Inside Impulse Space's Factory with Founder Tom Mueller (Full Tour)

[00:00] I was a founding employee at SpaceX. I led the development of the propulsion systems for Falcon and Dragon and also started the origins of what became Starship. My proudest development was the Merlin engine which is currently flying on Falcon 9, the most reliable rocket engine ever developed, and also the highest thrust weight of any rocket engine ever developed. I worked on Starship for the last six years at SpaceX so when I left I had the plan of launch mostly being solved or being solved. Like the next big opportunity is to move all that payload all that cargo around in space. [00:30] space transportation as opposed to from Earth to space transportation, which was SpaceX. We feel like the next step is in space transportation, including to the moon and landers on the moon. [00:49] Tom Mueller, welcome to Sorcery. Thank you, great to be here. [00:52] Okay, so before we start, I believe we're in a very special part of the office. What are we standing on right now? [01:00] Oh, you mean right here? Yeah. [01:03] - This used to be my parking spot. - This was your parking spot? - But now we moved to new offices. Now I lost my parking spot. - No. Okay, so I wanna know, [01:13] And we'll get into impulse in a second, but could you name every car model that you own? [01:18] Oh, I have about... [01:19] Yes. [01:20] five 911s from various models, RSR, three GT3 RSs, [01:28] to our race cars

[01:30] and '64, like a first year. [01:34] How many was that? It's like more than five. [01:37] two Ferraris, a F40 and a [01:42] - Four, five, eight. [01:44] A Lotus, '66 Lotus Elan. [01:48] Um... [01:49] three 9-6-2s, one in pieces being put back together. [01:53] Porsche 962 race cars. [01:57] Two Tundra trucks. [01:59] No, our two Raptor trucks. [02:01] Used to have a Tundra. One's in Idaho. And a whole bunch of dirt bikes. [02:07] I probably missed a car too. Oh my the one that used to park here my Taycan which is probably going to get lemon laud right now because it's broke for the third time. Love that car except for the braking part. Well I think you're properly warmed up so we're in Impulse right now. You have another facility here in LA but right now I'd love for you to walk us through and tour the [02:30] impulse for us. [02:31] Okay. Okay. So this is the [02:34] Avionics clean assembly area. [02:37] So this is where basically all the avionics that go on our spacecraft get assembled and tested. You can see there's a vault right there that's getting ready to go into maybe thermal back or [02:47] Electrical test, something going on with it. [02:51] So this is set up for high production rate for Mira and the other spacecraft. Let's go talk about Mira. Okay. So you also in your garage make your own rockets.

[03:03] I do. I have been doing that for a long time. Yeah, in fact, that's how I met Elon was through the amateur rocket stuff that I was doing. Yeah. [03:11] But I like to make prototypes for, you know, just to get some pre-production data on things that we're doing here. [03:18] Done a few things in my garage. Yeah. One of the things that I learned from our interview with Eric and research ahead of time was, [03:25] how fast Impulse has for iteration and testing and development. Yeah. It's really core to the business and vertical integration. Yeah. Being extremely vertically integrated. [03:36] I have an [03:37] you know, the machine shop right here, you know, with the assembly areas. [03:42] and just really rapid [03:43] rapid prototyping test area out back so we can build, assemble, test and iterate. Mm-hmm. [03:50] And so as we walk through, what are we going to see today? Could you just give us a little run through? Yeah, the assembly area out here where we'll see spacecraft like these being assembled, and then our next product which is Helios which is right over there in that area being assembled. [04:05] Then we'll go back to the machine shop where the actual parts are made. [04:09] and then in the back is the propulsion assembly area where the propulsion products are put together [04:15] and then tested back there in the test area. We're going to see a rocket engine test today back there. And yeah, should be pretty cool. - Amazing. - All right. Well, starting here, this is the Mira. So this is a... [04:28] This is our mirror prototype. So this is one we can have out here that you can actually touch. So

[04:33] Mira is up. [04:34] highly propulsive orbital transfer vehicle. So it's got these these thrusters actually, these these [04:41] thrusters two on each corner so it's got eight of these thrusters. They're about six pounds of thrust each so total of almost 50 pounds of thrust. [04:48] which for a small spacecraft like this is a lot. That's good enough to move around in space very fast. [04:53] Inside the chassis we've got three propellant tanks, two oxidizer tanks, and a fuel tank in the middle that feeds these thrusters. [05:01] It's got the solar panels which unfurl and they're steerable, so you can point the spacecraft wherever you want and still have the power positive pointing at the sun with the solar. [05:11] The avionics that we just talked about being assembled in there are all in this box down here, we call it the vault. [05:15] We keep them all in one box. And it's got, these are reaction wheels, which are basically flywheels with a motor that you can actually very precisely steer the spacecraft with. [05:27] And then there's [05:29] a bunch of antennas and stuff that go on this, which you can see on the flight vehicle in there. So, [05:34] They've got red hats on them, red removed before flight covers on them. But those are antennas so we can talk to the spacecraft. [05:40] It's got the multi-layer insulation or MLI, that gold foil. [05:45] that thermally protects it. [05:48] keeps it, it kind of makes it [05:51] like a thermos bottle. It's really, when that gets into vacuum of space, it acts like a thermos bottle and really thermally isolates the spacecraft. [05:59] Here's some flight solar panels right here that are assembled and ready to integrate onto the spacecraft.

