IREN Weekly Space - Horizon 1 announcement imminent? - Leopold x IREN - July recap + looking foward

Hosted by @Frans Bakker · 2026-08-02 · Tags: IREN

TLDR

The conversation analyzes recent IREN stock volatility caused by a targeted short position unwind by Leopold Aschenbrunner and discusses upcoming catalysts like the imminent Horizon 1 handover to Microsoft. Speakers cover financing strategies for GPU deployments, market mechanics, and long-term AI infrastructure growth.

July price action driven by leveraged short attack reaching 45B notional
Horizon 1 completion and handover expected imminently around August 3
GPU-specific debt at blended 3 percent with 45 percent prepayments de-risking funding
Discussion of Citadel assuming Leopold positions and potential ongoing market impact
Shift toward full-service AI cloud model versus pure colocation plays
Satellite imagery shows accelerated construction at Sweetwater and Childers
Long-term excitement around multi-tenant enterprise deals and hyperscaler expansion
Concerns over equity dilution via ATM balanced by operational cash flow ramp

Speakers

Frans Bakker — Hosted the space and provided detailed analysis on July price action, Leopold's short position, imminent Horizon 1 announcement, and GPU versus infrastructure financing separation.
₿itcoin ₿utcher — Discussed macro factors like Treasury yields, credit risk, long-short strategies, and compared IREN's full-stack model to colo players while emphasizing situational awareness.
Speaker 2 — Asked probing questions on funding math, CapEx differences between GPU deals, and Microsoft cabling validation requirements.
Mark — Expressed patience and excitement about GB300 deployments and future revenue ramps while urging calm amid volatility.

Notable quotes

“it was like a targeted short, of a, a portfolio of Leopold Aschenbrunner that brought the entire industry, at least the part of the industry that he was invested in down” — Frans Bakker
“reached as high as forty-five billion, I think I read, and then that leverage grew to four x leverage” — ₿itcoin ₿utcher
“I still believe Iron will announce a Horizon One handover, on Monday the 3rd of August” — Frans Bakker
“they have a blended average of three percent on ninety-six percent of the, Of the G- no, I think it's the, yeah, the GPU debt” — Frans Bakker
“nothing's changed fundamentally and you're just Kind of waiting for things to turn around where it looks like it was more mechanical in nature” — ₿itcoin ₿utcher
“Horizon Four's, dry cooler, installation is already vertical as well” — Frans Bakker
“when we look forward, to more GB300s and of course, the Vera Rubens, I think we all just need to take a deep breath” — Mark

Transcript

Frans Bakker: Good morning or good evening, wherever you are. I'm just gonna do a quick, sound check. am I clear? Would love a few thumbs up if it is the case. Bitcoin Butcher, can you hear me?

₿itcoin ₿utcher: Yeah, I can hear you well, Franz. Good evening. Hey,

Frans Bakker: thanks for joining me. how are you doing after this, turbulent week?

₿itcoin ₿utcher: Just like everyone else, we take it one day at a time. And, missed everyone the past two weeks, but we had a recharge, and it's time to get back to business. So good to be here.

Frans Bakker: Yeah, it's, Quite the eye-opener, to be honest. I mean, so just to give you guys, a little bit of background color on, what's been going on in the last, four to six weeks, I guess, particularly in the month of July. i-it's been like the price action that we've seen in iron has been so unexplainable, and everyone tries to explain the price action, basically every day, and there's a, a large group group of people that think it's a waste of time, and then there is a large group of people that think it's very, it, you know, very smart to do. But now we kind of, got a little, you know, a little insights into the what, what actually was happening in July, and iron was not, you know, the only stock that had this, you know, kind of price action. So I think from, my point of view, I'm, I'm like a one hundred percent fundamental investor. You always hear me talk about the asset base and, you know, the, investment thesis I have is, leaning on, the people and the assets for the most part, and, you know, everything that, comes on top of that, let's just call it the layers, you know, they, they lean on the, on the, on the first layer. So, it is just extremely it's interesting to find out that it was like a targeted short, of a, a portfolio of Leopold Aschenbrunner that brought the entire industry, at least the part of the industry that he was invested in down, with the exception of a few stocks which I still, scratch my head over. I'm, I don't have his full portfolio list here in, right now, but was he, do you know this, Butcher, was he invested in Hot Eight?

₿itcoin ₿utcher: I'm pretty sure he wasn't Franz, and I also think, purely speculation, but when the son of a sitting US president has a Ownership stake in Hut, I think it's prudent for people on Wall Street to maybe be a little nicer to Hut than their competition, but that's just, Speculation and that's what we're here for, for a little fun. But you said it best, the past six to eight weeks, you couldn't help but ask yourself out loud a few times, like Am I a tinfoil conspiracy theorist or is there actually something here only to find out that it appeared to be a targeted leverage hunt, with a guy who started with Approximately two hundred million dollars of capital and obviously got new investors as time went on, but reached as high as forty-five billion, I think I read, and then that leverage grew to four x leverage, meaning if you owned, For every four billion dollars worth of stock, he only had one billion dollar worth of equity, so it's a pretty-- You know, it's one thing to do make choices as an individual, and this is a young guy who clearly had a vision of the future with his white paper and Sharp guy that will rebound and, like, coming from personal experience, like, yeah, I've made investing mistakes. I think the only thing that surprises me with that is given that he was handling other people's money and had a fiduciary duty to those investors, I found it strange that No one tapped him on the shoulder and this kind of took on a life of its own. But we're kind of just speculating as we go, but I, I do think from some of the things I've read, some things circulated that some of his He had what was called a long-short strategy with, if you're long, you're obviously bullish on a particular security, whereas if you're on the short side of something, you're a little more, you're betting against it. So the common theme with this long-short play was that it was a trade-off that if AI was doing well, that software, stocks wouldn't do so well. So I, from something I read, and again, there's more to it, but I think what it came down to was They saw that he had a lot of leverage and there was an opportunity, and once the market took a hit on the front end of his AI plays, his software plays that he was betting against that he thought would serve as a hedge also started to turn against him, so it was getting hit in two different directions at once, and then it kind of takes on a life of its own and Then you see, there was a good interview with Martin, I, I can't pronounce his last name, but he worked for Jim Cramer and he was the guy who had the prescription, went to jail for raising the prescription drug prices and all that, but now he's pretty big on ACTs Shireli or Shirelli, I'm mispronouncing that. I think he's Albanian in ethnicity, so someone can correct me on how to pronounce that, but he had a good clip on a Podcasts discussing where, how these guys operate on Wall Street, you know, people talk and the word was that he was overlevered, so then kind of sharks smelling blood in the water and they could- Screw him in two ways actually, and he described this where he held a lot of shares of Iron. So as an example for Iron, if I was said hedge fund, again, I don't have any proof of this, this is speculation, but I do believe this to be true, that there were hedge funds that Initially were long iron, probably as we headed towards sixty dollars, then they hear about the guy's leverage position and they can first sell their position and de-risk themselves, which is one thing, but then on top of that, they can go and borrow shares and sell again. So they were selling essentially twice some of these funds and saw weakness and for whatever reason Leopold couldn't adjust and it, it just happened all so quickly and a lot of us were sitting through it too, where you're looking at your portfolio and you're like, nothing's changed fundamentally and you're just Kind of waiting for things to turn around where it looks like it was more mechanical in nature and not fundamentally driven. So those were just some of, the things I noticed related to that front.

