Sunday Night $iren 1.18.26

Hosted by @₿itcoin ₿utcher 🥩 🐑 🐷 · 2026-01-18 · Tags: IREN

TLDR

The discussion was strongly bullish on IRIN’s position in the AI infrastructure market, emphasizing its near-term grid-connected power, construction progress, and potential to secure additional hyperscaler agreements. Speakers considered xAI a plausible but speculative customer, generally viewed Google as a stronger fit, and identified financing, regulation, labor, and execution as the principal risks.

Speakers

Notable quotes

Transcript

₿itcoin ₿utcher: Hey everyone, we're going to get started in another minute or two. We're just going to give everyone an extra few seconds to join. Believe we have Franz this evening co-hosting. Wasn't sure if he's going to be able to make it or not. Franz, good evening. Thank you, Chair. I'm here. So we'll give it one more minute. I'm going to post a few articles on the nest that are related to the conversation. Okay, let's get this thing going. Franz, I couldn't hear your mic earlier. I don't know if it's because of the move this week, but how are you this evening? Yeah, I'm fine. Can you confirm that you can hear me or not?

Frans Bakker: Otherwise, I'll have to switch something around.

₿itcoin ₿utcher: I can hear you, but it's some people might be having trouble with your volume, so possibly might have to. switch, but we're getting going here. I'm on the Eastern Time Zone in Detroit watching the Bears-LA game, but we're here for iron. And I threw a few posts in the Nest. Primarily, the focus tonight will relate to the XAI rumors that are not necessarily linking iron to XAI yet, but There's the potential given-- I have in the Nest a post from-- I forgot the gentleman's name-- Edge of Power, speaking to the EPA, coming out with a ruling that looks to circumvance-- excuse me-- looking to-- Elon was using gas turbines to power XAIs. AI capabilities and was circumventing EPA rules. But now if he were to have to permit, get permits, I should say, to operate those, what does that look like for XAI's growth? So there was a few, I saw Marbles was in the crowd and he's free to come up to speak to what he pointed out. But where I came across and kind of lead things off, the EPA is an executive falls under the executive branch of the US government. So you have the judicial, you have the legislative, and then you have the executive. So from doing some quick research, I found that President Trump, through his power within the executive branch, is able to appoint the head of the EPA with Senate confirmation and also remove them without cause. which is why I believe, even though that CNBC article initially threw out the question, what's Elon going to do, I think the way Elon's neighbored up to the administration and was pretty instrumental in President Trump's victory, I find it unlikely that in the immediate term, it's going to change how he does business, because I believe President Trump would just replace the head of the EPA until he got the result he wanted. But I do think, and maybe we can use this as a talking point tonight, where it comes into play. And Franz also wrote an article about a subcontractor related to XI that he can detail more in a second. But my takeaway from that event is we're going to get into a world where We still have elections every four years in the US. And if the pendulum swings back to a more EPA-friendly administration, whether it's a more moderate Republican or most likely it would be a Democratic presidential nominee that comes out victorious in 2028. If that were to happen, I find it more likely that those kind of EPA rulings would have more teeth. So I think as a business person, you'd rather be regulatory agnostic and just have a business that's immune to regulatory, how it sways from party to party over the course of time. And that's why I believe these developments, along with how communities are reacting throughout the US, pushing back in metropolitan areas, including Memphis, where Elon operates, even though it's not grid-connected power, you're seeing these power-intensive, whether it's a methane turbine producing smog allegedly in Memphis, or if you're outside of Detroit in Saline, Michigan, which is neighboring city of Ann Arbor with OpenAI trying to open a new data center and locals complaining about how their power prices will be affected. You're going to see that more in metropolitan areas, whereas companies like Iron and Cipher might be in more rural areas, but that actually works to their advantage because they're more welcome in rural areas because those municipalities are just happy to have the tax revenue. So those were just some of my initial thoughts on that. And of course, there's the macro news. How can you not get through a day without this administration throwing a new curveball in some markets? And this evening, you'll see that Bitcoin and Ethereum were down, and futures were slightly down as it relates to the newest development with Greenland, which this administration is positioned as an area of national security significance, given the location in the Arctic. There's also, if we're just being honest with ourselves, there's an abundance of natural resources in Greenland. So there's also that incentive for, we're certainly just like Venezuela had a glut of oil, you can't help but think those natural resources would not be the desire of not only the US, but any developed country if they could get their hands on it in a world where resources are becoming more constrained and AI specifically is demanding more power, but also rare earth minerals to create the infrastructure and the GPUs to power it. So those were some, potential things we could walk down the road with. But Franz, I wanted to get your take on your posts. I had a chance to read it a little bit, but you were making the connection between a subcontractor that Iron uses and that was also teaming up with XAI. What do you see out of that connection?

Frans Bakker: Well, I'll be honest with you. I think it's a little bit reaching. I don't think there is a literal connection here, but, you know, it is certainly speaking of some kind of shared quality control inside AXA or sorry, Tesla. I mean, I can only speak for Iron. I know they are extremely ability to subcontract that they work with. And to see that this particular subcontractor is now looking for people to work with Dashlive, it's just, you know, it speaks of quality and, you know, of a certain ability to work with both Dashlive and hiring. I mean, you know, obviously there is not a direct link. that is speculative, but someone has to be talking to Iron for Horizon 5 to 10, Sweetwater, and we know that Iron's been saying there are multiple input parties and multiple hyperscalers. And we know that Elon is having some difficulties with obtaining new power or maintaining its current power portfolio. So, you know, I just thought this was a good opportunity to put a spotlight on this. It's not much for engagement, obviously. You know, at the end of the day, this is also helping Ireland to, you know, get their subcontractors some spotlight on the recruitment. So, you know, it's like a little bit of everything. And I think it's interesting to see that such a relatively small subcontractor is working with both Ireland and the recruitment work.

Speaker 3: Well, there is no direct link. I just think.

₿itcoin ₿utcher: So Franz, I'm getting some feedback from the folks in the crowd that you're like that arrangement might not work best tonight. So if there's an alternative, we would obviously love to hear what you have to say, but. We're having trouble right now. I'm going to kick it to Nick in the interim. Nick, you usually have a very well-rounded macro view, and I'm sure everyone would love to hear where your head's at with the market right now while Franz is figuring out his mic.

Speaker 3: Yeah, I can touch on a couple of things. So I mean, kind of the same par for the course as we previously discussed in terms of the macro environment, right? Like liquidity is loosening. I remember, you know, several weeks ago or a few weeks ago when price action was **** everybody was kind of super sour on the sentiment. And I think that it's starting to alleviate, obviously, as as the price goes back up. So clearly, you know, the emotions are going in the right direction. And as we all know, everyone that listens to these spaces and does the diligence on the regular the fundamentals have never been better. But what was particularly interesting to me regarding AI and the hyperscalers over the last week was this whole thing with this emergency auction with PJM that was arguably the White House wants to mandate that the Northeast grid operator does an emergency auction. And, you know, I When I first read that, I'm like, okay, let me do a bit deeper diligence on what the implications of that are and what the mechanisms are that triggers as a grid operator. As a grid operator, it's exactly as it sounds, they're an operator, but they don't actually build the power infrastructure. But what they do do is ensure that the grid is going to operate smoothly, that the resources match the future state demand. And so effectively, the grid operator can hold an auction, and in this case, I believe it was suggested from the administration that it was a 15-year long-term contract option aimed directly at hyperscalers, which means the hyperscalers would go into that auction, they would bid on the power, and then based upon that demand assessment, the operator would then go out and put that to market, make the market, so to speak, if that's what you want to call it. And then the developers would then come in to build the power. And so like I'm just thinking like first principle, second principle, third principle, like what does this do? Well, you got to imagine that the current administration in the White House is hearing nonstop about how there's not enough grid capacity to service the AI market. They know that they they also know that there there is no there's no losing this race, right? Like it's not an option for them to for for the current administration or for the United States to lose this race to other global superpowers such as China. And so they know that they have to do everything that they can. But in the political arena in America, I think that they're probably fighting substantial headwinds, right, which is why you see all of this all of this press and things coming out about, it uses a whole city's power. They're going to boil the oceans and the power prices are going to go up for individuals. And so you have this ecosystem of forces, right, where some people are pushing back, everybody's fighting their own political agenda. The current administration wants to be pro-business. It's a national security threat. And so they're trying to basically force the hand of the grid operator to hold this auction to basically market make and project forward future state demand to then have the hyperscalers go directly in and provide, you know, the quasi cash flow to fund the development of that infrastructure. So what does that do to our market as a whole? Well, we obviously know that even if they did all this tomorrow, I mean, if it was like a Goldilocks scenario where everything went right, it would still take, what, two years at least to bring something online, depending on the proximity, maybe a little quicker, but probably not. So I think this is all to say that I interpret this as bullish. I interpret it as kind of like the snowballs rolling downhill, right? Like it's coming, the news is now becoming a household type of story at the dinner table, right, which is indicative of building a really strong and perhaps long-term sustainable bull market, which we know that we're early to in the AI data center infrastructure development. But I just, I see all of this as bullish. I see all of it as creating momentum and I'm really curious how it's going to play out and what the administration's going to do. And in regards to the EPA, Butcher, I think you're right. I think anybody that stands in front of this train is going to just get run over and replaced and they're going to do everything that they can in their power to do so. Because again, I mean, there's no option to lose this race. And there's trillions of dollars already allocated to winning the race and it's only going to get more. So that's kind of where my head's at. I'm curious to get your guys' thoughts.