[06:05] Thank you. [06:06] So we've got, I think, three spacecraft being built up in there right now in various states of assembly. [06:12] We can see the one back there if we walk this way. [06:16] And then we can talk about Helios. [06:18] Yeah, here's another one that's [06:21] Oh, wow. [06:23] Got a team working on it. Yeah, wow. [06:28] This I believe is the next one to go up here. It's going up later this fall. [06:32] So you have three in orbit? We have three on orbit right now. We did a rendezvous last year between two of them. Got within [06:38] 1200 meters of each other, which is [06:39] Really damn close in space. Yeah, got some cool pictures. [06:43] Yeah. [06:43] We know right where it's at. They're all alive. We have GPS, so we know where they're at, what their state is. [06:50] Yeah. [06:51] So, [06:53] Mira is our space-- [06:55] a maneuvering vehicle that can stay up there for five years, but then we have a vehicle to get you there and that's this. This, [07:01] This is Helios. [07:04] So Helios is a basically a tank of propellant that [07:08] That tank right there holds 12 tons. [07:11] of liquid oxygen and liquid methane. [07:14] and then we have a very high performance engine that goes on the bottom of it [07:18] that you'll see back there. [07:20] that provides 15,000 pounds of thrust. [07:23] This can take [07:25] a large satellite, like a four-ton satellite, [07:27] from low earth orbit at a couple hundred miles up, [07:30] to geosynchronous orbit at [07:32] at 22,000 miles. [07:34] So basically escaping most of the gravity well of her.

[07:39] very high energy transfer. [07:42] It can take [07:43] You know, I'll [07:44] five tons of payload to the moon on a Falcon 9. [07:48] Um, [07:49] He can [07:50] improve your payload to Mars by up to a factor of five. [07:55] um yeah just it basically adds a third stage to a two-stage rocket. [08:00] For those who don't know your origin story for Impulse and why you decided to go after this after SpaceX, could you share more on that? [08:08] uh sure yeah i i was um a founding employee at spacex i developed i led the development [08:15] of the [08:16] the propulsion systems for Falcon and Dragon, and also started the origins of what became Starship. [08:23] Probably my proudest development was the Merlin engine which is currently flying on Falcon 9. [08:29] the most reliable rocket engine ever developed. [08:32] and also the highest thrust await of any rocket engine ever developed. [08:35] But I worked on Starship for the last six years that, that, [08:39] SpaceX. [08:41] So when I left, [08:42] I had the plan of, you know, [08:44] launch mostly being solved or is being solved it's like the next big opportunity is to move all that [08:50] all that payload all that cargo around in space so i started impulse space to do [08:54] in space transportation as opposed to from earth to space transportation which was spacex we feel like [09:01] The next step is in-space transportation. [09:03] including two [09:05] to the moon and landers on the moon, which we're also bidding on.

[09:09] I'm going to the moon. [09:10] Going to the moon, I'm all about the moon. Yeah. And so one of your engines is over here. Yeah, let's go look at it. [09:17] So these are domes that go on... these are tent domes. They're not alien spacecrafts. [09:22] They're not alien spacecrafts? No. They look like flying saucers don't they? Yeah. [09:28] So three of those are in that tank to make up the liquid oxygen and the methane tank. [09:33] Yeah, so this is the Deneb engine. Deneb is the name of a star. [09:37] Rigel, another engine that we're developing that you'll see back here is the name of a star. We named all the little the little engines that you see on Mira. [09:44] are safe. Those are really cute. S-A-I-P-H which is also a star in Orion. So we name our engines after stars. So this is Deneb. [09:55] which is a which is a blue supergiant star. [09:58] And this is our [09:59] our supergiant engine. This engine makes about 15,000 pounds of thrust. It's pump fed, so it's got the turbo pump there on the back. [10:07] and it is it's aux rich stage combustion so [10:11] All of the liquid oxygen off the pump, [10:13] comes out of this line, there's a valve missing here, but through the through the main aux valve into the pre burner. [10:18] where it gets burned with a little bit of the fuel, liquid methane. So we burn all the oxygen, a little bit of fuel, which makes makes hot oxygen. [10:26] at about 4000 psi, which [10:28] which goes into that housing, drives the turbine, [10:32] which provides [10:33] 1500 horsepower to run the turbo pumps. [10:36] Then the hot oxygen off of the turbine comes across this transfer

