The Space-Based Data Center Hype Machine Is Already In Orbit (ieee.org)
- Reference: 0184264182
- News link: https://science.slashdot.org/story/26/07/02/0230210/the-space-based-data-center-hype-machine-is-already-in-orbit
- Source link: https://spectrum.ieee.org/orbital-data-center-hype
> Consider this: There are roughly [3]14,500 active satellites in orbit . Musk's Starlink constellation accounts for about [4]two thirds of those . Both the launch cadences and satellite-manufacturing capacity would have to scale up astronomically to deploy a million orbital data center satellites. For context, there have been [5]roughly 7,000 orbital launches in all of human history . To loft 1 million satellites into low Earth orbit on SpaceX's Starship, which is designed to carry up to 60 satellites per vehicle, would require 16,666 launches exclusively devoted to satellite deployments. Considering that SpaceX launched a record 165 orbital missions in 2025, even at 10 times that cadence, it would take a decade. And how long would it take to build 1 million satellites, given Starlink's [6]current pace of around 4,000 per year and a generous tenfold increase in capacity? Short of a manufacturing revolution, try 25 years.
Dissipating heat in space also requires enormous radiators. As IEEE Spectrum editor Dina Genkina [7]noted , startup Starcloud has sent only one Nvidia H100 GPU into orbit, and "their radiator was too weak to let the chip run at full power." A single 700-watt H100 would require about 1.4 square meters of radiator area, while a 100-megawatt data center could need 2,500 radiators measuring 80 square meters each.
So, why are the hyperscalers hyping orbital data centers? Answer: because it's lucrative. "The Elon Musk part of it is honestly genius because he's got xAI building the data centers, SpaceX sending them to space, and Tesla building solar panels," Genkina says. "It's almost like he's paying himself."
[1] https://spectrum.ieee.org/orbital-data-center-hype
[2] https://slashdot.org/~xetdog
[3] https://satfleetlive.com/blogs/how-many-satellites-in-orbit/
[4] https://spacenexus.us/blog/how-many-satellites-in-space-2026
[5] https://planet4589.org/space/gcat/data/derived/launchlog.html
[6] https://www.advanced-television.com/2026/04/13/analyst-spacex-making-340-satellites-per-month/
[7] https://spectrum.ieee.org/orbital-data-centers-heat
"Yay" (Score:2)
Another bubble. Yippee-fucking-woo
Rax the Tucking Fich! (Score:2)
More evidence we need to tax the rich, they have way too much money and are chasing really stupid shit. After you have 20 yachts then #21 doesn't have the same ego kick, so you look for pie-in-sky investments.
Re:Bet against Elon if you like (Score:5, Interesting)
The counter argument, is if they've gone to these efforts to develop chips so thermally & energy efficient for their workload they can run in space then why bother launching them into space in the first place?
Re: Bet against Elon if you like (Score:1)
Solar panels work better in space, you donâ€(TM)t have on going cost for electricity, real estate. Size is free unlike on earth. Downside is of course to launch, weight is not free but it is on earth. This means unless you are an expert on space launch economics, you dont know if space data center is better. Guess who has the data to do forward predictions. So I am guessing they have a target pricing on $/lbs that they can hit to make that business viable.
Re: Bet against Elon if you like (Score:5, Interesting)
> Size is free unlike on earth
This is the kind of assumption that people are making which shows how this is wrong
* the bigger it is, the directly proportionally higher chance some part of it it will get hit by something, which has a decent chance of having a cascading effect on the other components
* you've pointed out weight, which obviously comes with size - but you haven't pointed out that size =>weight => fuel needed to maintain orbit. Even with ionic thrusters using electricity and very very low fuel use this will matter.
> Guess who has the data to do forward predictions.
people who are working on optimizing the energy needed by AI models - specifically the Chinese AI researchers and probably Google. If model sizes can be reduced with very limited loss of performance then the costs of putting them in space - especially latency as opposed to a local model - will be hugely damaging.
> So I am guessing they have a target pricing on lbs that they can hit to make that business viable.
There's one business model which is absolutely crucial and everyone needs to understand that Elon is fundamentally sucking on the government teat, whatever people pretend. Military people need to run AI models in situations of compromised ground communication. Running, for example target identification and data selection in space makes 100% sense. Elon will be doing this in the knowledge that he's got a series of guaranteed government contracts that will pay for all the development he's doing. The risk in this case is that if a non-corrupt government does ever return in the US, they may audit his contracts and punish him for getting them through what they will consider to be illegal influence.
