Anpanman - Europe SatCo JV, AI Data Centers, Launch Time

[00:00:06] Speaker A: This is the AST SpaceMobile Podcast.
[00:00:09] Speaker B: It will just basically come to your phone and be seamless regardless of where you are. We don't want the user even to know that it's connected by satellite. Listen, the opportunity that we have is very, very, very large.
[00:00:27] Speaker C: Hey, what's up everyone? Hope you guys can hear me. I'm gonna bring Jacob up here too.
[00:00:33] Speaker D: I can hear you.
[00:00:38] Speaker C: So someone— yeah, I can hear you too. So someone was saying, oh, you're gonna front run Kook Space. Yes, I am. We are. It's gonna be like, um, we're gonna, we're gonna go through Kook's, uh, tweet thread. for the week, and then when he comes on, he won't have anything to say. It's almost like, um, that rap battle in 8 Mile when Eminem makes fun of his own self, and then the, uh, the guys he goes up against, they, they have nothing to say. But, uh, no, uh, I just figured we'd talk because there's like so much stuff going on, and then Kook Space is so late, um, I thought we could have a pre-show, like, uh, tailgate before The big, the big show, the event. So anyway, Jacob, how you doing? Are you walking your dog? Oh, I guess he left. Anyway, um, but yeah, I, I know Kook is going to cover some of this stuff, but, um, yeah, there's been a lot of stuff going on just over the course of today, but also this past week. And so I just wanted to talk a little bit about some of the stuff that happened today, which is kind of nuts. Let's see, Jacob's back and then bring him back up. But yeah, this morning, so I guess starting yesterday and then really, I guess, peaking today, but who knows where this is going to go. Elon Musk has just been hammering the EU. He's obviously upset about the fine that they levied on him. And I'm not going to get into the politics of it, but as probably most people have seen, this is a pretty bullish development for AST SpaceMobile. And you could argue for Rocket Lab and basically any other space player that is not named SpaceX. within the SpaceX complex. But yeah, Elon, you know, he speaks his mind, you know, good for him. He's a pretty political dude. And so he, this morning, was retweeting some content that basically likened the EU to Nazi Germany. And then, you know, he was talking about abolishing The EU and yeah, I mean, again, not going to the politics of it, but the key thing, the reason why people, you know, that this has caught people's attention is that for those that don't know, ASD Space Mobile has this joint venture that they recently formed with Vodafone earlier this year. It's called Satco JV. And the way that that joint venture is structured is that it's controlled and primarily has majority ownership by European entities, primarily Vodafone. And eventually, once other European MNOs are brought on board, they're probably going to take some ownership stake in it as well. But AST SpaceMobile is probably, my guess will end up being a 49% economic owner, but then in terms of voting control, will be a minority. And so for all intents and purposes, that entity is European controlled, European owned. When the constellation goes over Europe, that entity will have sovereign control over the constellation. And so they'll have a kill switch as they call it, and they'll have the ability to beam down service, cut off service, know, run the data, all that stuff. But yeah, I think the key thing, and you're seeing, you know, this reaction from AST SpaceMobile folks and Cook will talk more about this later, but this is almost kind of like the Trump tantrum between Trump and Elon. You know, this is like 2.0 when obviously that relationship kind of soured and then you saw AST SpaceMobile and Rocket Lab rally significantly when that relationship kind of broke down. And as we say, I tweeted this before, but some of the stuff that Elon has been putting out there about the EU, like you can't, the genie's now out of the bottle. Like you can't put that stuff back in and what's been done is done and can't be undone. And so why that's important is that EU regulators are going to, you know, next year there's going to be a ton of regulatory activity on what to do in terms of allocating MSS spectrum over Europe. And EchoStar has their renewal coming up. I forget exactly if it's like 15 by 15 megahertz or maybe it's 10 by 10, but yeah, they're going to go to regulators and say, hey, we're, well, positioned to serve Europe and we're going to do that with Starlink. And so please give us a spectrum allocation back. And then I think the regulators are going to look at them and say, well, first of all, you've done nothing with it for the last, I forgot how long the concession was. I think it was like 18 years. You've done nothing with it. And now you're partnering with this company and this guy who basically says we shouldn't exist. So why should we give it to you? That's going to be a really tough hill to climb. And I think it's also important to note that EchoStar, the transaction that they entered into with SpaceX is actually just for US spectrum. So, um, I know, and it's a little, it's been a little unclear, um, when you, when you hear like EchoStar talk about it and SpaceX and It takes some digging through the spectrum purchase agreement, but that deal is only for US spectrum. It's not for global spectrum. Now, obviously, and that's AWS-4, AWS-3, and parts of the H-block. And so when SpaceX paid $19 billion for that first slug of spectrum, that's just for US. And then they paid another, what was it, $2 to $3 billion for some AWS-3. But SpaceX is actually in the same shoes as us. They actually need to go around the world and by either wheeling and dealing with companies that already own spectrum or perhaps entering into partnerships with them or going to regulators and procuring spectrum, whether that includes some amount of money or commitments, or sometimes that's just free. It's going to be kind of a jump ball situation. It's a race, right? For all these parties to go get spectrum. That said, just to give you an idea of like what that could look like, Saudi Telecom, who entered into a 10-year MNO agreement, definitive agreement with us for $1.8 billion minimum, $175 million prepaid revenue. And there's minimum revenue commitments there. They actually already own, because they won this through an auction, they already own 60 megahertz of S-band spectrum.
