Building the First Data Centers in Space

1. 0:00 – 1:06

   First thing every space company should do

   1. PJPhilip Johnston0:00

      First thing every space company should do is book the first available launch they can. [laughs]

   2. GTGarry Tan0:04

      Before they've built the thing that they're going to launch.

   3. PJPhilip Johnston0:05

      Before they've built the thing. Before they probably even know what they're gonna launch. Booking a launch is such a good forcing function for a space company. So we founded the company January 1st, 2024. January 2nd, we booked the first available SpaceX rideshare, uh, launch, and we were like, "Okay, something is gonna be on that rocket. I'm not sure, 100% sure what it's gonna be at this point, but something is gonna be on there." [laughs]

   4. GTGarry Tan0:27

      It's gonna happen. [laughs]

   5. SPSpeaker0:27

      Sounds like the same advice for all the software companies. You just gotta keep launching and launching.

   6. PJPhilip Johnston0:30

      [laughs]

   7. SPSpeaker0:30

      Same thing. [upbeat music]

   8. GTGarry Tan0:39

      Welcome back to another episode of The Light Cone. Today, we're sitting down with Philip Johnston, the co-founder and CEO of Starcloud. Starcloud is building data centers in space to address the energy bottleneck that AI is creating here on Earth. Earlier this year, they raised $170 million, led by Benchmark, and became the fastest growing unicorn in YC history just 17 months after demo day. Philip, welcome to The Light Cone.

   9. PJPhilip Johnston1:05

      Thanks so

2. 1:06 – 1:40

   Why data centers in space

   1. PJPhilip Johnston1:06

      much for having me.

   2. GTGarry Tan1:06

      So to start off, why data centers in space?

   3. PJPhilip Johnston1:10

      So we are very quickly running up on constraints on where we can build new energy projects terrestrially. And so by building them in space, we get access to this almost unlimited low-cost energy in the form of solar. Um, of course, there's other costs, like the cost of launch, and the cost of launch is actually very rapidly trending down, you know, with new launch vehicles coming online, the Falcon 9 program and, and Starship on the horizon. Um, and so that's the reason we're building data centers in space.

   4. GTGarry Tan1:35

      How did you come up with the idea? It's not sort of like the typical idea that if you're,

3. 1:40 – 2:59

   The Starbase trip that started it

   1. GTGarry Tan1:40

      um... [laughs] people come up with.

   2. PJPhilip Johnston1:42

      [laughs] I'd actually been working, uh... I was with McKinsey for a few years working with the space agencies of different governments, and I could see that the launch cost was very rapidly trending down. And actually, in m- early 2023, I decided randomly on a weekend to take a trip down to Starbase Texas, where they're building the Starship launch program, even before the first launch, and not many people were, were, were paying attention back then. And I could see that they were just building a ma- enormous capacity. So they're building these two Starship gigafactories. They're designed to produce something like three Starships per day, and because Starship is reusable, you know, on a sort of three or four-year timeframe, that could lead to us having thousands of times the capacity to get things to space. And so I started thinking, "Okay, let's just, uh, run the clock forward. If the launch cost was 10 times lower than it is today and the launch capacity was maybe 1,000 times more than we have today, what would make sense that, that currently doesn't make sense?" And so started thinking about some of the concepts from sci-fi that I remember as a kid. You know, like space-based solar, where you have these huge solar panels in space. You beam that power down. Um, and essentially, yeah, just reached out to, to a few s- folks I, you know, I knew in the space industry and got connected with a few others and started, you know, ideating on, "Okay, what, what would make sense if the launch cost was much lower than it is today?"

   3. GTGarry Tan2:54

      I'm curious, are there any other ideas that occurred to you during this process when you were, like,

4. 2:59 – 3:25

   Asteroid mining, space hotels, and the ideas they passed on

   1. GTGarry Tan2:59

      thinking of, like, what, what people could build in space-

   2. PJPhilip Johnston3:01

      Um

   3. GTGarry Tan3:01

      ... that was cheap?

   4. PJPhilip Johnston3:02

      Yeah, tons. I mean, we, we definitely looked at manufacturing in space, and there's Varda. We looked at astro- asteroid mining and, uh, you know, obviously YC has AstroForge. Um, we looked at space hotels. We looked at... We looked basically anything where- which could make sense if the launch cost was a lot lower than it is today.

   5. GTGarry Tan3:17

      And how come data centers was the best choice?

   6. PJPhilip Johnston3:20

      Data centers makes a lot of sense to be the first thing you do because you don't need to reenter a product.

5. 3:25 – 4:43

   Why space-based solar doesn't work

   1. PJPhilip Johnston3:25

      So with things like, um, uh, asteroid mining or with, uh, m- manufacturing in space or space hotels, they require this very expensive reentry process. Initially, what we actually started doing was space-based solar, so huge solar panels and then you beam that power down, and it's, it's a concept from, I think, like the '60s. Even Asimov in the '40s was writing about it. The problem with space-based solar is you actually lose 95% of the energy in transmission from space to Earth. And so w- for the first two months of the company, that was literally what we were doing. We were called Lumen Orbit because of that, that reason, the luminosity in orbit, or in orbit. We were thinking, "Okay, once we get that power down, what are we gonna be using it for?" And even back in '23, most new energy projects were being built for data centers. So we were like, "Okay, well, either directly or indirectly, this power is gonna go into data centers. Let's rerun those numbers." So i- initially, we'd run the numbers to know what's the launch cost breakeven that makes sense for space-based solar, and we'd come to a number of, like, $50 a kilo. So then we reran the calculations. We were like, "Okay, let's spend a month figuring out what would the launch cost breakeven need to be to make, uh, data centers in space breakeven?" And we came to a much closer to reality number of $500 a kilo, and that's what we think currently that the breakeven launch cost is. Um, and so then we rapidly pivoted the company towards that. [laughs]

   2. GTGarry Tan4:38

      [laughs] So Philip, you launched Starcloud One in November of 2025,

6. 4:43 – 5:41

   Launching Starcloud-1

   1. GTGarry Tan4:43

      uh, and that was a real emotional moment. Like, can you walk us through what that was like?

