This startup founder is building the first data center in space.
The first is we're doing like edge and cloud compute for other satellites. So, we're having these like small compute nodes that provide compute to other spacecraft basically.
This is Philip Johnston. He's the founder and CEO of a space startup called Starcloud, a Y Combinator backed startup that's trying to build a fully fledged data center in space. What makes like that advantageous versus just plopping a data center in Dallas, Texas and just saying, "Hey, we're going to use a ton of water to like run and cool this thing."
Yeah, so in space the ambient temperature is 3° K, which is about -270° C. So, it's like extremely cold.
Philip recently joined us on the Vesting Show to discuss his plans to make data centers the restaurant at the end of the universe and the challenges that come when undertaking such a difficult mission. I'm your host Bryson and this is Vesting. I want to jump in because before you were building these crazy data centers in space and meeting the prime minister, I want to jump back and just learn more about Philip as a person, as a human, your background. If you could give everyone listening the who is Philip Johnston, what would it be?
Yeah, my background, um, well, if you to go right way back, um, from the UK, I'm the youngest of five boys. I've got an identical twin brother. Um, I lived in South Africa immediately after being born, we moved out there for like 6 years. And then my dad was a diplomat, so we were out there with the British government. And then moved back, was in the UK for ages, and then left to go to grad school in the US, studied math and physics. And then I would went I did like algo trading for 5 years for the with the bank. And then, uh, went back to grad No, then I did VC and back in South Africa for a year, and then Yeah, I've been all over.
Was there anything in your like childhood that made you gravitate towards space? Like did you Were you like always an enthusiast of what exists beyond the frontier.
Yeah, definitely. I mean, like when I was a kid I wanted to be an astronaut and like all the things like, you know, yeah, I was I was big in space. I wanted to do nuclear fusion. I wanted to like do research on nuclear fusion cuz I was math and physics were the only things I was good at. So, I was like maybe I'll become a physicist or a mathematician or something and nuclear fusion seemed like the cool thing to do. Yeah. What I love about what we're doing now is it's like potentially much more impactful in that we could generate way more energy than you can generate doing nuclear fusion doing what we're doing now. So, I think that's kind of cool.
You said growing up you spent time in South Africa. You kind of like you kind of like traveled a bit. Like you were kind of moving around a bit and trying different things and it seems like that may have yielded also you being willing to move around and try different things into your professional career. Was there ever a moment pre Starcloud where you were like, "Okay, this thing's cool and I kind of just want to go in that direction or learn more about like space or data centers." You said you had this nuclear foray. I'm just wondering what in your early adult career kind of gravitated you towards thinking this whole space thing was possible?
That was probably yeah, during my time at McKinsey is where I started to see that the launch cost was coming down and then to be honest, I just used to consume a lot of YouTube content about the Starship program that SpaceX is building. Yeah. And there was a YouTuber called Tim Dodd. His his name on YouTube is The Everyday Astronaut and he made this video about going down to Starbase Texas and he was like he showed you around. He was like, "Look, you don't need a tour. You can just come. You can just drive around." Like here the roads where you stay is where you can get drink or whatever. So, I was like, "Shit, I could just go and visit Starbase Texas." So, in 3 years ago, I just got a flight on a weekend like not for work just like for fun. Just pulled up.
Yeah, I just literally pulled up on my own in South Bay, Texas and I just like was looking around and like like it was before even the first Starship launch. Like nobody was really talking about Starship except for like the the YouTube nerd community. And like that from that moment onwards I was like, "Holy what they're building here is going to change everything." So, that that's when I was like, "Okay, look like I need to I need to get involved in this somehow." And we started with my co-founder looking at different ideas. Like one of them was space-based solar, which is where you have huge solar panels in space and you somehow beam that power down. Um
A, you want to stop right now. [laughter]
Another is well, the problem with that is you lose most of the energy transmitting it from space to Earth. So, then we started thinking, "Okay, well like if you had a cheap way to get whatever you're going to consume the power with to space, that would mean you wouldn't lose all of that. And I mean you wouldn't need to send so much solar panels into space. So, like the most obvious thing then is data centers cuz almost all the new energy projects on Earth are being built for data centers. So, that's when that became the what the basis of a white paper that I put together with my co-founders Ezra and Addy and and then that became the basis of the company basically.
