execution of a lease with an, with an HPC tenant. This brings Mara's overall potential power pipeline to approximately 4.8 gigawatts if we include the pending long range and energy and power close. And as you said Colin, at 600 million hitting all milestones and, and at 600 million for the deal hitting all milestones, that would imply 600k per megawatt at 1 gigawatts and 300k per megawatt at 2 gigawatts. 300k per megawatt at two gigawatts is a very good deal for industry standards if I'm, if I understand them right.
- [07:08]
A
So to, to answer some of the, one of the questions that I just asked I consulted Jamie GPT as you were going through that Charlie and so this is a greenfield HIF issued a notice, has issued a notice to proceed on a switchyard connecting the site to the grid with access to 1 gigawatt by 2027 and 2 gigawatts by April 2028. So it seems like the ball is already rolling for that. The reason I was wondering if there was already, if there were already assets on the site is HIF is an E Fuels and synthetic fuels company and so this is a site that they are currently preparing and I guess they decided well if we can get a $600 million payday we might as well just go ahead and do that. So it seems like they already have some of the groundwork in place to access that 2 gigawatt pipeline. So clearly a good move on Mera's part if they can realize that gigawatt pipeline.
- [08:06]
C
So Mars CEO Fred Thiel issued a statement alongside this quote. This transaction advances our strategy of securing strategically located infrastructure assets capable of supporting high performance compute and bitcoin workloads. This acquisition meaningfully expands our long term development pipeline and strengthens our ability to support high performance compute and maximize the value over time. Conspicuously, Colin from Fred's statement does not say AI and I'll even observe that the press release itself is really light on AI. In fact, actually arguably hints more at Bitcoin in the flexible workloads language more than it does AI. It does have say AI HPC high performance heavy compute Workloads. But given that this is 2027 and 2028, Bitcoin could be very different regime then. I don't know if this is a good thing or a bad thing. But this does allow Mara to start. You could argue now you can start pricing Mara like an AI more along in these like AI HPC stocks now. Yeah. And I would also say, yeah, I
- [09:23]
A
think the HPC line kind of covers the AI angle.
- [09:27]
C
Right.
- [09:28]
A
I mean if they don't use this for that, then you're going to see that 18% wiped off very quickly.
- [09:35]
B
Yeah.
- [09:35]
C
But I'm comparing this to other statements which conspicuously do not mention Bitcoin. So this retains some flexible workloads and bitcoin reference. We'll see a lot can happen in two years.
- [09:50]
A
Yeah, for sure. It would be interesting to see if they once they had the infrastructure in place, if they did try to just throw some containers on there to monetize part of it. I doubt they will do that.
- [10:00]
C
But yeah, it's a common strategy. We'll see how successful it is to use Bitcoin as like a bridge load if you are sitting on your hands waiting for a tenant or you don't want to sign a deal with a tenant or you have to consume load today. Bitcoin mining sometimes can be an option for that. All right, anyway, so Mara pigs can. They can squeal at us for having been critical of Mara. And here you guys go. I'll throw you some slop. There's a positive Mara news. Look at how high the stock's going. Yay, marathon.
- [10:46]
B
Okay.
- [10:46]
A
You know, I do think just going back to one of my original questions with the interconnection part secured with this, that's, you know, one of the bigger hurdles. It's unclear whether or not they have the actual PPA in place, but clearly one of the biggest moves that Mera has made so far into this AI and HBC pivot. You know, the Xion acquisition arguably was a bit of a flop in the sense that it's not in a tier one market for this compute load. And it's actually kind of unclear as to what the advantage of that is considering they don't own the company 100%, they have the majority stake. But with this, I think to your point, Charlie, leading off the segment, it's like the furthest inroad they've made for the HBC pivots to date. So we'll be keeping an eye on it and as always we'll have updates once they hit the Wire if they do.
- [11:40]
C
All right, we're going to keep rolling. We're going to dive into the two media stories, ERCOT and Bloom Energy here. Before that, a word from our sponsor, CleanSpark.
- [11:51]
A
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. We are setting the standard for what's next. Learn more about about the intersection of energy and bitcoin@cleanspark.com if Bitcoin's actually the best money and it's the thing that people should accumulate and it's the best risk adjusted asset, I lose zero sleep
- [12:28]
B
about whether or not that's gonna happen.
- [12:30]
A
I just ask the question of when it's literally matrix math that you're running
- [12:37]
C
on large pieces of data the bitcoin
- [12:39]
A
miners can absorb or that energy.