   2. PJPhilip Johnston4:48

      Yeah. I mean, it was incredible. So just to tell people a bit about Starcloud One, so we'd set this launch date, um, you know, 18 months in the future, and we initially started with something quite unimpressive. Um, like we were gonna fly some NVIDIA Jetson chips. They've been flown in space before, and we were gonna try and do some new, new workloads on there. Um, yeah, so it was Addy that came on and said, "No, we need to do something way, way cooler than that." And, uh, we actually ended up putting five GPUs on there, two... three from NVIDIA, two from Arm. Um, but the most interesting one was this NVIDIA H100. We did some crazy things, like for example, you have to put it through th- thermal cycling. Um, and we didn't have time to... We'd, like, book the, the vacuum chamber, the thermal and vacuum chamber, and but we needed to know, okay, if we heat up this phase change material and cool it down and heat it up and cool it down, is it gonna crack anything? [laughs] And so at, like, 5AM the day we had to ship it down, we'd been working through the night,

7. 5:41 – 6:24

   Cooling an H100 in an ice bath at 5AM

   1. PJPhilip Johnston5:41

      Addy and Ezra are d- like dunking this thing in an ice bath [laughs] to cool it down. [laughs]

   2. GTGarry Tan5:46

      Oh my God.

   3. PJPhilip Johnston5:46

      And then we pull it out of the ice bath, and I'm like, "Are you sure it's gonna be fine with the electronics?" They're like, "Don't worry about it. Yeah, it's totally fine." [laughs] They pull it out of the ice bath, then we had these, like, h- like hairdryers to heat it up and melt all the wax. [laughs]

   4. GTGarry Tan5:56

      Yeah.

   5. PJPhilip Johnston5:56

      Like, or like, uh, hot air guns, and then we're dunking it back in the ice bath. [laughs] Uh, it is a miracle that it works, to be honest. [laughs]

   6. GTGarry Tan6:03

      You can just do things.

   7. PJPhilip Johnston6:04

      Uh, you can just do things. Yeah, this... So this is the kind of thing you can do as a startup because we actually had a quote from one of the primes on what it would cost to do Starcloud One, and they said $75 million to $100 million.

   8. GTGarry Tan6:13

      Oh my God.

   9. PJPhilip Johnston6:13

      And we did the whole of Starcloud Two, including the launch, for $2 million.

   10. GTGarry Tan6:16

       Yeah.

   11. PJPhilip Johnston6:17

       So yeah, leading up to the, to the launch- I mean, we, we got the most incredible separation video. I think we're gonna play it now

8. 6:24 – 7:25

   The separation video

   1. GTGarry Tan6:24

      Starcloud-1 separation confirmed

   2. PJPhilip Johnston6:26

      So half the time as it deploys, it will be behind the shadow of the Earth, and you don't see anything. The other half of the time it will deploy not into the silhouette of the Earth. And so, I mean, it just is the most perfect, uh-

   3. SPSpeaker6:37

      Oh, yeah

   4. PJPhilip Johnston6:37

      ... like deployment video I've ever seen. And this was played by Jensen as he walked on stage at the GTC conference this year, this five-story tall like, uh, screen behind him playing the Starcloud-1 separation video.

   5. SPSpeaker6:49

      Amazing.

   6. PJPhilip Johnston6:50

      Um, so no, it was the most incredible feeling seeing this, uh, deploy.

   7. SPSpeaker6:53

      Where was the team for this? Did you all like go-

   8. PJPhilip Johnston6:55

      So we took the whole team down. Yeah, there was only, uh, 12 in the team at the time of Starcloud-1 launch. So we took the whole team to Florida, and we had like 100 friends and family, investors. We, uh, some of our team brought their family and kids down, and like it was a real nice, uh, really nice day. We're gonna release a, a, a behind-the-scenes video of some of the like family moments in that, uh, launch.

   9. SPSpeaker7:15

      How long was it from like when the rocket lifted off until the thing deployed, until you powered it up and were able to verify the thing actually worked?

   10. PJPhilip Johnston7:23

       Yeah. So it lifted off.

9. 7:25 – 10:00

   First contact and the satellite that kept restarting

   1. PJPhilip Johnston7:25

      We were all on what they call the banana bleachers. It's like the famous from the Apollo era. But it was in the middle of the night, so it lifted off at around midnight, and then at 1:00 AM they closed the bleachers and everyone had to get on a bus and leave. And we're like, "Oh, shit, I thought we were gonna like be able to watch the deployment." And then I'm sitting in the back of a cab, like watching the live feed on spacex.com literally. [laughs] And I'm just like, "Oh my God, it's deploying. I can't believe it's deploying." Um, and then it took about 12 hours to make first contact, which is not bad actually. It usually takes at least 24 hours to get first contact, and then it's like a two-week commissioning period, and we had a whole bunch of software issues we had to work through. Um, like it kept re- the, the whole satellite kept restarting every two hours. And what was happening was one of 20 different, um, like failure, uh, like triggers was triggering, and we didn't know which one it was, and we can only get a ground station pass every, like at least every hour and a half, maybe longer. And so we had to manually turn off each one, turn back on everything else, and wait for a ground station pass to know which one it was. So that took like, you know, that took like three days just to figure out that software issue. And then, yeah, two weeks after that, then we started commissioning and turning on all the different payloads, and then we trained the first model, ran the first version of Gemini. We did the first fine-tuning of a model on s- in orbit. Um, did the first high-powered inference on, on satellite imagery and all this kind of stuff.

   2. SPSpeaker8:42

      So Starcloud-1 right now as we speak has just left Chinese airspace, I think.

   3. PJPhilip Johnston8:47

      Yeah.

