Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Economics of Orbital vs. Terrestrial Data Centers

andrewmccalip.com

Post reply on HN
Print view View original

121–130 of 282 posts

Re: Economics of Orbital vs. Terrestrial Data Centers

#121

Post by enderfusion »

wmf wrote: post #73
xnx wrote: post #50

François Chollet https://x.com/fchollet/status/1999982683708150014 : "Datacenters in space" make for a catchy narrative and an interesting demo, but the math simply doesn't work. When considering factors like launch cost, maintenance complexity, and the cost of high-bandwidth communications (latency included), there is no realistic set of economic and engineering assumptions under which orbiting datacenters become co…

So the guy who shows his work says 3.4x and the guy who doesn't says 50-100x.

OP here.

man who quotes big number - often made a fool

Re: Economics of Orbital vs. Terrestrial Data Centers

#122

Post by enderfusion »

SyzygyRhythm wrote: post #33
zokier wrote: post #12

tbh this feels lot like people throwing Drake equation around. You put in whatever random numbers together and you can get any result you want.

We know the upper bound for most of those numbers. SpaceX already achieves internal marginal launch costs of ~$1000/kg, for instance. We know their rough costs per satellite. In contrast, we know little to nothing about the inputs to the Drake equation. The numbers don't quite work out in favor of orbital datacenters at the current values. But we can tell from analyses like this what has to change to get there.

OP here. Thanks SyzygyRhythm

Re: Economics of Orbital vs. Terrestrial Data Centers

#123

Post by Figs »

ViscountPenguin wrote: post #112
compiler-guy wrote: post #27

Earlier quoted context omitted.

Unregulated casinos don't need this kind of compute. Most could be run on a decent raspberry pi or two. At any rate, one basic communication's satellite worth of compute would be more than enough. No need for TPUs.

A cube-sat casino is an interesting concept for sure.

Hmm. Has anyone ever flown a pirate radio blimp?

Re: Economics of Orbital vs. Terrestrial Data Centers

#124

Post by inejge »

JumpCrisscross wrote: post #43
zozbot234 wrote: post #19

Will these space-based data centers run on rad-hard silicon (which is dog slow compared to anything on Earth) or just silently accept wrong results, hardware lockups and permanent failure due to the harsh space environment? Will they cool that hardware with special über-expensive high-temperature Peltiers that heat the radiators up to visible incandescence so that the heat can be shed with any efficiency? There's zil…

At this scale could you do shielding?

There are ways in which shielding in space can do harm: really energetic particles get trapped and produce a shower of daughter particles and rays over a greater area. So you'd need even more shielding. Or you accept that such things will happen and use rad-hard parts, redundancy etc. When you have the whole atmosphere above, it's much less of a concern.

Besides, that's even more mass to be lofted. Pushing the economics further into the ludicrous end.

Re: Economics of Orbital vs. Terrestrial Data Centers

#125

Post by rekenaut »

wmf wrote: post #81
cmgbhm wrote: post #59

Earlier quoted context omitted.

Doing like an 8xh200 server ( https://docs.nvidia.com/dgx/dgxh100-user-guide/introduction-... ) is 10.2kW. Let’s say you need 50m^2 solar panels to run it, then just a ton of surface area to dissipate. I’d love to be proven wrong but space data centers just seem like large 2d impact targets.

Yeah, you need 50m^2 of solar panels and 50m^2 of radiators. I don't see why one is that much more difficult than the other.

Especially if with the radiators you can just roll out as rolls of aluminum foil, which is very light and very cheap.

Re: Economics of Orbital vs. Terrestrial Data Centers

#126

Post by anonymousiam »

This is an interesting analysis, and I like the sliders that let you instantly show the impacts of system trades.

The one glaring hole that I see is the challenge of moving the data to/from the datacenter while it's on orbit. Bandwidth to/from space isn't free. FCC/ITU licenses are required, transmitters/receiviers/modems/DSP/antennas all add to SWAP (size, weight, and power). Ground-stations are needed to move the data up/down, but those have recently become a commodity too. Still, they're not free. (see: https://aws.amazon.com/ground-station)

There is also the added latency between earth-based users and space-based datacenters, which may be a deal breaker for some applications.

Another issue I don't see covered are the significant differences between space-based hardware and terrestrial hardware. The space stuff needs to be radiation tolerant, and that usually makes it a lot slower and a lot more expensive than the terrestrial stuff, all other things being equal.

In the end, space-based datacenters are highly impractical even if you assume that Starship can put anything into orbit very cheaply.

Re: Economics of Orbital vs. Terrestrial Data Centers

#127

Post by sigmar »

daemonologist wrote: post #79

> This is all to say that the current discourse is increasingly bothering me due to the lack of rigor; people are using back-of-the-envelope math, doing a terrible job of it, and only confirming whatever conclusion they already want. Calculating radiation and the cost of goods is not difficult. Run the numbers. > References: Gemini, Gemini, ChatGPT, ChatGPT, Gemini, ChatGPT, Gemini, ChatGPT, Grok, Gemini (There are s…

>This isn't about talent. It's about integration... Vertical integration isn't a nice-to-have. It's the whole ballgame.

I'm going to assume there's tons of logical errors and oversights in the math, considering the author couldn't even be bothered to write the text of the post himself.

Re: Economics of Orbital vs. Terrestrial Data Centers

#129

Post by jmyeet »

jofer wrote: post #39

What really worries me is that I keep hearing "cooling is cheap and easy in space!" in a lot of these conversations, and it couldn't be farther from the truth. Cooling is _really_ hard and can't use efficient (i.e. advection-based air or water cooling) approaches and are limited to dramatically less efficient radiative cooling. It doesn't matter that space is cold because cooling is damned hard in a vacuum. The artic…

I've done some reading on how they cool JWST. It's fascinating and was a massive engineering challenge. Some of thos einstruments need to be cooled to near absolute zero, so much so that it uses liquid helium as a coolant in parts.

Now JWST is at near L2 but it is still in sunlight. It's solar-powered. There are a series of radiating layer to keep heat away from sensitive instruments. Then there's the solar panels themselves.

Obviously an orbital data center wouldn't need some extreme cooling but the key takeaway from me is that the solar panels themselves would shield much of the satellite from direct sunlight, by design.

Absent any external heating, there's only heating from computer chips. Any body in space will radiate away heat. You can make some more effective than others by increasing surface area per unit mass (I assume). Someone else mentioned thermoses as evidence of insulation. There's some truth to that but interestingly most of the heat lost from a thermos is from the same IR radiation that would be emitted by a satellite.

Re: Economics of Orbital vs. Terrestrial Data Centers

#130

Post by jiggawatts »

indolering wrote: post #101
JumpCrisscross wrote: post #35

Earlier quoted context omitted.

> Just a cursory search shows launch costs are around $50M per launch I think datacentres in space are predicated on Starship bringing launch costs down. Way down.

Spinlaunch is also promising drastically reduced cost per launch. The payload size for their first launcher is pretty small and they appear to be struggling to get the kinetic launcher online.

Spinlaunch is an outright scam. Their main product is taxing people for inadequate knowledge of basic physics. And engineering. Also… common sense.
Post reply on HN