Re: Bet against Elon if you like (Score:2)
If it's not LEO, the orbit maintains itself.
Also if it's not LEO, there's a lot of... uhm... Space :-)
Re: (Score:2)
LEO is used for most satellites because it's by far the cheapest orbit, which you can easily reach with a two stage rocket. Everything above LEO needs much bigger rockets and is thus much more expensive to reach.
Re: Bet against Elon if you like (Score:2)
Yesz but LEO is also crowded, unstable, lacks minerals, and has a very narrow horizon.
It really boils.down to what your application requires.
Re: Bet against Elon if you like (Score:2)
That's where starship comes in.
Re: Bet against Elon if you like (Score:5, Interesting)
> Military people need to run AI models in situations of compromised ground communication. Running, for example target identification and data selection in space makes 100% sense
That seems like a benefit of relying on satellite communication rather than a benefit of putting the data center in space.
What's the benefit over running the computation on the ground in a "normal" data center, beaming the results up to a satellite constellation, and then beaming them back down to those who need it? Starlink/Starshield already enable that.
Re: (Score:2)
> The counter argument, is if they've gone to these efforts to develop chips so thermally & energy efficient for their workload they can run in space then why bother launching them into space in the first place?
I think that "why" question is one we should be considering much more closely.
Disclaimer: I'm staunchly anti-space-datacenter. It's asinine. TFS summarizes it well.
Here's a plausible answer to the why: Gobs of money have been dumped into giant terrestrial data centers. If those get taken out (they're huge targets), then what? If someone has a fallback in space, then the plebs will be unable to revolt and burn that datacenter down or take it over. They're ensuring that the working class is unable to seize th
Re: Bet against Elon if you like (Score:2)
It's also worth considering who may or may not have legal jurisdiction in orbit.
Re: Bet against Elon if you like (Score:2)
A space-based datacenter is far more vulnerable than an earth-based datacenter (an earth-based datacenter can't be de-orbited by rouge control signals, for example), and once size exceeds anything considered 'small' the possibility of space trash corrupting the unit increases.
Of course, this all assumes the insane cost per Kg of lifting anything into space comes down wildly making these space-based datacenters in anyway practical at any meaningful scale.
Re: (Score:2)
> A space-based datacenter is far more vulnerable than an earth-based datacenter (an earth-based datacenter can't be de-orbited by rouge control signals, for example), and once size exceeds anything considered 'small' the possibility of space trash corrupting the unit increases.
How many satellites have been de-orbitted by rogue hackers thus far? And if there were a million of them spread across the sky's, you couldn't take them all out (at least not without a much more sophisticated attack, and this would already require a sophisticated attack).
Meanwhile, a mob of angry people with tiki torches could take out any earth based datacenter. That's what I think the mega-wealthy would want to protect against. A distributed super computer in space would be unassailable by your average Jo
Re: (Score:2)
> How many satellites have been de-orbitted by rogue hackers thus far?
Your parent poster didn't say rogue, he said rouge .
People are being far too blase about the very real possibility of Color Wars.
Re: (Score:2)
> Your fantasies aside, if he wants to try it despite your and IEEE's points, let him. It's his risk, not yours.
This isn't like letting your neighbor build his own car in his backyard. The risk goes far further than just him. That money wasn't earned, a million satellites would impact us all, and the launches aren't exactly good for the environment (you know, where we must all live). Scale makes this everyones problem.
Re: (Score:2)
"That money wasn't earned"?? If you have an electric car, it's because of him.
Stop trying to come up with reasons to live in a cave. This is how we progress.
Re: (Score:2)
Why not both? Even if you have chips that are more efficient for *some* workloads, that's unlikely to entirely cancel out the headwinds facing people trying to build datacentres on the ground - training still seems to need as many power-hungry GPUs as you can throw at it. But putting some workloads in space could help. Someone providing AI compute as a service, which SpaceX already does, would need a combination of orbital and ground-based compute, it's not one or the other.