₿itcoin ₿utcher: Yeah, I tend to agree, Nick. I was trying to invite Small Cap because he had a good tweet that I posted in the Nest regarding PGM specifically. But that tweet, it's references PGM's framework and I was hoping he could speak to it a little more. But similar to Senate Bill 6, I believe it's called, and Texas, the idea that the grid provider can require curtailment from those parties, I think we're going to find that it becomes almost universal. How much they actually use it, I'm not sure. But I think industry-wide, you're going to see CapEx go up for redundancy. I mean, we're already seeing that. with Iron, part of it was the tier three. Once Franz is back and up and running, he can speak to it a little better. But I think it's going to be, whether it's solar or battery storage or diesel generation, that it's going to become the data center provider and or the hyperscalers doing it on behalf of their co-location providers or whether they do a triple net lease. But a lot of this, it's going to work to the benefit of the administration as best they can. How they do it legally is another question. But for President Trump to already pull in the hypers and say, you guys are paying for this, not the people, it's pretty obvious that is rhetoric that's related to midterms. But I am, I think it's still just early days where it's really hard to tell. But I know one thing, I feel pretty damn good about having just under 2 gigawatts ready to roll this year with Iron, specifically the 450 at Horizon and the 1.4 at Sweetwater 1 and Sweetwater 2 should be coming pretty soon as well. I had a comment in small cap, one of his posts from another follower of his regarding Sweetwater's approval, which Franz can speak to you later. But everything points to things being ready by April. We had Xcap, if he's up in the room too, he had found an ERCOT presentation where they were specifically referencing the approval of SweetWater, and I believe they even mentioned it by name. So I have no reason to believe that SweetWater is in jeopardy of any approval. But if you can contrast that with other providers that have these 2028 and 2029 projected approvals with so many grid requests and that outnumber the existing power that's operating right now. It's going to be interesting how ERCOT scales up to meet that demand. While I'd like to think Texas is pro-business and can handle it better than any other municipality, these are still people that are, for the most part, going to be working 40 hours a week, 50 at best. I don't think there's a sense of urgency that you would see in a private sector job. So those are just some of my initial thoughts, Nick. But if you had anything to add, feel free.

Speaker 3: I mean, I think we're kind of rowing in the same direction here. Small cap's in here. Maybe bring him up and see his thoughts.

₿itcoin ₿utcher: Yeah, I tried inviting him, so I'm not sure if he might just not be able to come up on stage right now, which is fine. Franz, is your Are you on your phone yet, or do I need to? Moved him and he can send me a DM when he's ready to go. But we're still, I mean, I love hosting these every week, but it is, we did see Dan retweet the Horizon progress report that I believe Mike Power, whoever, someone in IR from the company account showing that the satellite imagery that Franz had was correct and that they're ahead of schedule. I think that plays into Iron's ability to, while they did pay an acceleration fee, It shows, though, that they are making progress. And I think the quicker they can hand the keys over to Horizon One, that's going to create momentum in itself, and that we've said it de-risks the business. But as there becomes more regulatory and more supply chain-related friction, those who plan accordingly, which we believe and we're positioned with iron because we have confidence that Dan and Will and Kent and the rest of the team are ready for those challenges. Once they deliver upon those, they will separate themselves from everyone else. But it looks like Franz is back up. Hopefully you heard some of the questions I had for you, Franz. Otherwise, I'm happy to repeat them.

Frans Bakker: Hey guys, how's this? I hope this is better. Yeah, so it was a bit of a struggle to get back on. You had to manually remove me as a co-host because I couldn't join from my phone. So I didn't hear any questions. I did hear Nick talk and I could repeat what I said about the job listing. So basically SWEX or Southwest has been working for Iron for a long time. I believe they've been a subcontractor all the way through a block two, all the way up to where they are now. And I'm quite certain they are also working on Horizon. So I just thought it was really interesting to see that a couple of weeks ago, the same account on Facebook posted a job listing for Iron at Childress. They were recruiting a lot of electricians and people to really start building a big data center project. So that's obviously Horizon two, three and four and beyond, I guess. But it was particularly for Childers. And then just recently, like a couple of days ago, they had this job ad for Tesla. And I just thought that's very interesting that a relatively small subcontractor that's been working with Iron for a while is now all of a sudden, you know, recruiting to work at Tesla. I know there is no direct link here. I mean, it's it's, you know, speculative, of course, but I think it does say something about quality management of these two companies that they are both you know, trusting their core data center engineering to this company. So, you know, we've been analyzing private jet flights between Memphis and Childress and stuff like that. And obviously, there is also it's still, you know, the connection is possible, but this can't be confirmed. And the same the same thing applies here, I guess. But I thought, It would just, it was a good moment to shed some light on this. There is no confirmed connection, but you know, at the very least, it gives a little bit of spotlight on these guys' recruitment page with over 30,000 views, or no, I think it's over 50,000 impressions now. Maybe someone will get a job there, so at least we helped Iron's subcontractor out at the very least, right?

₿itcoin ₿utcher: Now that we have a better functioning mic, Franz, I would like to circle back to-- Nick and I were speaking to Elon and the EPA. And while we believe that Elon's going to do what Elon does, and his financial influence and his social influence helped this administration get into power, so we have no reason to believe in the next year or two that that's going to change how he operates with turbines, because the head of the EPA can be removed. But at the same time, there is an election in 2028, and it's possible that someone that's politically opposed to Elon gets into power. So I was suggesting earlier that it's likely, even though Elon's obsessed with vertical integration, I think he's even more obsessed with progress of his AI. And in the event that the turbines require permitting in the future, and if the EPA wanted to, under a new administration, wanted to mess with him and pull those turbines offline because of permitting, I see the opportunity for co-location and xAI's future, whether that is Sweetwater, Horizon, We could get your thoughts on that. But I also, small cap sniper can't join us for another half hour, but he had a post that I haven't had the chance to look at it yet. That's why I was hoping to get him to speak to it. But I believe it's likely related to Senate Bill 6 in Texas, which requires additional power sources behind the meter in the event that the front of the meter, that grid connection, if you're asked to curtail, then as a data center provider, you have to have a backup source of power. And that's where I believe-- I was hoping, Franz, you could speak to what you think Iron's doing to address that at Horizon specifically, or what they're going to do in Sweetwater, given those are both ERCOT-related properties given I believe you had made some connections with potential battery storage and solar providers. So I hope those give you a little context of what we were hoping to get your insight on.

Frans Bakker: Yeah, so basically what I can tell you high level is that IRIN is ready for curtailment defense at Horizon. They've got a very, you know, they are going for tier three or equivalent. And what I know is that they are confident they can deliver more than four nines in uptime, even though they are specifically, I think, contracted for four nines. I'm not entirely sure, but so they are confident they can potentially deliver five. And, you know, that's looking at it from another angle. We know which brand of generators they're using, and we also can make a pretty good estimate of how much downtime from the grid they can tolerate. I think that the site on itself is, you know, is set up in a way that they are able to endure an outage for over a day. You know, I am and there is no talk of battery storage here yet. This is just the way that the site has been configured with generators and UPS's and you know, these kind of objects. But we do know that for Sweetwater, they have included battery storage and solar into their tax abatement agreement. So they have mentioned specific battery storage and solar vendors and suppliers. They are singled out as a separate entity for the tax abatement agreement because, you know, under these agreements, they need to source a part or as much as possible from the county. But obviously, Fisher County doesn't have a battery storage producer as far as I'm aware. So that's why they have made this line specifically for battery storage and solar. I've pointed out that the site is so incredibly large that there's ample room for solar and battery, but it doesn't mean that it's actually going to happen. I just think that they included it, you know, in case it would be needed. But there is no such thing for as far as I'm aware, for Childress. And I also don't think that, you know, they're going to drop best there or something like that. So I just don't think it's in the plans right now. I think that the site is ready for like long-term outages. I don't think that they're going to monetize you know, like power prices with battery storage in Childress. If any anywhere I would expect this to happen in Sweetwater, which is a lot bigger. And, you know, it's it's also situated in a in a area where there's really nothing around. So it would make sense for them to, you know, use this land to build out backup generation in, you know, with, for example, solar and battery storage. We also know of a future development site where they're working together with a development team that's not inside of Iron, but that's trying to establish a reinvestment zone somewhere. And those guys are originally solar and battery storage. producers, well, not producers, developers, I guess. So I think this topic is going to be more, you know, dominant in Iron's future sites, potentially including Sweetwater. But I think for Childress, it's just the site's not big enough and I don't think it's needed. I think they got everything they need with the generators and the way that the site's been set up. But that's just my assessment based on what we've seen, what we know from talking to people at Iron or around Iron and, you know, the uptime. I think you can't, you know, confidently sign with Microsoft unless you have all your backup in order. And I think as far as what we know, they got it figured out. So yeah, that's all I can say without going into much details or confidential stuff potentially. So I'll just pause here.