   4. SPSpeaker8:47

      About to enter-

   5. PJPhilip Johnston8:48

      It's heading to Japan

   6. SPSpeaker8:48

      ... Japan. Yeah.

   7. PJPhilip Johnston8:49

      Yeah. [laughs]

   8. SPSpeaker8:50

      It's going quite fast, actually.

   9. PJPhilip Johnston8:52

      Yeah. So it travels around 17,000 miles an hour. It's, uh, it's about an hour and a half to do one circumnavigation.

   10. SPSpeaker8:59

       So I mean, by s- Starcloud-4 or so, this won't be like one, it'll be a giant network sort of covering the Earth, actually.

   11. PJPhilip Johnston9:08

       Yeah. From Starcloud-3 onwards we'll have 88,000, and they'll all fly. This one is w- in what they call mid inclination orbit, so it doesn't go over the poles. Um, Starcloud... And, and that's another point here on how janky this one is. It doesn't e- even fly in the right orbit, but at least it, it, it, uh, ena- enables us to test the hardware. Um, the next one will fly over the poles, and that gives us 24 hour always in the sun.

   12. SPSpeaker9:32

       In order to do that, you need to put thrusters on the satellite to move it into that orbit.

   13. PJPhilip Johnston9:35

       In the end state we'll be deployed in that orbit, but you need thrusters anyway to maintain the orbit because you have l- you know, drag from the upper atmosphere there.

   14. SPSpeaker9:42

       What are the engineering and physics trade-offs you have to make this happen? Because what you basically are doing with Starcloud is you're getting infinite energy because the sun is basically the natural fusion reactor. It's just a giant mass that generates a nuclear reaction with the

10. 10:00 – 11:30

    The two hardest problems: heat and radiation

    1. SPSpeaker10:00

       giant mass, and you have infinite energy, basically abundant, and lots of space. But as a result of that, which we- we're constrained of that in Earth, you have a bunch of other really hard problems to solve, like things around interconnect, how you send all the data back and forth back to Earth. How do you handle cooling? People think that space is cold, but actually there's no, unlike Earth, there's no air to cir- circulate it back. So you need some crazy solutions with cooling. There's things around radiation to flip the bits in, in, in processors, and there's probably other things around how to get the solar panels to become really big and expand, and many more. I mean, these are the ones that I could think of. I mean, what are... There's like so many things that are just so hard to build.

    2. PJPhilip Johnston10:44

       Yeah, you are correct. There are many engineering challenges to, to, to be solved. The two biggest ones that are outstanding and, um, most of our engineering, uh, our, our engineering team is split basically 50/50 down these two lines is number one, as you mentioned, how do you get rid of this heat in a vacuum? And then number two, how do we make the chips work in a higher radiation environment? Some of the other problems are being solved by other people. You know, the interconnect, we've just signed a contract with SpaceX for our next 20 satellites to have a Starlink laser terminal. That gives us very high bandwidth, low latency connectivity. Other people are solving some other parts of this. The, the core ones we're solving are that. And so for the first one, for the thermal management side of things, um, we are building a very large low cost and low mass deployable radiator. Now, radiators in space are, are not new. It's not

11. 11:30 – 12:43

    Testing GPUs in a particle accelerator

    1. PJPhilip Johnston11:30

       a new physics, um, you know, solution. It's, it's more of a manufacturing and engineering challenge 'cause the, you know, the International Space Station has had a radiator which dissipates, you know, lots of heat already. Very similar mechanism, so, uh, a, a liquid cooled loop. Well, the challenge with it is making it cheap and light. So our radiator is at least 10 times less mass per watt of dissipation than the ISS radiator, and, uh, 500 times less cost per watt of dissipation than the ISS radiator. So very low bench- very [laughs] easy benchmark to beat on cost for the ISS. But, um, yeah. And then on radiation, it's just a lot of ground testing in different particle accelerators. So we've done, um, several rounds of testing at the Brookhaven National Lab particle accelerator for heavy ions, and also at the, there's a cyclotron for he- uh, high velocity protons down in Knoxville. So we ship all of our hardware down there My co-founder Adi runs in and out wearing this protective thing. [laughs] Adi, uh, is a lunatic when it comes to this stuff.

    2. SPSpeaker12:25

       [laughs]

    3. PJPhilip Johnston12:25

       He's like, "Oh, yes, it's the same as flying transatlantic. Don't worry about the radiation." [laughs]

    4. SPSpeaker12:30

       Yeah.

    5. PJPhilip Johnston12:30

       So yeah, w- basically you expose it over a 24-hour period to the same radiation dose that you would have in a five-year mission, and then that, all of that telemetry and data then informs our choice on both shielding and software to mitigate bit flips. I think we're the only people in the world now that

12. 12:43 – 13:49

    Automotive parts instead of space-grade

    1. PJPhilip Johnston12:43

       know where both an H100, a B200, H200 will fail if you blast it with high velocity protons and heavy ions.

    2. SPSpeaker12:49

       Do you have to change any of the underlying electronics to solve the problem, or you just build shielding around it?

    3. PJPhilip Johnston12:54

       Uh, both. Both. Um, so we do build shielding around it. Um, th- there's a, a lot of it is just component selection. So for example, all of the SSDs, all of the power delivery system, all of the power converters and everything else, we, we test 10 components, and then whichever the best one is, we pick. We're trying not to use space-grade rad hard, uh, components. We're trying to use off-the-shelf, like automotive style components, because it's way cheaper, and most people have never tested those in radiation chambers.

    4. SPSpeaker13:21

       I think that's one of the secrets of, uh, how SpaceX has actually reduced a lot of the costs of, uh, a lot of the satellites a- as well. They use not space-graded electronics, which would be exorbitantly very expensive, but instead you take regular components that there's a supply chain and run through all the tests and make sure if it works. And I think actually this is, uh, one of the things that Stratos does as well to make sure that telecommunication satellites actually work in space. Because they're, they're

13. 13:49 – 14:24

    "The dumbest thing I've ever heard"

    1. SPSpeaker13:49

       also in, um, in GEO, which has more radiation.