[10:41] tube and into the main injector. The fuel, other than the little bit that goes into the preburner, goes through this main valve. [10:48] into a bunch of cooling channels inside the combustion chamber to cool the combustion chamber. [10:54] and then goes in as hot gas into the injector and gets burned with the oxygen. So, [10:58] That's how the cycle works, Oxrich stage combustion. [11:02] uh [11:03] This engine [11:05] will probably be the highest performing hydrocarbon engine ever flown because it's it has a very high area ratio the weight the way you get the way you get ISP [11:14] You have very high mixing efficiency, which this is designed to be a very, very good mixing gas gas main injection, very fine injector. [11:22] and then it has a huge skirt. When you see the engine back there, it's got a big carbon skirt on it. So this is only half the engine by height. [11:28] It'll have a carbon skirt that'll glow white hot [11:32] that gives it area ratio, that nozzle converts pressure into velocity, which is pure performance. So the more expansion nozzle you can have, the more performance you get. [11:42] What is the development process like been with this? I know you have land in Mojave too, so do you have to do something special when you're testing it there? A lot special, the uh difficult engines. Oxrich stage combustion is a difficult um [11:56] very difficult development. [11:58] So we're right now we're finishing up the turbo pump development and just just starting to get into engine testing so we're just getting there. [12:06] Yeah. [12:07] exciting. [12:08] How long have you guys had the land in Mojave?

[12:12] Couple of years now. [12:13] We started with Rigel there, a little 200 pound thruster, and then moved onto this [12:19] uh, [12:21] in the last year. [12:22] What's your favorite part? What's my favorite part of this thing? The main director. I'm pretty proud of the main director. It's... [12:29] Very fine mixing. I think we're going to find out it's going to [12:33] That's going to be a [12:35] real high-class engine. We've got the design team, the test team, the manufacturing team, [12:43] Uh... [12:44] We've got the valve team, it's a bunch of pieces that come together. [12:49] There's even an avionics box, there's an engine controller, so there's even avionics contribution to this. [12:55] Yeah. [12:57] But this is the heart of the stage. And this engine will probably become our prime mover. [13:03] Very high performance, high thrust engine. [13:06] Imagine if we built a bigger version of this with multiple engines or say in the farther future, say a human lander. [13:13] would have several of these engines to land on the moon. [13:17] You need something like this to do these big payloads to [13:21] you know, to space, so. [13:23] This is going to be our [13:24] our prime mover. [13:25] So should we see how it's built? [13:28] Yeah. [13:30] It's all, in fact, while we're here, it's mostly 3D printed and you'll see... 3D printed? Yeah. Really? It's 3D printed right here. [13:37] Wow. Yeah. So I'll show you the 3D printed part. It's actually easier to show you the parts that aren't 3D printed. Okay. This part right here is hydroformed. This part right here is machined.

[13:49] Everything else is 3D printed. [13:51] So this part, this part, this part, all these parts, [13:55] and this part and this part are all 3D printed right back here. This is the igniter that lights it, 3D printed. When did you get into 3D printing? Is this just before? As soon as we started the company within the first year we bought it. [14:06] Really? Yeah. [14:07] When did you? The little thrusters on there are completely 3D printed. You'll see them back here. I heard you make little party cups out of those ones. Yeah, yeah, yeah. They make little shot glasses. [14:19] That's amazing. I'd love to get into more of the knowledge that you've built along the way. So you prefer 3D printing. What are some other aspects that are core? Let's talk about 3D printing a [14:37] It's so easy to make a high performance rocket engine with 3D printing. [14:42] because the rocket injectors and the cooling passages are pretty [14:46] small and complex, [14:48] So to make a rocket engine by machining and assembling and welding and brazing is really difficult. [14:55] With 3D printing, [14:56] You just... [14:57] basically draw [14:59] the fluid passages you want and then put a wall around it and just print it. It's whatever you draw you know within within reason you can make so it just makes making rock engines really easy. [15:09] But we still have to do a lot of machining. [15:11] But before we get to the machine shop, I want to talk about this. [15:16] These are composite overwrapped pressure vessels, or COPVs.

[15:21] So these are the lightest way to make a propellant tank or a gas tank, a high pressure tank. [15:27] Because we wrapped the carbon, which you can see happening on that machine back there, [15:32] the carbon fiber is extremely strong. So we have an aluminum liner that seals the propellant in, [15:37] and then a carbon overwrap that holds the pressure. [15:42] So a properly designed COPE [15:45] will be about half the mass of a state-of-the-art all-metallic tank, like a titanium spacecraft sphere. The carbon's so much stronger it'll actually be about half the weight. So we brought this in-house early on too. [15:57] along with 3D printing. [15:58] just because a major portion of the spacecraft mass is the tanks for the propellant. [16:03] and we want to be able to control the [16:05] You know. [16:06] the quality and the quantity and the cost. [16:08] So here they are, we're making, these are tanks for the Mira spacecraft. The bigger one there is a tank for VAST space station. We provided the whole propulsion system for VAST, including the thrusters. And then the one that's getting wrapped there, it will be a very high pressure, like a 5,000 psi helium. [16:26] tank for for the um [16:29] Helios VK. [16:30] to pressurize the tanks. [16:32] so [16:33] Everything from 1,000 psi for the spacecraft up to 5,000 psi. [16:38] for all. [16:39] for the helium tank. [16:41] Wow. [16:42] These things are really hard to get from [16:45] typically from industry, like if you wanted to buy one of these it might take [16:48] After you order it might take a year or two to get them