Frans Bakker: Yeah, so it was particularly interesting to see that he bought puts on, the semis and then he went long on the, more the infra, plays and, the market just, completely, you know, backed up him, you know, they went long semis and short, the infra stocks, the clouds, et cetera, and, and Bloom Energy of course. But it's just, you know, the, the whole, psychological- Psychology behind this, where we were like, questioning our fundamental thesis when there's just a, a, a few trading desks conspiring against a, twenty-five year old, fund manager, you know? it's just, it's just hilarious to, you know, to a certain extent, but it was also very tough to, to go through. I, I'm sure there are, are people, that are listening right now that have experienced, significant drawdowns, in the last- Six weeks and, have really questioned themselves, blamed, things that aren't related. I mean, I've, I've, I've lost a bunch of subscribers that went pretty much from full, blown bulls to the biggest haters of Iron, including the management and, most notably the ATM. I would, you know, I would argue that, having an ATM in place does, invite a certain, short you know, sentiment, but I wouldn't, I wouldn't like blame the ATM on the price action, and neither do I think that the, remuneration package or, or something like that was actually a real cause of, you know, drawdown. I mean, there, it, it must have been like a s-small, part of it. W- I, I do have to say Iron was front-running it a little bit more than some of the other stocks, right? I would say the drawdown in Iron really started, earlier than I think, Blue Energy or Nebius or something, did. And so there's probably something related to the fundamentals, but if you look over the, over the period of July, it was mostly in line with the other stocks, you know On average, weighted basis. So I, I, I think it's just, a very, interesting phenomenon to happen, and I'm glad that we can put this in the rear mirror and move on, because at the end of the day, there are mostly positive things to look forward to, with regards to iron In particular, at least. I can't, speak for the other stocks, I don't know if Bloom Energy has any catalysts coming up, or, or for that matter, any of the other holdings that he had. I think Neebys was really his biggest, single name position, right? so I, I, I think it's, Kinda like sad to see that he was like, had to, had to file a thirteen G or thirteen F for, being more than five percent shareholder of, of Niveus, and now he just had to liquidate the entire position and, hand it over to Citadel. Now, Iron has always, oh, always, has in the last eighteen months had a very high percentage of, hedge funds, a market maker kind of, shareholders. And we've not always been very happy with that. And based, you know, judging the price action of Friday, I would say that, you know, if Citadel was Dumping all their holdings that they took on, f- on the cheap for, from, from Leopold. Why was, Nubia's doing better than Iran, right? I mean, so I, I think there was a, a lot of different forces at play. I think probably Citadel wouldn't be just, you know, oh, we, we got a discount, let's, let's just throw it back on the market immediately. I mean, I guess they, they would take a little bit more time and, be a bit A bit more prudent with that. at the end of the day, my personal opinion of the Friday price action is, it was a broader market that was pulling back, and, on top of that, there was the weekly, options expiry. I, I won't, I wouldn't say that we are have seen the last of Citadel, yet, if there is anything to come, if you catch my drift. so of course, all of this is speculation and, we don't know what their, what their plans are. I do know that, I am glad that Leopold didn't have a five percent stake in Iron, looking back, because, you know, if, if I would have to- you know, consider right now that Citadel would have his shares, yeah, I'm not so sure if I would be happy with that, so yeah, so, that's, that's, probably, enough on this topic. I mean, w-w- there's a lot of speculation, with regards to this. I don't know if you have any, final thoughts on this, or any comments on what I said, Butcher, otherwise we'll, move, move on.

₿itcoin ₿utcher: I think part of our job here, we, we're certainly iron balls, but you have to have, the right term should be situational awareness, right? That would make sense given the, but jokes aside, I, I think we have to be honest with ourselves, with our thesis right now, and the macro- We'll see how tonight holds up, but I think part of what's going on too is if you consider, again, the conflict that seems to not be ending, but maybe there's progress, but with oil prices being higher and now they're down overnight, back under eighty bucks a barrel. But I think the thing that I've been looking at more closely lately is the long end of the, Treasury curve, meaning the US government raises funds through the US Treasury market and the cost to fund our government because of this. perception of an ongoing forever war keeps going up, so that impacts if the US government is issuing what's known as the risk-free rate That impacts the price of all that, and you see that trickle through into, like, for, as an example, Galaxy tried to price Senior secured notes, I believe the instrument was, that was almost ten percent because they were doing business with CoreWeave. So it makes-- there's two things, it makes you appreciate that Iron has secured really good debt terms in the past But it also signals that there may be challenges ahead, and part of that has to do with like CoreWeave has a really weak balance sheet because they have a lot of debt, so it's harder for them to pay back all their obligations. So as a creditor, you would want more money in exchange via higher interest rate payment. And similarly, Oracle has had, they've had their own issues, and the best way that you can look at this is there's this, CDS, which is a credit default swap, you're seeing that go up, and that's just the price of insurance to ensure the bonds. So if the bond were to default, then a bondholder is still made whole, even though the Counterparty can't pay back. So there is this problem with credit risk or a perception of it, and then it doesn't help, in my opinion, like I jokingly had a post after OpenAI came out with a funding backstop from Nvidia, and it appears we are speculating, but I'm speculating as part of the deal with Nvidia that NVIDIA will likely serve some sort of backstop as Iron starts dealing with smaller AI labs and enterprises, 'cause part of the main attraction, even though the economics weren't ideal, when they signed the bare metal deal with Microsoft, they were able to sign a triple A credit rated company, and that's why they were able to finance those GPUs, in the low sixes or might have been the high fives, I don't remember exactly off the top of my head, but I was just talking about earlier how to finance data center debt for Galaxy for CoreWeave was closer to ten percent. So That's all to say that there is kind of this question up in the air, like I don't, after listening to earnings fronts, and this was kind of my attempt to transition into this past week of earnings, like you hear good color from Applied Digital, and then you hear all the hyperscalers speaking, most notably, Google had spoken to more third-party data centers to fill their capacity, not sure that would be iron, but Indirectly that helps Iron, because Iron has such vast, secured power, where if Google's using up more of Cipher or TerraWolf's power, even if they don't do business with Iron, Iron's remaining power is more valuable, just as examples. And then similarly, you saw Microsoft themselves echo that sentiment, as well as Amazon with the last hyperscaler. being Meta, their problem and why their stock was down was there's this perception that the big four hyperscalers, and then sometimes people include Oracle or CoreWeave as hyperscalers, they're investing all this capital and CapEx But they want to see a return on investment, and Meta kind of, their financials weren't as strong and results weren't as strong, so there's this perception that they're throwing bad money and not seeing the return on investment as opposed to someone like Amazon who quickly showed that they're converting as is, Microsoft and Google. So those were just kind of some Similar, category, items that are related to AI, also the memory players reported and they did pretty well, Samsung, SK Hynix and I think they may have missed their numbers because there's this distinction between spot pricing immediately in the marketplace where since there's such a shortage for high bandwidth memory or even, DRAM, or excuse me, CPU memory, there's a distinction there. They could attract higher spot prices right now, but to de-risk their business, they're taking on longer-term contracts, which I don't think the market fully appreciated. So that's kind of-- that was actually something I think is bullish for the sector, where in the past memory tends to be a lot more cyclical and crash, where now if they're Securing these long-term contracts with the hyper-- or, I should say, the chip producers or in some cases even car OEMs that makes their business a little more viable going forward as they try and expand and accommodate AI demand. So those were just some kind of a mix of macro and some of the industry, but I know no one's gonna- Top view fronts as far as Iron specific, knowledge here, but I, I try and bring a little outside perspective and things that I'm looking at just to kind of circle it back to the business, and I think it does impact how Iron trades though, and unfortunately, there's this funding gap right now where until Iron, hopefully this quarter, and we'll, we'll talk about it in more detail, whether it's this week or next week, until they start showing operational cash flow, the The burden is on Iron to prove how they're gonna fund their business and, but I think that's a good segue and we can talk about this next, Franz, if you don't mind. Agrippa and God Particle had some good analysis on the press release that came out I don't know if it was last week or the week prior regarding the economics on the most recent deals and why we can feel a little op-optimist, more optimistic on some of the things we were speaking to in March when we were asking for higher guidance finally came to fruition and are showing that there's better prepayments, which means there's less of a funding need up front for their GPUs and When they decide to use the ATM, it most likely is being dedicated primarily to infrastructure. So that touched on a few things. I don't know if you have any reactions. I kind of did a shotgun approach, and that's what happens when we take two weeks off, but that's kind of what I saw these past two weeks.