    2. PJPhilip Johnston13:52

       Yeah.

    3. SPSpeaker13:52

       Are you in GEO as well or in LEO?

    4. PJPhilip Johnston13:53

       No, we're, we're in LEO.

    5. SPSpeaker13:54

       Okay.

    6. PJPhilip Johnston13:55

       Yeah.

    7. SPSpeaker13:55

       When you first told people that you were gonna put data centers in space, what were people's reactions?

    8. PJPhilip Johnston13:59

       [laughs] Uh, for a very long time their reactions was, uh, "This is the dumbest thing I've ever heard." [laughs]

    9. SPSpeaker14:07

       [laughs]

    10. PJPhilip Johnston14:07

        Um, so the, yeah, we, we, we tried to... We, we actually applied to YC once and we got rejected. We, we then tried to raise, and it took, we tried to raise 2 million at a 10 post on a safe. It took us three months, and we got rejected from at least 100 VCs. And then even after YC, um, when we went out to raise after demo

14. 14:24 – 18:32

    What 100 VCs got wrong

    1. PJPhilip Johnston14:24

       day, I think we got rejected from at least 20 VCs before we got to the first, uh, the first check.

    2. SPSpeaker14:30

       This is one of these ideas that was, like extremely unpopular with investors, and if you had believed the in- if you had listened to what all the investors were telling you, you would have quit and not ever done this idea. I'm curious, why did people tell you that they were not investing, and what did they get wrong?

    3. PJPhilip Johnston14:45

       People had a hard time visualizing a world with low-cost launch. And, and to be fair, we're still not out of the woods on that front. Um, you know, lots of things need to go right with the Starship program for that to make sense. So that, that's one thing. I think it just sounds too sci-fi-y and too... Yeah, it's a con- To make this work, there's a convergence of two factors. One is, um, the launch cost coming down, and people have to believe that. The second is it becoming way harder to build stuff terrestrially. Maybe two years ago, that wasn't quite as obvious as it is today. Now people are like, "Okay, well, they've just banned building new data centers in New York. They're gonna ban it in like seven other states." Now that tho- those two things I think converge to make it more investable, essentially.

    4. SPSpeaker15:26

       And I think the other thing, after hearing you explain all the engineering and physics, it actually is doable. It doesn't sound like sci-fi from first principles.

    5. PJPhilip Johnston15:34

       Yeah.

    6. SPSpeaker15:34

       It's like it's actually buildable.

    7. PJPhilip Johnston15:35

       The reason I had confidence in it, um, is because we had such a strong engineering team. So if, if you have, you know, the world's best space engineers saying, "This is possible," and to be frank, my background isn't, uh, as a space engineer, but, uh, you know, I, I, I have full confidence in my, my two co-founders, one from SpaceX, the other was building satellites for NASA, saying it's possible then, you know, uh, it's, that's a, it's a, it's good enough for me. [laughs]

    8. SPSpeaker16:02

       Once you had the idea, and then as you, once you were in YC, how did you sort of break the problem down into like a, a plan for what to do? Obviously in software-

    9. PJPhilip Johnston16:11

       Yeah

    10. SPSpeaker16:11

        ... you know, you, you think of your MVP, and you launch quickly, and you get feedback, but hard to do that with data-

    11. PJPhilip Johnston16:16

        Yeah [laughs]

    12. SPSpeaker16:16

        ... centers in space. [laughs]

    13. PJPhilip Johnston16:18

        I have a very tangible piece of advice for all space companies. [laughs] The first thing every space company should do is book the first available launch they can. [laughs]

    14. SPSpeaker16:25

        Before they built the thing that they're going to launch.

    15. PJPhilip Johnston16:27

        Before they built the thing. Before they probably even know what they're gonna launch. Booking a launch is such a good forcing function for a space company. Um, so we founded the company January 1st, 2024. January 2nd, we booked the first available SpaceX rideshare, uh, launch, and it was cheap. It was like 300 grand, you know, which, which-

    16. SPSpeaker16:44

        Which was, how far out was the launch?

    17. PJPhilip Johnston16:46

        Uh, it was 18 months out, and then it got pushed to 21 months out. And we were like, "Okay, something is gonna be on that rocket. I'm not sure, 100% sure what it's gonna be at this point, but something is gonna be on there." [laughs]

    18. SPSpeaker16:55

        Something's gonna happen.

    19. SPSpeaker16:55

        Oh, that's interesting. Sounds like the same advice for all the software companies. You just gotta keep launching and launching.

    20. PJPhilip Johnston17:00

        [laughs]

    21. SPSpeaker17:00

        Same thing.

    22. PJPhilip Johnston17:01

        Yeah, yeah. [laughs] And what, what we initially planned to launch was very unimpressive actually. It was like we were gonna put a Jetson chip, which has been flown on many satellites before, and we were gonna use it for, for kind of edge processing and things. We actually had a co-founder change, uh, quite early on, and when we got Adi from SpaceX, Adi was like, "Oh, that's, that's lame. Like, let's, let's put an H100 on there." We're like, "Adi, you can't put an H100 on a, [laughs] on a satellite." He's like, "Sure you can." [laughs]

    23. SPSpeaker17:28

        [laughs]

    24. SPSpeaker17:28

        [laughs]

    25. PJPhilip Johnston17:28

        The amount of people who said it was physically impossible to run a p- like a, a, a chip that's that power dense, um, in orbit. Like, people just thought it was like literally impossible to run data center grade, you know, GPUs on orbit.

    26. SPSpeaker17:41

        What would have caused it to be impossible?

    27. PJPhilip Johnston17:42

        There's two things. One is the thermal side of things. They're p- very power dense. They produce a lot of heat. And the second is they'd never been tested in the radiation environment of space. Um, and both of those... So for the first one, [laughs] like it's kind of a wacky solution, like we submerged the entire thing in this phase change material.