[00:08:04] Speaker D: That's right.
[00:08:05] Speaker C: Yeah, S-band spectrum. And they, and by the way, like L-band spectrum in Saudi Arabia is not even allocated. And so in that instance, we're already partnered with the MNO that owns all the S-band spectrum in Saudi Arabia. country. And that spectrum is for satellite use. And so if you put one and one together, that 60 megahertz is going to be brought to bear with our constellation, which is why I think Saudi Telecom views this as so strategic. And that doesn't include the L-band spectrum, which again, hasn't been allocated yet. So I think I mean, that's kind of the danger. Like Elon clearly is a super smart guy and can execute and inspires people around him, but he's super political too, right? And so if you can imagine, if you are on the regulatory team, like I think Dave Goldman at SpaceX, you know, he's the famous guy who called AST SpaceMobile own— he called out the foreign partners giving SpaceX a hard time in one of his complaint letters. And obviously we all found that funny because Deutsche Telekom or SpaceX's key partner, T-Mobile, is like, you know, owned by Deutsche Telekom, which is a foreign entity. And, you know, there's like political, there's been some political hay about whether or not it makes sense for a German company to own such a large quantum of US spectrum. which is highly strategic, but we're not going to talk about that.
[00:09:43] Speaker D: Anyway, um, quick question, quick point of clarification. So the spectrum that they, that Starlink already is using for the, um, fixed wireless, um, do you know off the top of your head which, I mean, that's like upper, that's like 30,000 plus, right? Or 3,000 plus, right? Okay. So that, that, but that's, it's a similar like path through all the regulatory environments, right? And so I'm wondering, I guess my, my only question was, is Did SpaceX have some sort of leg up there in the sense the company has already gone through, regardless, you know, take Elon out of the equation, just at what they already have? Does that lean to the idea that it's going to be at least somewhat easier for them to get this next bands of spectrum for direct-to-device?
[00:10:26] Speaker C: That's a good question. It could ease it a bit for countries where they already service Ka and Ka-band. spectrum, which is what you're talking about, which is much higher. It's, it's, you know, high, high band spectrum. Um, but when it comes to MSS, um, I think in particular, and this is primarily for Europe, I just, the fact that EchoStar and SpaceX and Inland are tied at the hip, I think they just, I think literally the last few days kind of decimated any hope to get that renewal. Um, that said, It's not as if SpaceX owns that spectrum because that spectrum in Europe was not included in the transaction. And so one thing that I'm actually, and this is like for any EchoStar bulls, this is something to really think about because I, this morning when I saw these tweets, I was like, oh my God, this could be a good shorting opportunity for EchoStar. But then when I looked at the deal docs, it's like, okay, it's not contingent. But, you know, if I guess taking this step back, if you were SpaceX and you did this EchoStar deal and you paid a nice price for the US spectrum, you probably have ambitions and you're talking to Charlie Ergen and Ergen's obviously very incentivized to work with you because he's going to own a lot of SpaceX stock. Your next plan is to go get that European spectrum, right? And the key thing of course is renewal because And this is kind of like how all these things come together and it's important. That renewal's in May 2027. If you look at the transaction, the deal doesn't close actually until 2027, which I think lines up with the maturity of some of its debt. But there's, it's like a 2-step deal where the US spectrum gets put into a trust and then it's there, you know, This regular, by the way, for like for EchoStar, the regulatory process to actually get this spectrum approved back from terrestrial to satellite use, space use, is a very long road. And then on top of that, it's not, I think previously the spectrum was approved for, maybe it was like GEO use, but now they have to get it approved for LEO. So Moving back from terrestrial to space and then from space, from GEO to LEO. And so there's a lot of steps there. And so as a reminder, like this stuff, the deal is actually not going to close until 2027. And that's what they've guided people to. I think it's like November 2027. And so part of me was like, when I was reading this stuff, seeing what Elon was writing, I was like, oh, you know what? Like SpaceX basically has this transaction is an option. It's an option. The transaction with EchoStar is an option to see if they'll get regulatory approval. But then within that timeframe, you're going to have the European reallocation process happen and you're going to know before the deal, this, you know, the current deal closes, whether or not EchoStar is going to get that spectrum. And so that's actually a pretty, it's the deal's not contingent on the European spectrum getting reallocated, but If you're SpaceX, you might have some level of buyer's remorse or say, hey, you know what, Charlie, like part of the, our, you know, not in written, but our verbal agreement was like, you were going to get that European spectrum and you didn't get it. And so maybe we'll cut the price or whatever it is.
[00:13:55] Speaker B: Right.