- [12:41]
C
And in many ways this feels like
- [12:43]
A
a second bite at the apple to build a new Internet. All right Charlie, two meetings coming today out of Texas. The most important one is the PUC meeting, the Public Utility Commission, there's another one with ercot, but that's more of a procedural thing to make sure that paperwork for the new Batch 0 rules are being filed accordingly. But so we'll focus mostly on PUC and I'll touch on the one with ERCOT specifically regarding Batch 0, but I'll get this up right now. PUCT meeting at 9:30am Central Time in Austin, Texas. Then also we've got ERCOT's own meeting taking place which I believe is actually is a web meeting later today or today at the same time. And the most important part of this is the PUC element. So the PUC is meeting today to decide whether or not they are going to push forward with finalized rules and requests for comment on specifically how large loads pay for grid infrastructure and whether or not to move from 4cp to 12cp. So the 4cp program as we cover a lot on Here, is this 4 coincidental peak program where during the hottest months of the year, June, July, August, September, there are these events known as the coincident peak where power prices spike and oftentimes large loads, specifically bitcoin miners and some data centers will try to curtail in anticipation of those so that they don't share an outsized cost on their electricity. Considering this is the time when residences are at home ripping their ACs, businesses are ripping their ACs and it typically there's a lot of stress on the grid at these times and power prices spike accordingly. So one of the things they're debating is moving this towards a 12 CP program where they do these, these coincident peaks will occur every month. And part of the reason for this, as Texas has moved increasingly towards renewables, the winter months have gotten just as dicey as the summer months because the wind, the wind don't blow or blowing too hard or the sun don't shine. And so you don't have anywhere from like 20 to 30% of the electricity producing assets in Texas actually operating during those moments. Winter storm URI is the premier example of when this can go very poorly for the Texas grid. Winter storm URI caused immense destruction to the Texas grid. People were without power in some cases for I believe a couple of weeks with, with the blackouts that happened as a result of that. But the other thing that is on the docket for this meeting is an elimination of interconnection cost allowances for large load customers. This is very critical because typically before this AI and bitcoin mining boom that happened in the state, utilities would cover the cost, or at least most of the cost for transmission to end users in the industrial consumer bucket. So if you were building a large site in Texas, the utility would throw you a bone and throw you an allowance to build that transmission to you. That is going to be done away with entirely. I think, I think it's safe to say that the commission will decide that that can't be the standard going forward since there's so much demand in Texas. It is a seller's market for energy right now. The utilities have a lot more leverage than they have in the past and they're going to demand that these data centers cover some of those costs to get transmission and electrical infrastructure to these data centers. The other piece of that is requiring large loads to pay a portion of system upgrades costs. So this goes even beyond the interconnection costs that they will likely have to pay. The utility commission is saying, yeah, in addition to that, you're going to need to help a little bit more with paying for maintenance on the transmission and the infrastructure that we already have in place. So a double whammy there in terms of overall transmission costs for the data centers. And then the other one is requiring annual updates to class allocation factor values using proceedings like the transmission cost recovery factor. So kind of a, you know, trying to make sure that everything's running smoothly and making sure that they have clear views into the cost these data centers are having to foot the last one also very big, a minimum billing demand applied to Large loads. Now this is in the orbit of the 4 CP12CP rule change proposal as well. As I understand it. What the PUC is basically saying is if you do curtail and you're on a demand response program that allows you to do so, you're not going to be compensated as much as you had in the past. And I believe that they are going to be setting basically a floor on which you will have to pay something to participate in those programs. The rationale behind this is that if they at least pay something, the grid is going to not make sure that they don't lose their shirt in these instances because we're talking about gigawatts worth of load at this point. That could just cut off like that. And if the data centers aren't footing some of the bill for the generation that they're not consuming at that time, then that gets passed on to the utility or in some cases the residents in terms of higher power costs. And that's the entire rationale behind this. Governor Greg Abbott issued a mandate to the PUCC and to ERCOT to get to basically have a joint statement or a joint understanding on what these new rules will be going forward so that the residents of Texas are not footing the bill for all of these data centers indirectly as they move into the state. I believe they have a. July. July 17th is when Governor Abbott wants that memoir. And then there is a deadline for the new rule book to be firmly decided by December 31st. So the end of this year. I've got a few more things to cover just to make sure this is holistic here, Charlie, but before I do that, I want to throw it to you for second thoughts.
- [19:16]
C
Yeah. I think if you're watching this and you're listening to this, you really don't know much about ERCOT or the Public Utility Commission of Texas and that much about the 4CP program. Miners and data centers have been going to Texas both because they're cheap energy, but also because of the of how they're able to play the power markets. And that has been because they can trade these power markets, they can really optimize the overall load profile of their data center. And that is kind of downstream of this type of regulation. The 4CP, which is now on the. It might be changed. So if it's changed, that could significantly affect the entire regime of how profitable a lot of these strategies have been. I don't currently have a position, but now basically I think you have to become much, much more sophisticated. Bitcoin miners are a very different load profile than AI training and inference. And so but they're swept up in this, in this, in this regulation because I believe like the, the Cutoff is like 75 megawatts, of which like a significant number of bitcoin miners back most of them. The most large ones in Texas are over 75 megawatts. The thing is, bitcoin miners tend to be more agile than AI HPC right now and so they might be able to adapt better. But as far as I'm aware, reading the language and listening to you talk about it, it doesn't seem like there's differentiation between industries and agility of doing the variable load, which I think is one of the. It's funny, this is kind of at the intersection of everything we talk about on and previously the mining pod and now on the show. I'll toss it back to you to keep.