Frans Bakker: Yeah, so, I agree with you that, Iron's, you know, debt financing to date, have been very good. they have been, you know, the rates have been very favorable compared to peers. that, that being said, I do think that, especially on this platform, within the retail community, the concerns around the equity part of the financing Are, you know, of valid concerns, to, to the extent that, we talked about earlier, where an ATM overhang has a certain, you know, impact on the price action as well as it invites desks, trading desks to short, or to, drive momentum. I think we can, after this whole Leopold, story, I think it is, you know Basically a fact that, whatever, happens, you know, with leverage and people playing IRE, or, or massive, option positions, the, the market sees all this, they can read the tape and they will, you know, abuse it where they, when they can. within, my subscriber group, I have mentioned, the word fourth C. So Iron has introduced the three C's and I've added a fourth C, which is basically Stands for conspiracy, but conspiracy is just, a, a multiple, you know, parties working together to conspire, against or in favor of their own portfolio, or, or against the others. And that's basically, I would say what, what has happened. I, I can't imagine it was just Citadel shorting everything, that, Leopold had. So what I'm trying to say here is not like the tin foil hat conspiracy theory kind of thing, it's just that in, in this market, there are forces at play that, Ride on sentiments that, that, that go against leverage, that, that liquidate people. There's, algorithms that are trying to, hunt for stop losses. It is just the way, this market works. So, I would say the indirect results of, having an ATM in place are probably, more evident than, just You know, the pure fact that you have it. Like, I would say that the amount of shares that Iron would probably, distribute on the market or, or like tapping the ATM on a day-to-day basis, I, I don't think that is, going to impact the, the share price that much. I mean, they, they probably have the instruments in place to prevent that from happening. And from my communication with the, company in the past, it's been, You know, more or less confirm that they, they, they intend to raise on green days, right? So, raise into strength. And, so what I'm, what I'm saying is, is, is because for the listeners of this space, the equity part is, is something we have to touch on. So now I will just give you my, reply to what you said about the debt, part. yeah. So, having said all that, I think there- Where, their GPU financing has, have been solid. I, I really liked that lease structure they put in place as well last year. can't actually, mention that again at Rays, that, they have a significant amount of, something along those lines, of, lease, instruments in place for the GPUs as well. And I do think, I agree with you that the, the depth, the GPU specific depth is, is that, that's what it is. I, I think going forward it will be probably Iron's, focus to Get debt for GPUs and finance or pay for, infrastructure, the data centers, et cetera, with, cash, with convertible notes, proceeds and with ATM. so, of course, th-this is my anticipation, this is my estimate what's gonna happen. I think they're going to direct repayments towards GPUs, as they have said, we have received, forty-five Five percent of the GPU CapEx in prepayments, I think they are going to, going forward, I think they're going to, you know, separate this to a certain extent where they wanna show like, "Look, we are doing AI, we are doing AI cloud, we're doing infrastructure as a service, GPU as a service." instead of colo, but look what we're doing with the additional CapEx, we aren't paying it from our pocket, we are having our customers pay for it, we're having, financing, in place, GPU specific debt with low rates. I mean, they're touting a blended average of three percent on ninety-six percent of the, Of the G- no, I think it's the, yeah, the GPU debt, right? For, for the Microsoft contract. So it's just, it's just Iron's way to show that by not doing colocation, they aren't inviting excessive, dilution. that being said, of course, you can also, look at the other side of that. If they would be doing colocation, would they not be able to get prepayments for just colocation contracts that could be diverted to the infrastructure? I don't know, but by isolating the GPUs and specifically touching on that, I think, it's, it's starting to become clear that, you know, they have an internal colocation fee and now they're starting to, you know, they have projects, e- data, of eighty-five percent, and they are going to start to finance the GPUs, separately, like at, at least from an optics, standpoint, which means that, you know The perception of the market could be that there's not that much higher risk at, at Iron if they are able to demonstrate that they can, going forward, keep on doing these, relatively low rates, a GPU debt, including prepayments, etcetera. Okay, go ahead.

₿itcoin ₿utcher: I would just say that, yes, the GPU financing and by having the higher percentage prepayments that we've both talked on does de-risk the Iron business, but I'd still- I think both things can be true that being a full service cloud and trying to own more of the stack, the upside is Way greater, but I would say the difference between Iron and some of the Colo players is the Colo players are coming out of pocket with less Equity, and it is less dilutive on the front end because they're able to finance the data center upfront and pledge future lease payments against the data center, and there's some companies that are putting up as little as five percent equity to ten percent equity, but I think Cipher and Wolf specifically tend to do twenty percent. So on a data center that's, two hundred megawatts that costs, roughly two and a half billion dollars right now, two to two and a half billion dollars, you know, they only have to come up with five to six hundred million dollars as opposed to Iron paying that complete data center out of pocket right now, which will have a future life and they won't-- they'll have the optionality down the road, so I- I'm just trying to be upfront with everyone where Iron is taking a harder path that is more dilutive immediately, but the whole point of it is down the road that upfront dilution, they will be able to generate way more value per megawatt than their colocation counterparts who, while those, their paybacks are five to six years instead of seven years now, like the economics are improving. For Colo, but they're also locked into those agreements for anywhere from ten to twenty years, so they do cap their upside. So that's the, that's the trade-off, and that's the, by owning this, that's what you're, That's what you're betting is that Iron's able to generate more value despite that upfront investment in the infrastructure. But having said that, that reminds me From Microsoft specifically had mentioned they were extending the useful life of their data centers to twenty-five years, and Google was extending the useful life of their data centers on their balance sheet to thirty years. So as an example, Horizon one through four for Iron is only a five-year contract, but if that data center has a useful life now, especially given that it's Microsoft specs of twenty-five to thirty years, that means there's twenty to twenty-five years of value that, and we know that the higher upfront CapEx costs make it, generational. I mean, certainly there's gonna be a third or fourth generation But it's fair to say for at least the next ten to fifteen years that they don't have to change the build out, and Franz can speak to that better, but that's the point is after five years, if you see more moratoriums like the state of New York right now, or if you see more shortages like PJM, those existing data centers that are built will have more value because it's gonna be harder to build new capacity and the fees The thesis on AI is that the growth in that AI usage rate, whether it be through training or inference, is gonna be exponential while these hurdles that are facing the, whether it's the grid or it's behind the meter, people trying to figure out how to do that effectively, because that's mostly theoretical at this point, and we haven't seen outside of Colossus a large scale behind the meter, project, or there might be a few others Like Meta has one that's being built in Ohio. But all that's to say is with Iron, you are putting in money up front and you might not see the immediate reward, but down the road, if things-- it's just having a worldview and being willing to wait and see it through.

Frans Bakker: So I don't wanna, make a big comparison, space between colocation and, GPU as a service, but I'm, I am trying to, say that I think Irons going forward will try to provide a clear line between the difference in how they finance a GPU as a service business model versus, say, pure colocation play. And I would like to, I, I'm wondering, I mean, I'm, I'm like zero percent, In, you know, invested mentally into these, colocation plays, but I know you are pretty well versed in Cipher. these, these deals that Cipher, has, are they typically involving pre-payments to a certain extent where it's like similar to what Iron is able to, to get now for, for example, they had twenty percent from Microsoft, now they're getting forty-five percent on the GPU specifically from the- The, the last, you know, iteration of their, B three hundred, deployment, w- you know, how does that work with, for example, a cipher, and then compare that to if Ironwood have Horizon One operational, they could, do like a, an asset-backed financing on top of it. Is, would, would all these things together, like the prepayment from Microsoft plus the asset-backed finance, would that not exceed or at least rival something? That Cipher is doing with a pure colo play.