    28. SPSpeaker17:59

        Mm.

    29. PJPhilip Johnston17:59

        So the, the, uh, everything on Starcloud-1 is, is submerged in this, um, like phase change material. So it's like, um, immersion cooling for all of the components, all of the power delivery, memory, everything is submerged. Um, like nobody's ever considered doing anything like that in space before. It, uh, it's not a particularly scalable solution. It's not like you, you have quite a low duty cycle. You have to wait for it to melt and then, uh, and then, um- And then it solidifies every so often. But it enables us to prove that you can run the H100 in space. The, the next one is this directed chip like we're calling architecture, that we can run continuously.

15. 18:32 – 20:25

    The path to 20 gigawatts in space

    1. PJPhilip Johnston18:32

       Um, but yeah, I mean, I, I give full credit to my co-founder Ali for coming out with some pretty wacky and crazy [laughs] ideas.

    2. SPSpeaker18:38

       To get to a data center that's, like, commercial scale, where you can actually, like, make money on it, it's actually, like, profitable, it's gonna require many steps. How did you, like, think about, like, sequencing out, like, the journey to get there?

    3. PJPhilip Johnston18:49

       So I would say first step was book whatever the, like, first launch lowest MVP we can have, uh, available. Second step is producing something which is a commer- like a, a, a product we can sell to customers. So that's StarCloud 2. That's a 10 kilowatt spacecraft that we can sell compute to government and military satellites. And then the third step is the product that we're gonna be able to sell to hyperscale, uh, data centers. So that's the StarCloud 3. We can... It's 200 kilowatt, three ton spacecraft that's six meters long. With that, we can fit 50 of them per Starship, so about 10 megawatts of new compute capacity per Starship, and we've just filed with the FCC for a constellation of 88,000 of those, so that means on the order of 20 gigawatts of new compute capacity.

    4. SPSpeaker19:32

       Wow.

    5. PJPhilip Johnston19:33

       And we can fit up to 10 terawatts, uh, in the Dawn Dust sun synchronous orbit. So that's like 20 times the entire US power grid.

    6. SPSpeaker19:39

       For context, what's the largest, uh, data center deployment on Earth? Because you're talking about-

    7. PJPhilip Johnston19:43

       Yeah

    8. SPSpeaker19:43

       ... 22 gigawatts. Oh my God.

    9. PJPhilip Johnston19:46

       Yeah.

    10. SPSpeaker19:46

        What's, what's the largest one here on Earth today?

    11. PJPhilip Johnston19:48

        Uh, they're about a gigawatt is the, is the largest.

    12. SPSpeaker19:50

        Yeah. [laughs]

    13. PJPhilip Johnston19:51

        Yeah.

    14. SPSpeaker19:51

        What's the timetable for all of this to become real?

    15. PJPhilip Johnston19:54

        So yeah, it's very dependent on the launch cost, and so, um, you know, the st- the estimates for Starship, if you believe Elon, it'll be end of this year, but let's say, uh, could be end of next year, early 2028 for Starship, uh, deploying, uh, you know, ramping up the launch cadence. They're actually phasing out Falcon9, so, um, they're gonna be moving over commercial customers to Starship. There's also a whole bunch of other launch vehicles coming online. Stoke Space has the Nova rocket. Um, Rocket Lab has Neutron, Blue Glen, um, Blue Origin has New Glenn. So

16. 20:25 – 21:10

    First customers

    1. PJPhilip Johnston20:26

       end of 2028, I'd say, is when we're starting to ramp up for the terrestrial business, competing with terrestrial data centers.

    2. GTGarry Tan20:31

       So your customers are actually anyone who wants compute, and then-

    3. PJPhilip Johnston20:34

       Yes

    4. GTGarry Tan20:34

       ... initially-

    5. PJPhilip Johnston20:35

       Initially it's government and military, but then as Starship ramps up, so for the next two, three years we're, uh, we've just won four contracts with various DOW and government entities.

    6. GTGarry Tan20:45

       And then initially also, um, it's people who want that compute actually, like, for other purposes in space then.

    7. PJPhilip Johnston20:51

       Yes.

    8. GTGarry Tan20:51

       Is that right? Yeah.

    9. PJPhilip Johnston20:52

       Yeah, exactly. So right now many satellites are very constrained with the amount of data they can downlink. And so using optical terminals, we'll have three optical terminals on our second satellite. They can ship enormous amounts of raw satellite imagery, um, and SAR data, synthetic aperture radar data, to us. We then process that on orbit, and then we can just downlink

17. 21:10 – 22:50

    Designing a GPU for space with NVIDIA

    1. PJPhilip Johnston21:10

       very quickly the insight from that, and the insight might be the coordinates of a vessel or something like that.

    2. GTGarry Tan21:16

       In terms of Nvidia, I mean, are you, how are you partnering with them around chip design, and, you know, are there changes that you have to make to the design to, I mean, as you reach scale, like being, making it more radiation resistant or, you know, what does that look like?

    3. PJPhilip Johnston21:32

       Yeah. So we're, we're working very closely with Nvidia. Um, they, we recently announced this, uh, Ruben, Space Ruben 1 chip that we're working with them on. Uh, so that's designed for s- the space environment. We very heavily modified the H100 to make it work in space. So on the mass side of things we stripped out a lot of things like the, um, AC to DC converters, the, um, the direct, the, uh, cold plates and everything else we sort of stripped out, and then we stiffened the board and made it radiation tolerant. So because of that, we, they, they're very interested in, you know, we're the only people basically that have any data on how one of these high-powered GPUs operates in space. Um, so yeah, we're working with them very closely on, um, designing this new space chip. 'Cause y- you can actually do very simple things to make these chips way more effectively, effective to run in space, and so that's what we're doing.

    4. GTGarry Tan22:19

       YC's next batch is now taking applications. Got a startup in you? Apply at ycombinator.com/apply. It's never too early, and filling out the app will level up your idea. Okay, back to the video.

