Muon Space CEO on data center satellites, the launch cost curve, and why software is the hidden bottleneck in space

barriers to entry have come down in a lot of ways but a lot of that complexity remains and our kind of goal is to really um abstract a lot of that away from a customer that wants to do something in space and doesn't want to have to go build that full vertical technology stack for every new use case that comes up and I think that was a a big lesson we took out of Skybox is we kind of built two companies under one roof satellite company and a data company and that the future should really about be about you know a data company being able to go build a data business without having to deal with all the complexity on the the space, the ground, the operations, the hardware, uh the integration side and that's really what we're trying to solve.

I want to know about uh the potential of data centers in space. It sounds like a crazy idea. I know some folks are working on it. Just kind of like what's your high level take of the progress? We've been tracking, you know, uh dollar per kilo to orbit.

It's been falling, but recently there's been setbacks and different programs and there's more competition and there's a lot of thing different dynamics going on.

Um, what do you think, uh, the key milestones are to get us to a future where we're really doing, you know, mass manufacturing, big, you know, mega scale projects in space. Yeah. Well, let me let me start off. I mean, I think you guys are hitting on the right metric, right?

It's like the hardest part of this is what does it cost to get a kilogram in space? because everything else kind of deres from that ultimately.

If you want a certain amount of power, if you want to be able to put um an aperture in space to do communications, whatever it is, like that's really driven by what it costs to get it there in the first place. And I think it's important context to kind of see how far we've come down that path.

I mean, largely driven by SpaceX over the last decade. And you know, I I like to tell the story of, you know, about 15 years ago at Skybox when we were trying to launch these small satellites, we were literally going to southeast Russia and launching satellites on converted Russian ICBMs.

I mean, they were they were popping out of the ground and putting satellites into space. And that was the only way for like a Silicon Valley venture-fed startup to go put something in space. Wait, wait, hold on. Did I hear you correctly?

Like the the the ICBM actually launches from one of those missile silos in the ground and makes it to orbit. pops out of the ground and goes to orbit. I mean, there's videos on YouTube that the you search on YouTube, you can see this and it's it's crazy.

I mean, you know, and that that's that's what it took before SpaceX kind of like revolutionized the launch business. And even at that, we were spending something like 10 times as much per satellite or per kilo is what we can go buy on a transporter launch today.

So there's been at least one and arguably two orders of magnitude improvement in the last call it decade on kind of what it means to launch things to orbit.

I think if you imagine that happening again another order of magnitude um it you know again it's going to dramatically change the way you think about feasibility of some of these things like putting very very large power hungry things in orbit.

Um, we know how to do the solar, we know how to do the structures, we know how to get, um, you know, we know how to make electronics work in the radiation environment, which is always a concern, and do it reliably. Um, and so really, ultimately, I think it's going to come down to unit cost.

Um, and and what does it actually take to do that? The great thing about space is you have virtually limitless power. The sun, you know, you can you can go into orbits where you're in the sun all the time. So, it's not like solar on Earth where you're going in and out of eclipse or in and out of night.

Like, you can be on all the time. And then you have this cosmic background of three Kelvin cold sync that you can go dissipate all the thermal energy you need from running your electronics and stuff. So it's actually in a lot of ways like sort of an ideal environment to do this if you can actually get there.

Um so I yeah that's the kind of way I'd think about it. It's interesting. Um what it feels like SpaceX has commoditized launch and is kind of the power law winner there. Obviously there's other companies that are competing but uh they've they've standardized uh launch.

Uh they've also standardized Starlink and then we talked to the Endurosat founder about standardizing a satellite bus platform and kind of getting to the mass manufacturing less bespoke, less customized, less handbuilt pieces.

What else are you seeing as kind of critical pieces of the space supply chain that need to be standardized or you might be trying to standardize? Where wh where are the pieces in the supply chain to do the things that you want to do?

um where it's still like, "Oh, that has a long lead time or that's handbuilt or that's way too customized. I get what I want, but it's too exquisite system. It's too expensive. " Talk to me about the supply chain. Yeah. And I'm I'm actually going to give you the answer you asked for and then an answer I want to answer.

So, I'll do both. Yeah. Um on the supply chain side, you know, I think there's a lot of progress been made. Like our satellites look very much like, you know, um especially things like the electronics look very much like what goes into an EV right now.

the batteries, a lot of the electronics, a lot of the individual kind of semiconductor parts heav he heavily leveraging other commercial industries. So it doesn't look like a traditional space supply chain. I think that problem more and more is is is solved.

Um I think Enduroat obviously is is doing very similar things in that way. Um some of the places where we really see that that that has not yet happened one is in more on the payload side. So things where like you know you think about the spacecraft bus and durosats building buses.

Um you know I do think there's a path to that becoming very standardized but every mission you put in space has a bus and then it has a moneymaker. It has something that's actually doing what you need it to do whether it's communications or remote sensing or beaming power if you're trying to beam power.

If you're putting a data center in space it's got a payload of of compute or whatever. And what we see in most of the kind of traditional space missions is that that now has become kind of the hardware bottleneck.