₿itcoin ₿utcher: Well, pure colo, we were talking about earlier, Franz, there's two things. The deals have been smaller, for, although Terwillegar Specifically has two newer projects that are gonna be a gigawatt plus, one in Eastern Kentucky and one that just got approval in Maryland. Most of these projects that have been leased out so far are two hundred to three hundred megawatts, some as low as one hundred and fifty megawatts, but just for round math, if you assume Like Patrick in an interview, I, I had the opportunity to speak with him, but he also said this on Mcnally that they're, they have a different business model. Like They are less cap intensive in the sense that their job is to simply build a DC, have it temperature controlled, make sure that there's plumbing to get to the racks, and make sure that the racks are powered to five nines, and whether it's FluidStack, whether it's Anthropic with Terawolf's newest deal, or whether it's, Amazon with Cipher as an example, all those guys, all of their tenants are bearing the burden of the tech obsolescence with the inside of that, shell. So if they want to upgrade the racks to So as, as long as there's a power connection, anything else that is going on inside of there is the responsibility of the tenant, and if they're just building for- they typically guide, I mean, it's going up a little given the labor costs, but most of the guidance for like cipher and Terawolf's been ten million to twelve million per megawatt. So if you have, two hundred megawatts of critical IT, you're looking at, you know, roughly two billion to two and a half billion dollars, I was saying as the example, but they just have to come up with essentially twenty percent equity. So Now the goal of both of those entities down the road is they keep developing, and then eventually that net operating income that they're generating from those leases creates their own flywheel, and then they have these predictable stream of payments that they can use to, use for other data centers and create that equity and that cash flow or Potentially even once these sites are stabilized, even sell them, like in the case of Wolf, like Wolf just signed, sold their joint venture that they had with FluidStack at Abernathy, so now they have five hundred million dollars that are readily available that they can use for one of those two gigawatt sites that they, are gonna be developing. So it, again, it's just a different business model. I like having exposure to both of them because- Because while I think IRNs is the superior model right now, given it's easier for Wall Street to value and because it's more predictable, so that's why I think you've seen outperformance from some of these players this year, even though the Cash flows haven't necessarily hit yet, they kind of can say at this point in time we're gonna get these rent payments, and it's, it's just like valuing a, a REIT to some extent, and eventually, they'll get the valuation premium of like a DLR or an Equinix. So I just personally like having exposure to both methodologies, and you can kind of play the pairs, and when one's running a little hotter, you can trim and reallocate it, like, if If you were to see Wolf or Cipher get back to thirty bucks and Iron's still at forty, forty-five dollars, like probably would be a good time to, reallocate to Iron. So those are just some of, I hope that answered your question.

Frans Bakker: Yeah, I think the business models are completely different, and so will be the cash flows and, the tools that they have. I just, I can't help but wonder, Iron typically mentions things for a reason, so I think that their focus on the ninety-six, ninety-six percent blended average of three percent all over the GPU CapEx is a very specific, mentioning of the separation of the GPUs and the infrastructure. So Going forward, I would, seriously expect Iron to focus on GPU financing and, f- you know, self-fund, or, or- To, you know, like, a convertible notes, ATM, a cash operating cash flows going into AI infrastructure, so data centers, substations, et cetera. so that being said, I do think that the, the start of a potential, flywheel, could be that the, like I said, when they can start to finance Horizon with asset-backed financing and pour this money back into- Future, or ongoing, infrastructure upgrades. So there, there will be a point in time where, you know, let's, let's, go fast forward, twenty-four months, and you are able to finance Sweetwater one, six hundred megawatt of, gross power or like four hundred megawatt IT load, you have Horizon one to six, financed This could, like compound, and then if they will continue to be able to get high prepayments from customers for the GPUs, you don't have to worry about that part, because they will be able to pair these prepayments from the, the end users together with, The, GPU depth, specific GPU depth, and then there may or, may or may not be a backing, or a backstop from Nvidia or whatever, that I don't wanna speculate too much on that because it's, it's, over my head, but, If there is such a thing, I think they, they will consider it, but Iron likes to minimize, third party, risk and reliance on, external, so I'm, I'm not sure if this is an absolute, requirement, but I do think, like what you said initially, at the start of the space, that if they will do more deals with smaller enterprises and, smaller end users rather than, an anchor tenant, I do think it would make sense. They would look into a structure like that to be able to get, that, you know, lower-rated GPU debt, because that's ultimately what it comes down to, to keep this flywheel spinning, they will need to increase the GPU, prepayments, so the specific GPU, CapEx prepayments from customers, and, they need to keep on, having lower, rated GPU debt. Which could be based on the end user. So yeah, the, the good thing to point out here is that Kent Draper said at REI's, a, a couple of times that The, sub, sub-investment grade, counterparties, sub-investment grade, customers now, are, are, you know, there's an appetite in the, in the capital markets to, to finance these GPUs, for those end users. So I don't know how that structure will work in, in reality, like, like how does, how does it really work, but if he says this on a public- you know, in a public, gathering like that, I do think that this, this is what we're gonna see going forward. So that now, if we like bridge this to operations, the big question is, where will they be deploying these, end users, these sub-investment grade, customers? Will it be in Sweetwater or will it be in, Childers? So that is- This is the big question, and, I think it's going to entirely depend on NVIDIA, because ultimately, the Sweetwater is a DSX, You know, reference, like a DSX reference design and a flagship deployment. So I don't know if this, is, even suitable for multi-tenants, enterprise or if it is, designed with, an anchor tenant in mind. So I would, I would say that, both, Horizon five, six and beyond, is probably Vera Rubin, and, the same will ca-- will account for Sweetwater one So it's not just as, you know, it's not about, which GPUs, or, or what quantity, because, of course, Sweetwater One is a two hundred megawatt IT load deployment, the first phase, and then, Childress will be a one hundred, megawatt IT load, deployment for, Horizon five and six. You, you know, maybe it will be as easy as Microsoft takes five and six and, multi-tenants in, in Sweetwater. I don't know. but at this point in time, I am not one hundred percent convinced that the deal, for Horizon five and six is a certainty with Microsoft. I mean It, it just, f-from my communication with the, s-people that are involved in the build out of this project, the, the terminology of Horizon has been mentioned in conjunction with, mining buildings, right now as well, as in Horizon seven and eight, for example, are not necessarily a liquid cooling deployment at this point in time, but more a talk about A segment of the site, if that makes any sense. So, what, what I'm trying to say is, it, it makes sense of course that, Horizon five and six will go to Microsoft, because of the, why would they demolish mining buildings or disassemble them, in favor of, liquid-cooled data centers, at that particular, spot on the site, right? Why, why would, why do they have to neighbor Horizon, three and four so much. Like they're, they could just build new data center somewhere else and, or, you know, they could use Sweetwater, Sweetwater one's first primary substation for whatever they wanna do. But no, the, geographical location of Horizon five and six must be exactly next to Horizon three and four. So You tell me, does that, sound like Microsoft? Yes, of course it does, but, you know, by the same token, you know, it, it could, it could also be s-someone else or maybe a multi-tenant. I mean, so, so it's, this is, this is a part of the, you know, of speculation that I, I just can't give you a, a very definitive answer. Somewhere between these two deployments, there needs to be a multi-tenant, enterprise, or maybe, like a Fortune five hundred kind of customer There's just no doubt about it. I mean, I just, I think with the whole Morantus deployment, sorry, Morantus acquisition with the coordinates, AI, with the, open-weight, a-open source push, with the whole, movement around this I, you know, and the way that, Iran's, people have been speaking at raise about this, I think they are gonna service the whole bunch, so the whole range of, Hyperscalers, large, language, model providers, the, the, the closed ones, but also, they're also gonna service, the, open, open-weight users, like, like for example, Together AI that is, focusing a lot on this. They're gonna keep on expanding capacity with, Fireworks AI, YouMay, all these guys. But at the same time, I think they're going, to You know, to include these smaller end users as well. At the end of the day, like Fireworks AI is a middleman, right? I mean, the-of course, Iron doesn't have these ca- these inference capabilities in-house, so they will probably keep doing things with a Fireworks AI, but I think going forward, it's probably, A certainty that they are going to add the end users of these companies, into their portfolio. So, yeah, it is going to be a big, you know, I think Iron will have to inform the market very, well in the coming earnings, if the analysts would have a good question, if I could propose a good question to ask is, I have a, you know, I have a question for Ken Kent, you said at Raise that, there is an appetite from the money markets for, from the credit markets for super investment grade end users. how do you see the rates for these, GPU deployments and, how do you see the prepayment amounts and, where do you see these customers ultimately ending up? Like something like that. That, that, that would make them be able to speak on this. I think that would be extremely valuable because they have been touting about fat margins from end users, right? Fat margins, The high appetite, the, adoption rate is still very low, but the demand is overwhelming, and, You know, d- they just need to explain how they're gonna finance that. are, is it gonna be like very high amounts of prepayments? Maybe, maybe, let's, let's just say that what if- what if lenders for debt are saying, as long as your GPU specific prepayment has a sixty percent threshold, we will be able to give you a, a very low, like a almost like an investment grade kind of rate on your GPU debt. Like, does it make any sense? I mean, this is super speculative, but we are seeing forty-five percent now on, on twenty twenty-six deployments. What if, Iron just, sits down with all these guys that say, "We want your GPUs," and say, "Look, if you can pay seventy-five percent of the GPU, specific CapEx or, or, you know, up-front," We will give you, this big of a cluster in this quarter, or, or something like that, right? I mean, this is ultimately the only way that I can see it happening, unless it is a backstop from Nvidia and makes, which makes some kind of like a, investment wrapper, right? Like, I think you spoke about this as well, Butcher. So, yeah, so why we originally came to this topic is because, rate scares are, are, you know, we, we see them almost, every week, but especially around these FOMC meetings, and I, I think it's, it's You know, iron is high be-high beta, but it reacts excessively to, you know, worries about rates, and I think it's not entirely justified, especially if on iron's end they are able to negotiate with prepayment am-amount that will, you know, have a much higher e-effect and impact on, their, their debt rates as well, because the more GPU prepayment, amounts they can get, the lower the absolute amount they need to borrow. And, you know, if there is a, a, a few basis point, interest increase on a, on a smaller absu-absolute amount, I think it just, you know, it doesn't really do that much, by comparison, to the, negotiations around prepayment amount. So yeah. sorry for, for, for giving you guys this big rant. It's, it's really something that, you know, I, I, I watch these, Morgan Stanley, watch on the market YouTube videos like every day, and it's, it's always about, you know, it, it always gives me a very bullish, iron, feeling, you know, like, the way they, they talk about, credit markets and where the appetite Ideas, it just, it's really in line with what we hear from, from Iron, especially, the way that Kent has been talking at Raise. So, Like the analyst question from the last earnings, I don't know if you remember, one of the first questions was, "How are you gonna finance five gigawatts with, with Nvidia?" Like that was, I mean It's a, such a, such a stupid question, you know, like, like, it's like a, a shotgun approach that just makes no sense. How can you answer that as Iron? You're gonna like have to break down all these little parts of the, the, the GPUs and, and everything, right? And the equity part. I think it's a, a lot better that, that Iron is now starting to talk about GPU, specific prepayments and, touting their, ninety-six percent The, you know, financing, blended average of three percent. this is a great way going forward, but now we're gonna, have to see them talk about end users, enterprise, customers. so yeah financing is a big topic for Iron, it's a very capital-intensive business, so it's always going to be a big topic in these spaces. So, I would, first give Butcher a, a, an opportunity to, reply if, if you have nothing, Butcher, we can go to, Stock Analyst Pro.