When you look at kind of the undertaking of taking an idea that's very pie in the sky, data centers in space, the whole idea I'm guessing is like, "Hey, you can you can get rid of the heat, you can get rid of things a lot easier, um and you can run them more efficiently because, you know, in space you don't have to deal with some of the elements that you would be dealing with here on Earth." Did you always have the belief that you could turn this into a company immediately or were you kind of just figuring out how far this this thing could go?
The first is we're doing like edge and cloud compute for other satellites. So, we're having these like small compute nodes that provide compute to other spacecraft basically. So, yeah, that's like a business that we can do with with current launch cost that will be profitable. It enables us to develop and build the technology as we go though. If the launch cost were to never come down, that would be a pretty good business just in itself. Um nowhere near as big as competing with all data centers on energy cost.
When you think about kind of the stepping stone of you know the long-term business model and I guess the way to break it down would be when a company is smaller, maybe they have an idea of like what a business model would be to get them to stage two or stage three. But for you guys, I mean the goal is eventually like how do we just service companies that need to run their data in space, whether it's you know big loads, training LLMs, where we're going to which we're going to get to. But how have you guys gone about designing business strategies and and and pricing models for to your point like earlier stage products that you can provide that eventually yield the results needed to get you into space consistently.
And as I said, as you mentioned, in the end state, yeah, we're planning on serving everybody, everybody that needs kind of compute in space will be serving. Um in the in the short term, we're basically targeting the highest value workloads. So that's to be frank, mostly like defense department as well as like anybody that's willing to pay more to do their compute in space. Um so like earth observation satellites are willing to do that. Um there's some crypto people that are super keen to do stuff in space for some reason.
[laughter] I don't know what you I maybe I don't know, maybe you're mining Bitcoin in space. I'm not sure what
We will. No, no, we literally will be on the second satellite. We're going to be mining some Bitcoin in space.
Is mining Bitcoin in space or doing having an H100 train an LLM in space, do you see these as like just landmark marketing moments that show off the capabilities of like where you guys are going?
So like for example, yeah, training nano GPT from Andrej Karpathy. If you go to the Starlink 1 tab, you'll see a couple of nice videos and things. Yeah, click this click this sound button on the on the bottom left. It'll It'll play. Starlink 1 separation confirmed.
Ah, so beautiful. So cool. Starlink 1 separation confirmed. So give So walk me through this, right? Like how how big is this payload actually? Like when you guys were like, "Okay, we're going to send this into space." How big is this payload? Like how much does it actually cost to catch a ride on a SpaceX rocket?
It turns out it's like pretty cheap. It's the size of a small fridge. It's like this big.
It costs The launch cost was like 400 grand, which is like dirt cheap. Um like compared to what it would used to be, which would have been like $5 million per launch.
Yeah. Cuz of SpaceX, cuz of the rideshare program. Um like the next if like literally the next uh rocket up would have been like $5 million. So SpaceX is about 10 times cheaper than anyone else. Um so that And then the satellite bus is about a million dollars. And then everything else we did inside that was about a million dollars. So you're talking about yeah, like 2 and 1/2 million dollars to to launch that thing. We raised like a $2.4 million pre-seed at 10 post.
Cuz we took us It took us 3 months to get that together. Nobody wanted to give us any money. That was before we got into YC. We got rejected from YC twice. Um like so sometimes people think we have an easy time fundraising, which we do now, but we didn't always. Yeah. [laughter] Um like I know what it feels like to bash your head against the wall trying to fundraise. Um but like I I used to tell our investors, "Hey, we're going to do this whole spacecraft for 2 and 1/2 million dollars." And they're like, "Nah, that They I think they they thought they would fund the pre-seed and then would have to raise again to get this up. But no, we did it for 2 and 1/2 million dollars. And one of the reasons is my co-founder is basically built the entire payload from scratch. Like not the bus, but like all of the compute modules. So like Ezra and Adi basically built this whole thing from scratch. Um like soldering with their hands and like yeah.
How big is your team and how much of that first payload going out like how much of that was like very like high risk, right? Where you had to decide there's certain components and things that might not work in space. Like were you 100% certain that this would work or were you kind of taking a risk as well?