₿itcoin ₿utcher: Yeah, thanks for clarifying that and my mistake everyone with confusing the batteries related to the tax credit with Sweetwater, with Childress, and I appreciate that Franz. compromising your position, friends, and you can just respectfully decline. But I think I know the answer to this, but this is more for newer listeners. Given your intimate knowledge of the company and the relationship with the company, can we at least commit to saying that we have no reason to believe that given the success of Horizon 1 through 4 right now, and the development of, most likely, Horizon 5 through 10, as well as Sweetwater 1, is it fair to say that all of the long lead items, most notably generators now that are becoming harder to come by, that the company, through its prescient planning, already has those under their possession or can at least give themself enough time where they're in the right space in line where it won't compromise their build out.

Frans Bakker: Yeah, so I think with regards to long lead time items, if you're specifically talking about UPSs and generators, I think that this has all been locked in prior to signing the Microsoft deal. And I think the whole acceleration fee is partially going towards, you know, the sourcing of these items. I think that all the other items, you know, are vendor specific or already present on site like transformers and, you know, shock breakers and stuff like that. I think they got the whole high foldage step down to data center voltage already figured out. This was, you know, this is their core competence as far as I'm concerned. So, you know, on a data center level, I think that it was contingent on some of these components. And yeah, I am very confident that the reason why is they are accelerating hiring. We can see based on satellite imagery that, you know, on a Saturday, the site is like completely packed with cars. They've they've created like, I don't know, five new parking lots in the last two, two months. And it's it's, you know, it's full like it's it's becoming a very busy site, very crowded. And I think that the reason that they are increasing the you know, hiring all these people and having so much stuff on site is because they have everything they need to build. So they got all the components ready. And from my talks with the people that have been involved in building Horizon One shells, what I heard is that basically after the first shell was sort of completed, it became like a copy and paste thing. So it's basically running on autopilot now. And Those are not the kind of words you would, you know, hear if there was any constraint in terms of items, long lead time items or other. So, yeah, no, this project's going to go faster. And I think that that's why I'm very confident in, you know, making posts like that where I'm saying, you know, they are accelerating, they might even deliver earlier. There are no material constraints And as far as I'm aware, and I can confidently say that I think that they're going to deliver on time at the very least, as they always have for Microsoft. And I think this month there is some handover between construction and ops. And I think that this is going to, you know, do very well for the rest of the delivery of Horizon One. There's just one big question mark is, the chips, the delivery of the chips was set to be for March for the first batch or the first tranche, I think that Horizon One, Shell One, I don't know how to call it, is probably ready to operate right now or next week or the week after. So, you know, it would be, if I was Microsoft, I would just bring my own chips for a while, like not to fill the building, but like to test everything. Maybe they would like send like, I don't know, a batch of 24 or like a one NVL 72 GB 300 rack, you know, like one, just one to test out everything. And that can also de-risk things, you know, and they would, Iron would probably tell this to institutions and that could be very, you know, reflecting in the stock price. So maybe they will tell us at earnings, You know, like, oh, we've completed the first building and we've run some tests and everything is according to plan. You know, that will massively de-risk the entire thing. That's the great thing about Horizon 1 to 4. As soon as you deliver one, you know, it's like extremely easy to do the other three as well. So I think it's going to accelerate, it's going to de-risk, it's going to, you know, it's all going to go according to plan and then maybe even earlier completion. So, yeah, that's that's what I can say. You know, I don't know exact delivery times of long lead time items, but it doesn't look like they're slowing down, obviously.

Speaker 3: Hey, Franz, I'm curious. I don't know if you had a chance to read Agrippa's latest latest substack, but he has a very strong probability or his highest conviction is that Childress and Sweetwater one may be prime targets for Google in regarding the chips integrating perhaps some of their GPUs because they have existing facilities in Haskell and Armstrong, which are both about like 85 ish miles away. from those sites. So, you know, his theory is that Google may want to start building a corridor in there. I'm just curious to hear your thoughts if if you got a chance to read his research this last week.

Frans Bakker: Well, I'm I'm subscribed to Agrippa Substack, but I haven't read it yet because I'm I'm currently in lack of time for a lot of things. I'm behind on on YouTube and all kinds of things I have to read back, including Agrippa's work. I told him this, but I will make sure to read this one because it does ring some bells for me. I've been talking about a Haskell site for a long time now. I think I was one of the first people to find out that Google was actually building there. I got tipped off by someone who knows the previous owners of this site. I think it's, it was, maybe it was even in 2024. But anyway, it was a Musselman tech that used to have this site. I have a long post about it with, it's like a thread about the whole Google Haskell site. But yeah, so the proximity to Sweetwater 2 in particular is very interesting because Google's Haskell site is like the first substation next to Iron's Sweetwater 2. So like they're directly connected basically. So yeah, that made me also wonder if this could be something, you know, since Iron's also making a direct fiber loop between Sweetwater 1 and 2, why not just extend it towards Google's, you know, I mean, Thinking from Google, that would be something relatively easy to do. You would add two gigawatt of connected power and infrastructure, IT infrastructure to your own one gigawatt site, which could potentially mean you have a three gigawatt cluster. So I don't know exactly how Agrippa phrased it as a corridor, but the proximity is striking. And if there is some way that this could work, then I'm sure that they will look into it. I, I just know that Iron would love to work with Google. I would actually say that it's probably a priority for them to add to their customer portfolio. So, you know, it's all circumstantial again, but yeah, I mean, I'll read the post and I'll see if I have to, you know, comment on it or on X maybe, but I mean, I would also love to have Google as a customer, but I'm just not really interested in colocation of TPUs right now. I don't think the economics would be worth it for our flagship site, you know, so let's see.

Speaker 3: All makes sense. Okay, cool.

₿itcoin ₿utcher: So Franz, let me, we'll pick on you a little while everyone else is getting some courage to come up, but I'm particularly-- I haven't had the chance to ask you, but I'm curious with-- we have two sites right now that are primed to be signed, the 450 megawatts at Horizon 5 through 10, and then you have the 1.4 gigawatt at Sweetwater 1. Can you kind of give everyone your understanding of how the construction team at Iron, how it, are there separate teams at Sweetwater and Horizon, or is it kind of one uniform team where everyone kind of gets distributed accordingly? I ask this because I'm wondering, in the event you sign a new deal for Horizon 5, I can't help but think you're still committed to completing Horizon 1 through 4 for Microsoft, and there's got to be some premium from space on one site at one time where it feels like that you would develop the site sequentially where you would do 1, 2, 3, 4, 5, 6 through 10, which I think you alluded to with 1 through 4. And you still have Bitcoin mining on the site, and while Bitcoin isn't $1,000,000 yet, it still generates free cash flow for the company. So I guess my question to you is, how do you see if, how do you think it plays out given these construction constraints potentially at Horizon? Do you think it's more likely that even though they might sign a deal at Horizon that they might dedicate more resources to getting Sweetwater built quicker while they're finishing Horizon one through four. Do you mind speaking to that?

