Anybody Who Thinks Orbital Data Centers are a Good Idea
Is Suffering from AI Psychosis, Experts ArgueDoesn’t Understand Basic PhysicsTwo main problems with data centers. Power and cooling. In space the power is doable. The cooling is a major pain in the ass and always has been. There are only three ways to get rid of heat. Conduction, convection, and radiation. The first two don’t work because of the vacuum thing. The third is horribly inefficient. Just look at the ISS and the giant fins that only dumps about 70 kW of waste heat through radiator “wings” that weigh several tons. A single rack in a high density compute rack can generate 100kW by itself.
So yeah given the expensive and how inefficient it is, it’s a terrible idea.
Edit: I’m a system architect so dealing with data center heat is something I’m familiar with.
You’re just too small minded to comprehend the grand vision of business genius™ Elon Musk!
And Bezos apparently…
These chucklefucks are just trying to get some sucker… er I mean investor to fund the whole thing so their respective space companies can do the job.
And after doing to some very cursorary research other issues like…
- Radiation and reliability - apparently cosmic radiation is a bitch
- Lifecycle costs - I should have thought about this one. It’s not like when a drive dies in space you just drive down to the DC to replace it. And not mention we recycle out compute about every 5-6 years or so.
- Connection - somebody mention this already in the thread but yeah you ain’t hanging a fiber cable to a satellite
Cant you see, were just around the corner for AI Humanoid Robots in FullSelfDriving™ Teslas heading towards their data center job in space
/s
There’s also the very real problem of data transfer.
On land you just lay down another fiber optic cable and you can double your data transfer rate.
In space, you have to deal with cross talk and interference on a limited spectrum.
Free space laser communications are possible, but even then you are only talking about 10s of GB/s, and you cant add more lasers or receivers on a satellite already in orbit.
But if you have a constellation of thousands of LEO communications satellites to leverage…
Doesn’t help, your laser (or RF comms if you are using them) can still only send out a fixed amount of data per second, it doesnt matter if it is being sent to the ground or another satellite, once it is launched there is a hard cap on how much data can flow into/out of it in a given time and there is no way to improve that.
That’s what the lasers are for. It’s a solved problem.
Not really, because it can’t be solved, just worked around.
Lasers are still subject to the inverse square law, but with a slightly different multiplier.
Also, lasers still have the bandwidth issue of not being able to double up the communication lines due to cross talk and other fun physics issues.
There’s a reason why fiber will never go out of style.
If it was a solved problem it would be widely used, but it isn’t. Ever looked at the reports of starlink speeds? It’s not reliable at all, everything other than a fully clear sky with cold weather (meaning less moisture and particles in the air) affects the communication. It physically can’t be a good or better alternative to fiber (or anything else that isn’t wireless).
Yeah… Terrestrial 5G towers with a fiber backbone for some proportion of them… are… stupendously more cost effective at getting a decent level of internet to a lot of people.
Also doesn’t cause Kessler Syndrome, which is, you know, good.
Now, such a system will still suffer in more abberant atmospheric conditons, but to a far lesser extent.
Literally the only actual ‘use case’ I can think of where StarLink ‘makes sense’ as a better solution is … you are a boat that is actually moving most of the time.
If you’re a house boat… terrestrial 5G probably exists near your mooring.
Either that or you truly, truly live far away from civilization.
… but we already had satellite internet that did those things.
Agreed on the downlink.
I thought this was about the node to node communications. Blue origin and probably others are also using it for in orbit communication.
What if we run a really long tube down to earth to send water back and forth? You gotta think like Elon to be innovative.
iirc the power is not very doable, You’d need hundreds of times as many solar pannels as are on the ISS to power a single modest data centre.
And at that point wouldn’t the solar panels act as a sail and fuck with the data centers orbit?
Solar sail effect is going to be dwarfed by regular atmospheric drag in low earth orbit. At perfect right angles the radiation pressure on the panels is 4.5 micro-Pascals. Meanwhile, in low orbit there’s enough residual atmosphere to generate a dynamic pressure (for drag) of 5 milli-Pascals, give or take (and strongly depending on the space weather).
So atmospheric drag is around 1000 times more than photon pressure. And the drag is big enough to be noticeable over weeks and months, requiring regular boosts to stay in orbit.
Maybe? Probably? For all I know, a solar-sail effect might help it stay in orbit.
Send it to a cold moon like Europa. Free cooling, plus A.I. is kept at a safe distance
Ummm….
“All these worlds are yours except Europa. Attempt no landing there.” - Arthur C. Clarke 2010
Even better, the A.I. problem is solved. Hooray!
“give me a recipe for dinner”
…
…
…
(2 hours later for signal round-trip)
“Here you go:…”
Yeah but space is cold. Just put the hotness out into space and it’ll freeze just like in the movies.
/s
The radiators would be about the same size as the solar panels. Both would have to be huge to run a rack full of GPUs.
Radiators work because they have something to radiate heat into. Space is famously empty, so a radiator the size of a planet would only work as a heat sink until the total heat in the system was high enough to make everything glow like a heating element, at which point you dump waste energy as visible light.