    5. SPSpeaker22:32

       I think the thing of, uh, getting GPUs to now natively run on DC, you avoid the whole-

    6. PJPhilip Johnston22:39

       Yeah

    7. SPSpeaker22:40

       ... conversion of energy AC/DC, because that's what makes solar panels not as efficient, is the whole conversion.

    8. PJPhilip Johnston22:45

       Exactly, and-

    9. SPSpeaker22:45

       You lose a lot of power, which is-

    10. PJPhilip Johnston22:46

        Yeah

    11. SPSpeaker22:47

        ... very critical

    12. PJPhilip Johnston22:47

        ... and our, our solar panels generate DC immediately, so

18. 22:50 – 23:41

    A Bitcoin miner in orbit

    1. PJPhilip Johnston22:50

       there's no point doing... We, we just need DC to DC converters. Um, but that's much better than AC to DC converters.

    2. SPSpeaker22:56

       So it could unlock a bunch of, uh, companies also just building GPU data centers just using solar power, period.

    3. PJPhilip Johnston23:02

       Yeah.

    4. SPSpeaker23:03

       Whether in space, under the water, or on Earth. [laughs]

    5. PJPhilip Johnston23:05

       Yeah, I mean, yeah, actually people can definitely use this tech for other things, yeah.

    6. GTGarry Tan23:09

       Can you talk about StarCloud 2 a little bit more? Like, is there actually a Bitcoin miner in there as well?

    7. PJPhilip Johnston23:13

       [laughs]

    8. SPSpeaker23:14

       [laughs]

    9. PJPhilip Johnston23:15

       Yes. We will be flying a Bitcoin mining ASIC next, next year. Which the crypto community is very, uh, excited about. There's much more enthusiasm than I was expecting, to be honest, about that. So [laughs]

    10. SPSpeaker23:25

        I have a feeling Elon's a good narrative.

    11. PJPhilip Johnston23:26

        Oh, the, the number of people that want me to launch StarCoin, you have no idea. [laughs]

    12. SPSpeaker23:33

        [laughs]

    13. PJPhilip Johnston23:33

        There is actually a StarCloud meme coin, I think, out there as well.

    14. GTGarry Tan23:35

        Yeah, don't buy it. [laughs]

    15. PJPhilip Johnston23:37

        [laughs]

    16. GTGarry Tan23:37

        It's not official.

    17. PJPhilip Johnston23:38

        Do not buy the StarCloud meme coin. [laughs]

    18. GTGarry Tan23:40

        Yeah, yeah.

19. 23:41 – 25:18

    Why hard tech suddenly got popular with VCs

    1. PJPhilip Johnston23:41

       So yeah, we'll, we'll, we'll, we'll be launching, we'll be, we'll have one of those on there. Um, we'll have a whole bunch more H100s, BlackRock chip. Um, we're also partnering with AWS on launching the outpost hardware. That's their on-premises server blade that they give to customers to run a local instance of EC2, and that'll be useful for, particularly for our military customers.

    2. SPSpeaker24:00

       Something we were talking about earlier is, uh, at YC we've just observed over the last few years it's been quite hard for hard tech companies to raise funding in general. There's just not been much investor appetite for it. It certainly seems to have changed over the last six months.

    3. GTGarry Tan24:14

       Great news, yeah.

    4. SPSpeaker24:15

       From your perspective, I'm curious here, like, how was it raising the seed round in, sort of during YC, and then this huge round you raised from, um, from Benchmark? Just sort of maybe tell us the inside story of kind of how that all came together and, and what it feels like being a hard tech founder raising money today.

    5. PJPhilip Johnston24:32

       So when we first raised, uh, end of 2023, we tried to raise $2 million at 10 post, and yeah, that took three months to get together. About 100 VCs said no. Yeah, deep tech and hard tech in space was definitely not a cool thing to invest in back then. Um, then after YC, yeah, we, we... It took us quite a while to raise. So I think we got rejected from at least 20 VCs before we got our first, uh, first check on the demo day raise. And things have definitely s- swung around since then. Um, it's become a lot, uh, easier to raise for deep tech, I think. I think a number of reasons for that. One is people think that software doesn't have a moat anymore, which I think is probably accurate.

    6. SPSpeaker25:11

       That sort of hive... It's funny, like the hive mind just switched, like-

    7. PJPhilip Johnston25:13

       Yeah

    8. SPSpeaker25:14

       ... it seemed to me like very quickly. It was just like the SaaS stocks. Like Claude Code came out,

20. 25:18 – 26:37

    Inside the Benchmark round

    1. SPSpeaker25:18

       SaaS stocks, public company stocks all went down, and then suddenly it like... Was it... I, I think the like fastest I've seen sort of things in the public markets like directly affect demo day, where suddenly everyone's like, "Oh, like we need to like invest in hard tech now."

    2. PJPhilip Johnston25:30

       Yeah. That's fa- that's fascinating. Yeah. I think, yeah, people are just much more open to investing in hard tech now. Yeah, like I think we were the first space investment Benchmark have done.

    3. SPSpeaker25:38

       Yeah, I mean, whatever you can share, like how did that round kind of come together and, um, and yeah, what do, what do you think sort of Benchmark saw in, saw in you guys?

    4. PJPhilip Johnston25:45

       So to be frank, that round wasn't as easy as it looks from the outside. Um, there's a lot of VCs that have big SpaceX positions for one thing, and SpaceX like midway through the round came out saying they're doing exactly what we're doing, and it wasn't public then. And so anybody with like a conflict policy then was like, "Okay, we're conflicted out." Um, but no, I mean, towards the end of the round, it, it certainly came, came together much faster. Um, I think, to be frank, I don't think... So Chaitanya is our partner. Chaitanya's awesome. Um-

    5. SPSpeaker26:15

       Amazing investor.