Frans Bakker: Yeah, I will tell you what I think is going to happen, and it's based on my own assessments of what I heard and what I read. I think I don't think it's going to be either or. I think they're going to increase the pace on both sides in parallel. So I think they're going to build out the staff at Sweetwater, and I think they're already building out the team at Childress. I don't think they've topped yet. I think that you can see they're still hiring. They did get a lot of new people at the start of January, but I think they're still ramping up. This is also why they keep building parking lots and these lunch tents and these, you know, like sort of like places for the workers to eat and to, you know, have. So the site is still, you know, increasing in in volume in terms of people. So what I know about Sweetwater is that they are very confident that they can ramp up the workforce there quicker. And they also think that they will be able to get people there easier because it's more close to Abilene and other towns in that area. So Childers is a bit out there in the middle of nowhere. And the Sweetwater one side is much closer to some bigger cities and towns. So I'm not sure if I could call them cities actually, but you know, bigger towns, I guess. So no, I don't think it's going to be something that, you know, if they ramp up the workforce and Sweetwater is going to go at the detriment of Childress, I think that's not the case. In fact, I think they're going to increase the pace on both sides in parallel. So I can't prove it, but from what I've heard and what I'm seeing is this is the case. So what we know from Iron is that they always deliver on time, but I think they're going to surprise us in how quick they're going to build in 2026 and 2027. I think building fast is going to become the number one thing that we'll see from Iron with the extreme emphasis on the speed. So that's not something we talk about very often. We did during the whole Bitcoin mining construction period, we did. It was, you know, very good to see that they always you know, started building these shells like six at the same time. But I think we're going to see that on multiple sites now for AI data centers in parallel. So yeah, some exciting times ahead. And at the same time, you know, the things we don't know that are happening in Canada are also happening at the same time. So they're already ramping up the workforce at Mackenzie. There is a buzz going around for a single cluster at Kennel Flats, and we know that they are building, that they're actively constructing in Prince George. And then there is also the intel we have about a possible new site. So yeah. I can only tell you this much, but it's not going to be just one or two sides at the same time. It's probably going to be more like four or five or more at the same time. So let's just say that it's going to be very interesting to hear them talk about the guidance at the next earnings.

₿itcoin ₿utcher: Yeah, most definitely. I mean, it's going to be a case study the whole year. how they're able to orchestrate so many moving parts in varying geographies, even though two are in the same state, I mean, they're still a relative distance away. And to coordinate all those resources, and most importantly, the human capital right now, just finding team members to scale up will ultimately speak to their ability to scale quicker than anyone else. And I think over time, that's going to become their competitive advantage. So we appreciate that, Franz. Looks like small cap can join us now. I briefly touched on PJM from your post earlier, small cap, but just feel free to take the floor here on any thoughts you have on the market and what's on your mind.

Small Cap Snipa: Yeah. Hey, fellas. Good to be back. I yeah, I'll just say my part a little bit because I've been thinking about XAI and, you know, all the stuff that's been happening regulatory. So and I'm sure you guys already touched on it, so I'll try not to go for too long. You know, on paper, I think it looks like a strong fit, right? XAI is without a doubt the most aggressive hyperscaler right now. really focusing on Colossus and Memphis, massive power needs, right? Like they're pushing 2 gigawatts now for Colossus one and two. And when you look at the constraints, you know, everyone is dealing with constraints as far as power, but XAI specifically, huge constraints, right? The grid limitations. And then that's why we saw them using the methane turbines. they really need high density power and they have very, very fast timelines. So Iren, right, extremely, extremely impressive pivot in the industry from pure Bitcoin mining to, I mean, I don't got to talk about it. We all know it. So it makes sense, right? Three gigawatts of power capacity, big deal with Microsoft, multi-year, almost 10 billion worth at Childress. And liquid cooled facilities all across their pipeline with Nvidia ties too, which is a big thing because I mean, not that iron would be, you know, running it for them. I mean, who knows? Maybe they would or they I mean, they probably would to some extent if something like that were to happen. But anyways, it makes iron irons are really credible miner in the space right now. So in theory, I think that a collaboration makes a ton of sense, right? Iron supplies power. secure, pre-secured power, data center capacity, a great team to help and help XAI to expand beyond Memphis. And where is it now? Mississippi, I think, was the one that they just closed on a couple of weeks ago. And I mean, I love it, right? In theory, it gives us another it gives Iron another high profile, probably the most rapid growing hyperscaler on the planet right now. And It's a logical extension in my eyes of where we're seeing, you know, former miners becoming key players in the supply chain. But, you know, I got to be the devil's advocate here. As much as I would love XAI to be someone that would cut a deal and have Elon in the mix now with iron as well, specifically, you know, XAI, if we look at the three large hyperscalers, I think there's three main models in right now. And I, the first one is XAI, which is extremely, extremely homogeneous model, right? They're relying on primarily one location. I mean, if I said it a month ago, it would be really homogeneous with one location, but now they're kind of, you know, going off to Mississippi as well. But previously, it's the biggest thing that they're doing is in Memphis. So it's one location. They're relying on one technology, Nvidia, right? Blackwell's, they got some hoppers in there, but mainly Blackwell's. And the networking is NVIDIA as well, right? They're using NVLink, Ethernet or whatever network connection is. Everything's through NVIDIA. It's all through one thing. So it's giving XAI a really good opportunity to scale and scale fast because it's not as complicated when you're utilizing different technologies. So, but there's a risk in it, right? You're relying on one location. You're relying on one technology. You know, what happens if you have power problems and with what is it? It's I think it's Memphis gas power and water, whatever the the grid that they're connected with there is. So I would love to see XAI kind of expand out and be a little bit more of a heterogeneous model if we call it like OpenAI, right? OpenAI has got a five or four or five different locations. Not all of them are out there yet. They're using a variety of different different chips. They have, you know, AMD and Nvidia accelerators. I think I'm pretty sure Broadcom is doing some stuff for them as well. And then the third model is Microsoft, which is like transforming previous data centers in all around the world. They got like 400 around the world into now these AI high performance computing. So that's why I would, you know, I think it would be a great fit and a great thing for XAI to do. But if we think about it again, The approach for XAI has, and they've been very, very like outward with this and Elon's emphasized it, that vertical integration and just having direct control over the entire stack is a big thing for them. You know, they've built, they built Colossus so fast. I don't, I don't know how if anyone's going to, I mean, I'm sure it'll happen as we progress, but as of today, no one's built like a supercomputer that fast. 100,000 GPUs, they were the first one to really do that as well and super, super aggressive. They own the stack, handle their own power, secure like large financing by themselves as well. So it's not a company that's like leaning on third party colo or hosting providers. They're just, they're just kind of like constructing their clusters under one roof because scaling fast and optimizing like efficiency and having as little latency as possible is just that's just what they are doing and that's what they've showed us. So that's part of the reason why I get a little just a little hesitant as far as thinking or like getting too, you know, caught up in the idea that it could happen. It definitely could happen, right? The things that have curtailed over the last couple, the last week, couple weeks have definitely maybe added some fuel to the fire, right? They're having issues with the EPA, with their turbines that they've, the methane turbines they're relying on, they were using 35 and now they're only allowed to use 12. Federal government's putting some pressure, right? Although I think that's kind of just a smokescreen, like smoke and mirrors. I think that although Trump is going to do whatever he can to protect the Americans, right, and not allow that cost to affect the bills of Americans. He's also going to do whatever he can to be the, you know, in quotation, right, the greatest economic president of all time and then really push AI and try to beat China and just do everything that he can to, you know, be the best stock market president of all time, right? So that's a positive overall. But we just, we got to go back to like the monolithic idea that XAI has. Massive deployments in Memphis, right? They're expanding Colossus currently to targeting 2 gigawatts. Tons of joint ventures and like self-generation, all with themselves. On the other side, Iron is kind of like the polar opposite. And obviously, they're not a hyperscale. You can't compare it. But as far as just like the business model, you know, their strengths are spread out sites. heavy power, renewable and liquid cooled builds that are amazing for hyperscalers like Microsoft that and even open AI, although I don't want to see open AI with our and I didn't don't quote me on that. I just as far as the business model goes, right, heterogeneous, spread out. It's really good for guys like Microsoft and Google being diversified and having like a multi-tenant infrastructure. That's I think what fits, you know, hyperscalers that are, you know, fit into that description, I think is what fit Iron best. But, you know, XAI's like needs are, I think, more extreme and favor in-house development rather than outsourcing. But again, everything that's, who knows, right? I'm kind of speaking outta my *** right now. And then I think the last thing is just that, The competitive landscape is crowded right now, right? Like all of, every Xminer is getting into it, getting into AI and computing AI deals in some way, shape or form. It might not be their whole business model now, but everyone is involved, right? Core Scientific, Cipher, HUD8, everyone's chasing AI deals. And I think XAI definitely has options through its ecosystem. right? Like Tesla synergies for energy storage, Nvidia relationships, nuclear-- One for the best crypto after. Well, Bitcoin's the best crypto after. What's the second best? There is no second best. Was that me?