- [21:15]
A
Yeah, and on that note, there really isn't a differentiation, although there is a new, there are new rules for ERCOT evaluating what they call large loads, which is anything over 75 megawatts or above, which would include obviously probably any data center in the state, barring some smaller ones and a lot of the medium to big sized bitcoin mines. So this is called batch zero. This is the new implementation of new rules for how these large loads connect to the grid. These rules are already in Place and ERCOT's meeting today specifically is about looking at how the paperwork deadline July 10th for this batch zero, is coming through the pipe. And it's not a vote on new rules or anything, it's really just a procedural thing to figure out. Okay, is this, this new paperwork process working properly? And the important thing to also note about this Batch 0 rule change is that instead of looking at each load individually, any load of 75 megawatts is now studied system wide alongside every other qualifying project because there's just been so much demand that they now have to look holistically at everything that's going on rather than individually to make sure that the grid's not going to break. Going back to the importance of the PUC meeting and to be clear, what's being pushed forward today is the commissioners are voting on whether to formally publish this, these proposed rule changes and then submit those for public comment. If it passes, the PUC will have formally launched its rules to make all of those changes that we just discussed and to get an idea for what's at stake. The Texas Grid spends about 5.5 billion annually on transmission, maintenance and other costs and about 44% of that can be attributed to being passed on to ratepayers in residential areas. And this is why Greg Abbott is saying that we have to put an end to, to this postage stamp system, as it's called. Because ultimately the vast majority of upgrades and maintenance going forward will be to the benefit of these data centers, also benefit the residents to some extent. But the majority of demand is not coming from commercial or from residential areas currently it's coming from these industrial large loads in the far flung reaches of Texas and outside of the city centers. And so what Abbott ordered is again, he wants that joint member by July 17, but he also wants to actually start cutting those residential transmission costs by July 31st. And so they're right within that window with this PUC meeting. And then if everything goes according to plan, there will be new rules that are on the books by the end of the year.
- [24:15]
C
Look, Texas and the Texas grid may have a revenue problem. They've got to figure out how to, how to respond to their grid changing dramatically over the past four years.
- [24:32]
A
And the silver lining to all of this is, I think probably, I'm obviously not an expert on public utility responses.
- [24:43]
C
We're talking heads, we're experts on everything.
- [24:45]
A
But you know, I would imagine that the PUC and ERCOT have moved quite quickly on this. And to me, I think it's an excellent case study for how a grid and government can adapt to certain white swan events like this, where now they have this kind of good problem of all these businesses want to move there, but you want to make sure that you're not screwing your residents and screwing your populace. And so in the context and in the conversation of AI driving up energy costs, if they're successful at mediating all of this and getting these rules implemented in a way that doesn't piss off the data centers and doesn't, you know, railroad their residents, then this will be a very good example to point to, to say, you know, it's almost a policy choice if you do let this impact your communities, if you're letting these companies come in here and then they suck up all the energy and the energy prices go up, Texas could serve as a case study for that. Doesn't have to be the case. It doesn't have to be a dilemma.
- [25:49]
C
I'll point out, if they manage not to totally hamstring the market and find some kind of happy medium which does strengthen the revenue towards the grid and also not totally disincentivize existing data center players, is this a win for deregulated energy markets because because pretty, I mean if you look at these things, these things, you know, they, they think on a 30 year timeline. So to be able to respond this quickly and institute new like pretty transformative regulation and pricing is an opportunity. I don't, you know, there's probably a lot of people who are, who have very strong feelings about this.
- [26:33]
A
There's a lot of pearl clutching on that. I was having a conversation with one of my friends, friends from Europe a while back and I was talking about the surge in energy prices around the US and using Texas as an example and looking specifically before they implemented massive battery shortage or storage, the penetration of renewables was causing a huge problem in the summertime and it was actually causing energy price spikes. And he was saying, well, is that because Texas has less regulation? And I said actually no, because the fact that ERCOT is a less regulated market and a more free market for energy. Texas has historically had some of the lowest energy costs in the nation. Giving rate payers the choice to choose between utilities is pretty amazing when you think about it. Most residents in the U.S. operate under a monopoly. I have one choice here in Oregon. I have Pacific Power and they suck. I would love to have more than that, but I'll leave it there.
- [27:35]
C
Look, tell your Europore's scoreboard man. What's your cost per kilowatt hour? How much does it cost? Can you, can you buy a house?
- [27:43]
A
Funnily enough, this was a Belgian friend. So if I say scoreboard, he's going to point to the four one way.
- [27:47]
C
Oh yeah, we need to wait.
- [27:51]
B
Yeah.
- [27:51]
C
Oh yeah. Shoot, that does get us well. Okay, we're gonna move on. We've got another very, very meaty story again. Collins put in some legwork on this. This is the, the next one is we're going to tackle Bloom Energy and by we I mean Colin. And we're going to dive into that right after a word from our sponsor, Luxor.
- [28:16]
A
This episode of Block Space 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 five minutes and tests your fleet's power settings against live energy and hash rate markets. Ercot backtests show 10% improved profitability with intelligent mining versus binary mining. Commander Pro is $100 per megawatt or a 25 basis point pool fee adder and you can try it for free for 60 days. So if you're interested and want to learn more, go to Luxor Tech for free. Forward slash, commander. All right, Charlie, I'm going to start this by playing a short clip from Sammy Koppelman, publisher at Hunter Brook Media. And this is with regards to a short report that Hunter Brook Media put out regarding Bloom Energy. Now, I do want just to disclose that up front, Hunter Brook Media is the media arm of Hunter Brook Capital, and Hunter Brook Capital currently has short positions open against Bloom that resulted from this investigation that they did. So I'm going to play this clip really quick.