Re: (Score:1)
Why? Because it is hard to unplug our new AI/crypto overlords when you can't reach them with a ladder, and the power source, the SUN, will last a few billion more years. It's also difficult for a Marshall to deliver a tax bill off world. Once these orbiting datacenters are controlling Neuralink or animating our sex robots I don't think they'll have much trouble mopping up what resistance remains.
I'm no fan of our current big gov't, btw. Until we get our fiscal house in order (balanced budget amendment anyon
Re: (Score:2)
You could even pitch it as a ground-based satellite You could make a box that has compute, batteries, starlink and a few acres of solar panels. Pack the whole thing onto a semi truck and you could plonk it down anywhere there's available land. There are loads of places in the US where you could do that, and I think most the anti-datacenter crowd would be placated if the facility didn't need water and generated its own clean energy.
Re:Bet against Elon if you like (Score:4, Insightful)
Another counter argument is that Elmo doesn't give fuck about damaging the environment so shoving oodles of his tat into space won't bother him, either the pollution of the launches or filling the skies with his tat. And it isn't clear why we need all them datacenters other than to make him and his buddies richer and more powerful. Their idea is that money = power and with enough of it, they can direct the lives of everyone else. Just like a good little Nazi.
Re:Bet against Elon if you like (Score:4, Insightful)
Bingo. They have already put the paperwork in place to build pipelines to ship in the nat gas so that the launch cadence is 1/day. Currently they truck in I think I saw around 650 thousand gallons of methane for the fuel for a launch. I figured it is about 1% equivalent of all the gasoline burned in tx for transportation for a day. And because elon really likes no dependencies, he has his own wells as well. I still think it is beyond stupid as an idea, but the leon lovers lick it up. It is the heat problem in which there is no solution. And as others have said, if you can get the processing so efficient as to reduce the heat, well then the power requirements drop equivalently and you'd just do it on earth as the load on the grid wouldn't be an issue.
Re: (Score:2, Insightful)
> All the Elon haters
If you don't hate Elon after what he did to USAID, which has killed thousands abroad elsewhere and also resulted in the screwworm showing up here, you're stupid.
> You're going to look at what are the most practical workloads for space AI, what are the most efficient chips for those workloads in terms of tokens per Watt.
Fine words from someone who has done neither.
> And so on and so forth - it's called engineering.
What you're doing is called simping.
Re: (Score:1)
Ah, ad hominem attacks, the sure sign of someone who has a really strong argument.
Re: (Score:3, Insightful)
> Ah, ad hominem attacks, the sure sign of someone who has a really strong argument.
You don't even know what an ad hominem attack is, since I didn't make one.
Learn what words mean before you use them, kid.
Re: (Score:2, Insightful)
An ad hominem would be "Elon is wrong, because Elon is awful". Saying "Elon is awful and also is wrong about space datacenters" is not an ad hominem.
Re: (Score:1, Troll)
No. Cutting our foreign funding didn't kill anyone. That is fallacious logic. We are not responsible for the welfare of people in other nations. Likewise, cutting funding for the UN didn't mean the UN couldn't have reallocated funds for that screw worm program. That's on the UN, not the US.
The argument you should make is that cutting USAID funding reduced our soft power. That's the valid one. "We stopped doing someone else's job, and they failed to do it themselves", is not.
Re: Bet against Elon if you like (Score:1)
Maybe I'm being thick (I can accept that) but on earth we cool equipment by dissipating heat into the atmosphere (heatsink, fans, etc), but in space there is no atmosphere. How would you cool a processor without atmosphere? Would you have to create an atmosphere to absorb the dissipated heat, then somehow get rid of that heat?
Yes, I'm aware of liquid-cooling, but at the end of every liquid cooling system I'm aware of there are fans and heatsink/radiators that dissipate the heat into the atmosphere.
It would
Re: (Score:2)
> How would you cool a processor without atmosphere? Would you have to create an atmosphere to absorb the dissipated heat, then somehow get rid of that heat?
[1]Black body radiation [wikipedia.org]. The heat is emitted in the form of infrared light (and other wavelengths, but mostly infrared).
Most satellites need to be carefully engineered to maintain their temperature range without getting too hot or too cold.
[1] https://en.wikipedia.org/wiki/Black-body_radiation
Re: Bet against Elon if you like (Score:5, Interesting)
Giant black-body radiators are required. This is the the number one reason why space data centers are not practical. The radiators would be many times bigger than the satellites themselves. Every watt of energy generated by the solar panels has to be radiated into space. This is not something that can simply be engineered around, as the OP seems to think.