₿itcoin ₿utcher: No, I'm sorry about that. Chad, you can't answer that.

Small Cap Snipa: I thought that was me for a second. I was like, what did I leave open? What did I leave open?

₿itcoin ₿utcher: Yeah, you'll have your chance, Chad. Please respect the speakers.

Small Cap Snipa: No, I'll wrap it up really quick. Anyways, where was I? Like, it just seems like XAI doesn't really need to partner with a miner when they're already raising like billions and billions to... just scale extremely, extremely fast on the vertical like integration and in-house side. So in short, the overlap is real, right? Strategic overlap and intersect, the idea of intersecting becoming a little bit less homogeneous is interesting and it makes sense to me, but I just, I personally just think that like XAI scale requirements and how fast they're moving and just the plans that they have with for the Blackwells. And I mean, what are they? They want to go to a million GPUs by the end of, by the mid 27 or possibly by the end of the year. I think that their self-reliance kind of just makes a partnership with Iron, I don't want to say impossible, but slightly unlikely, more towards the unlikely side than likely. Just the race is moving so fast. And I think that you can, we can project synergies, but execution paths might not align. And that's why I still, you know, the more that I've listened to your guys' spaces, listened to you talk, Butcher, listened to Franz talk, and really everyone talking, even just the posts that you guys talk about, I really do think Google is the best fit. Not just location, but just, it's kind of like it's the similar model to Microsoft, spread out. And it just makes sense. So I'll let you guys kind of feed in from there. I'm not sure how what you guys were thinking. My mic was, my listening was going in and out as you guys were talking about it. So I'd love to kind of just hear your thoughts on what I had to say.

₿itcoin ₿utcher: Chad was pretty enthusiastic. So we're going to go to Chad and I will. follow up small cap with.

Small Cap Snipa: Oh man, and I love Chad. I love Chad and Bitcoin AI guys. So I mean, I'm enthused. Listen, I love it, the idea. I just have a feeling it's going to be Google or Microsoft again. But let's see. I'd love to hear more.

₿itcoin ₿utcher: Yeah, Chad, let's your turn, Chad. And then we got Marbles follow up. How are you this evening, Chad?

Small Cap Snipa: Yeah, I mean, absolutely. It's an institutionally driven.

Speaker 3: Cycle, not to part of why Bitcoin has stuck a lot of the air out of the room in terms of altcoins. There's, of course, been a few small rallies and some memecoin stuff, but this is an institutional driven market. Very few institutions are interested in anything other than Bitcoin. When they are, it's been Ethereum, maybe a little bit, and then maybe a little bit of Solana.

Speaker 5: And that's sort of it in terms of like large scale demand. And so it's concentrated things a bit, the focus on Bitcoin and the retail has largely been uninterested, even at 100K. I think a lot of us thought 100K would be a trigger point for bringing in more retail, but like search.

₿itcoin ₿utcher: We're going to go to Marvels.

Speaker 5: Hey, what's up y'all? Can you hear me?

₿itcoin ₿utcher: Yep. What's going on Marvels?

Speaker 5: Yeah, nothing much. I just wanted to kind of dovetail off the small cap. You made a few interesting points. So I don't really think it matters whether xAI wants to be vertically integrated or not. I think I mean, this is something we all generally know. Can they get the power online in a timeframe that they need to? And I think there was a comparison made to OpenAI about how XAI is ostensibly, possibly less risky, but I don't think that's necessarily the case because XAI is burning a **** ton of cash. And if you look around, what are they good at? How are they more interesting in Anthropic or OpenAI? At least OpenAI has massive consumer mindshare. They have the most weekly active users of all the LLM companies. right? So at least they're winning there. And then Anthropic, we all know they're winning in coding. I think the other thing is OpenAI is winning pretty substantially. Well, they're not winning, but they have an edge in the scientific arena, which is enterprise and also kind of for lawyers. I follow a few accounts and they offer Codex, which is a model from OpenAI over cloud, right? So all these LLM companies, they try to specialize in one area. I think the mode isn't the model itself, it's the data. And if you can get in the data earlier, you can actually get customer lock in earlier. So I think the name of the game right now is speed and execution, and that's something that's really missed. A lot of people just talk about power this, power that, right? And they don't even think about when that power comes online. And I know Mike Alfred, for instance, is particularly bullish on Cypher, and I totally understand that. But look, Cypher doesn't have anything coming online, so like... end of 2026, '27, '28. So it's like, that's interesting. I'm sure they'll get a deal. But power right now is infinitely more valuable. And I think I made a post about it. If you think about why, it's-- well, it's because, look, it's sort of existential. For example, xAI, right? They're burning cash every week. every month, right? And a lot of the money they're making is through funding. And in order to get more funding at a low cost of capital, you need investor confidence. And to get investor confidence, you need a better model. So it's actually even more existential for XAI and OpenAI because they need to show the world that they're somehow in the lead. And this is not something that matters as much for Google or Microsoft because they have so much cash. So I I thought that was an interesting point. I think at the end of the day, it doesn't matter if XAI wants to be vertically integrated or not, because at some point they're gonna have to deal. And I think the other thing is it doesn't matter whether Iron gets a deal with XAI or not, right? I think they're gonna get a deal with someone because every hyperscale needs that power. So I think we can talk about, Hey, look, we're gonna sound XAI. I don't really think it matters. I just think that Iron, of all the companies, all the newer clouds, of all the Bitcoin miners, they're just uniquely positioned because they have so much power now. Not only do they have so much power now, like Franz said, I think the execution risk with Iron is substantially lower than all the other companies. And at some point, execution will become important. So just wanted to kind of like add my thoughts and chime in there because I thought that was a point that was missed.

Speaker 3: Marbles, who do you think they contract with? Who do you think the next deals with?

Speaker 5: I have no idea. I really don't know. I'll leave that to Franz or Small Cap Sniper. All I can say is I'm confident one of these hyperscales will sign with them.

Speaker 3: So hey, Franz, Nick, I just got a quick question here. So do you see a world that they do cloud with diversity of hardware via TPUs on the next, say, in one of the deals that happens in the near future, right? One of the deals that happens this year.

Frans Bakker: What's that to me?

Speaker 3: Yeah, I think you're probably most qualified to answer it.

Frans Bakker: I would give it 90% probability that Iron signs anything with Google in 2026.

Speaker 3: With specifically with deployment of TPUs, but not Colo actually purchasing the TPUs, making the hardware sale for the for the hardware revenue side of Google, and then and then structuring the deal as a cloud deal.

Frans Bakker: I think I think it would be more like a hybrid between Nvidia and TPUs where Iron would be doing the GPU as a service with GPUs from NVIDIA and Google would probably be delivering TPUs for sort of like a co-location, a hybrid with a GPU as a service. I mean, I have no idea really. I haven't looked into the GPU economics. I don't really know how the supply chain works. I haven't really dug into Google. I'm just trying to assess the probability of them signing with Iron. But how that view would work out, I don't know. I just know that Google would probably also need GPUs, you know, in conjunction or at least you know, besides their own TPUs. I don't think that everything can run on just TPUs, but maybe marbles could comment on that.

Speaker 5: Yeah, no idea. But not to change the subject, not really sure, but I do think Google will deal with Iron at some point. I think one thing that Agrippa mentioned, and I know we brought this up earlier in the conversation, is the Haskell site Google has, right? And then I think the other thing that wasn't mentioned was the OpenAI has the Abilene site. right? That's going to be 1.2 gigawatts. So really, if you think about it, Iron is sitting on a landmine of massive opportunity here where hyperscales will want to bid for that power. Because if you don't bid for, it's going to go to your competitor. And I just want to illustrate kind of how existential this is if one of those companies loses that power. Because if you think about it, right, the Rubens are coming. It's 5X better training for for the next Gen. AI models. And then you have the power, which if you get 1.2 gigawatts, let's just say it's 33% more of OpenAI's portfolio. So 5 times 1.33, that's like 6 to 7x more training power for your model. If you think about it in the grand scheme of your AI model, if you have a model that is 6 or 7x times more powerful or has more compute than your competitor, it's game over, right? Particularly if you're OpenAI. Because if you lose your lead, it's utterly horrific. It's utterly horrific because you need to invest the money, you need to invest your confidence. So I just think at some point there will deal, and I do think there is going to be a point where the hyperscales in that region, because you can also have a direct fiber loop to Abilene from Sweetwater, and you can have a direct fiber loop, just something Franz mentioned earlier, to the Haskell site. I just feel like it's written to stars. It's not going to happen yet, obviously, but I just feel like this is such a huge opportunity that people are missing. And at some point, at some point, not because hyperscale wants to, I just feel like they will have to do it. So I just feel like it's early days and it's a particularly exciting time because we still have not seen a Google deal and we still have not seen these gigawatt deals. And I do think one day they're going to come.