- [29:31]
B
Yeah.
- [29:32]
A
When we talk about rare earth materials in general, I mean, China controls so much of the marketplace. So are there other places in the world where a company such as Bloom could be getting scandium oxide and specifically enough based on its backlog?
- [29:47]
B
We ran the numbers, we built a model, and we don't think that without China, Bloom could be getting enough. What's important here is that Bloom is a trust me story. This is a $70 billion company that for 20 years has essentially had no success, and now people are betting on it to power AI data centers around the world, when right now, it's not powering a single AI data center, despite having a market cap bigger than Vistra Energy. And so I think the burden of proof at this point in Bloom's history is on Bloom to show that it can actually deliver. And I think the fact that they're lying about this raises the question, what else are they lying about?
- [30:23]
A
So how do you respond to the response from Bloom Energy? They say they're going to correct the record.
- [30:28]
B
I thought the response was fascinating. We reached out for comment, as journalists do. They posted a blog post the night before we published, essentially claiming that they can power 25 gigawatts. Show me the scandium that enables that. And I think that our evidence is very clear. It's on our website. Our process is meticulous and anyone can scrutinize it. Bloom is telling people to just trust them. And the truth is, if you look at Bloom and its history, it's big promise after big promise, and they have never actually delivered on any of it. And you look at their customers now, customers like Oracle, and Bloom's buying that demand. They gave Oracle a huge stake in Bloom, and so I'd love to see independent people independently vet out Bloom's supply chain. I'm sure that every single one of them is going to come away with the same conclusion that we had, which is that this company makes big claims that it can't actually back up of themselves. And I can't imagine a company with a bigger story and a bigger market cap than Bloom with less evidence to back it up. And look, I don't actually know what's going to happen with Bloom. We can just report the facts that we have. Yeah. But I think it's worth asking, if there's a Theranos of the AI moment, who is it?
- [31:47]
A
Little insinuation there. So, you know my intro, I said. They said that Bloom is the Theranos.
- [31:52]
B
They.
- [31:52]
A
He didn't go out and say it, but he's basically implying it. Right.
- [31:55]
C
With.
- [31:56]
A
With that comment.
- [31:57]
C
Okay, before you go into the details, I think a little visual aid could help. This is what scandium looks like. This is a metal. And why. What does the scandium go in? It goes in the Bloom Energy fuel cells, like, inside. So this is kind of the important context.
- [32:18]
A
Yeah, that's good context. These are solid oxidized oxide fuel cells. They basically take natural gas and convert it into electricity without combustion, which sounds like that doesn't even exist.
- [32:30]
C
It does.
- [32:31]
A
And one thing I want to push back on with Sammy Koppelman. He made it sound like they haven't delivered any of these. They have. There aren't that many active currently. And we'll get into those numbers here in a second. But Bloom is building this technology, and it's.
- [32:44]
C
Yeah, I see pictures of them online, like I'm going through, and there's actual pictures. They look like they're deployed not in some kind of warehouse or display case, but, like, out in the wild. And I'm looking, oh, there's some deployed in.
- [32:57]
A
Yeah, it's truly incredible technology. And I think that what Hunter Brooks analysis boils down to is basically saying the technology's there. Is it there enough? And is Bloom executing enough to justify a $75 billion market cap? And I think that that's a legitimate question. So let's get into the meat of the report here. There are basically five claims from Hunter Brook that says that Bloom is not the story that it is trumpeting itself to be. Number one, Bloom Energy uses Chinese supply but denies this publicly. Number two, there's not enough scandium to meet Bloom's 5 gigawatt projections. That's. That's a 2030 projection. Number three, circular demand from joint ventures. Basically, the idea is that Bloom sets up these joint ventures with other companies, and then these joint ventures end up purchasing inventory from Bloom. Number four, there's a huge backlog gap in terms of what they're actually booking for revenue right now versus what they say they have under contract number five is a flagship project slipping. So we'll take these one by one. And the first claim that's super easy to address is that they do or do not use China for supply chains. So the CEO has stated that Bloom has quote, no China supply chain, end quote, and doesn't depend on China for scandium. I'm just going to go out and say this right now that, that it just strikes me as totally patently false. China according to West Point, according to one of West Point's research arms controls 90% of all scandium refinery refining. So if that's the case. Yeah, let me the right here.
- [34:54]
B
Yeah.
- [34:54]
A
The modern War Institute from West Point puts China at controlling over 90% of global refined scandium chemical production and effectively 100% of all metalized scandium used in advanced semiconductor applications.
- [35:06]
C
I also, it's also like it should be really simple to prove to the, to the opposite that you have other like scandium suppliers. Like you should be able to like here's the purchase orders here, disclosures like and, and parade out some dude who sells you scandium who's not Chinese and is like I have an American or South American company or something like that, you know, 100%. Yeah.
- [35:30]
A
And, and anyone who knows enough about how much China controls the critical mineral supply. This has been something we've been talking about for the last year or so. Fact of the matter is, depending on what mineral and metal you're looking at, China has anywhere from 50 to, to near 100% of the refining capacity. And these are, these are all of the metals, the Alphabet soup of metals that are the backbone of this entire AI revolution, the green energy revolution. Most electronics. Like you said Charlie, you learned about a new metal today. You could probably wake up every day and then do research on a metal that you've never heard of that is critical to the AI and the green energy and electricity electrification pivot. So the fact of the matter is, even if you didn't know anything about Bloom Energy, the CEO saying that they source none of it from China could reliably be easily dismissed from the fact that you don't have anywhere else to go.