It's like how do you actually get the payload you need to solve your ultimate business problem um built in uh in quantities large enough to deploy in a constellation at cost low enough that you can do it. And I think there's a lot of movement in the right direction in that way, but it's uh we're not there yet.

And and the the the other question that I kind of I'll I'll answer that you didn't ask is I think the other big bottleneck outside of hardware is still software. So like you know we we spent there's a lot of talk right now on hardware supply chains across a lot of industries including space.

Um I think there's paths to address that. What is still very true in space is that these are very complex autonomous robots with global networks that are having to communicate um in in inside communication outside of communication and software is really the glue that holds all that together.

And in a lot of ways the integr integration of all these things and the software integration the hardware integration is still the hardest part. It's really making these large complex systems work as a whole. And I really think that's a place that there's not being enough emphasis put in the industry right now.

It's something that we're really trying very hard to solve with kind of the the kind of core hardware software stack that we're building.

So that instead of you know for every new use case you're trying to solve in space, you're going and from the beginning having to do this crappy new integration of a bunch of parts of hardware and software that don't actually aren't actually designed to talk together.

You have a platform that much like a data center rack today everything from the lowest levels of hardware to the high level application software is designed to interoperate. It's designed to be pluggable.

And so you know when Google's putting data centers in the data center they can go take a bunch of stuff off the shelf. It does every rack doesn't look the same but they have a bunch of parts that are interchangeable. They can go throw that in plug them in the software works the hardware works etc.

And we've got to get to that with space too. Yeah. Yeah, I mean we've done that a ton in data center like cat like cat 5 cables, cat 6 cables like these are all standardized. Even Facebook open open source their their blade design for their rack mounts. Um and there's a lot of other uh you know ecosystem tools.

Um last question from me. Um are you tracking the SpaceX Starship progress? Um there was news today that Elon Musk is winding down his relationship with the government as a special government employee. That was a 130day mission.

Um, probably spending a little fewer nights at Mara Lago, more nights in the SpaceX factories. Um, are you are you tracking that? Are you excited about that? And I guess the big meta question is Elon has always seemed interested in getting to Mars, but he's willing to go to Mara Lago if that's what gets him to Mars.

He's willing to go to Starbase, Texas if that's what get him to Mars. So, it feels like this might be a moment where he's shifting his focus. The mission's the same, but he's shifting his focus from uh from regulation to engineering challenges. Mars is back on the menu, boys. Mars is back on the menu.

But what what has your take been on the on the recent star uh uh Starship or SpaceX news? Yeah, I mean, I guess I would start by saying, look, it you know, building rockets is really [ __ ] hard. Like, I mean, I think anybody that says it's not is crazy. It is rocket science.

I think we we've almost been lulled in complacency by how successful SpaceX has made it and like how how easy they make it look. But I mean Starship is a crazy complex complex rocket unlike any that's ever been built.

So like at some level I feel like these setbacks should be expected and people shouldn't act so surprised.

I I think there's I don't have enough inside information to know if like forward progress is being made or sideway pro progress there's more focus whatever but like this stuff's really hard so I don't think it's that surprising that there's setbacks. Mhm.

Um, you know, I think that the the key thing for the launch unit economics, which I think ultimately, you know, Starship, we're hope we're all we're all hoping it's going to get us another order of magnitude on unit economics. So much of of it is not even about scale or size or technical performance or capability.

It's about launch cadence. It's about rate. It's about flying. It's like airlines, right? If you bought a 737 and flew it five times a year, nobody would be able to afford to fly. But because it flies 10 times a day, all of a sudden that the amorization of that fixed asset over those flights makes a ton of sense.

So I think for Starship to really go, what we need is we need applications that require Starship like Starlink has for Falcon 9 that drive us to launch it three, four, five times a day like you fly 737s. And I think if that happens like it it will we will get another order of magnitude in the unit economics.

And I mean maybe tying the loop back to the original question, you know, you start thinking about applications like deploying data centers in space where you're putting thousands of things that are each many many tons each into space and there's like a really core first order economic driver requiring that needing that to scale AI training and these type of workflows.

That's the kind of thing that I think could drive the demand that you need for something like Starship to really get another another sort of order of magnitude unit economics. So, I don't know if I really answered your question, but that's the way I think about it. No, man. Makes sense.

Uh, well, thank you so much for stopping by. This is great. Come back on again soon. Yeah. Yeah, you bet. It was great talking to you guys. We'll talk to you soon. Cheers. Bye. Uh, and while we're waiting for our next guest, let me tell you about Linear.

Linear is a purpose-built tool for planning and building products, meet the system for modern software development, streamline issues, projects, and product road maps. And I've been lucky to use for over a decade now. Over a decade on linear. That's crazy. I'm surprised the company's even decade old.

Seems seems No, sorry. Half a decade. Half a decade. Because a decade ago, you were in middle. I was about to say the better part of a decade. Yeah. Yeah. Okay. Uh well, we have our next guest. David from Retool is coming in the studio. I'm very