Speaker 3: Marbles, do they have any sort of disclosure around how they've structured the infrastructure on the Haskell site or on the Eveline site in terms of, well, obviously, Eveline site's all GPUs, but Did they deploy any TPUs, or are they planning to at Haskell? Do you know that?

Speaker 5: No idea. Franzina.

Frans Bakker: I know Haskell is in early stages of construction. I think they're just starting with civil work. So I think it's way too early to tell. I've posted one of my posts about Haskell. Google site in the comments. For some reason, I can't share it with the nest. Maybe you can do it for me, Bitcoin Butcher. That would be a good read to catch up on everything with regards to that particular site. But I have no idea if they plan to deploy TPUs there. There is not much to find about that, as far as I know.

₿itcoin ₿utcher: I believe I added that post to the nest. And I would say related to all that, I still think we're in a cost of capital game where when we talk about Google, Amazon, Microsoft, Meta, all of those parties, I can't speak to XAI's financials as well as I can those prior three hypers. And the biggest challenge iron is going to have is there's execution risks that we spoke through that there's no reason to believe that they have. But then there's also the financial risk. How do you fund all of it? And I'm of the position that whether it's institutions bidding up your share price, if you're going to use convertibles or if it's prepayments from people that have cash on their balance sheets like Microsoft did, or the debt markets looking for who is your anchor tenant on your flagship site at Sweetwater or Childerist, they're already taking enough risk by owning the GPUs as opposed to going the co-location route. I support that as a shareholder, so I'm not against that. But what I would say is if you're going to sell to the market as to why this isn't going to be Core Weave 2.0, then I think it's really important that they get the financing straightened out on the front end and tell a really compelling financial story on how yes, we know that they're going to not have co-location fees because they own the power and the infrastructure and the data center. And everyone can say that till they're blue in the face. But we're still waiting on Canada to spit off cash flow and Bitcoin mining. Bitcoin, while we're not going to do it in the future, still was supposed to be a big part of what was going to help fund all of these ambitious plans. And Bitcoin's still sub-100K right now. So just playing devil's advocate or the bear case here, I would like to see if we're not going to do Colo at Sweetwater, then at a minimum, you have to have the right brand, per se. The economics are going to matter, and I want to see a higher GPU rental rate than we did Microsoft. And there's no reason to believe we won't, given the evidence with the H100 pricing, and just the fact that they're going to have delivered the data center, and Microsoft's going to be operating with no hiccup, and there's going to be no headline of anything going wrong. So that makes them more valuable. But personally, there's plenty of people within our internal group that would favor Google for that reason. And then certainly Amazon with their ambitious goals of growing, I think, five plus gigawatts over the course of this year. I don't know how you do that without Sweetwater. It's going to be pretty difficult. And even if you do sign it with someone, whether I know Galaxy has new approved space or someone mentioned, I think Marvel's mentioned Cipher earlier, but as a Cipher shareholder, They have 370, 200 in Ohio, and another 170 in Texas. All the roads go through iron. And I'm almost positive that, yeah, we're going to have a deal, but that's one thing. But when people-- I just still think we're in a unique position here, where we kind of see the chessboard better than your average Wall Street analyst or institutional investor, because They have four or five times the power of their nearest competitor, or if someone has a similar power portfolio, they still haven't built a data center yet, so they might as well be six months to 12 months behind because they're going to go through the same growing pains that Iron went through with Horizon. I mean, Horizon took 9 to 12 months to build Horizon One. Now, we're in copy and paste mode, as Franz was saying, because they were able to create the right construction processes that now they can duplicate, and now they can use modular pieces, and now they have a process where I just think all these other guys are talking out of their *** at this point until they deliver a site. Or if they are competent, like a wolf or a cipher, they don't have enough power in the immediate future to be as compelling as Iron. So those were my thoughts on that.

Frans Bakker: I just shared a LinkedIn post in the comments. I'm not sure if it's worth to share in a nest or not, but it's just a random thing I see when I open LinkedIn. I follow all these data center development companies. And I think, if I'm not mistaken, Tracked was involved in the Stargate expansion in Texas. But even if I'm mistaken here, it doesn't really matter because they are supposedly delivering a very big data center somewhere near Austin, I believe. But it's interesting to see that these companies are still buying land, like they added like 1458 acres, and then they translate this to, it now has the space to support up to four gigawatt of data center capacity, positioning it among the largest data center parks in Texas. But just to, if you zoom in, then it turns out that they are only going to deliver 250 megawatt in the first quarter of 2028, you know, with an initial 360 grid connected expected to be energized that year. You know, it's just so funny that, you know, we're in 2026 now and these companies are still adding land like it's going to directly translate to power. you know, they make sure to mention four gigawatt at the beginning of the post and then later on it turns out that it's actually just going to be 250 to 360 megawatt in 2028. And then imagine you're iron and you're sitting on 1.4 gigawatt grid connected power to be energized in just a matter of a few months from now, right? I mean, the, the, the, the, comparison is just, there is none. You know, there's, I don't think there is any company in the entire world that will energize 1.4 gigawatt at the grid level in the first half of 2026. I just don't believe there is anyone else. So, you know, this is super exciting. I think there's going to be a huge premium on this site. I think that you know, the ability to scale on a site that big with that much power is going to come on a significant premium. And I think it's going to be a complete waste to just throw it away with a TPU co-location. So I expect fireworks from Sweetwater One. I think it's going to be their rightful flagship site. And I think it's going to be very profitable in terms of margins. I think the economic is going to, you know, blow the Microsoft deal out of the water. And I think we're going to look back at the Microsoft deal a year from now and realize that it was never about the economics. But yeah, more about that some other time. But I just think it's very interesting that, you know, we're talking about, you know, the gas turbines and the EPA rulings and whatever. But at the same time, all these companies, the peers of iron or the competitors are still, you know, building on this hopium of delivering on-site gas generation in 2028. You know, I don't want to, I don't know anything about Cipher, so I can't really comment on that. But I do think that Everyone, including the market, is still vastly underestimating what iron has in their pipeline. Grid-connected power will prove to be more valuable than gas turbines and on-site generation. And I think we're just going to have to see this play out. But exciting times ahead. And I guess we'll know a lot more in a couple of weeks from now.

₿itcoin ₿utcher: Okay, I got Morgan on stage. Let's see how this goes. Maybe a little better than Chad. Morgan, how are you?

Frans Bakker: Hey, can you hear me?

₿itcoin ₿utcher: Yep. Thank you. So I had a follow-up to the comment about the economics of Sweetwater versus Horizon 1 to 4, which I agree with, although obviously I'm I defer to your judgment. But I was curious about how you feel about the economics of Sweetwater versus Horizon 5 to 10. What exactly do you mean by the economics being like, what do you think is going to be the better deal financially for Iron, or how do you think it's going to be structured? Is there something-- can you be a little more specific? Or just like the margins, like you, Franz had the margin comment. I mean, I don't know, I don't wanna, I actually missed the first part of the space, so I hope I'm not like, although I generally listen to like the whole space every week, so I don't wanna just kind of come in sideways into the conversation. But like Franz had the comment about the margins of Sweetwater versus Horizon 1 to 4. So I was just, Like do you think that the larger size of Sweetwater is going to drive most of that margin premium? Or is it more just about like the evolution of pricing throughout the space because Sweetwater will be contracted significantly later than both Horizon 1 to 4 and Horizon 5 to 10?

Frans Bakker: So I don't know if that provides a little bit of color.

₿itcoin ₿utcher: Franz, what do you got?

Frans Bakker: I think it's threefold. I think you're right on the money there with the timing. The power crunch will intensify. That's why we're seeing all this chatter about the gas turbines now, because everyone is just, you know, looking for power as soon as possible. So I think that the Roberts Brothers are holding the, you know, the cards. close to their chest with, you know, they're not going to sign something early on. I think that Mike Alfred's been commenting on this a lot as well, that it's better to, you know, know what you what you have and don't don't sign away your capacity for a long term colocation deal early on. All the deals that we've been seeing recently have been better than the ones before. I think that's still a statement that holds. I think that This is the first part why I think SweetWater One will command a premium. The second one is, like I said, the ability to scale for a customer. So a customer could say like, okay, I want 250 megawatt, but I want the ability to go up to a gigawatt or 750 megawatts, for example. No other site can guarantee this power in this time window. And the third one is obviously on Iron's side, the operational leverage that they can, you know, throw down there. I think that this site will be able to generate higher margins because they will be able to, you know, scale up without, you know, I think their OpEx going to go, is going to go down on a, megawatt hour basis as the site develops. So, you know, that would lead to higher margins. These are the three things why I think that this is going to be, you know, the site that will be much more juicy than Childress over time.