[00:13:55] Speaker C: And so that's something to be aware of because the transaction is structured in a weird way. Like it's, the deal's not closing until late 2027. And so you're going to have this risk of Europe not giving EchoStar that allocation, which technically is not part of this deal, but it is, I would argue it is part of the deal. But then, you know, if I'm Ergon, I'm like, hey, you guys, Elon's warpath against the EU kiboshed our ability to get the renewal. And then, you know, you might get into this huge argument. I guess I would say is that in terms of If you want to do a deal with somebody, an M&A deal that has a very long tail on it in terms of timing, Elon Musk may not be the best partner because we all saw that with Twitter where how he threatened to cut the price or walk away. But of course that deal got enforced in the courts and then he decided to just go through with the deal. And I'm not saying that's going to happen here, but it's a consideration. I think as the political, he turns up political rhetoric with the EU, the ability to get that renewal, the risk goes up. And then I think also just getting MSS spectrum globally, that risk goes up as well because you're dealing with someone who one day is supporting one political party, the next day might be sort of supporting another one and undermining your country, right? Because he's got a huge social media platform to do so. And you saw that in spades today. So anyway, I mean, yeah, it's, it's, but it's super bullish for AST because I think, you know, Iris probably gets some allocation of MSS spectrum. And then if you've got like Vodafone and a bunch of European carriers like Orange and, you know, Altice and any, and Telefonica, a whole host of other carriers. come to you and say, hey, we want allocation of this. And by the way, like this constellation's going up and it's going to go up well before Iris ever even starts putting together a solar panel, then I think that's pretty compelling, right? And it's European, it's a, you're going to create, Satco JV is going to be a European champion, right? And it's going to be able to provide consumers with connectivity and it's going to ensure From a military point of view in defense, a communications capability, whether it's, well, not just communications, but other non-communications capabilities that will help the continent defend itself against Russia or whoever else. But on the flip side, there is that, I'm sure Europeans, there'll be some concerns about the fact that this constellation will be shared with the US and the company is, you know, it's a US company, but I think you can structure around that where at least you will have sovereign control when it's over Europe and, you know, you'll ring fence the technology such that it's not going to be, you're not going to have these conflicts. But anyway, but yeah, that's something to keep an eye out. And I don't know. It seems like that whole European versus Elon is going to get worse. And so in a perverse way, that's better for us. And quite candidly, it's like for every space company, like if you're Rocket Lab or some of these other companies that are competing with SpaceX for business outside the US, that it just opens more daylight for you to go pursue contracts. Because I think the bigger controversy that Elon kind of musters up, like that's going to give governments and private industries, it's going to give them a second thought.
[00:17:54] Speaker D: But yeah, so— Real quick, last thing on the spectrum. I can't, I'm trying to, I was trying to look it up while you were talking, but somebody early, maybe beginning of summer, or something like that, or over the summer talked about, um, it was there. So basically the, the EU regulators stated, or somebody expert in the field said, there's no presumption of renewal. So especially for something like EchoStar, who hasn't, who's had this spectrum license for what, 10 years plus or whatever it's been, and then they haven't actually done anything with it. Like there, there was a, I don't know, it's, I almost feel like that's gonna cause like them to be pissed off. Like, well, you know, you've had this, you've had this spectrum for however long and now, and you know, you didn't do anything with it. Why should we renew it or give it to you for renewal?
[00:18:33] Speaker B: Yeah.
[00:18:33] Speaker D: I don't know, just a— I wish I could find the reference, but I felt like it was probably you or somebody at AST that definitely shared it this summer.
[00:18:40] Speaker C: Yeah, no, that's right. I mean, they haven't done anything with it. And then I think Viasat's the other player. They provide, with some of the MSS spectrum they have, they actually provide airlines with connectivity. But then funny enough, Starlink has actually started taking all that business in Europe. And so I'm not even sure they can justify keeping their allocation. But I think there's been, the EU had asked participants who were seeking spectrum, how would you divide it up? And I think it is going to be probably allocated to 2 or 3 players. But I'm almost, it's almost guaranteed is my guess is that Iris will get a decent allocation. And then the key question is like, who else gets that? And I think AST makes a very compelling case. And I'd be surprised. I mean, I don't know if they'll just completely lock out EchoStar, but I'd be surprised if they get a large allocation. And if they do, if I were the regulators, I would make it a requirement, obviously, that if EchoStar ever decided to lease that spectrum or to sell it, that you would have the ability to, you know, basically effectively a poison pill, put a veto on it. And then if they weren't, if they're, you might have like some kind of build-out requirements or usage requirements, and if they don't use it, then they lose it. And then, you know, that gets allocated to somebody else. But yeah, 2026 in particular for that situation is going to be pretty damn interesting. And then, yeah. And then And I guess, Jacob, we can talk about like AI data centers, which is like the soup du jour. It's like—
[00:20:23] Speaker B: Yeah, yeah.
[00:20:24] Speaker D: I was like, I'm like, oh, we're gonna have a spectrum conversation. I was like, I'll try.
[00:20:28] Speaker C: Yeah. Yeah. No, no, no. The AI data centers, it's kind of funny because I think I mentioned this before, but it's like one of these, when you look at like what StarCloud proposed and I know, and what's great, by the way, everyone should go go back and I think it's like 3 weeks ago that Jacob hosted StarCloud CEO. And so that was a pretty informative space. It was unfortunately one that I missed because I was at my son's soccer practice, but it's definitely worth going back and listening to. But I think one of the key things that the CEO talked about is, or a key problem is dissipating heat and figuring out radiator solutions. And so that, but then, you know, I, as I was brushing my teeth, getting ready to do this space, it must have been brutal for this guy because, and because we all faced it back in, was it September 2022? When it's like, hey, we're leading, we're doing this thing. And then Elon comes out and says, I want to do it too. And then he just drops the hammer, which on the one hand, it's great because this StarCloud CEO, it's like he's gotten a ton of attention and followers and everyone's like, oh my God, this is real. Just like me when I kind of, I saw the YouTube video of the architecture he's putting up and I was like, holy shit, that's like got Kessler syndrome written all over it. But then, but then because Elon Musk is talking about it and people are now doing work and then they're realizing, oh, Google's been talking about this and everybody, you know, there's been this like discussion already. And he's kind of, I'm not saying Elon's like late to the game, but he's validated it by acknowledging like, hey, SpaceX is going to do it. And of course, like it's perfect timing, right? Like he's, the company is having a tender and they're going to buy back stock from, they are, well, they and then outside investors are going to buy stock from insiders. And that's going to happen at an $800 billion valuation. There's, it's not as if The timing of the whole idea of discussing data centers is perfect because it gets people going, like they say in, what was that movie?