You can radiate heat into the vacuum of space, it’s just extremely slow compared to doing it into atmosphere. Vacuum is not a perfect insulator in this regard.
Think of it this way, if a vacuum was a perfect insulator, how would the sun radiate heat to Earth?
Your car radiator is actually using convection to convect heat into the air.
The spacecraft radiators use radiation to dump heat by emitting infrared photons. Photons do not require a medium. This type of radiator works by maximizing the area of hot surface exposed to empty space (which has an effective temperature of 3 K). They have to be pointed into a dark area and away from the sun. There’s no advantage to fins, because you want to maximize area perpendicular to the dark sky.
Both devices are called radiators, but they are different kinds of devices.
Do you have a podcast? I saw a podcast clip on tiktok saying almost verbatim the same thing
Not to my knowledge. But I assume this is nothing new and any reasonable person could come up with the same thing. I did google the ISS thing so that part may have come from there.
they are trying ti in the ocean, they have to deal with corrosion , animals gettin encrusted on the surfaces, plus transportation and logistics.
Basically they’d need about as much in radiator fin surface area as they would have in solar panel area. The ISS has 8 solar array wings, 35m x 12m, that can produce about 30 kW each, or 240 kW total, in sunlight (which is only half the time). The ISS has a complex cooling system, but relies on 4 radiators about 3.1 m x 13.6 m to reject up to 14 kW of heat each (56 kW total) for cooling the solar arrays themselves. The main cooling system uses 6 radiators, each 23.3 m x 3.4 m, to reject 70 kW of heat (from this report it sounds like each radiator may be capable of rejecting more than 1/6 of the heat but that the system as a whole needs to be kept under 70 kW of heat rejection).
So that seems like about 650 square meters of radiators can provide about 120 kW of heat rejection.
Today, a 72-GPU Blackwell server is 130 kW in a single server rack. The next generation rolling out now has 72 Rubin GPUs in a 230 kW server, in a single rack. And that’s not even a “data center.” That’s just a single (albeit very powerful) server. How many can you string together, with networking equipment beaming data connections back down to the ground, before the ratio of solar panels and radiators to the actual ship size becomes unworkable?
That said, it’s technically possible, especially if you can radiate the heat at higher temperatures than the ISS does, as the Stefan-Boltzmann law shows that the hotter the radiator, the more heat it can reject. Just completely infeasible from an engineering and economical standpoint, for any data center that hopes to be relevant in an age of 100+ MW data centers.
All they’re sending up is 1 server rack. 125kw avg, 150kw peak.
Radiator needs to be ~twice as big as ISS, but we probably have improved their efficiency in recent years so maybe not twice?
Question
With space being so cold, doesn’t that keep temps low? How does it all work
Space really isn’t cold. Temperature is a measure of how fast particles (atoms, molecules) are moving.
In a perfect vacuum with no particles at all, you literally couldn’t define a temperature, because there’s nothing around to jiggle
In order to understand why space is “cold”, you have to understand how we, as humans, perceive temperature. What we feel as hotness or coldness is just how excited the molecules in the atmosphere are. Molecules very excited? It’s hot. Molecules very still? Cold.
Space is vast. Incomprehensible to our puny human minds that have evolved to exist on this tiny mote of dust. Most of space is devoid of matter. Sure there’s hydrogen and helium out there just floating around, but not enough of it for us to be able to feel. So space feels like cold, and indeed, is quite cold. But as the above poster explained, losing heat in space is fundamentally different than how we lose it on earth.
Your body generates heat to keep your squishy organs running smoothly. The way we prevent ourselves from overheating is we rely on perspiration and evaporation, but that only works if we have Earth-like conditions where airflow can carry that excess heat way from our bodies.
In space, there is no airflow. Your skin would freeze while your blood boils.
The same issue is present with our technology. Radiating heat is very, very inefficient in space because we need something to carry it away from the source that is generating it faster than it can generate it. At least with the tech, we can turn it off to let it cool slowly over time, but it doesn’t solve the fundamental problem of having no way to efficiently cool an entire AI datacenter that is meant to be used to fill a continuous demand here on Earth.
The solution, then, is a downsized dc.
I mean, yes, but specifying a more specific cause of the delusional thinking is useful.
I don’t know how that even got past the brainfart stage. AFAIK nobody has actually demonstrated how that would really work.
- Despite SpaceX’s advancements in regards to things like resutable rockets, shooting stuff into space is still prohibitively expensive.
- Server clusters are exceptionally heavy.
- Server clusters run hot, cooling is not a triviality considering you can’t just rely on convection in space, so more mass for alternative solutions.
- Datacenters need regular maintenace.
- Logic boards won’t do well with the radiation in space.
- Despite SpaceX’s advancements in regards to things like sattelite internet, getting large datacenter level quantities of data from earth into space and back, and at low latency, is no triviality.
Not saying this won’t ever be a thing. But not in the lifetime of anybody on earth right now I don’t think.
People don’t understand just how difficult it is to cool stuff in space. Half of the shit sticking out of the ISS that people think are solar panels are actually radiant cooling systems, and the ISS will generate WAY less heat per volume than a data center.