₿itcoin ₿utcher: Yeah, I had not considered that about the OPEX. That's food for thought. Thank you. Yeah, Franz, you were saying earlier with Cipher specifically, like I had mentioned, they have the new Ulysses site that's 200 megawatts in PJM in Ohio, and then it was another 70 and 100, respectively, for Stingray and Reveille. But it's such a-- this is why you have to trust your management team and why I believe we're going to be rewarded. But I self-admittedly, after the Microsoft deal, while the CapEx caught me off guard, and then I got a little more hesitant going forward on the infrastructure as a service model, given the rate that they were able to attract, then you say, OK, it's their first deal. And we've all been past that. But I think what's most revealing right now, especially after this EPA ruling, and what Franz was speaking to, is this immediate need for power given the scaling in real time where a few months is like a few decades in this new space and how much development there's been. And now that Wolf and Cipher have already signed off all of their sites, they don't really, like Cipher specifically has a one GW site called Colcus that may be approved next year. And Tyler Page has said you can pre-release that, and he's not wrong. He could still sell a deal, but right here, right now, Iron's the only one that you can go to and immediately build, and they already have their construction process down pat where there's a reasonable-- you can have confidence that in a few months time that you could potentially have a data center done and be able to plug in. And that's a really unique selling point in the marketplace right now. And I formerly held the opinion that Sweetwater had to be a hybrid site given the capital constraints where there would be some co-location. And that's not necessarily out of the picture, but I think it gets back to just Dan and Will being that confident with the hand that they have right now, where they can go to someone and say like, yeah, you could use a Colchis or you could go to Helios or one of these other large sites, but you're going to be waiting and we can do it for you right now. And by the way, we need you to pay a big upfront payment to secure that. And I'm sure that's the dance that's been going on the past few months behind the scenes, while as shareholders, we were a little frustrated that we weren't hearing much since the convertible following the Microsoft deal. But if this power crunch is as obvious as we think it is, There's no reason to believe why they don't have their choice to do what they want when they want at Sweetwater, assuming they deliver. And they haven't given any reason to believe that they wouldn't deliver up until this point. And that's why you've seen this uptick. People are trying to front run a deal right now. And we're back from 33 to 58. And it's an exciting time. And I just hope speaking out loud, we spoke to it earlier that the macro doesn't take the air out of the balloon right now, but we'll see what happens.

Frans Bakker: One thing I'm very interested in is to find out what the electricity rate is going to be at Sweetwater, because you're looking at 1.4 gigawatt of power, right? I think that's going to be you know, significant amount of electricity costs on the on the, you know, as a cost in the in the in the revenue. So I have no idea how this will how this will work because they're buying spot in Childress. They have they can draw 750 megawatt there, but now they're going to work with a different utility. a transmission only utility. So it's not the same as AEP that they're working with in Childers. I just I have no idea how this is going to be as a factor, you know, because at the end of the day, if they are going to go with infrastructure as a service, the electricity costs are going to play a part in their margin. Right. So I think it's going to be I would hope that it would be some something like maybe like a bulk discount or something like, oh, if you buy more than a gigawatt every month, you know, like this sounds kind of kind of simplified, but we have no idea about this. But if, you know, if it would work like that, then you can come on a higher premium per GPU hour because your site is so big and you have the ability to scale for your customers. And at the same time, your opex can go down through various ways, including potentially your electricity costs, right? I mean, as we've seen with the Microsoft deal, every little bit counts towards your unlevered IRR. So I think it will be very interesting to see. I'm almost curious how they can pack all this information into one earnings call and report. You know, at some point, they're going to have so many sites, they're going to have so many things to talk about. Let's not forget, they're building a liquid cool data center in Prince George as well, which is basically about to go live for the first ten megawatt. You know, how how are they going to explain the market when they have five, six sites? They're going to announce a new site. You know, it's just going to be very interesting to to see how they how they are going to translate this into a, you know, an understandable earnings report where you can distill all the details from the profit and loss statement. You know, I personally really don't like all these very fake reports from some of the peers where you don't know anything with regards to the rack density or the GPU count. Not going to call any names, but, you know, for example, who I'm talking about. I think it's going to be very interesting to see how they're going to, you know, bring this information to the market, how they're going to tell us about everything that's happening. Especially now there's no more monthly updates, you know, but the amount of information is going to increase. So instead of informing us more regularly, they're going to drop down to quarterly information and then everything they're going to tell us is going to be so much more, you know. I'm just curious how they're going to do that, but kind of rambling here, so I'll pause here.

₿itcoin ₿utcher: Nick, you had something you wanted to add.

Speaker 3: Yeah, so Franz, to the point on the price per kilowatt hour, I mean, I would think that that's almost a pass-through cost, but that's really not the point that I came up to speak about. It was to ask you guys, did you see the press release that AWS increased their rental cost on H200 clusters by 15%. So I mean, we have a very scarce asset and just supply demand economics would indicate that they should be able to drive prices up to some degree, right? So I don't know how incremental or how substantial the price per kilowatt hour is going to be in terms of the increase. That they're going to release, but I don't think it probably shouldn't be that big of a concern.

Frans Bakker: Yeah. I don't think it's a literal pass through cost, right? I mean, in your in the Microsoft deal, this is still considered an expense because you're selling a GPU hour. So everything that you have on your cost end is factored into your margin, right? As long as you're not doing colocation, I think you have to consider this an expense. Unless I am mistaken here and Iron is factoring this in their internal colocation fee that they have for Iron Cloud, I'm not sure, but it should be relevant at some point, I think. We've seen in Canada that they have 99% margins on their GPU hours, but you know, obviously that doesn't factor in depreciation and everything related to the GPUs and the data centers. But I just I'm just curious because at the scale of 1.4 gigawatt, if you're paying three cents or five cents, it's going to matter at the end of the day, you know, so just that that was my point. And I haven't seen anything about the margins from the H200s, but Iron's not been very transparent to the public about their GPU hour prices. I think if you really want to find out, you got to like sign up and inquire with them. But I have never, I've never not done that myself so far. It's yeah, I guess it's possible to find out.

Speaker 3: It's not mathematically possible to find out based upon the disclosure of the parameters of the deal and backing out the amount of power that they're using. It's just that's it's too opaque, so to speak, to be able to do that.

Frans Bakker: Well, that depends. So far, they've always, I mean, disclosed it right with regards to Bitcoin mining. It was always in the monthly ops and the same thing for the GPU hours in Canada. But I don't think that they've specifically mentioned it for the Microsoft deal, but it could be wrong. Maybe it's just part of the internal co-location fee. But anyway, it doesn't really matter. It's probably a small margin, you know, improvement, if any, for SweetWater. But I think it's just an example of something that could scale in their favor for a site as big as that.

Speaker 3: Yeah, I got it. I guess the only other thing that I would add is, is JP, you know, if you look at it was in I think Agrippa actually put it in his research report too. Somebody was circulating around the JP Morgan report of their latest revision of of gigawatt demand. I don't know if I have it in my photos. I'll try to drop it in the comments, but curious if you guys saw that and I think they stretched out to 2028. So it's just. It's more of everything that we know in my eyes, right? I mean, we've got the only site that's got 1.4 gigawatts plus coming online this year. I don't know, Franz, maybe you want to talk about Oklahoma or something. I saw you going back and forth this week. Kind of got me excited. I want to learn something.

Frans Bakker: Yeah. So with regards to Oklahoma, all I can say is that we'll probably update the market by the end of this month. So far everything that we've been researching is you know only France proprietary but I can tell you that there is probably something to tell the market just before earnings.

Speaker 3: Well that's outstanding.