    6. PJPhilip Johnston26:16

       Yeah. He's, he's great. I think basically he saw a really, really strong technical team doing something that if it worked could be incredible. And I think when he came to visit us, you know, he spent a lot of time doing DD on other things, you know, on the backgrounds of all of our... We've got the best eng- we literally have the best freaking engineering team in the world in Redmond

21. 26:37 – 27:48

    Their hiring strategy

    1. PJPhilip Johnston26:37

       building this stuff. You know, half of them from SpaceX, other half from, from some of the hyperscalers. Um, so he spent a lot of time with the team. He spent a lot of time, um... But I, I don't think if he'd, I don't think if he'd come across a report that said that cooling in space was impossible, that that would've like thrown him off. I think he was like, "Okay, this is a good enough team that they can figure it out."

    2. SPSpeaker26:57

       How did you build such an awesome team?

    3. PJPhilip Johnston26:59

       So my co-founders, Adi and Ezra, um, I... Well, with Ezra, I, I reached out to him after being down to Starbase Texas, and I was like, "Have you got any ideas that would make money if, uh, launch cost was 10X cheaper than it is today?" And we started with this space-based solar idea, um, with Adi, who was introduced to, through, to a friend, through a friend, and he, he actually had already been thinking about this. Um, he already had registered the domain name stellarcloud.com [laughs] before. And then with the team, we were just ruthlessly slow at hiring, uh, a- and like picky about hiring. After YC we'd raised, uh, like 11 million at 40, which is like a decent round in YC, and even then it took us six months to make our first hire. Um, and we got like the most kickass space engineer as our first hire. We've raised like a lot of money now. E- even this round that we announced in March, we've raised a much larger

22. 27:48 – 30:29

    How to be contrarian but right

    1. PJPhilip Johnston27:48

       round since then. We're still only 20 engineers. Like the... And we have one commercial guy who came from the Space Force who was working on Golden Dome. Um, we are so picky about who we hire. Like it is, uh, to a point of being very frustrating, to be honest, but it's the whole game. It's the whole game, is having an, a kickass engineering team.

    2. GTGarry Tan28:08

       What was it like to have this idea, have people sort of poo poo it, like didn't really get it, and then suddenly, I mean, to have much more commercial validation that like, you know, Google, um, SpaceX, the, the stalwarts of the industry sort of embrace the idea and realize actually like some crazy percentage of future data centers likely are going to be in space.

    3. PJPhilip Johnston28:30

       Yeah.

    4. GTGarry Tan28:30

       What was that like to sort of see the, the sea change and be early?

    5. PJPhilip Johnston28:35

       It was surreal. So if you go back through my Twitter feed, I've been like shouting from the rooftops-

    6. GTGarry Tan28:39

       [laughs]

    7. PJPhilip Johnston28:40

       ... for the last like three years about this.

    8. GTGarry Tan28:42

       Yeah. Contrarian, but right.

    9. PJPhilip Johnston28:43

       [laughs] Yeah. Every day I would post something about like why data centers in space are gonna be, uh, are gonna make sense, and then slowly and slowly and slowly people would start to drop hints about things. Uh, like Elon at one point was like, "99.999% of the energy in the g- in the solar system is, is, is from the sun." So I'd repost that and I'd be like, "We need to build data centers in space." [laughs]

    10. GTGarry Tan29:08

        Yeah. [laughs]

    11. PJPhilip Johnston29:08

        Like I would like... E- everything I would find a hook for, like... [laughs]

    12. SPSpeaker29:12

        [laughs]

    13. SPSpeaker29:12

        I think this is one of the things we see with all the best hard tech founders. You're essentially living in the future and are almost like a prophet that brings the future back to now us underlyings on the Earth that don't quite get it yet.

    14. PJPhilip Johnston29:26

        I'm gonna tell my mom, who's very religious, that [laughs] YC said I'm a prophet. [laughs]

    15. SPSpeaker29:31

        But I think that's cool. I mean, a- and the other thing that's fascinating to me is that, um, your background, if you go on your LinkedIn, it's not very legible that you would be in space, but you had a background previously doing software and physics and then went to business, but nothing there that said space.

    16. PJPhilip Johnston29:52

        Yeah. Um, so for people who don't know my background, um, spent the first five years of my career as a, a- an engineer on the software side, studying math and physics before that, um, undergrad and master's, and then I went to McKinsey more on the commercial product side. But I actually ended up doing a couple of projects, um, with national space agencies on different satellite missions, and that's where I really started to notice the launch costs coming down very rapidly. I actually had another, uh, startup before this, um, and yeah, when I left that one, decided, okay, if I'm gonna do another startup, it has to be something I'm like very passionate and, uh, a- about. And, and yeah, for my whole life

23. 30:29 – 31:25

    Recruiting co-founders before having an idea

    1. PJPhilip Johnston30:29

       I've been very passionate about space. Because I don't have a space engineering background, realized, okay, I'm gonna need the world's most kickass space engineers, and actually this is where I give credit to YC. I was kind of blindly following the YC advice. I was like, okay, the first thing we need to do before coming up with an idea is ask- Get the most incredible space engineers in the world, um, on the team, essentially.

    2. SPSpeaker30:49

       This was before you even knew what you wanted to do in space. All, all you knew at that point was that like launch costs were gonna come down dramatically and there was gonna be cool stuff to do in space.

    3. PJPhilip Johnston30:57

       Yes.

    4. SPSpeaker30:57

       And you wanted to find some space people to do it with.

    5. PJPhilip Johnston30:59

       Yes. Literally. That's literally it. [laughs]

    6. SPSpeaker31:02

       Yeah.

    7. GTGarry Tan31:02

       But it wasn't a vibe of like, "Oh, can you please build this crazy idea that I have?"

    8. PJPhilip Johnston31:06

       No, no.

    9. GTGarry Tan31:06

       It's like the reverse even. Like-

    10. PJPhilip Johnston31:08

        Exactly

    11. GTGarry Tan31:08

        ... I'm going to figure out the commercial part.