- [36:28]
C
Yeah, it does doesn't pass the gut check to me. Happy to be proven wrong.
- [36:34]
A
Yeah, maybe they can find the other 10% in other countries and that's enough for their appetite for these fuel cells. But according to Hunter Brooks data, they don't even have enough. They won't even be able to get enough from what is currently produced out of China. We'll get to that in a second. In Hunter Brooks reporting, they actually went to a supplier in China that claims to do business with Bloom. This is Hunan Oriental Scandium, which claims to ship scandium oxide directly to Bloom's Delaware plant. Or it claimed to ship it directly at least four times in 2023 through 2024. The sales rep told Bloomberg, quote, we are also Bloom Energy's or, sorry, told Hunter Brook, quote, we are also Bloom Energy's largest supplier of scandium. And on how materials reach the US under Beijing's export controls, they said, quote, not exported directly. The export controls are really important. We'll get to that in the next segment. So there are three avenues through which these Chinese companies reportedly get scandium to Bloom Energy. One is through Thailand, three circle groups, Thai subsidiary shipped 154 metric tons of ceramic electrolyte membranes to Bloom with raw materials sourced from China. Per three circles own Hong Kong IPO perspectives. They also apparently have avenues through Japan and South Korea. Pretty common. Kind of reminds me of the ways in which ASICS would make their way from Southeast Asia to America.
- [38:00]
C
That was my first thought, is like, oh, he's purchasing the scandium through Malaysia. I mean, bro, we all know that.
- [38:07]
A
We've all been there. Who hasn't tried to just say it,
- [38:11]
C
you know, if you're being pressed.
- [38:12]
A
Yeah, so, and this is important too, when you consider at least a few years ago the fact that China had the largest tariffs on its goods coming into the US I don't really know where the tariff regime is now. The news cycle moves so fast. I mean, we're back in the Iran war when we were reportedly out of it. Tariffs were a huge stinker in the beginning of 2024 and now no one talks about them. Are they even still in place? Point being though, if you wanted to get around those tariffs, one way to do it, like with asics, where do you ship or offshore your manufacturing from China somewhere else and then ship it in through other Southeast Asian countries. All right, so claim number two. Claim number two is that there's actually not enough supply of scandium to meet what Bloom is saying it plans to deliver through 2030. So according to Hunter Brooks supply and demand model, which they backed out looking at Bloom's patents and basically had a rough figure for how much scandium needs to go into a single unit for these energy cells and then took that and extrapolated out to the 5 gigawatt figure that Bloom hopes to produce by 2030 and then looked at global supply of scandium. So According to their estimates, bloom needs roughly 220 tons of scandium oxide annually by 2030 to hit Wall Street 5 gigawatt expectation. This is against projections of total global supply of 240 tons and total global demand of 310 tons annually. This includes scandium is used in weapon systems too. So Bloom's competing with the likes of Lockheed Martin and Raytheon here for scandium. And importantly, there are industry estimates that say that global capacity for scandium oxide was over 90 tons in 2050 and sorry, in 2025, and global production only totaled 80 tons with China being the leading producer. So if you look at that and then you take again into. You take into account the War Institute at West Point's estimate that China controls 90% of the global refined scandium chemical production and almost 100% of the metallized Scandinavian scandium production. There's almost no way to get around the fact that the only place to get this is from China, and there might not even be enough for Bloom alone, let alone the rest of the world. And this is really critical considering the whole fear around China's control of these critical minerals and why West Point even has studies on it, is the fact that these things are used in weapon systems, are using AI, they're used in electric cars, they're used in solar panels, etc. China now has export controls where these Chinese companies have to get permission for whom they're selling to from the Chinese government now. And Beijing could cut off the supply chains like that if they wanted to. So if you're looking at that, this is the whole reason why Bloom probably has to push back heavily against this and claim that they don't use China at all, because that is an extreme threat to the entire business model. It's predicated on China just not shutting down the whole supply chain, which they can do because they control the whole supply chain currently. So that's an existential threat to business outright right there. All right, moving on to the last few claims. I think those are the biggest ones. The reported deficit of scandium supply and also the fact that Bloom Energy is actively saying that they don't source from China when they actually do. If I'm being charitable, I think the idea is that, well, actually the shipments come from Thailand and Japan and South Korea, even though some of them may have come in 2023 and 2024, Bloom is no longer using those avenues. They're skirting export controls by going to other Southeast Asian countries to get their supply. But those countries are sourcing it from China. All right, the circular demand piece is really, really interesting here. So according to according to I believe Bloom's own Financials and Brookfield reporting 74% of Bloom's Q4 2025 revenue 700574 million of 778 million came from joint ventures Bloom part owns with Brookfield and entities that didn't even exist until August 2025. Bloom's own footnotes concede Counterparty's quote may be a project finance affiliate rather than the ultimate end user. The Brookfield Fund's designated anchor tenant is Radiant, which is Brookfield's own cloud company which reportedly has no CEO