[16:51] and be hundreds of thousands of dollars a piece. So how do you find the talent to build them? Um, we learned like we mostly mostly it was guys that hadn't done it before. [17:04] so we bought the winder and learned and then we've since hired people have run the winders before and stuff and some people have [17:10] carbon experience, but we mostly [17:13] learned. We had a consultant come in and show us how to design properly. [17:17] and a lot of cut and try. You saw the one that we burst back there? Yeah. Yeah, that's, so you make one, you blow it up, find out how you did and then, you [17:27] design from there. [17:28] Where does software come into the development piece for you? I know like I would even love to [17:35] learn about the long tail of that. I mean, clearly you've seen a couple eras of space. [17:41] Well... [17:42] In 2026, you can't [17:44] you can't have modern [17:45] hardware without software so [17:47] The spacecraft has, you know, has autonomous software on it. [17:52] um [17:53] We use you know, we use AI for coding and for design. [17:57] Um, [17:58] We use pretty powerful generative software. [18:02] for 3d printing so you know everything you know like the machines here all are running on uh [18:09] software to [18:10] you know, to do the machining. So. [18:12] A lot of software. Yeah. [18:15] Big, big IT department. [18:17] Okay, so what are we going to see through here? I know there's one machine we're not allowed to see, but... Yeah. Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three venture-backed startups in the U.S. Nearly 40% of startups fail because they run out of cash. Brex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, chipping away at it with fees, Brex is designed to help you spend smarter and move faster.

[18:47] combines checking, treasury, and FDIC protection into one powerful account. You can send and receive money globally at lightning speeds, get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one. With same day and even same hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, Service Titan, HIMSS, Anthropic, [19:17] That's B-R-E-X dot com slash sorcery. [19:47] Visit Turing.com slash S-O-U-R-C-E-R-Y. [19:52] Let's see what we got on here. There's a turbo pump part. [19:55] for the Deneb engine. [19:57] So this is what's called a mill turn. [20:00] But that table can spin fast and you can actually [20:03] run it like a lathe where you can just single point, [20:06] turn the bore, [20:07] Which I believe he's doing now. Yeah, he's putting a boring bar on it to turn the bore. [20:12] That way you can turn all the... [20:15] all the different bores in there, the different diameters,

[20:18] at the same time and have a very [20:20] precise and concentric. [20:22] machine. [20:23] Yeah, see that's the boring bar there. [20:26] So it doesn't spin, it just holds stationary and the part turns. [20:29] And that's a lead, basically. [20:30] Then on the same setup, it's a milling machine. So, [20:34] They'll replace that boring bar with a milling cutter. [20:37] and then it's a five-axis milling machine so they can go and they can cut the ports and the bolt holes and the ceiling surface and everything. [20:43] So in one setup, you can machine [20:45] you know, like three quarters of that part. Then you flip it over and hit the backside and it's done. [20:50] So really, [20:51] You've got two of these machines and another one similar back there, so a lot of these machines, because they're very, very handy for making parts like this. [20:59] I think we're going to go see a copper liner back there. Yeah. [21:05] Yeah, the the. [21:07] The inside of the dend of engine, the combustion part is made of copper, [21:12] just like the bottom of your cooking pans is. [21:15] because the heat transfer inside the combustion chamber is so intense, [21:21] that only copper can get the heat out without melting. [21:25] So we start with big forgings of copper like this right here. Wow. [21:29] and then they go on this lathe right here. [21:32] to turn them down, to thin them out, to make them very thin. [21:36] And then they [21:38] Can't see that machine. Then they go on that machine over there and they get slotted. [21:44] which Lee is gonna stop them from filming. - Yeah, we can see that. - So we start with these foraging that weigh about 700 pounds,

[21:50] and the final slotted [21:52] uh, [21:53] product weighs about 25 pounds. [21:55] Got some pretty strong employees. [21:59] And there's all the chips that we end up generating. [22:02] Oh. [22:03] Those are pretty. And then we have a big milling machine here. [22:09] to make big parts like that. [22:11] The Mira chassis, all that aluminum on the Mira is made right here on this machine. So this is our big mill. This is huge. How did you get this in here? You can machine a motorcycle on this thing. [22:22] Oh my God. Yeah, it's huge. [22:27] Did we see the 3D printers yet? Or are we gonna... We're right there. Okay. [22:31] Yeah. [22:31] So see this part right here? This is the payload deck for Mira. [22:37] So we call it the pegboard. It's got all these bolt holes so you can bolt anything to it. [22:40] That's done on that mill right there. [22:42] How did you get to this design? Was this original or did you have to go original design? This this, [22:50] This is like our sort of like for a development mirror, like our demo flights where you can mount anything. Like you have a whole bunch of hosted payloads. [22:58] Eh. [23:00] If we have a specific mission where they just have one payload in the middle, we actually redesigned this to get rid of some mass. [23:05] Okay. So it can be custom fit to whatever your payload is. This is the one size fits all, you can put anything on it. [23:12] Bye. [23:13] Payload Deck. [23:14] Okay, let's talk about 3D printing. [23:17] So this is a 3D printing lab. [23:19] So we've got three