Re: (Score:2)
I personally don't care who made the space based data center concept. Overall, it is fraught, other than edge computing or a way to create a mesh network.
It still has a lot of issues. Space is an extremely good insulator, so heat has to be dealt with via radiation. Power can be handled via solar, nuclear batteries and betavoltaic panels (assuming that tech has moved since being at best a prototype in the 70s). If something breaks, it isn't like you can send a L1 tech to go yank a hard disk, replace it,
Re: (Score:1)
Assuming Musk knows all this as well, and he would, is there any reason he shouldn't try if he thinks he can do it? It's his money.
Re: (Score:2)
> it's called engineering.
so what's your suggestion for cooling? dissipating heat in space is a tough problem with very limited options, none of which seem viable for a frigging datacenter operating nonstop.
> so you're going to look at which chips can handle running at 80 or 90 Celsius.
running at high temps makes radiation more efficient but you still need to dissipate any excess heat generated over that operating threshold. once you hit that threshold (whatever it is) you either shut it down for the radiator to catch up or it will melt. considering that such cpus would generate a lot of heat (precisely to oper
Re: (Score:2)
Data centers in space are more expensive than data centers on earth. There is no way to avoid that. It doesn't matter how much you like Elon, the data center in space will always be more expensive.
The only reason to have data centers in space is if you want to do processing in space, for lower latency or something similar. [1]And in fact, that is what the government wants to do [wikipedia.org].
[1] https://en.wikipedia.org/wiki/Space_Development_Agency#/media/File:RapidReactionLaunchProliferatedGroundC2.jpg
Re: (Score:2)
Silly assumptions? A matter of Engineering? What about physics? Maybe listen to real engineers for once. They've been showing us the actual numbers that state clearly this AI data centers are not possible. Sure you can get lots of solar power, but that's not the issue. The issue is cooling, requiring huge radiators that are far bigger than each satellite. Besides the impracticality of it, you have other issues like air pollution (already a problem with starlink deorbiting), light pollution (who needs t
Re: (Score:2)
A vacuum is a terrible place to try and radiate waste heat.
Some poor sysadmin having to deal with this. (Score:2)
I need you to suit up and ride the next rocket to space. We need to upgrade the RAM on one of the space servers.
Re: Some poor sysadmin having to deal with this. (Score:1)
We don't build high-density datacenters with plans to EVER touch failing hardware. When a node breaks, it's taken off-line, not repaired. When an upgrade is needed, the datacenter is turned-over/forklift upgrade - no one runs into a google datacenter and talks about adding more RAM.
At one time we talked about dropping watertight datacenters in the ocean - this is kinda the same idea, but it includes rockets! (And, relies on wireless technology and solar panels, ocean-based datacenters could be wired into a
So if you're wondering why such an obvious scam (Score:2)
It's the same reason Elon Musk made that stupid fucking tunnel. It's misdirection.
So this lets them guzzle down all your water and electricity so that you have to start rationing both of them while they go to public and tell the public that they have a solution to the problem.
As an added bonus Elon Musk can pretend SpaceX has a magic new customer that doesn't actually exist in order to justify the phony valuation while he extracts all the money from your 401k.
Re: (Score:2)
Do you not realize how silly that sounds? He wants to try and put datacenters in space in order to steal your water and power?
The rest of the world is not out to get you.
Re: (Score:2)
That's it, he is doing it to generate fake potential future business for SpaceX. In reality, what will probably happen is others enter the market and undercut Starlink and launching to LEO, and that business dries up. It's already happening with Tesla cars, which are constantly heavily discounted because rivals make better ones at lower prices. They are hanging on in the UK by somehow being a "prestige" badge along side BMW and the like... Actually I can see the connection there.
Oh it's not feasable (Score:2)
And they know it, radiation, heat and cost. Another problem is kessler syndrome and space debris. There is a limit to how much stuff we can put up there.
Re: (Score:2, Insightful)
Neither was half the other stuff people said he couldn't do but then did. If he thinks it's worth trying, let him. It's no skin off your nose.