or customers. So the idea here being is that Bloom is setting up JVs with Brookfield and then those JVs are purchasing their units and that accounted for 3/4 of their revenue. Hunter Brook also flags that Bloom appears to be recognizing revenue not just before deployment but before billing. A new 62.3 million quote non current contract assets quote line appeared in 2025 for their filings which is revenue recognized where billing milestones haven't been reached and invoicing is more than a year out, 78% of it is related party transactions. Deloitte flagged this J flag the JV accounting as a critical audit matter. I think I believe adding to Hunter Brooks concerns here and then Claim four also has to deal with revenue. Blue Markets has an unaudited $20 billion backlog. Its audited remaining performance obligations are almost 500 million which is a 40x gap. So the idea here being basically that they have what they claim is a $20 billion backlog but only 500 million of that is currently is currently in the works and there's no real way to audit this unless they bring receipts. So the question that Hunter Brook is asking is how much of that is reliable and even of the stuff that is on the books there are non binding use arrangements and there are 5 to 20 year of service revenues that customers can just cancel annually if they don't want to actually have those services rendered anymore. Last part Charlie, and then I'll throw it to you is this claim from Hunter Brook that these flagship projects are slipping from Bloom. Oracle's Project Jupiter which is up to 2.45 gigawatts in New Mexico, has no approved air permit and gas pipeline is stuck within regulation. Semianlysis pushed the first power for the site to 2029 and it could even slide into 2030. AEP has 2.65 billion orders tied to a Cheyenne data center campus that lost CRUSOE as developer and that's been pushed to no later than year end 2020 or pushed to no later than the year in 2028 and even no later than 2030. The 2028 was from a February deck and the 2030 was from a May deck. And so those are two of their primary customers in terms of very large orders. And those projects are up in the air. Currently, Bloom does have 75 megawatts active at Equinix across 19 data centers. There's roughly 30 to 40 megawatts installed at a New Jersey site with Nevius and roughly 1.6 to 2.8 megawatts installed with a Core Weave site and Volo, Illinois. So again, fuel cells are out there. The question is, do they actually have enough customer demand for all of this backlog that they're claiming that they have? And also, can they even end up scaling to the size that expectations place them at by 2030? I'm not taking a stance on this. I will say though, they had a revenue of 251 million in Q1 2025 and I believe that's Q4. Let me go back to the Q4 numbers. 778 million in Q4 2025 and the market cap is at 75 billion currently. Which is it roughly? The trailing twelve month revenue price to sales is like 32.5x. Ford is 22.1. That's pretty rich valuations and I think that's at the heart of Hunter Brooks investigation here is basically, is this company just grossly overvalued based on the fundamentals right now and what they can actually do to reliably scale into the future? Bloom Energy has been one of the most booming stocks over the last few years in terms of the AI boom, partially because the technology is pretty incredible. So I don't have a stake into whether or not Bloom is just absolutely fudging everything because quite frankly, Charlie, I don't want to get sued. But I will say Hunter Brook raises some really, really good points in terms of what's happening with this stock currently. And I believe the shares of Bloom fell like 12% after they published their report.
- [47:26]
C
Yeah, Bloom Energy is still doing okay though, I think overall. I mean it's technically up on the five day, still up on the month, up on the half year, which I'll note is around when AI investment wonderkind Leopold Aschenbrunner started disclosing that his fund, situational awareness had purchased Bloom Energy at one point. As of the end of Q1, 2026. It was situational awareness is largest single holding at 875 million above core.
- [48:09]
A
That's crazy. I wonder if that was on account of just not rebalancing. Because Bloom energy is up 1,000% over the last five years and 824% over the last one year.
- [48:21]
C
Yeah. So, yeah, I don't know if he still owns it because we only get quarterly disclosures from Leopold. But yeah, so we'll see. Bloom Energy, you know, is it. Is it a good clean energy battery fuel cell story without a little bit of spicy potential scandal without a little bit of drama? Without a little bit of drama. And also can we just appreciate that it's called Scandium? I mean there's this. These headlines write themselves. They.
- [48:57]
A
They really do. And I will just say this much about the Theranos allegation. Theranos didn't have a product.
- [49:05]
C
Yeah, you. Theranos is way more.
- [49:08]
A
Well, maybe it wasn't less obvious and more of just they.
- [49:12]
C
They what?
- [49:12]
A
They said they were building. They literally didn't build and couldn't build it. It wasn't possible. Bloom is building what they're saying they're building. So I don't think calling it Theranos is the fairest. I don't think that's a very fair accusation. And obviously again, Hunter Brook is shorting the stock, so they're going to try to make it as apocalyptic as possible. The question to me just again, resides entirely, are they able to scale to the heights that investors are betting on? That's why the valuations are so high. They're betting that this company will continue to or will will produce up to 5 gigawatts by 2030. I think that's a fair question. Calling it Theranos though, when they actually do have the technology. And I don't want Bloom to fail. I want them to be honest. The technology sounds really freaking cool. You know, I mean not turning hydrocarbons into electricity without combustion is incredible, but whether or not they can deliver on the full expectations of the market is another question. Question.