₿itcoin ₿utcher: Really quickly, the only thing I would push back on, on Horizon one through four, if you were to take out a mortgage per se against the DC, the difference between that and a colo project is With H1 through four specifically, it was roughly three billion dollars to build and another five point eight for the GPUs, which is eight point eight billion, and then the proceeds from Microsoft are roughly, I think, nine point four or nine and a half, so Over the five years, there's eight hundred million that's in excess, but then you're, you have to take into consideration the debt interest for those GPUs. So that's a long-winded way of me saying that you would need- Cash flow from Canada or Texas via the, air-cooled DCs to support any debt that you would take against that data center, because all of the revenue associated with Horizon one through four is paying down those GPUs, given that they were signed for like two ninety-one an hour. So Every time we're pretty passionate about getting GPUs plugged in and at higher rates, it, it has, ramifications that kinda Compound in a good way or a bad way, and that's why we've just were, we'd like to see things, and everyone knows that they're working as hard as possible, but the moment that they're able to plug in those GPUs, especially in Canada, given that the DCs are Paid for, and a lot of those GPUs were paid for with cash. All of that extra cash flow that, comes from those operations can be used to service, You know, data center level debt, or they can use it to, invest into those new projects, whether they be in Australia or Sweetwater. So Thanks for that, Franz. if that's it, we can go to Stock Analyst Pro. How are you this evening?

Speaker 2: I'm doing good, and thanks to you both for covering, in depth. So I have a few questions on the funding. So, yeah, in the overall twelve, thirteen billion market cap of Iron, you know, the cash and cash equivalents are seven point eight billion or something that was from June thirtieth. And, we get some of the revenues because of the ramp up. Now, so McFly has done amazing job in, he, he published an article just, some time ago. Around, the 7.30 megawatt build out possible in, in 2027 and additional one in the 2028, right? For the 2027, for 7.30 megawatts, the rough math was around twenty-seven billion to thirty-five billion for the CapEx. The Ireland may have around seven ish, and there will be some revenue coming up from the current build out. so Fran, since you're going a little deeper on the, on the Subject of GPU financing. So do you know what was the difference between the previous Microsoft deal and the new deals which were signed? because there is memory also which is expensive, so when we combine memory, CPU, GPU, and some of- The other, the server build out, the overall cost is around seventy percent of the overall hundred percent cost, roughly. So, is Aaron's deal, the next deals which were signed, were only focused on GPU or it had GPU plus the server component?

Frans Bakker: I would have to pull up the, announcement, so I don't have that answer ready now, but it, I think it was specifically mentioned as a GPU related, CapEx. if you, yeah, I, I, I would need to, let me just pull up the, Announcements, I'm just opening it now. There was one line, so eighty-five percent is under contracts. And, oh yeah, so also include customer prepayments representing approximately forty-five percent of the associated GPU capital expenditure, reducing Iron's net funding requirement for those deployments. and then it says, there is a footnote three, customer prepayments represent amounts contractually payable by customer in advance of service delivery under agreements executed since June one Twenty twenty-six, expressed as a percentage of the estimated capital expenditure attributable to the associated deployments. Prepayments terms vary by contract, and there can be no assurance that future contracts will include pre- okay. yeah, well, it says specifically GPU, but as far as I'm aware, memory is part of the, the package, right?

Speaker 2: I think so memory comes as a part of the server package because server includes both GPU, CPU and memory and storage. And that occupies close to 70% of the CapEx. It's okay, I, I, I think they are okayish

Frans Bakker: because-

Speaker 2: It's

Frans Bakker: a, it's a very good question. I mean, it's a very good question for, for IR, right? For IR. I mean, it, it, it would, typically this would be, explained in a footnote, like they did in the past where they said, including storage and, like the, the whole Dell package, right? It was, always like, "Oh, why is your GPU cost so high? It's because it They are focused on the GPU specific part, but I agree with you, if that excludes memory and, and, you know, CPUs, it would, be less impressive.

Speaker 2: Yeah, I mean, the reason I'm going there is because, okay, if you need around twenty-seven to thirty-five billion, the CapEx you have around seven point eight, right now maybe they will generate a little additional in the next couple of quarters. And they may not need all twenty-seven billion or thirty billion on this single quarter. It is, it will also be, you know, spread out. Hopefully it's fine, but, yeah, if they can make a better deal, which includes GPU, CPU and the storage and memory, that would be amazing.

Frans Bakker: Yeah, so I don't know if they have to make a better deal, maybe they just need to explain it a little bit better. I think that, there is a, you, you could perceive this, announcement in two ways, the bullish one and the bearish one. I would, say it's better to check with, IR and revert, because right now, it's, pure speculation. On, on my part, if I would, you know, give you an answer, it wouldn't be, you know, I can't give you an answer right now. So, now I have you here, y-you work, you work at Microsoft, right? I know you can't speak on, on behalf of Microsoft or anything like that, but, are you familiar with, with Microsoft's, Cabling, requirements, upon a GPU deployment handover. Let's say, let's say hypothetically you're Iron and you wanna hand over, let's say, seventeen thousand, GPUs. that, that comes with, with, like checks and, particularly cabling, checks. Is there like, like a fault tolerance or like a, a, a threshold where typically, like, like, like let's just say like hypothetically, right? You could speak like high level, don't, we don't have to mention anything or an actual deployment, but Do, do hyperscalers such as Microsoft, need a one hundred percent like, validation or is there like, we will accept a ninety percent, like, and then, and then the rest can be fixed over time? Do you have any, any thoughts on that? Any experience? Anything you can, tell us about that?

Speaker 2: Yeah, I, I, I generally don't work in that particular division. of data centers, but I, I work on the little top layer on, on building different solutions on AI and, and data. But overall, any service, right? Whether it's internal or Microsoft, I, I think if Microsoft is giving out some of the services, The SLAs, they generally look for ninety-nine point nine nine. Oh, Iron, Iron or any other, the new clouds are building their, their fault tolerance and the failover or BCDR. So They have to adhere to that standards, otherwise this hyperscaler gets penalized, generally for not available.