[00:22:41] Speaker D: Blades of Steel.
[00:22:43] Speaker C: You say something provocative to get people going and that's AI data centers. I think it's kind of silly in the beginning when you start talking about it, but then when you go through the numbers and you think about The limited resources here on Earth and how much stuff it's, how much energy and cooling, like water and things like that, it's going to take here. And the fact that, you know, if everyone's looking at their heating bill recently, it's like, oh, my electricity is like gapping up and going straight. It's going vertical. It's not sustainable. And so at some point, and who knows, like this AI bubble, it's probably a bubble to a degree and you're going to have like some pullback, but It's in this inexorable rise, like that's going to continue whether it pulls back and then keeps going up or a year from now it does that. But it seems like that's going to be the trend. And so I think like, I think putting, so it's kind of, I think it's important to kind of separate 2 things, right? Like it's not as if data centers on Earth are going to be completely replaced with data centers in space. But I think for particular applications, at least in the beginning, and we'll see where things go. But, you know, we talked about this in one of the chats, I think with Jacob, where, you know, if you're going to put up Golden Dome, you're going to need AI processing out in space, and that's going to have to be fully independent and cut off from Earth, right? Like if you want this Golden Dome to be smart enough to like stop threats, It cannot be dependent on anything down on Earth, right? Because those things on Earth, like if you need someone to make decisions or you have a data center on Earth that's going to enable, like it's got AI processors and it's going to enable a response. What if that data center's gone? You cannot rely on any ground infrastructure to be successful, right? For it to be resilient because in a conflict, If your adversary has already lit up a few of your cities and the data centers are there, or they, in a smart way, they target data centers on Earth, maybe population centers aren't targets anymore. It's going to be infrastructure, right? Then you're going to need AI computing capacity out in the— not in the cloud, but out in space. And so that's where, when people look at StarCloud, I know there are some folks who are like, skeptical and they're like, oh my gosh, that's 5 gigawatts. Like, no way. It's like, no, no, no, we're not. That's further out and we'll figure out. I mean, who knows? Technology's going to change. And yeah, I don't think we should be putting up massive structures like that in space because they probably already get smashed by a meteor or whatever it is. Right. And then that's bad news for everybody. But you are going to need distributed computing out up in space. And you might need, I joked about like a DataBird, right? But imagine like an AST SpaceMobile Bluebird, but you take out all the antenna elements, you replace the ASICs or FPGAs that are focused on signal processing, beamforming, and doing all those calculations. You replace those with GPUs or TPUs. But you've got Elon Musk today, he said, hey, the way this would work is you might need a satellite that has 100 kilowatts of Well, yeah, a Bluebird Block 2 has 100 to 120 kilowatts of power now. And you don't have to orbit it at the same inclination that we've been talking about. You could actually do it in a sun-synchronous orbit or who knows, right? There's various ways to send those birds up and have them focus on the sun completely, or maybe they're in and out. But, um, I think the key thing though is that, I think the key thing is that you, you will need computing capacity out there. Um, I know some folks have asked, well, these Bluebirds, like, do they have optical inter-satellite links? Because I don't remember seeing that. Um, yeah, actually Block 1 does. Uh, and that was part of the testing regime that was required by some of these defense awards. And my guess is FM-1 and FM-2 have them. And my guess is going forward, all of them are going to have them. But I think having inter-satellite links, as I mentioned before, you take out the antennas and you take out anything related to direct-to-device, direct-to-cell communications. You keep the QV feeder links that will be able to communicate with the Earth. And by the way, I think the StarCloud CEO was saying, well, we're not going to communicate with the Earth because we don't want to deal Spectrum licenses and having to go through all that technological, the hoops to get there. It's like, well, oh, these repurposed Bluebirds already have the high-throughput data links down to Earth. And so yeah, I think you could, I mean, you have the building blocks, right? And there would have to be design changes, but I don't think the design changes are that Herculean. And so you add battery, I mean, there's already batteries, but Perhaps you add more batteries. I talk about this idea of not having storage on the satellites and just keeping storage on Earth, but at least in the Golden Dome type of application, you actually want a decent amount of onboard storage because again, perhaps the things that you're relying on down on Earth might be gone in a conflict. And so these things need to be fully self-sufficient. self-reliant and autonomous. But yeah, like I was talking with Jacob and of course, I think there are some other companies that are well positioned, like I think Rocket Lab in terms of end-to-end space systems, they've actually built— it's pretty sad because like Jacob and other Rocket Lab bulls have talked about this ad nauseam and I guess I just didn't pick up on it, but they did, you know, Rocket Lab did buy this. He's laughing. They did buy this solar company that's focused on space that has a unique technology that can build solar panels that are basically gallium arsenide, but much thinner films. Like it doesn't have the backing, so it's much lighter even. But, and I think those, I think the efficiency is like pretty high. Maybe it's like 30, 40 to 45% or even higher. But it seems like those sort of panels are, or the sort of cells and modules are for highly specialized applications because they're very expensive. But who knows, the price will probably come down.