[23:22] 3D metal printers in there. Wow. This machine right here prints, uh, Inconel, which is a nickel alloy, pretty common nickel alloy. [23:30] And it prints the little safe thrusters, you can see them there. [23:33] In fact, over here is an even better one. [23:37] Right here are the safes as they print on a plate. [23:40] and then we cut them off and machine them and add the igniters and the valves and everything. [23:45] The all 3D printed, the injector, the combustion chamber, everything's all printed in one piece. [23:51] And then these two black printers here, these are Velo's and these are printing [23:56] a special alloy that's like ink canel but it's burn resistant in the hot oxygen in that site in that cycle that we have on the denivator so all the turbo pump parts [24:05] the main injector, [24:07] and anything that sees the hot oxygen is printed with this ally called noveli. [24:12] which is has the strength of high strength of [24:15] uh bond [24:16] in canal but is very burn resistant so proprietary alloy. [24:21] I would say casting could be faster, but... [24:24] You can't [24:26] you can't cast really tiny fine passages like the injector on Deneb has you know hundreds of elements that are really really small [24:34] and you can't cast that. [24:35] and casting does not have the strength that a 3D printed part has. The casting has big grains and [24:40] Yeah, it's just [24:42] It'd be really hard to make these parts of the casting. [24:44] My castings are fast. [24:46] now. [24:46] like, [24:47] raptors cast because they're big parts they don't have a lot of fine [24:51] They print some of the fine stuff, but the big parts that they've

[24:54] They're calcium. [24:56] What was it like to see the iterations of Raptor over time to get to the latest version, which just looks a lot. It was amazing. That one chart is so great. It was amazing. And I always said it took three versions of Merlin to get it. [25:13] to be really tight. [25:14] and it took three versions also of Raptor. I always say it takes really three, three iterations to get to a really, really tight product and Raptor's there. Yeah. And Merlin's there. [25:25] Now, [25:28] All right, so we're entering the back area. I heard this is the secret area. Yeah. [25:33] *gasp* [25:35] This is shipping and receiving, so boxes everywhere. [25:38] Now we're in the assembly and test area. [25:41] so [25:42] In these clean rooms here, they assemble things like valves. You can see the valves for dinner right here. [25:48] That's the ball valve right there. You can see the ball. Oh yeah. [25:52] There's a pre-valve. [25:55] some actuators, [25:57] So these all get assembled in this room. [26:00] Here's the thrust chamber assembly. [26:03] So that that copper piece that you saw back there is inside there is the inside of this and then the outside is is this inconel jacket that gets brazed to it. [26:12] So the copper, basically the high parts except for the channels of the copper are stuck [26:19] You're braised to the... [26:21] to the outer shell. [26:23] And then that forms grooves, you know, channels that the fuel flows through to cool it.

[26:28] So that fuel will come into this combustion chamber at [26:31] you know, minus 250 degrees F as liquid. [26:35] liquid methane and it'll go into the main injector at plus 500 degree gas. [26:39] could have picked up so much heat from the combustion. [26:42] And here's the whole den of the engine. So like I said, you've got this big, [26:46] carbon ceramic skirt. [26:48] This is not carbon ceramic, this is carbon ceramic. This is actually carbon epoxy, just a mock-up. [26:55] But the final part, it'll look like the brakes on my car, you know, that's that carbon ceramic material that can run red hot. [27:02] Actually white hot. This thing will run at over 3000 degrees F. So it'll be probably the hottest [27:08] skirt, your nozzle ever flown on a [27:11] This thing's gonna break a few records, probably the highest performance and the hottest [27:15] nozzle that's ever been flown in space. [27:18] Have you thought about putting one of these on your car? [27:21] That'd be awesome. [27:24] That would be awesome. [27:26] 15,000 pounds of thrust on a [27:28] 2000 pound car. Yeah, that would get up and move out. [27:31] Thank you. [27:32] All right, let's [27:34] Let's see ya. Let's see ya. [27:36] A rocket firing. [27:38] Are you guys ready to do a firing? [27:40] Yeah. [27:41] Ready when you're ready. [27:42] All right, let's come over and I'll show you what we've got. [27:47] This is one of the [27:50] one of the safes, the six pound thrusters that we use on Mira. [27:54] It's in a vacuum chamber and it's sitting right now at... [27:58] at about 6 tor, atmosphere 760 tor, so we're