Frans Bakker: Yeah, no, sorry, this isn't what I mean. I don't mean uptime, I'm talking about, acceptance of a cluster, where clusters internally have, wiring, cabling, that can be validated. so we're talking about fabric validation. so ahead of a, like, let's say a hypothetical Horizon One Handover, there would be like software where, as Microsoft, yeah, I see the gold particle laughing, I know why, where, Microsoft could, like look into the software that, Ironwood, like, demonstrate, like, look, we have installed this many, GPU nodes and, we have, connected this much cables and, our validation tool says we have a ninety-six percent, you know, a completion or, or a validation. what, my, my, w-won-what I'm wondering is if Microsoft could get GPUs today But then, there is a little bit more work to do on the cabling, like maybe some cleaning, maybe some, like faulty neighbors, stuff like that. would, would Microsoft be like, you know what? We'll just take it, because we can use these GPU, GPU hours immediately, right? Or, you know, is it like, it needs to be perfect, otherwise, w- you know, they'd rather wait a week and, for a hundred percent, y-you know, it's, so it's not Or, a node or, or cluster hygiene or something, or, or cabling, stuff, yeah. So, I know you work on a layer above that, but ultimately it comes down to software tools that check these things as well, right? It's like validation. Yeah, yeah. Yeah.

Speaker 2: Yeah. Yeah. I, I think they would like rather delay by a week or two than taking a Semi finished product is why, I mean, that's how I, yeah, I'm correct, but I, it's like similar to IR, right? I know things But I don't know things at a low level where I can comment on it. You know what I'm saying?

Frans Bakker: Yeah, I get it. Alright, so this doesn't give us any useful information. But I'm trying

Speaker 2: to get into a role soon, hopefully, if, if I- then I can answer, maybe. I don't know yet whether I get the role or not. We'll see how it goes internally.

Frans Bakker: Alright, alright. Well, at least thanks for, trying to answer it. I know it's, it's gonna be very, difficult, to ask without seeing what I'm seeing. So I, I'll just, say that, so this, this space, has a main theme, it, it's actually not GPU depth. I was going to talk about Horizon One announcement imminent, and I still believe that that's the case. So, Me and a few others have done some, very extensive, research over the, over the weekend, over the past two weeks. I know, I know everyone has, ha-has heard about the nineteenth of July, where there was a celebration, from one of, Iron's subcontractors. They had a dinner in Childers in a restaurant where they had like t-shirts that said "Horizon One Completion." I mean, you can't make this up. They, they actually, went so far as to print T-shirts for a, Microsoft, data center. I mean, it's a, a Microsoft deployment in an iron data center, I get that, but, you know, I mean, it's kind of like-- and I'm posting it on Facebook, like, and tagging the, the subcontractor's name, it was like kind of asking for scrutiny, right? And so they, they, they got that, but now, we are two weeks further, a little bit more than two weeks even, and- Now, I am, you know, I've said, 19 is an internal completion date, so it's, there's, I haven't been wrong so, so far. and my final deadline for an announcement from Iron would be, Monday, the third of August. So I am going on the record here, because it is being recorded, I still believe Iron will announce a Horizon One handover, on Monday the 3rd of August. But if it is the case that stock analyst Pro is right and Microsoft is a extreme, Karen with, accepting, cabling mistakes that could just be, switched over in a, in a heartbeat, right? If that is the case, and there would be a, a longer, a transition period, I know I'm trying to like-- it's not like I'm trying to excuse myself, because I want the full credit for, for, the Monday the third, if it happens. But force majeure or, Microsoft being a complete, little bitch about, some cables, it could, it could be, a little, you know, a little bit later perhaps, but We are at the, we are at the absolute, final moments of, Horizon One being handed over. And, if you, if only you knew what I knew. And, there is a few people in the crowd here, including the God, God Particle, that know that I'm not talking bullshit. maybe God Particle can come up and, say that, he, what he thinks, because I know that even within our little, in our- Our little, due diligence group of the Horizon One handover, there are still different opinions, like, for example, Jim, Jim Liu is also in there and he thinks there is a fifty percent chance it will be one week delayed, as in the tenth of August someone else is more, leaning towards Wednesday the fifth, and then there's me, I am basically the most bullish and I'm saying it's gonna be today, well, for me it's today, Monday. So, The gold particle, if you, want to come up, you don't have to, like take any, any specific, date, but, I'm just wondering if you're leaning towards, my, and/or Jim's, maybe next week. So, so yeah, that's, that's basically the, the biggest alpha, That I, you know, would possibly have for you all, so yeah, the horizon, one handover is, imminent. So, yeah. we have Mark, here as well. Mark, you have any, thoughts?

Mark: well, I think it's gonna be very, very soon. I tend not to get too anxious, about it happening 'cause I'm a firm believer that it's gonna happen. I guess I'm just more patient than a lot of other people, I guess, but but I think it's obviously very exciting. I mean, when you think about it, how many data centers right now out there are gonna be GB three-hundreds? Not many, right? Not many at all. Just a couple of isolated ones, perhaps. But I haven't seen any tout the fact that they're gonna have GB300s at this, at this number. And then of course, when we look forward, to more GB300s and of course, the Vera Rubens, I think we all just need to take a deep breath. Because I do sense a lot of anxiety on these calls, I think, I think the space is so much fun, but I do think we get too anxious here on the space, and I think we just need to take a deep breath and relax. The Microsoft guys have definitely been, in Horizon. We know that it's coming. And again, a lot of the anxiety and a lot of the- You know, the tumultuous, environment that we're, that we're feeling we're in, all of this gets solved when it's no more just thirty million dollars of cloud revenue when they report. So when is, when is Iron gonna have their first one billion dollar quarter that they can actually report? Not, it's coming, not, we'll get there, but when they're gonna report it? And I think that's Either February or May of twenty twenty-seven, correct? I think we'll split the difference, I'm happy either way. But it's coming, guys, and I want this Group to, to feel confident and assured, but of course, the space is very entertaining and it's full of lots of information and, and I love the, I love digging into the nitty gritty and what you guys do an amazing job doing that. But I think there's gonna be a lot of smiles and a lot of, anxiety that will dissipate once the revenue reporting quarter over quarter and year over year starts. Moving in the right direction. I think this group has been waiting a long time, and that's understandable. I think, We just take a deep breath and, it's, it's gonna be amazing. And I think whether they report that the handover is this week Or they report that the handover is next Monday, hey, it'll be great if it's tomorrow, right? Or I think that's already today for you, Franz, right? but it's awesome, we have so much to look forward to. The-- these data centers aren't gonna be your father's and your grandfather's data centers, depending how old the people are in this group. I mean, these are gonna be unique data centers. They're, they're gonna be data centers that are gonna- Have the ability to be optimized to produce levels well beyond even the gigawatts and megawatts that are gonna be employed. So this is really gonna be very impactful to the entire AI world that we're gonna be living in, and there'll never be enough compute. This is gonna be fifty years And it's gonna be ongoing, and there's always gonna be the trepidation, and there's always gonna be those naysayers, and there's always gonna be, "It can't be done, it can't be financed, it can't be this, it can't be that." No, it's, it can be. I'm a can-do guy, I always have been my whole life, and this is my little sermon, but it, it's gonna happen, and it is happening. Jensen's the man. His product, by the way, is the greatest. Data centers that are using TPUs and Trainedium aren't gonna perform to the levels of this product. That's why I don't want Iron, I don't want the Roberts brothers to be concerned about hardware-agnostic systems. I want them to be focused on Nvidia systems, because Nvidia systems are what are gonna take them to the top of the pyramid. And I think we already have- Well, I, I wanna,

Frans Bakker: I wanna push back on that because, optionality is never, you know, it's never, a problem. No, I mean, it's possible. Look,

Mark: but, but no one, no one is showing us, no one is showing us a product that is better.