₿itcoin ₿utcher: The other thing I'm looking for at earnings is I hope they let not that Mike Power specifically would speak to it but I hope they update what they're doing with respect to institute, maybe not institutions, but the index is specifically NASDAQ. And then if they were to have to, you know, even if they hint at moving their headquarters to the US, if that was required to qualify for the S&P 500, I would like a little bit of more transparency related to that personally. But that, I mean, we've kind of, feels like we've covered every single angle of this. It's just a matter of the old meme, one deal, and then waiting for earnings. To Franz's point, how do you pack so much information into a press release and a 20 page slide deck and then speak to it in 20 or 30 minutes and then address analyst questions in a call that might run an hour, 20 minutes total. It's going to be either really well done or there might be some confusion on the back end. I'm like excited for it, but I'm also nervous because there are so many moving parts to this where I think we're going to understand it better than most of these sell side guys that are covering it. Because, Nick, you were speaking to, I don't have the chart, but I know what you're referring to with the Grippa's post, like I've read it. And the fact that JP Morgan sees this power shortage yet their own sell side that covers iron still has us rated at $39 like the disconnect of information and and we've talked about the gamesmanship before and there was probably a lot of institutional buying this past week given the volume spike and they suppressed it long enough and now everyone you know is it's after the 1st of the year and the institutions got their respective allocations or they can start their new P&L for the year, but it's. It's still a big, there's just always a lot of questions and I think sometimes why something like Cypher might resonate more with some of these institutions is because it's easier to understand. And while I'm still hold iron as my biggest position, like there is something to be said about just communicating this and there's different ways they can do it. So just questions like, How are you guys going to fund this? And then with reporting with the sites like in the future, are they going to split out Childress, Sweetwater and Canada and just simply show the profitability of each region? Like how transparent are they willing to be or is it in their best interest strategically to just blend it into all one cloud category? Like those are. The things that I'm looking forward to seeing and the only way to be able to tell is to just get to earnings here.

Speaker 3: Yeah, I'd put that applying to men chart the JP Morgan 1 to 2030 in the comments there if anybody wants to have a look.

Frans Bakker: I mean, Reggie is just a complete clown. I think that he's completely disingenuous and I think he should be faded at all costs.

Speaker 3: He's a paid CIA agent, France. You didn't know that.

Frans Bakker: I don't care what he is. He's just a, you know, he's a waste of time. He doesn't know 10% of what I do about Iron, and he gets a podium to give us a price target. It's just ********.

Speaker 3: That's what they do to try to manipulate the price, right, so they can get their position. I mean, it's psychological warfare.

₿itcoin ₿utcher: He's kind of warming on me and Franz because I was able to open those $50 calls for May when we were at $42. So there is something to be said about. I feel bad for some of the other retail people in here, but I'd like to think that since they have the purview to listen to your good research and decipher ******** from the masses. At the same time, I get less upset about it than I used to, because now just seeing iron in the 40s, knowing with what we had coming, it was a very simple decision for me to swap shares for shorter dated calls. I mean, it's still May, like I got a little I originally had the position for March, but that's asking for a lot to happen in a few weeks or a few months. But yeah, the last week was a good week. And like I said, we just, I'm hoping that I saw a post, I'm not sure if it was from Alan or someone else in our diligence group that there's indications that the European Union may come to the table and not necessarily retaliate, but I feel like right now it's not iron that's in their own way. It's just the macro personally. And even so, we're going to win over time, but that would be our luck for them to come back from. MLK on Tuesday, get the steel that we wanted, and then have the market tank 2% or 3% because of something that's just out of our control. So that's my-- I wouldn't say it's a fear, but that's my biggest, probably, frustration right now. But that's just part of the process. And over the long haul, it'll figure itself out. It's more I'm frustrated for people who were positioned accordingly and could benefit from a nice spike in the short term for their options, or maybe they want to hedge accordingly going into tax season or whatever, but they might have to wait a little longer with this macro stuff. But what we're going to add, Franz?

Frans Bakker: Yeah, well, I didn't want to like go on too long about Reggie Smith or anything. I think sell side analysts just serve a certain goal, which is not always the goal that we as investors are, you know, we're not always aligned with them. Sometimes we are, but, you know, doesn't really matter. There is always things that can impact the price. Like you said, it can be Reggie or it can be Trump. It can be anything, basically. But I thought it would be interesting to sort of give one little detail about the Oklahoma site that would maybe just be like a teaser for what's about to go down there potentially. So Iron used to have a president back in 2020, 2021, I think. called Lindsay Ward. I think he was at some point let go or he retired or he went to do something else. But it turns out that they have got him back for this project in Oklahoma. So I think it's this is a very senior guy in the infrastructure renewables sector and I think it's very telling that they've, you know, made him come back to be the head of this new project. I think, you know, the people that know him from back in 2020-21 probably agree with me that this is a pretty significant move. So I just thought that would be interesting to disclose here as a teaser for what's to come for the new site.

₿itcoin ₿utcher: I forgot to ask you earlier, Franz, was there anything in the recent satellite imagery of Sweetwater? Like we have a pretty good understanding of how Horizon at Childress that there's going to be 10 DCs, the 750 megawatts, 75 apiece. kind of configured this similar manner. Do you think there will be more Horizon-like data centers at Sweetwater? Or is there any reason that you have to believe that given the size of the site, that there might be bigger data centers? Or would it be anything that you've seen so far to suggest if it's going to be similar or a different design than Horizon?

Frans Bakker: Um, yeah, I think it's going to be a different design than, uh, Horizon. I think they're going to borrow some, uh, some features of it. Uh, like Iron's, uh, completely a, uh, best practices, uh, company. So they're going to take the best practices from Horizon and, uh, implement it in, uh, Sweetwater. But I think it's probably going to be more modular of a design. I think it's, uh, also going to be very customer specific. Um, But I think that they're going to try to dominate the design from their own best practices. I don't think they're going to. I think Microsoft was very largely driven by the customer side, but I think this is why I believe it's going to be a much more economical, profitable data center shell where the CapEx per megawatt is going to go down dramatically compared to the Microsoft deal. So these are all basically assumptions. So, you know, not nothing that I've said is constitutes as investment advice or financial advice. But I think it's going to be a new new design, like a not completely new, unless it's going to be like something that Google specifically wants, for example, if it would would be Google. But I think there is going to be a new design for Sweetwater. And I think that that's going to be the breaking model for data centers for all their future sites. So I think that what we've seen in Childers is not the blueprint for the future. That's my take on it.

₿itcoin ₿utcher: Great. Small cap, you're still on stage. You got anything to add? Probably start wrapping this up. We're approaching 2 hours here and no one else was trying to come up. If you were trying to come up, we got probably another 5 to 10 minutes at most. So this is the last call. But I would just say that we do have Martin Luther King Day that, so the markets are closed tomorrow, which might be for the best while the US and the EU make up with each other and probably try and come to some sort of agreement to stabilize markets. I think that's, personally, there'll be some sort of tweet or truth or Besant will come back tomorrow night and say, oh, they're great and we're talking to them right now or something of that sort. Heading into Tuesday would be my prediction, but always be prepared for the worst. And with iron specifically, like I'm just so confident in this that in some ways I already feel like I'm looking at how many people are in here right now, 361. I think it's fair to say most of you own own iron. And in a lot of ways, you've already won. And it's an exciting time that we're here before. Yes, it's gone from 33 to 58, and that's a pretty nice move. But a lot of us were here from 5 to 75, and now I'm just-- I'm looking forward to breaking 100 and maybe even go up to 150 this year because But that's also not just price action. That's more just reflecting fundamentals that have been underneath the surface that a lot of us have taken a lot of time to research and try and talk about, and we enjoy it each week. And now the general public will start to recognize that, but they're going to have to pay a higher price than we did. So small cap or friends, anything got to close up. Otherwise, it's probably time to wrap this thing up. And I appreciate everyone's time this week.

Small Cap Snipa: Yeah, real quick, last comment. I guess I think Marvel's made some really good points about XAI and, you know, the bull case for it.

₿itcoin ₿utcher: Hey, Small Cap, we're having a problem with your mic there.

Small Cap Snipa: So hopefully... Pressure, should I say the EPA on XAI? And you know, it definitely could push them too. If you just look at the post that I put up, those three key components, grid restrictions, right, grid policy changes, the EPA, and then obviously the federal government. putting some pressure on these data center uses. So those are three really good, you know, reasons for XAI to kind of expand and be a little less of a homogeneous model, more heterogeneous. So I don't know, you know, we'll see what happens. It's man, this is a really awesome time to be in the market and the space is amazing what you two have built. So pleasure, blessed to be able to join, talk to you guys, get to know everyone and keep gaining knowledge. So let's see what happens and let's get to it on Tuesday.

₿itcoin ₿utcher: Thanks, Smallcap. Franz, closing words.

Frans Bakker: Yeah, same for me. I'm looking forward to the coming week. I think it's going to be an interesting short week of the market. I'm personally kind of hoping for a little dip on Tuesday. to be able to position myself for some boot spreads. Yeah, so let's see. And there's a very interesting two or three weeks ahead of us. So you know make sure to benefit of the volatility and I will I hope to host next week's space when I am back on a solid ground. So thank you for hosting Butcher and it was a pleasure to join this one and thanks everyone for joining and see you the next time.

₿itcoin ₿utcher: That's a wrap folks. Good night and we'll talk soon everyone. Have a great week.