[28:02] six, seven sixties of atmosphere. So it's got a pretty good vacuum in there. [28:06] And when they give us the count, [28:08] You won't hear anything because it's in vacuum. [28:10] but you're going to see a flash in that gap between the nozzle and that pipe. [28:15] and then it's going to start glowing. [28:16] Give us a count when you're going. [28:19] 30 Seatwick. [28:20] Pop firing five. [28:22] four, [28:24] Three. [28:25] dude. [28:26] Why? [28:30] I didn't even see the flash that time. [28:32] Yeah. [28:34] How crazy that you can sit this close to a running rocket engine. Yeah. [28:42] So the way it's designed, the part that's glowing [28:45] Um, [28:46] Red hot. [28:48] is uncooled nozzle just like that carbon skirt will be. That's called radiation cool. It just radiates the heat out to space. [28:54] The forward part is cooled by the propellant. You see that pipe on the top that comes down to the nozzle? That provides the oxygen, nitrous oxide, [29:02] and it flows through little channels to cool it. So that part runs, it's still red hot, barely glowing. [29:08] But it's much cooler. [29:10] And it's carrying the combustion chamber pressure too. [29:15] it has more stress on it so it's running at a lower temperature where it can actually take the stress. [29:19] The skirt is that's so hot it doesn't have very much strength, but there's no pressure load on that skirt. There's no stress on it other than thermal. [29:27] So it actually worked out pretty good, the 3D printing. [29:31] and the cooling scheme made this a pretty efficient little tiny thruster.

[29:36] What was it like when you tested the first one? [29:39] It worked. Really? Well, the first, the absolute first one didn't, but we just, and it was a machine one, it wasn't even printed. [29:46] The one that I machined [29:49] on our lathe here. [29:51] And after a couple tries, we tried a few things and we got to light and we were super excited. [29:55] And that wasn't in we didn't have a vacuum chamber that it was out. [29:58] It was sitting out in the shops. It was loud. We were wearing hearing protection. [30:01] Goggles. [30:03] Then we got this vacuum chamber and we did all the development qualification and this is applied acceptance test. [30:08] of them so hundreds of these have now already been through tests on this on this stand. [30:14] So we're measuring on that. [30:15] You see it's mounted on that pedestal right there. There's a wire, the black wire come up the back. That's a load cell, so that's measuring the thrust. [30:23] And then in the propellant feed, we're measuring the flow rates. [30:27] The thrust divided by the flow rate is the ISP. So we're measuring the ISP, which is the performance metric of the engine. [30:34] This engine has about a 290 second ISP. [30:38] whereas the Deneb engine will have almost [30:42] almost 400 second isp about 285 or three i'm sorry 385 second isp [30:47] how are you like how are you measuring all the data off of this i'm assuming this just [30:52] fuse off data. Oh yeah. [30:54] Yeah, you can see the thermocouple, the yellow things are thermocouples, the [30:59] There's little pressure transduce, we're measuring three pressures, a bunch of temperatures, the flow rates, the thrust. [31:05] and then all the temperature, these are test stand temperatures, we're measuring temperatures on test stands.

[31:11] probably 100 channels of data that's being recorded. [31:15] And then they look at the data after every run. [31:19] Just make sure it's running in family. [31:23] He's a little longer on three minutes. [31:25] It's getting really far. So vast there now. [31:29] Now it shut off and now it's going to do some pulses. [31:32] Oh, wow. [31:36] So as you can see, this engine can do anything from running steady state to pulsing. [31:41] And on the Mira spacecraft, the way we steer it is all [31:45] all eight engines light, [31:47] and then to steer we off pulse. [31:50] So if we pulse one at like, [31:52] say 80 duty cycle it'll have about 80 thrust and that'll that'll allow it to steer so [31:57] They'll be pulse, some will be pulsing, some will be running steady state to steer the spacecraft. [32:01] That's kind of how we do it. [32:03] So it's it's. [32:05] Instead of using continuous throttle, adjusting continuous throttle like Dendip can do, [32:09] This is called pulse whip modulation. [32:11] to set an equivalent for us. [32:13] Have you had to maneuver across any asteroids? [32:17] Not asteroids, but we've had to do collision avoidance on the first spacecraft. No. Wow. [32:25] Like in the first, I think within the first three or four months, we had to do a collision avoidance. How did you get warnings of that? What was it? Air Force tracks everything and you get warnings. There's actually some commercial companies like Leo Labs tracks things and they just tell you, hey, you've got to. [32:40] you know, you've got a potential