[00:29:33] Speaker D: Yeah, I was going to say, I looked at, I mean, they're definitely going to get some pricing advantage with their new, like, so that part of the CHIPS Act, they got $25 million or so and then the state of New Mexico added another $25 million so they could upgrade their, um, basically the reactors. The thing we talked about earlier, the epitaxial layer, the growth of the actual cells, those, uh, the tools that do that, machines to do that, they are purchasing new ones, the newest models from Veeco, who's a big name in the industry, semiconductor industry in general. But, uh, so that'll drop the price. But I was gonna say, like, the, the primary, like, use case for those solar cells is like, oh, we're gonna go to— it's like deep space missions. They're very— they're radiation tolerant, you know, or that's what's going to be used on The Ingenuity Mars helicopter, stuff like that, or the Solar Parker probe. So, like, real specialized stuff that they cannot fail, and they need a high efficiency for lower mass because of the distant shots. That's like the bread and butter of where that comes in. Stuff like Starlink or even AST SpaceMobile, the cells they're using, because everything lower Earth orbit, lower lifetime, shorter lifetime, stuff like that, they haven't needed that advantage. So just in pure terms of pure scaling, yeah, that technology is not— that type of cell would, I mean, wouldn't be the optimal, at least the first choice. I mean, maybe because it's internalized, they would have enough of a cost advantage to make it make sense. But yeah, they're not going to go out and be selling gigawatt levels of power out of using this type of technology.
[00:31:06] Speaker C: Right.
[00:31:07] Speaker D: At least I don't think so.
[00:31:10] Speaker C: Yeah, no, that makes sense. I mean, it's a very specialized and really expensive technology. And so sending that to, like, I was reading this guy, you know, of course I got sucked into the Twitter hole over the course of the day, even though I had all these other things to do. But I was reading this guy's post who was saying like, no one's going to be able to survive on Mars because there's so much radiation. And it's like, even if you build structures to live in, they're going to be out of metal, heavy metal. Like eventually they're going to become radiated. And then people are like, well, you just like Elon Musk can just use Boring Company and we're going to live in caves at Mars and get around it. But I could see like how having a vehicle on Mars with solar, it's, you're going to need obviously something more resistant and And efficient, but resistant to radiation and can survive really harsh environments, which is what Rocket Lab has, right? Versus the silicon that, you know, so for those that don't know, like AST SpaceMobile and Starlink use basically consumer, like silicon-grade or silicon-based solar cells and modules. And then the companies themselves actually harden those. And so you use Some types of resins and then also maybe like a glass that can block radiation and you put that over it and that's good enough for low Earth orbit. But then as Jacob mentioned, like as you get out to higher orbits and then of course in deep space, that stuff's not going to fly, especially if it's like mission critical. Because for Starlink or AST SpaceMobile, if you have parts of the solar panels, that fail, that's okay. Or the entire satellite bricks, then I mean, it's not great, but that's okay. But then if you're sending 1 or 2 satellites on a mission to Mars, like nothing can fail. So, but yeah, I mean, Jacob, like what's your thought about, I guess, Rocket Lab and how it's positioned for potential data centers in space?
[00:33:17] Speaker D: Yeah, I mean, I have got, I've definitely like done a 180 on this whole topic since, I don't know, since the summer. I mean, Even just leading into that interview with the StarCloud CEO, like, I was, I don't know, 2 weeks before or so is when we got the message and Francisco was able to connect us for that interview. And so I started reading his white paper, started reading more into it, and I was like, all right, well, I'm not. And then I started to like move to neutral. And then, you know, in the time since then, I've kind of gone into the more bullish stance on it. But so, like, total cost of ownership is what I'm like for specifically Rocket Lab, like, their approach, like total cost of ownership of the, of the, um, you know, of a data center. Let's, let's not assume it's a giant single unit. Um, it's, it's more of a— gonna be like distributed architecture kind of thing.
[00:34:05] Speaker C: Yeah.
[00:34:05] Speaker D: Then, then it's like, you know, they, they can build the buses. They, you know, they already have the tech there. They, you know, all the, all the reaction wheels, star tire, everything that goes in a normal satellite, they already produce in-house. Um, and then the quote-unquote payload for this would be, would would be the GPUs, right? I mean, that's the biggest payload. You know, if they— as soon as they close Minaric, they'll have the optical satellite interlink. So like, they, they own the whole stack to be able to go into this, obviously, other than the GPUs themselves, plus launch. And so, you know, they're getting internalized launch— well, will have internalized launch costs for that. I, I mean, to me, I'm— if it's— if this is actually a real revenue-generating, like, good business model, and that'll kind of get proved out over the next couple years, I feel like they're, they're well positioned to take advantage of it. You know, just because of the, you know, they already touch everything that's on a satellite. They already build enough different kinds of them that like, if this becomes something that whatever customer wants to actually have up there, then I don't know. It feels like they're well positioned to be able to do that. Some of the areas where I'm like, hmm, they're either going to need to make some acquisitions or come up with some like, you know, internal R&D kind of thing for it is the an area that AST has an advantage on is the radiators, dissipating the heat. Sorry, I'm bringing it up, but you brought up one of the patents they have around basically each of the microns has their own— dissipates the heat locally at each micron. And so instead of having a rack of GPUs and then having to use some liquid cooling loop through the whole thing, um, you could have the— you could have the rack of GPUs, quote unquote, distributed out over the entire satellite, which I don't know, that— I mean, that at least on the surface seems like a, uh, basically instead of, you know, it's like you're distributing the architecture onto the satellite and the same idea of distributing the, uh, the entire AI data center over several satellites. I don't know, I like the idea of that. So, um, I don't know, at least if it works, if it's a good business model, I, I mean, and, and it's like something that's could be really revenue generating. I mean, Rocket Lab seems really well positioned for this. They're the full vertical integration there. AST SpaceMobile, same thing. The only piece that they, you know, that Rocket Lab would have over that would be the launch in-house. So I mean, yeah, I guess it's still the use cases. I definitely saw the thing that the StarCloud CEO said that initial cases would be for things like, you know, the data that's already created on orbit, processing that data on orbit and then passing down the actual results so that, you know, the inference of that stuff. So that's all the Earth observation groups. Because right now, like Planet Labs, they basically inactive over the oceans or that all that imaging processes is lost and or something like that. And so there's like, there's tons of data that's up there that for a while now, you know, RF is not good enough to send all that data down. It's too much, you know, you need— so that's why the whole—
[00:37:10] Speaker C: Oh, I see.