Frans Bakker: No, but you have to follow your customers, and if your customers are, diversified outside of NVIDIA for whatever reasons they have, then, keeping an open mind and an agnostic view, especially when you're geographically, trying to diversify into Europe, South and Australia, you don't know what kind of a GPU provider will be adopted by the, sovereign, A European, you know, plan that they are going to roll out. Maybe, they will, proactively avoid NVIDIA for whatever stupid Euro-European, regulation they have. You, you don't, you don't know, right? I mean, the- Well,

Mark: none of

Frans Bakker: us

Mark: know. For sure, for in, in that regard. But I, I have a feeling that Jensen's going to win, win them over.

Frans Bakker: Yeah. So the, the thing, with having a feeling is, ju-just as you started your, speaker with, you know, we have to take a deep breath and, not get too excited. Those are also feelings, you know? And I think, being excited, at the, eve of a, a handover of Horizon One is, I mean, I, I kinda like, announced that I would, include a "wake up early tomorrow" message into- This, space and, you know, wake up early tomorrow is if you know, you know. And, so, I am not going to agree with you here, Mark. I will, like to keep the spirits high and the excitement even higher because, it is our time to moon and, we, we can be very, patient and we have been and we will have to be going forward, but right now, you know, We are, we are going to celebrate that Iron has, completed this, problem child of a Horizon One. I mean, it's been a long journey, and if you're, you know, following, my content, specifically the, OnlyFans content, you know that it's been, one, hell of a project to, complete. And, well, I do agree with you that these aren't just normal data centers. In fact- Iron has been pioneering their way, to complete, you know, the, the Horizon One deployment. But the, the good thing is, Horizon Two and Three and Four are going that much faster that, the, the, you know, I can, I can tell you that I was, the first, one of the first, or as far as I know, the first to see that, Sweetwater One went vertical, including the Dry cooler, facility. But there is something that, most of you don't know, and it's Horizon Four's, dry cooler, installation is already vertical as well. There are fin fans up At Horizon 4. I mean, that's, unprecedented blazing speed. It's so incredibly fast, if you see that they have now reached a peak, a peak momentum in terms of, peak cadence, it's incredible. I mean, full parking lots on a Sunday They have twenty, they have night shifts, they are going so incredibly fast, there are almost more cranes than cars around the Horizon data centers. I mean, it's, you can get like, if you position your camera properly, you can get like eight cranes in a single shot. it's just, i-it's, it's amazing. And, so they are going to complete this project in twenty twenty-six. I would even go as far as saying they will probably be done before Christmas Christmas, and I think that the, these guys will be sent home early for Christmas. and they, you know, maybe Dell will need to finalize a few, supercluster, cables here and there, during the Christmas days, but, you know, the construction is going, going to be done. And, I have the, you know, almost day-to-day updates around that. And, so there is really a genuine- cause for excitement and, sometimes, especially after being manipulated to twenty-nine or twenty-eight dollars because, Leopold had to be liquidated, I think we deserve a little, relief space where we are, leaning back And eating some popcorn while, the iron folks are, fixing some bad neighbors in the, in the racks, and, flipping some, some cables or cleaning some cable ends and, pressing validate and Microsoft is, going to start paying very soon. I mean, two dollars and ninety-one cents for a GB three hundred, you, you're not getting that anywhere right now. Anywhere. And, so my opinion, and, that is, it is an opinion, I think Microsoft's gonna take it and they will just start to use it, and maybe that means that there is a little underutilization or maybe it's not optimized for the end use case of a supercluster. But if they can, run individual nodes or individual racks or even pods, you know, they, they can still deploy this and, onboard it as inference, right? I mean, I, I think that is likely because who would say no to a, seventeen thousand, GPU deployment? In this, scarcity of, GPU availability, right? I mean, so I know that it's going to be, a bit of a disappointment if-- for some reason I'm wrong, but let me just say that I believe I'm right, and, because I believe I'm right, and I'm on the record That's why there's excitement here, and, you know, sometimes it is, you can take a deep breath and be very excited as well. So, you know, it's not mutually exclusive. So, right.

Mark: can I just-- I do-- I also want to say is, I actually am in contact with a lot of the people that are always in these spaces. My point was really just, I'm very excited. My point was that a lot of these guys in the space- Talk about disappointment, they talk about being frustrated. My only, my only, my message was, try to not get frustrated, be very excited, hey, I, I would love it if this is the handover tomorrow, and I would celebrate. And Franz, if you didn't notice, I actually, I actually do subscribe to your space, so.

Frans Bakker: Oh yeah, yeah, so you, you know what's going on. Wow, that's great. so, yeah, before I'm go- I, I would, I was going to share a post about, the, the, from the guy who drove past, Mike, who drove past Horizon, sorry, who drove past Sweetwater. We, we'll give Bitcoin, AI guy, the mic for a moment. what's up?

BitcoinAIGuy: Hey, I'm at the gym, so I won't talk too much. nothing. I'm excited. hopefully, hopefully the announcement is tomorrow, right? I've historically missed every major announcement 'cause I've overslept. the alpha for today is like I, I slept in today. so maybe The universe is telling me to wake up early tomorrow 'cause I'm well rested today, and we might see an announcement. I'm excited, like, you know, for the short term stuff, but also long term, right? I mean You know, when, when I talk to IR, when I don't talk to them frequently, you know, the, the tune has changed, right? Like, you know, e-even them, like, they're, they're saying they wanna be the next hyperscaler, right? that, that took a, it's a big thing, right? Like, it took a long time for me to digest that, right? Like, A-AWS was doing forty billion dollars a quarter in, In cloud revenue, right? So, you know, we're excited about like, yeah, the validation with Verizon, but if, you know, the people who are buying the stock at seventy, they're gonna be buying at a hundred and twenty, right? They're investing for like three, five, ten plus years, right? And I think, you know, some of us are still gonna be here when Iron's gonna do forty billion dollars in AI or cloud revenue per quarter, right? Like, that's what I'm excited about. You know, we'll see what happens first, you know, does Bitcoin hit 500K, 300K, where, where is Bitcoin gonna be? I, I know it's not a Bitcoin space, but I, you know, I'm thinking long term too. It's like, man, we are, we're pretty early in this. And, you know, I don't get too granular. I, I try to stick like pretty high level, you know, supply demand, number go up, You know, volatility, vitality. but it's helpful, right? Like, the, these stocks go down, fifty percent, two to three times a year, right? And, and I think this year has been insane. you know, but and as an early iron investor, I'll tell you, like, this This year feels like twenty twenty-four, up to early '25 to the deep seek moment, right? I think we're in that window, maybe Leopold had to be the sacrificial lamb, you know, and i-i-in that magnitude of event is similar to like a deep-seek moment relative to Iron Stock, But, but what happened after that? I mean, the stock went parabolic, the rest of twenty twenty-five. I, I think we're, we're inching there, right? And I think it, it maybe the Horizon One, announcement Could, could, you know, re-rea- When we had that thirty percent pump the other day, like that was just a taste, right? Like, I, I think we're gonna see more of those, not, you know, not just one, based on this one horizon thing, but I think it, it, the momentum's gonna start building up again. And I'm not a TA expert, right? So, maybe some, some, some other Dude can, can validate what I'm saying, or, or what I'm trying to say. but, but I think the good times are coming, right? I think, you know, we're all saying it, like, you know, the billion dollar Revenue per, per quarter, that, that's around the corner. and, I, I don't know, it's just, it's, it's, people, confuse The stock for the company, and, you know, the, the traders are really loud here, right? They'll tell you the, the price is what pays you, which is true, but You know, they're not following the nitty gritty, they're not, they're not here for the generational opportunity, they're here to sell, covered calls, right? or just to make, you know, twenty percent if they're lucky. you know, they're not here for the two thousand percent or the, twenty thousand percent. they're, they're missing out on, on those gains, and I think that's why we're here. Or at least that's why I'm here, right? This is gonna be my first hundred X stock. that's why I'm excited. and, yeah, I mean, keep up the spaces, I, I enjoy them. sorry, I can't talk too much, but, I'm on a treadmill right now and I wanna be mindful to other people around me. Thanks.