[32:42] And that's exciting for you because it's your job. It was kind of... [32:46] It kind of caught us by surprise because, okay, we got to do this maneuver. We haven't done very much yet, but we were able to do it and get out of the way. [32:53] Yeah, pretty cool. [32:54] Yeah. [32:56] And so what is this huge... [32:58] ginormous machine. This is uh that's this is electric propulsion so if you look in there that's a that's an electric propulsion thruster right there. [33:05] That's a Hall Thruster, just like Starlink has. It runs on Krypton. [33:10] And this this chamber [33:13] pulls down, this one will pull down to about one tor. [33:17] This will pull down to 10 to the minus 6, so about 1 millionth of a tor. [33:20] So the vacuum in this chamber is literally one million times lower vacuum than in here. [33:26] because because the [33:28] The electric propulsion needs, the plasma only works in a hard vacuum. [33:32] So this one really, that's why it's got so much piping and machinery to pull that back. What are you going to use electric propulsion for? [33:39] Um, [33:41] for probably for north-south station keeping, when you're in geosynchronous orbit, [33:45] the orbit gets perturbed, [33:49] and you want to take out that those perturbations so it's called it's called north south station keeping. [33:54] So, [33:55] if we want to preserve we want to preserve the the chemical propellant for when [34:00] the customer wants to make a fast move. [34:02] So by using just a small tank of Krypton and a couple of these, [34:06] you can do north south you can do the north south station keeping for five years and not use any of your chemical prop okay so it's yeah

[34:15] Most big geo satellites are all electric propulsion. [34:19] But they're not... [34:21] They're not. [34:22] orbital transfer vehicles. [34:23] We're an oval transfer vehicle, so we need chemical propulsion to move back. [34:26] Yeah. [34:28] What do you think about nuclear? [34:30] Do you think that's going to enter anytime soon? I'm a huge proponent of nuclear. I've, um, [34:34] I've been talking with NASA folks or anybody that would listen about [34:38] how the next step, the future in propulsion is nuclear electrics, I think. Really? Yeah. [34:45] Yeah. [34:46] What do you think the timeline for that will be? Well they're doing that SR-1, the Space Reactor 1 mission to Mars, so it's starting now. [34:53] Yeah. [34:54] I think [34:57] Nuclear electric NEP propulsion is going to be important for going out to the outer planets. [35:04] If you're staying near Earth or even between Earth and Mars, [35:08] probably solar electric is [35:12] is going to outperform it on cost and performance just because [35:17] Solar is now getting very light in space. [35:19] Especially if Elon starts doing these data servers, server the space, they're going to really optimize [35:24] the solar panels to be really light. So I think solar electric is still going to be a major form of propulsion for [35:34] you know, for the [35:35] like Cislunar, [35:36] Um, [35:38] But if you want to go to the outer planets where you get far away from the sun, [35:42] The example I always give is the New Horizons mission where they flew past

[35:46] Pluto and got those awesome [35:48] - Yeah. - Photos of Pluto. Well, that spacecraft was going so fast when it went by Pluto, [35:53] that they only got a few hours of pictures before it was gone, right? It went past. [35:59] with a big nuclear electric stage. [36:02] you could actually get to Pluto faster because you got high performance and you could slow down and enter orbit. [36:06] around Pluto. [36:08] And even if you may be out of propellant at that time, but [36:11] now you still have [36:13] potentially megawatts of power depending on how big your nuclear reactor was. [36:18] to do science so you've got all this power that you can do crazy science things with like you've [36:23] ground penetrating radar or whatever. [36:25] And you can keep it warm because you've got the nuclear heat. So you keep it warm for decades. So you can do science or you could run a very powerful radio for decades. [36:35] Like we know Voyager is they're shutting off experiments because it's so far out and the power is starting to fade. [36:41] with a fission reactor, you could do the same thing. You could run for decades and still have a lot of power to run experiments. So I think it just opens up [36:50] I think it's pretty far in the future. I mean, you know, and [36:53] maybe a few decades away before we're really doing big nuclear stages. Hopefully sooner than that depending on how much funding it gets but [37:02] That's the next step. Today's episode is sponsored by VCX by Fundrise, the public ticker for private tech, allowing investors of all sizes to invest in venture capital. [37:13] Learn more at GetVCX.com.

[37:17] Some of you may not have heard this yet, but our sponsor Public just launched something called Generated Assets, and it brings AI into investing in a way I've honestly never seen before. Here's how it works. You type in an idea like AI-powered supply chain companies with positive free cash flow or defense tech companies growing revenue over 25% year over year. Public's AI then dispatches a swarm of agents that scan every single US stock, evaluates them, and instantly builds a custom [37:47] why each stock is included. And before you invest, you can even backtest your idea against the S&P 500, so you're making decisions with real context, not just guessing. And beyond generated assets, Public lets you invest in stocks, bonds, options, crypto, all in one place. They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com slash sorcery. [38:13] Paid for by Public Investing. Full disclosures in the description. [38:17] Enterprise AI runs on Merge, the AI infra platform for integrations, agent tooling, and model orchestration, so your teams ship product, not plumbing. [38:26] Mistral, Dropbox, and Drada already trust merge in production. [38:30] start building at merge.dev. [38:33] Founders scale faster on Deal. Set up payroll for any country in minutes, hire anyone anywhere, get visas handled fast, and get back to building. Visit deal.com slash sorcery. That's D-E-E-L dot com slash sorcery.