[00:37:11] Speaker D: Passing it around on orbit with optical interlinks, which has much higher data rates, then you could process the information up there and send down a result kind of thing. So you would, you have a way of like basically not dumping all of this useful, potentially useful, um, data all these satellites are collecting. And that's like on the EO front. I'm not sure how this affects the comms front, but I'm sure there's, there's, I mean, as stuff gets more, I don't know, it's got to be a reason to have more compute up there, right? So I feel like those are the first use cases is is the data that's created on orbit, whatever that may be, defense, civil, then processing that data on orbit. I don't think we're going to have any sort of like, you know, a training cluster up there anytime soon kind of thing. But I could definitely see the near term using, you know, basically what the Starcourt CEO pointed out as their biggest first use case. Right.
[00:38:00] Speaker C: Yeah, it seems like this is really— there's going to be some initial use cases and it's probably going to be related to, like you said, Where if you have data-intensive applications and you need AI to, you know, if it's Earth observation to make sense of everything and you have all these satellites that are constantly ingesting information, imaging, especially in video to draw some insights, like it probably makes sense to do that in space versus put beam down certain things. that are limited because you can't, like you were saying, you can't beam down that much data. And so I think like going back to the Golden Dome example, you're going to need space-based processing to make AI to make those types of decisions. And so I think, and perhaps like, if it's the military, you're going to need some level of redundancy, command and control out in space versus if if something gets wiped out here on Earth. But wholesale replacement of data centers, that seems like it's going to be quite a few years away.
[00:39:11] Speaker D: Yeah. On that point, you were pointing to the different reasons why this could become a more useful case later. So we had a space last week, talked a little bit about it. So some of the gating factors will be launch costs coming down, the Then the power, the power basically trade-off, like it, it stuff starts to get more expensive on Earth for power, it starts to become more useful. And then the last piece that I was talking about before was that maybe it's not cheaper to put it in space yet, but what if it's faster? What if you don't have the regulatory hurdles of, of, you know, building the, finding the land, building it, hooking to the grid, all that stuff? If you could, if you could take all that and shorten it to be able to just launch, like, let's say it's a few years from now, the technology's matured, and then you could launch it. And like, then you look at the 2 paths to, all right, from starting at, you know, let's say it's 2030. So starting from zero, I want to build a new data center. Is it faster? Or, you know, do I want it fast or do I want it cheap? And maybe it'd be faster to do it on orbit at that point. I don't know.
[00:40:13] Speaker C: Yeah, that's a good point. And I think, yeah, it's the more, the more like I look into this, it seems like And this is the key thing, right? Like technology's going to improve and it's going to become cheaper. And of course launch is becoming really cheap. And so it's probably, we're going to be entering the next few years where all these companies are going to start partnering with space companies. And there's only a few that actually, I mean, you could go work at the prime, like at Boeing or one of the other Lockheed Martin, but then we all know how efficient those companies are in terms of cost and speed versus working with some of these new space companies, which, as I mentioned in the space before, it's not going to be just Google or Amazon or Microsoft. Obviously, Grok is going to be integrated with SpaceX, But yeah, it's going to be those companies, but it's also going to be the data center companies. Because if you're, let's say, an Iron or a CoreWeave, like you're going to want to understand what the economics are and the exposure is of space competition. And so you're going to probably take strategic stakes or joint ventures. You know, we saw this with American Tower when AST SpaceMobile was first put together and perhaps like a few years after development, American Tower took a strategic stake and got a board seat because they were like, oh, this company, if they're successful in what they're doing, they're going to put a big hint, you know, they're going to stop our growth in certain areas because people are going to then start making the cost-benefit decision of whether or not a terrestrial tower makes sense or whether or not in a less populated area, it makes more sense to just have a satellite cover that. And that's not really the tower company's decision. It's actually an MNO's decision. Like, do they actually want to pay to build a tower or do they want to have something covered by satellite? And if that's good enough. So as this narrative kind of builds, and this is why I wanted to talk about it, and I know Coop will talk about it even more because we discussed it over the last few days, but This is kind of the whole, this is a whole like Nvidia narrative, right? Like you, Nvidia was like a graphics card accelerator company. You know, you bought the card, put it in your PC so you could play Quake or Doom or, you know, Call of Duty, Counter-Strike, and that was it. But then as we started moving over to AI, it's like, oh, these GPUs are optimized and the best processors to do training and inference and all that stuff, right? And so then the company's narrative changed where, and this is kind of like for AST SpaceMobile or just space applications in general, when people kind of look at space companies, they're like, oh, that's kind of a backwater and it's nascent. But then now we're kind of moving to this new phase where it's not just Obviously communications is big and launch is big, but then there's been this desire to find other killer applications, which no pun intended, but then Golden Dome is probably the most visible third application for space that's going to have real dollars behind it. And then data centers is probably not that far off, at least in terms of hype and people talking about it and the possibilities. It's there. And so when you have that, how that impacts your investment thesis and how the market potentially discounts, pulls forward the future, that can change the story quite a bit. And we've seen this with like the SMR, the modular reactor companies for nuclear where 2, 3 years ago you couldn't sell that idea to anybody. Except for just