    12. PJPhilip Johnston31:11

        Yeah, exactly. My pitch to Adi and Ezra was essentially, "If you have any ideas that would make sense if the launch cost was 10X less than it is today, um, I can figure out everything else around that besides the

24. 31:25 – 32:26

    What if data centers get banned on Earth?

    1. PJPhilip Johnston31:25

       building the satellite piece." Uh-

    2. SPSpeaker31:26

       How did you, how did you find them? How did you recruit them?

    3. PJPhilip Johnston31:29

       Uh, so Ezra and I had grown up in the same place in the UK. I didn't... Actually, to be frank, I messaged, uh, like a, a lot of space engineers. Maybe 10 of them took a call, and two of them said yes. [laughs] And that's who I co-founded with, so.

    4. GTGarry Tan31:40

       That's all you need. You don't need like 100 people to do it. You only need one or two.

    5. SPSpeaker31:44

       So Philip, during the batch, the way I thought about Starcloud was that it could only work if you were able to make it cheaper per unit of compute than like ground-based compute, and I think you will. But it's occurred to me since then that like it's not actually clear to me anymore that that is necessary for Starcloud to succeed. I'm curious if you think about it the same way, because since the batch, we've learnt that AI data centers are becoming politically incredibly unpopular, and it seems not unlikely to me that it will soon become like basically de facto impossible to build AI data centers like anywhere, um, at least in like a democratically elected country. Like, I'm curious like how you think about that, how it affects your plans for the company.

    6. PJPhilip Johnston32:24

       Yeah, 100%, and, uh, this

25. 32:26 – 34:49

    Data center myth busting

    1. PJPhilip Johnston32:26

       has moved faster than I was expecting as well, to be honest. Uh, you know, we've just seen New York now blocking data centers' construction, um, you know, uh, for reasons which are not grounded in science really. Uh, you know, it s- it seems like more, more like vibes than anything else. Building these things in space will definitely be much easier from a, uh, regulatory perspective.

    2. GTGarry Tan32:45

       So let's do some myth busting. Um, you know, there are a lot of misconceptions about certainly water, uh, and data centers, but also power. You know, what are some of the things that come to your mind?

    3. PJPhilip Johnston32:58

       With water specifically, um, water is really actually just a function of power. Um, you know, you can cool data centers with no daily consumptive use of water if you're, if you just throw more power at the problem essentially, and that is what all of the data center companies are now proposing. It's very unlikely that any data center company proposes any, um, any, any data center that is gonna be consuming daily amounts of water.

    4. GTGarry Tan33:22

       And there is actually, uh, evidence that you can have closed-loop systems that consume very little water.

    5. PJPhilip Johnston33:28

       Yeah, that's definitely... Yeah, exactly.

    6. GTGarry Tan33:28

       And, uh, I think the analogy is like if you ran a burger shop, like a fairly large data center with like hundreds of megawatts would consume about the same amount of water as like a McDonald's.

    7. PJPhilip Johnston33:41

       Yeah. [laughs]

    8. GTGarry Tan33:42

       So it's like it's really-

    9. PJPhilip Johnston33:43

       That's about right

    10. GTGarry Tan33:44

        ... you know, a non-issue, but it's unbelievable that they're repeating this so much.

    11. PJPhilip Johnston33:47

        Yeah, yeah. They do use a lot of power though, and so most new power projects for data centers right now are being built with natural gas generators actually, and so there's certainly some people that are pushing back on, on those.

    12. GTGarry Tan33:58

        I mean, you know, one of the arguments is that, uh, as data centers get built, they sort of, uh, compete on the grid, which is kind of not true. I mean, people are actually bringing in-

    13. PJPhilip Johnston34:07

        Also not true if you, if you build a new power project for it.

    14. GTGarry Tan34:10

        Right, which is generally what people are doing.

    15. PJPhilip Johnston34:12

        Exactly.

    16. GTGarry Tan34:12

        You just can't get enough power from the grid, so.

    17. PJPhilip Johnston34:14

        Yeah.

    18. GTGarry Tan34:15

        There's also some evidence actually, um, that adding data centers, um, increases the capacity of the grid, and then prices actually come down over time. That being said, you know, it's sort of a perfect storm. Like we need intelligence. We need data centers. It's become basically a national security issue. You know, um, for America and for the West to have access to intelligence is, um, sort of the most important tech tree of the next 100 years. And so for tech policy to sort of put us back into the Stone Age and possibly lose the war for us, that

26. 34:49 – 36:12

    Advice for hard tech founders

    1. GTGarry Tan34:49

       puts Starcloud in a different position in terms of bringing compute and AI and power online.

    2. PJPhilip Johnston34:58

       Yeah, for sure. It definitely puts us squarely in the national security bucket at that point.

    3. GTGarry Tan35:03

       I mean, thank you so much for fighting for the West then. [laughs]

    4. PJPhilip Johnston35:07

       [laughs] We're doing our best.

    5. GTGarry Tan35:08

       Yeah. So as one of the fastest growing unicorns in YC history, uh, and to do it as a hardware company, do you have any closing thoughts for people who are, uh, you know, looking at space, looking at hard tech as a thing that they wanna work on? Like, you know, what have you learned, and what would you tell the, uh, you know, just starting out version of yourself based on where you're at now?

    6. PJPhilip Johnston35:28

       Yeah, I mean, I think there's just huge opportunities now coming in with space. We're really at the, like, very early innings with what's gonna be this enormous industry in space. Um, you know, the m- amount of capacity and launch that's gonna come online in the next five years is, is really mind-boggling. So yeah, I would definitely encourage people to, to step into that. Uh, in terms of advice, like the, the YC advice stands true. It's technical talent and founding team is the thing you need to solve for first, uh, and then everything else follows from there.

    7. GTGarry Tan35:57

       Philip, thanks so much for coming on The Light Cone.

    8. PJPhilip Johnston35:59

       Yeah, thanks so much for having me.

    9. SPSpeaker36:00

       Nice being here. [upbeat music]

Episode duration: 36:13