- [50:16]
C
Yep, we will have to see. Would love to get one of the. The authors of this report and. Or someone from Bloom Energy CEO perhaps on this pod. But we'll have to wait. The emails are being sent out as we speak. Our. Our booking minions are running wild trying to acquire more talking heads and experts on this topic. Before we go to our last and final story. Iron some shit shareholder pushback against the comp packages. We are going to go to our sponsor. Lygos.
- [50:58]
A
Hedge funds are getting Liquidated. Is your Bitcoin safe? It's not just Bitcoin's price drying up. Hedge funds and lending desk are reeling. With Bitcoin chopping just above 60k, it's more important than ever to know what your counterparty risk is if you're using a Bitcoin backed lender. And with Lygos Finance that counterparty risk is not what it is with competitors. Lygos Finance is our preferred Bitcoin backed lender here at Block Space because they use 100% bitcoin native smart contracts to make sure that you are always in control of your Bitcoin when you're lending out through the platform. With Lygos there's no rehypothecation that no bridging and no wrapping, just self custodial Bitcoin for your Bitcoin backed lending with rates as low as 8.5 to 10% APR. Go to Lygos Finance to learn more. All right, Charlie, we going down and down.
- [51:51]
C
We're going down and down. Okay. So on July 2, a compensation package granting Iron Energy Co CEOs William and Daniel Roberts, yes, they're brothers, 18.2 million restricted stock units which would be worth at Today's valuation between 700 and $800 million was approved. This represented roughly a 5% dilution for existing shareholders. A lot of people didn't like this, notably short seller Jim Chanos. So there's been a bit of a backlash and a 10% stock plunge. You know, just pullback, I'll say that coincided with a secular pullback in Neoclouds overall. And we have now a letter from the independent chair of the board, David Bartholomew, kind of justifying and explaining here's why we did this, here is the letter. I won't read the whole thing, but the highlights are basically that the board and the in this independent chair of the board, Mr. David Bartholomew has four, has three arguments for why they the board award chose to award these, these restricted stock units. So the first one is that this is not the first award they've given. They granted a smaller one in 2025 at a lower base and deferred the rest. And this is just because the share price is way, way higher. There's also a fixed share count, not a national dollar target. So both Dan and Will share full downside risk. This is also a time based and not performance hurdle. This is again a little bit nuanced here. So they claim that because the half of the 2025 awards hurdles were cleared too early, that actually kind of defeated. The purpose of them being a multi year agreement. So those are more price based and these are time based. And then these are shares which can't be sold. So they're restricted. They takes them nine years between eight to nine years to vest. And the argument is basically that the board wants to save the talent. That the reason that in this a very competitive AI race, you look to people who have a history of successfully executing and that is one of the scarce resources. The board believes that the brothers are very valuable. They want to align their incentives long term and comp them for what they perhaps weren't comped enough previously. They want to lock them in for a long time. So there are some criticisms to this. Jim Chano says that even if the stock halved over the next four years, the brothers would still be sitting on a collective $400 billion with no performance condition attached. So this is just time based. So even if like somehow the iron started performing terribly, this is still a massive, massive well into the nine figure payout. Some people would say that this entrenches them rather than aligns them. And then the dilution math, Colin is kind of a doozy. It's 5% dilution plus 17% dilution of projected net income through 2030. This is on the back of, you know, huge pay bomb record NBA sponsorship. And it's people might say that they're spending as if they've already won the race. And I don't think that's clear. Also on dilution, I'm sorry, on governance, while the founders hold just 2.3% equity, they still retain 21.8% of the vote in A via their dual class B shares. So combined they have 44 voting control that expires in 2033 when this final tranche unlocks. So I kind of zoom out really far back and I interpret this as like, how much do you believe in the magic? How much do you like, how much do you buy into the Roberts brothers, like expertise and ability to execute? And do they really have that je ne sais quoi, that magic to do this? And so do you believe in them? So it's almost like a little bit, almost like personality or cult of personality or cult of the mythos of the brothers driven, if I can like really reduce it to the brothers, the brothers Roberts. Yeah. So I mean like, you know, how much do you want to, how much you want to bet on them? Because if you really, really like them and you are one of these iron iguanas or whatever, whatever animal they are, Then maybe you're like you maybe like this. But my read on social media again I'm not really super dialed into Iron Twitter read is it's kind of critical. So.