really far out there VCs that are like, oh yeah, this is going to be important. But then as the narrative shifted and it's like, oh my God, AI is going to eat the world because you need all these data centers and all these processors are going to eat power. So where are we going to get that power from? And I think Morgan Stanley today, there was a report, they were saying like, there's going to be a deficit of like 40 gigawatts. in the next few years. Either it was like in the 20s or 40s. I can't remember. I think maybe the full demand is like 60 and we can supply 40, and then maybe the deficit was like 20, something like that. But yeah, that, so that shifted the narrative where it's like, okay, these nuclear technologies that aren't going to be viable for another 20 years, I'm going to discount that today to back to today and give people credit because I want it to get funded. And I want it to become a reality because we really need that. And so that, you know, whether that's Oclo or SMR, like these companies just took off, right? And some of these other earlier stage companies too. And so that's kind of the opportunity where, and somebody on Twitter kind of said it very specifically. They're like, we kind of stumbled into perhaps another growth vector, right? And that's exactly what Nvidia did. And I think that's kind of why everybody's been excited for the last few days where it's, you all of a sudden have this other pillar of growth and it's not something that's new because as I mentioned before, I remember talking to the portfolio manager at Hennessy and these guys are long-onlys, have been long-term holders of DSpace Mobile, perhaps like One of a few handful that have held since the beginning. But, you know, they specifically mentioned to me at, when I met with them for the Block.one launch, they said, well, one of the key areas of growth is the fact that these constellations are going to be massive power generating entities out in space and you can do a lot of cool things with that. And at the time I didn't really understand what that meant. Like, oh yeah, so power generating entities and they've got a lot of batteries. They specifically said like, you got solar panels, you got batteries. So there's a whole host of things you can do. And now it makes sense, right? I think Elon, as he so often does, can articulate things very well. And when he talks, like people listen because he is a visionary. And so today, like when he specifically said, well, you'd like to have a satellite that has 100 kilowatts of power, And it's going to be distributed. You can have these AI data centers in space that magically fits like what we're doing. And so then the mind can dream. The tech companies or data center companies, I think we're probably going to see more of those. And so I think the other thing is having the experience of being a space company, because obviously doing things in space is very different. Google in particular, for example, they're doing some initial early-stage testing with Planet Labs. Google has an investment stake in Planet Labs, and I think they're putting up 2 CubeSats to see how AI GPUs perform in space. And StarCloud also put up a CubeSat with the first NVIDIA processor to see how that would react in space. So you're seeing like some baby steps right now. But if you think about who has the flight heritage and the ability to manufacture and put up, I mean, in particular for AST SpaceMobile, like a 5,000 to 6,000 kilogram satellite with massive solar arrays, you know, a novel like unfolding array architecture that fits into a standard fairing, You know, you may not need, you'll, yeah, you definitely wouldn't need this massive control set probably. So you could save weight and maybe reduce the form factor. But, you know, also again, like Rocket Lab putting together satellites that can do missions to Mars, like this stuff is hard. And I kind of look back at like the StarCloud CEO and others and That's like AST back in 2020 and 2021. It's like stars in your eyes and you're like ripping, ready to go and like, I'm going to take on the world. And then man, it's like a grizzled path. But if you've got the fortitude and the capital and the belief, like people can do it. And I think that's, but I also think like the folks who've recently traversed that probably are pretty well positioned to be partners to help develop the market. And so yeah, it's exciting and I think it's going to become a building part of the narrative for space and it seems like it's going to be a big opportunity initially. It will enable space-based processing for those applications that sit outside of Earth, but eventually perhaps you'll see some data centers move up there. But yeah, I don't know.
[00:49:39] Speaker B: Do you have anything else?
[00:49:39] Speaker C: So yeah, I was just going to put it in there.
[00:49:42] Speaker D: So 2 things. One is that like the, the, I guess even the more bullish thing that I could see is just like with comms and everything else is when Elon gets involved, it's he wants the whole stack. And so it doesn't allow for partnerships. And then, you know, if he's going to be doing this primarily for his, his technology, his whole portfolio of companies, xAI, stuff like that, suddenly like it's going to be a rush to these other hyperscalers or other other big names in the— in that they're going to want this or need this, or you're going to be looking for that, you know, a non— a non-SpaceX partner. So, um, yeah, I mean, that's just a, I guess, a bullish sweetener. But then the other thing I wanted to say is maybe to throw some cold water just to give the bear idea, is it— so, so the 2 TPUs, the 2, uh, SATs that the Google TPUs are going on are— I think this is the new Alsat for for Planet Labs. So if they're going to put the GPU directly on the sat, it's taking the data in, suddenly my mind was like, all right, I don't know how much the power requirement would have to change or how big of a, you know, big design change these satellites, but what if you just had these distributed, um, GPU, TP, whatever, um, into all the satellites that are already getting launched up? Then, and then at least in the short term, maybe you don't need a massive, uh, like, you know, you don't even need a third, a third kind of satellite up there or secondary satellite up there that has just— that is just the data center one. And in the short term, maybe you could do all those— that, that, uh, computing on the satellites themselves. I was just in a thought I was having before is like, all right, if this whole distributed architecture— I'm like, why don't you just distribute the GPUs right onto the satellites that are collecting the data kind of thing? I, I don't— I haven't looked into that, just, just as a cautionary note on that.