Naysayers don't accomplish anything. (Score:1)
Yeah, there are a ton of technical hurdles to overcome. It won't be easy. Does that mean he shouldn't try? If he wants to take a risk on building something new, why shouldn't he? It isn't your money. It isn't your reputation.
It doesn't matter to him, if he listed to naysayers like y'all, the modern electric car industry wouldn't exist. The modern rocketry industry wouldn't exist. Hugesnet would be the only satellite internet provider. He's taking big risks for big rewards. Stop trying to make yo
Re: (Score:1)
This. There's a big group of /. moderators who just downvote anything that supports Musk, but the man has accomplished amazing things, and this petty, sour grapes nay saying is kind of pathetic.
Re: (Score:2)
"Pathetic" is an excellent description. Resenting someone else's success is pathetic. And very bad for one's mental health and the health of society.
Anticipation! (Score:2)
I look forward to Musk overcoming the limits of insolation in Earth orbit, the latency induced by the speed of light, and the Stefan-Boltzmann law.
This is just as good an idea as the submarine he ordered built for cave rescue: it appeals to idiots who don't give any thought to the problem but think the proposed solution is 'cool'.
I don't get the appeal (Score:2)
AI data centers are evolving rapidly. You want to be able to rip and replace gear all the time. That's expensive to do in orbit.
Managing heat in a data center is a huge issue. It's very hard to dissipate heat in space. It's a lot easier to dump waste heat into the air or a river.
Communication between nodes, racks, and rows is a fundamental limit to AI performance. Spacing rows hundreds to thousands of kilometers apart is going to add speed-of-light delays and bandwidth limits. I don't see how you maintain p
Re: (Score:2)
But it is easy to hype because most people don't understand the technical issues. And people are way more likely to hero-worship than to be rational. And it ties together all the tech pieces he has build (mostly), so it is fantastic amplification of his brand.
AI / Crypto overlords need more power (Score:1)
It's hard to unplug our new AI/crypto overlords when you can't reach them with a ladder, and the power source, the SUN, will last a few billion more years. It's also inconvenient for a Marshall to deliver a tax bill off world. Once these orbiting datacenters are controlling Neuralink or animating our s-x robots I don't think they'll have much trouble mopping up what resistance remains.
I'm no fan of our current big gov't, btw. Until we get our fiscal house in order (balanced budget amendment anyone?) they ar
Building a Chinese Data Center on the Moon (Score:2)
Didn't Heinlein write a novel in which someone built a data center on the Moon, out of off-the-shelf Chinese hardware?
Ah, got it: Temu is a Harsh Mistress.
So basically... (Score:3, Informative)
... it's just another pack of lies like everything else Musk hypes up.
Re: (Score:1)
Not lies, a carefully crafted scheme to defraud investors of their money
Re: So basically... (Score:2)
If that was the intent, it wouldn't really work due to Elon himself having more downside exposure than anybody, including all institutional investors combined.
Besides, some of this argument is over engineering problems that have already been solved.
Re:So basically... (Score:5, Insightful)
The scary part is Musk isn't alone. The insanity of the LLM data-centre build out has got far too much money thrown at it already. There's no way it will be recouped. It can only end badly now.
Bubbly turtles all the way down (Score:1)
Rumor is there's also a bubble in bubble insurance*. Does that mean it will double-pop?
* Or at least bubble hedges
Re: (Score:2)
[1]It's part of the Golden Dome [wikipedia.org] anti-missile defense. The architects have decided it's important to process missile tracking data in space, rather than sending it down to earth for processing.
[2]They have a whole network of things up there [wikipedia.org] processing and communicationg.
I haven't looked much at the architecture, but I suspect it's kludgy, completely lacking elegance.
[1] https://en.wikipedia.org/wiki/Space-based_data_center
[2] https://en.wikipedia.org/wiki/Space_Development_Agency#/media/File:RapidReactionLaunchProliferatedGroundC2.jpg
Re: (Score:2)
I think satellite data centers are colossally stupid, but I suspect the larger problem is the public's gullibility for big lies.
Now, which things ARE lies and which aren't has been delightfully co-opted by politics; what one puts on that list is *instantly* translated into political affiliation.
I can think of 3 big lies that would immediately get me labeled "stupid maga fuck".
I can think of 3 others that would likewise get me labeled "woke fag".
Amusingly, putting all 6 in a list would be cognitively negativ