Frans Bakker: Thank you, Darkpool, Max Payne guy. so, always, appreciate your, you rallying the, the troops with your, generic statements. No, it's, I, I'm more granular than you, but, I, I really like the, the high level approach as well. but, talking about high level, so I just shared my, post with the, satellite image, in the nest. it's, it's really high level, but it's also very granular at the same time, isn't that paradoxal? it's actually, thirty-five centimeter, in terms of the pixel density, so that's why it's, expensive satellite imagery, but it's taken from a very high level. but what I wanted to-- why I shared this one and the, the other post from Mike, when he drove past Sweetwater, is, typically, the engagement that, comes up in, on X is a sort of a, barometer for, the interest in In the stock as well. And it's, this is a very tricky one, but, typically, if, if I share something, it's not, it doesn't really get over a hundred thousand views, and I see that both my Sweetwater, my children's picture and the Sweetwater drive-by, have well in excess of a hundred thousand views. So what does it tell you? It tells you that people are very interested in the assets. people wanna see what they own, and they wanna see what is the progress of, of The construction, what is the pace of construction? I mean, that's why, there is a, a repeated interest in, my content as well and, what we do in these spaces. it's, it is what it's ultimately about, and, vertically integrated, just, sounds a lot better when you can see the data center going vertical or the dry coolers going vertical, and in this case as I just said, Horizon Four has a vertical dry coolers, and Sweetwater One has data centers, and dry cooler, you know, like a frame for the dry coolers already vertical. So, Iron is going vertical, in, an end of twenty twenty-six deployment in Childers, and they are already going vertical for a twenty twenty-seven deployment In Sweetwater. So this is, showing that Iron is now working, and printing data centers at two locations in Texas. And no matter how you, you know, how you flip this, whatever you wanna say about the stock or about the, the CapEx, needed, the physical data center build out is accelerating, and it's a hard fact because if you go back, six months in time, there was only One location where they were building, right? That was Childers, and now we are building in two locations. So that means there are cranes in Sweetwater that aren't needed in, in Childers. I mean, they have an excess of cranes now because they have, you know, on Aggregrad, they have more, cranes in use, and now there are two or, or three visible, in this, picture, and it will soon be five, and then, you know, this means that the, construction is, accelerating, and from what I've seen on, satellite imagery, they are starting to build, a lot of these tent structures at Sweetwater as well. That means they're gonna be housing more people, they need more lunch tents, you know? They need, places for the people to eat and to, you know, shelter for, for the rain or, like have their morning coffee or something. and, this, this just shows- So they're ramping up the teams and, ultimately it's gonna, mean that they are going to bring online costers faster and, so You know, the, watching the fundamentals and the, the asset base, we, we all share an interest in this, and I think it's for a good reason because it will, lead to revenues, which will lead to better earnings. So kind of a no-brainer, of course, but, I just thought it would be, you know, interesting to, point out this, interest in terms of the volume, of, of, of impressions for these kind of posts. other than that, I think, we, we have said all we need to say for now. We've talked about Leopold's, unwinding, the, transition of his portfolio to Citadel and the last Friday's, dump. I think, It is now up to Iron to decide if they are going to announce, a Horizon One handover, on the back of, you know, my, prediction or, or if they're going to, you know, frustrate me, now, of course, I'm, I'm not a factor in this, but, I, I think we're, we're gonna get it. so, you know, looking forward to August, w-where we are now, we can do that as a final, topic before I close the space. my personal, expectations are after a, Horizon One handover announcement and acceptance by Microsoft announcement They're going to, hopefully announce the GB300 Exemplar status and, the closing of the Morantaz acquisition. these are all things that I expect pretty imminent to happen, as in it could all happen on Monday or all in the coming one, or maybe coming two weeks. And then, in, roughly ten days, d-around the thirteenth, I expect them to announce the earnings date, as in, we are going to announce our earnings on, August the twenty-seventh is, for example, my number one, guess for, for this year, for August so that's, as far as I can say, what I'm looking forward to. currently I'm not really like expecting something as in a commercial, announcement, unless it's Horizon five and six or at least Horizon five, but From my current, due diligence, I am not so sure if they have a design for Horizon Five and Six yet. So I do know that they have, you know, they're, they're taking down these buildings, bu- these mining buildings in favor of building a Horizon Five, but I'm not so sure if it is, really like in a commercial end stage yet. So that, that would make you wonder if they have an LOI for, for it instead of a, you know, final, commercial, negotiation stage. I mean, w-would they really demolish, six mining buildings that they finished in June last year, without a customer in mind? I mean, that's the big question, right? but, Yeah, so can't really say too much about that. so Butcher, any, things that you are looking forward to that I haven't mentioned?

₿itcoin ₿utcher: I think this earnings versus the prior two or three, I think it's, you know, Mark can be excited, you can be excited, but I think overall the sentiment is A lot more sober, which if there is anything that surprises the upside, I think we're better positioned for that. I think we all know it's ahead, but I also think we've been Talking at length about how the disposition of mining equipment and some of the other, everyone's familiar with the count plan, all these other things that have been kind of noise for this last, end of the fiscal year quarter, you know, a Kitching sink quarter, if you will. I think expectations, with the exception of the Horizon One delivery right now, are kind of on the lower end, so I'm just kinda- Trying to keep, stay even keel. I'm obviously, optimistic for the future, but I also, having gone through the past six to eight weeks with things that were out of our control, I think it's just, It's not the most encouraging message for me, but it, it's more kind of like, you're just, it's your time in the market and nothing's gonna be cured tomorrow, and this is a long-term game, and the more that you can stack at these levels You'll be thanking yourself in the hopefully near term future, but even if it's next year, six to twelve months from now, I, I'm confident that we'll be in a better spot

Frans Bakker: Thanks, Butcher. Yeah. I think so too. and, yeah, I agree with, what you said. w-with regards to earnings, yes, it will probably be a point that the bears are gonna make, as in, I, I fully expect Jim Chanos to come up and say, "Iron had five times more, depreciation on their, mining- Building, sending equipment, then they had AI revenue, something like that, you know? It's probably gonna make like a, causal relationship between two things that have nothing to do with each other. the consequence, you know, this is the cost of doing AI business, you know? You gotta get rid of your, your Bitcoin, data centers that are unwanted, and, you gotta retrofit the ones that are wanted. I mean, ultimately, if you have a diversified customer base, you gotta listen to your customers and you gotta pick the ones That are paying for your expenses. I mean, if there would be no valid reason to, not retrofit these data centers but instead disassemble them and tear them up, then Iron would have definitely done that. I mean, it's not as if- You know, if you remember, there was an interview or a, a space with Dan Roberts, before where he literally said If we want to, you know, instead of repurposing the AI buildings to Bitcoin, that was a question I think, if AI would fail, would they be able to go back and put Bitcoin, miners in there? Instead, we'll just build new, new, Bitcoin, buildings. So, if you turn that around, they wouldn't just, start to demolish good Bitcoin mining buildings if they can be repurposed To AI, air cooled deployments, right? So there is a reason, and we don't know it today, but it's a, a very valid one, and that's all I can give you for now. So the, the five hundred and twenty million hit that they, that we're gonna see on the, impairment Post, in, in, the earnings in August, yeah, it's gonna suck, it's gonna absolutely destroy the EPS, but whatever, it is a non-cash cost, it has to be done, it is, It's, it is what it is, and it's already been modeled in by the, institutional side and the sell side of, of Wall Street, so, Except for Jim Chanos, of course, he's gonna, he's gonna run with this, and so will all the bears. so yeah, I see Ghosty wanna come up and speak, but I have no, I have no time for that, sorry mate. I mean, I know you're, you're bearish, Iron, you don't believe in, Catalyst, and its technical chart says it's gonna go down. So, that just sums up what you wanna say anyway, so I'll just do it for you. so with regards to earnings, we're gonna be doing a space, for, for the earnings, you know, like looking ahead at earnings, maybe two weeks before or one week before. So for now, I think it's enough, we can leave it here. And, unless anyone has any final words, I will close the space here and, yeah, and right now it looks like you should consider waking up early, but that's not financial advice, that's just, you know, it's also not medical advice 'cause I'm not a doctor, but, if you, if you prefer to be, around when possible new news may drop, and it's more important, to get a little bit of adrenaline than get some good, nice sleep, then maybe you could consider waking up early. so I'll leave it at that. Okay, I see no one, else coming up, so, thanks for joining, Butcher and Mark and, everyone else that came up and speak. And, we'll see you, in the next one. Cheers!