[00:51:25] Speaker C: Yeah, and I will say, uh, this is not I mean, this is very quick research. I just asked Gemini, like, how many NVIDIA processors, the most recent ones, could you theoretically operate reasonably on a 120-kilowatt satellite? And I guess the answer is, if it's a Blackwell GB200, you could run 72 of those. And then let's see. Oh, that's a lot more than I thought. If it's a hopper, it's 88. And so, so yeah, that's a pretty decent amount. But then, you know, you're not, again, you're not going to be replacing data centers on Earth, but at the same time, as we all know, like technology's going to advance and I think the capabilities that you'll be able to put out in space, yeah, it's going to, I think it's going to surprise people. But yeah, I, and that's, I think we're 2026 after kind of going through all this. It's like, this is going to be a pretty exciting year. I think you're going to have some of these large companies, Google, Microsoft, any number of data center companies too. They're going to, they're probably going to announce some types of partnerships to start developing technology and it's going to be with some of the large primes, but it's going to be a lot of these new space companies who quite frankly have figured out a better mousetrap and can do it cheaper and faster. And that's going to be in competition with SpaceX, which is good. I mean, competition is great. And I think there's going to be enough opportunity for all these different companies. And I think, for example, Rocket Lab, you guys mentioned they have very specialized technology for Solar cells and modules, like there's going to be demand for that. And then for companies that are already, I mean, yeah.
[00:53:26] Speaker D: So, so for the solar cell thing earlier, we were talking about that. Like the, the other part is they, they, you know, they, they don't, they produce all the paneling and stuff as well, all of that. So even if they ended up saying like, all right, it's not cost effective to use these high-end solar cells. I mean, they just buy that off the market that everybody else is buying off of it.
[00:53:44] Speaker B: Right.
[00:53:45] Speaker D: And using these If it's a LEO kind of thing. I don't know. Just in general, Rocket Lab guys are really excited because we've been beating our— hitting the wall trying to figure out what is this magic space application potentially that they could be doing, right? I mean, there's been so much. It's been years of like, well, what are you going to do? And the company's been really coy, but man, so much of what they've bought lines up with this potentially. Like, so it, I don't know, we just for on our end, it's exciting because, you know, I feel like we've been without a home for what our eventual application is going to be like. Everybody's been beating their head against the wall. So if this ends up really being something, it kind of lines up more with the timeline that Peter's been talking about where I kind of been putting like 2028+ kind of lens on it for the application layer. So I don't know, that's why I got overly excited about this personally. But I mean, but it is exciting for the industry because I mean, we've all been looking for what's the next big thing. Is it going to be on-orbit manufacturing, asteroid mining, or something else? What is the next thing? So if it is AI, if it's infrastructure, AI data center kind of thing, that's exciting. I don't know.
[00:54:52] Speaker C: I got a message from Cook, our producer. Cook, he told me he's starting at 8:35. So for real.
[00:55:00] Speaker D: All right, I'll shut up.
[00:55:01] Speaker C: No, no, no, it's okay. No, this has been good. It's good to have somebody else speaking versus just me rambling, so it's good. But yeah, I think Tuesday we're going to have our weekly space and by then I'm sure there's going to be even more news that will come out. But yeah, what an exciting time. I think it's kind of funny because I think for anybody who's invested in space, we probably were the laughingstock for 2022, 2023, 2024, but then all of a sudden it's like, hey, This is the place to be invested in. And it's like everything is coming to a head, which is great. But anyway, I'm going to cut it there. This was the early show and we're going to transition to the Cook Show Sunday Night Live. So thanks everyone for joining. He's going to run that space in about 3 minutes, so I'll see you guys there.
[00:55:54] Speaker B: Take care.
[00:55:54] Speaker D: That was good. Thanks, man.
[00:55:57] Speaker A: Yep.
[00:55:57] Speaker D: Take care, Jacob.
[00:56:00] Speaker A: Thanks for listening to the AST SpaceMobile Podcast. If you enjoyed this episode and you'd like to help support the podcast, please share it with others, post about it on social media, or leave a rating and review. To catch all the latest news about AST SpaceMobile, make sure to subscribe. Thanks again, and I'll see you next time.
[00:56:19] Speaker B: We're doing something very, very big, and I think with this technology we can really affect a billion lives. AST SpaceMobile is the only company that has proven technology to deliver cellular mobile connectivity directly from space to the everyday smartphone. People will just basically turn on their phone and be seamless regardless of where you are. We don't want the user even to know that it's connected by satellite. Our goal is to bring this into reality, always in partnership with the animals.
[00:56:57] Speaker A: Listen.
[00:57:05] Speaker D: Mmm, waffles.