[38:48] What are your thoughts on data centers in space? Data centers in space are a no-brainer. [38:54] Yeah? Yeah. [38:55] I think it's going to take a while before [38:58] It's cost. [38:59] straight up cost competitive with terrestrial, [39:02] But the real cost, [39:04] is [39:05] how soon you can get the power. [39:08] And from what I read, it can take up to seven years right now to [39:12] to get power for a new data center. [39:16] if you can [39:19] build like [39:19] Elon plans to do and get there faster, I think that's going to be the right answer. So I've actually been saying this for a long time that, um, [39:26] Compute will need to move to space at some point. That's my hot take. Really? Is compute needs to move to space. It's [39:33] it makes the most sense of anything to move to space. [39:35] to build megastructures. [39:37] because [39:39] All you need as an input is power. [39:42] and all you have as an output is data. [39:44] So, [39:46] you move it to space, [39:48] You have all the power you would ever need. [39:50] and you transmit that [39:52] that. [39:53] that data back down on a [39:55] on the laser beam. [39:56] terawatt laser beam. [39:58] So it's just it's solved. It's just so simple. [40:00] Compute. [40:01] Power for Compute is growing at [40:04] greater than 15% per year. [40:06] Some people say that it'll be [40:08] equal to all the earth base power right now in about 20 or 30 years. It's just it's gonna it's just it's crushing [40:15] our resource use. [40:16] on earth. [40:18] um so i think it's just important to move it up i'm glad that

[40:21] it's getting moved earlier, not wait until it's crushing us on total power use. [40:28] Are you surprised at how big SpaceX has gotten and especially [40:32] the Starlink business. [40:34] Yeah, I mean, [40:36] It's been amazing to be part of that company and see the growth and the [40:42] Yeah, just the [40:43] incredible like when we first designed starlink we all had spreadsheets and we're looking at the [40:49] Return on. [40:50] investment going. [40:52] we need to make and fly as many as we can. We all saw it. It was really funny to see [40:56] you know, all the naysayers online saying there's no way they're going to be popples, they're going to crush them. We're looking at it the other way, going make them as fast as you can and launch them. This is going to [41:05] You know, [41:05] this is going to hockey stick the stock which it exactly did. [41:09] Were there any secrets to working with Elon and... [41:13] handling his Elon-ness. Tons. Yeah. [41:19] - Oh, you mean? [41:21] Elon's great to work with. Generally he [41:24] He just energizes you. He's got this [41:27] infectious enthusiasm and get you just hyped up to go do crazy stuff and and we do. [41:33] And [41:34] You know, he's [41:35] Thank you. [41:36] It's all about urgency and getting it done. And he really gets the best out of everybody. [41:42] and that's why that company is so incredible. [41:45] He's also really good at finding and recruiting talent. [41:49] The key to success is having a great team.

[41:53] And we spend a lot of time to making sure we get good talent. This is, you know, we're basically we're kind of the same plan, vertically integrated, [42:02] hire the best, and move fast. [42:04] And that's what we learned at SpaceX. [42:06] If you didn't have your role as CEO, [42:10] Which job of all these jobs would you want? [42:16] CTO. I am CTO actually. I'm a better fit in CTO. I like designing the rocket engines and the avionics and the radios and the [42:27] uh, the, [42:28] telescopes and yeah. [42:30] Okay. I'm very technical. Well, cool. As we wrap up, any last remarks, anything you're super excited about? [42:37] I'm super excited about the moon. You know, NASA is [42:41] you know, [42:43] Jared Isaacman said, we're going to build a base on the moon and we're going to stay there. And that's [42:47] That's right on, man. That's exactly what I want to do. [42:51] I think the moon is more important than Mars in the near term. [42:54] and I think the same thing I was talking about earlier using the resources of space to offload the crush of resource on earth like people are predicting we're going to run out of copper here in the next few years. We're just [43:06] data centers require so much copper we need to start getting copper from the moon uh from the asteroids. [43:11] there's an almost infinite supply [43:14] in our solar system of material, enough to supply us for thousands of years. We just got to get ready to go get it. [43:20] Well, are you interested in photonic chips if copper is becoming the strain? I think we're going to have to flip to something else. I think it's going to drive that technology. Yeah.

[43:30] Okay. Well, thank you so much. Thank you. [43:34] Great talking. Cool. [43:35] Hey, it's Molly. If you enjoy our interviews, check out our newsletter, Sorcery.vc, where we deliver a once a week top deals and tech headlines email and also go deeper on our podcast interviews. Subscribe to Sorcery today. And don't forget to subscribe to the podcast on YouTube, Spotify, Apple or wherever you listen. Link in description to sign up.