- [57:48]
A
Well and that's what the chairman basically said at the beginning of the letter. It's that they view the board views them as critical to Iron success. And I will just say Iron was almost, I wouldn't say almost dead but you know I was covering stories in 2021 of them having to forfeit, I don't know if it was the majority, but a sizable chunk of their bitcoin mining fleet because they had defaulted on their ASIC backed loan which they took out at height of the market. So that was 2022 that they defaulted. They took out the loan at the height of the bitcoin market when ASIC prices were sky High in 2021. And this is a company that has since gone on to new heights. I mean I think he said in his letter that iron eclipsed $15 billion market cap 14 billion right now. So clearly it was the all time high, had it well above 14. And one thing that he points out in this letter is that if you look at other comparable industries, the Roberts brothers didn't actually have a sizable equity stake. If, if the math here from Jamie GPT can be trusted, I think each brother had roughly 3.9% before these RSUs after these RSUs, assuming they don't sell from from then on and, and depending on if they dilute further which I they probably will especially if the stock goes up. If they have access to equity financing at favorable stock prices, they're going to each have 6.5%. I would just say maybe some people are saying that's too much but for co founder and co CEOs having a sub 10% stake for a stock doing this. Well, I don't know man. It doesn't seem too crazy to me. I get the headline figure is a lot in Jim Chamo's Chanos Mode wants everyone to think that these companies are just enriching themselves and no doubt they're making a lot of money. But Iron has been one of the most successful, if not the most successful bitcoin miner to pivot to AI. I'm not necessarily a fanboy. I don't know if they'll do well in the future. I don't know if they'll continue to do well. I think there are a lot of questions as to whether or not they should continue the NEO cloud model considering other competitors like Core Weave and NEBIUS according to semianalysis, are leagues ahead of them in terms of performance. But the fact of the matter is they performed incredibly well. Bumping up their stake in the company to further align them with the success of the company seems pretty procedural to me and I think it made a lot of noise just because the headline figure was so big. But as the chairman points out in this letter, it could be substantially less when these options or when these stock units actually vest because it's a four year schedule, they can't sell for two years. Once the tranches vest, they've got 4, 24, 25% tranches. And so I get why. Also people were a little concerned about it being time based and not performance based. But I mean, in a way it's almost the same thing in the sense that, well, they obviously want the company to do super well and if they, they've already hit some incredible performance benchmarks. And as he points out in that letter too, they had performance based compensation in the past and the market cycle just made it impossible to hit those benchmarks. And I wonder if that's almost the, I wonder if this letter almost belies their true confidence in where they're at right now or where the market's at. Because you could almost read that as saying, hey, when these vests, we might likely be in a bear market for these stocks after the ebullient 2025 and 2026 phase, you may even be in a recession, the market could crash, who knows. But the fact of the matter is there's a very high chance that these shares will be worth less when these, when these vest if the market takes a downturn.
- [61:47]
C
I mean, if you co founded this company and you have that low of an ownership now as it sky as you after you like literally staved off like going under, you pivot at the right time, like you, you, I can get it. You, I see you want to get paid for the, for the absolute like back against the wall pivot execution you did in 23 and 24. So I think there's arguments to, there's, there's a bunch of arguments you made. This is for the investors to decide. I think that's the whole point. They're like, hey, investors, we're gonna do this, we're gonna lock ourselves in. How much do you believe in us? So, but isn't that the case of like every company, every corporate, all of these. So this is just, this is just markets, you know.
- [62:36]
A
So yeah, I see you've got a very good.
- [62:43]
C
I didn't know whether to bring this up or not because you want. You mentioned it.
- [62:46]
A
Well, I'm gonna, I'm gonna dunk on people who are mad about this. So go ahead and bring it up.
- [62:52]
C
Okay, so this is from last year, last summer when Dan Roberts bought. The headline is, Dan Roberts bought a 15.5 billion dollar beachfront house. Mind you, that's Australian dollars. So they're just not as much.
- [63:06]
A
How much is an Australian dollar worth, Charlie?
- [63:09]
C
Like, I think it's like worth like 60 to 70% of an American of a real dollar. So anyway, so they, he bought, you know, eight figure house off the coast of Australia, I think near Sydney. Anyway. Off the coast of Australia or on the coast of Australia. And you know, it's a bungalow, very expensive bungalow. And people are like, oh, you spent all this money on a house? And there's a whole hullabaloo about it last summer.
- [63:44]
A
What do you expect him to do? His company Just like 10x.
- [63:48]
C
Yeah.
- [63:49]
B
Why?
- [63:49]
A
You know, look, I get it and I'm not, I'm not usually going to be one to go to bat for these types of things, but anyone complaining about, like, I guarantee you a lot of the people that got in on iron early are not complaining about this because they made a ton of money too. So, I mean, what do you expect? Of course he's gonna do that.
- [64:06]
C
Yeah, there's other people. Yeah, look, there's other people who, I don't have the tweets pulled up. There's people who bought houses from their higher iron investments who are just normal shareholders. So. Yeah, you know the guy who actually did it. Yeah.
- [64:20]
A
He executed. He minted a bunch of millionaires in the process and now he's cashing out some of his stock to buy a nice house. Everyone would do it. Literally everyone would do it.
- [64:30]
C
I think my main criticism though is aesthetic because what they're doing is knocking down this bungalow and building one of these, like, you know, flat top, like more modern bungalow villas. And I, I don't know, man.
- [64:44]
A
Like, yeah, that's not it. That deserves ire more than, you know. Breaking news. CEO of wildly successful company becomes multi millionaire, buys nice house.
- [64:57]
C
Yeah, duh. Yeah, go, go buy that house. Get. Yeah, that, get that bag.
- [65:03]
A
King haters mad because they can't get that bag. Including me, because I'm poor, because I didn't buy.
- [65:10]
C
Yeah, you need to go bet back against the wall. Bet on. Bet on yourself. We're going 100x long block space. Anyway, thank you so much for listening to Black Space live. We do this every weekday at 1pm Eastern featuring quick hits on AI, data centers, emerging markets, tech and bitcoin miners. If you like what you hear, you'll love the newsletter newsletter Blockspace Media and subscribe to the podcast anywhere podcasts are found. This show is presented by CleanSpark. Nasdaq listed ticker CLSK. I'm Charlie. I'm Colin and we'll see you tomorrow.