Not to mention the power requirements would likely require more than solar unless they put solar panels up far bigger than anything put up there before.
Beyond the power requirements for the computer, there is also the power that will be needed to transmit back to earth.
A wireless link that can pass the amount of bandwidth needed (and do it reliably) will need huge power as well.
It will be far bigger. Its mostly solar panels.
Or one of those mini nuclear reactors they are working on
Im seeing the ISS needs to reject 70kw with a max ability of 84kw. The datacenter dishes will be between 125 and 150, so around double or less.
Not saying this won’t ever be a thing.
I’m saying it
This whole idea reminds me of the “putting solar panels on highways” idea that keeps popping up from time to time. Anyone who has ever built anything understands how stupid it is. Even if you could do it, it still wouldn’t make sense over just putting solar panels next to highways.
That, and solar windows.
Making an expensive solar panel that lets most of the energy pass through it, and is not mounted in a way to effectively collect solar energy, is a terrible idea.
I agree, that this is at the moment not a viable thing and especially the SpaceX “concept” is complete bullshit.
I do not agree with some of your points, since they are solved/irrelevant (e.g. “regular maintenance”, “low latency”) or could be overcome with reasonable tech advances (e.g. “rockets prohibitively expensive”, “radiation shielding”).
Let me steelman the argument a bit with this single bit of - sadly forgotten - “super cool and innovative tech”: “Underwater data center”, like project Natick (Microsoft) or the Chinese project:
Soooooo, if we will ever see something other than our current land based data centers, we will see millions of ocean data centers, before we will ever see a single commercial space data center.
Reasons:
- Delivery is super cheap (in comparison to space) at scale, thanks to the already existing wind farm infrastructure
- Weight is not an issue
- Cooling is solved
- Maintenance is not necessary, but replacement is. Easy on scale, because modular.
- No radiation shielding necessary
- Connection: data cable = no extra lag or quantity limit
Oh, and by the way, it is still not clear if even ocean data center will be viable. Just found this 😂
getting large datacenter level quantities of data from earth into space and back, and at low latency, is no triviality.
Latency is a huge issue, but so is bandwidth.
Land based data centers will have multiple hundred gig (and faster) fiber connections to the outside world.
Replicating that level of bandwidth on wireless links to a satellite in any sort of stable way is (as you said) no triviality. I would even classify it as near impossible.
Unless it becomes cheaper than having a datacenter on earth per quanity of compute, it won’t happen in any meaningful scale even if these issues are solved.
It will never be an economic thing. Only unpluggable skynet military thing. The weight is not an issue. though. It’s volume.
Weight is always the issue with lifting stuff into space. Volume might merely be an additional issue.
The $200/kg launch price target is based on 150 ton capacity. That’s a $30m launch costs target. Volume/foldability matters the most because that is the actual constraint that limits datacenter launch to a single NVL72 size.
Projected cost targets from SpaceX, especially for Starship are only losely related to reality. Weight is what determines the minimal required energy input to lift something into orbit. Independently from SpaceX number magic. Volume, like I said, can be an additional bottleneck but never undo the above.
Starship is huge. I dont know how tightly they can fold these expected dishes, but by weight, they can amd will do 60 starlink v3, and itd be 50 datacenter dishes equivalent. How many they can actually launch is going to depend on how well the solar and radiator folds down, so it might be a volume issue vs weight where they cant launch with the max weight capabilities of the ship.
They’ll manufacture it on the moon ofc. We won’t ship it from earth.
Orbital data center plan:
Step 1: Establish a slave labor force and advanced chip fabrication facility on the moon.
There is an unsolvable compute problem. The average PC on earth has multiple bit-flips a year from cosmic rays. The space hardened chips we use are 50nm and the chips used from inference are 4 to 6nm. 50nm is far more cosmic ray resistant than 6nm because of the transistor size. Are we supposed to think making H100s with a 65nm process is possible? The speed of light creates a die size limitation as well.
Either that or they’d have to triplicate everything and have a voting system.
this is the current real solution on the ISS - three main computers doing everything.
They are not using 6nm process chips in the ISS. The computers themselves were made before that process existed. An off-the-shelf space hardened computer system uses 65nm process. Cosmic rays is a very general term, it covers basically everything that flies around in space, that includes sources like the sun, which is hammering everything in the solar system with rays. Outside of the atomsphere there are so many more cosmic rays that non-space hardened computers can not even make calculations. Combined with the difference between the bit flip rate when you make transistors 10 x smaller is also fucked up high. One CPU cycle will have enough erros to make the computer useless. It’s a multi-faceted problem and when the largest limiting factor is weight & size, it can’t be solved with scaling.
How would that even work with inference where the expected output will be different between 3 runs with the same input?
The way I see it is they are doing inference, not transfiring bank account balances. I’d be curious to see some actual experimental data, but I’d expect LLMs to skip past bit flips same way you shrug and move on from spelling errors. At worst you can do your critical calculation in triplicate on your 6nm nodes (with redo upon dissensus) and reduce your bit error from 4/year (or 4000/year or whatever have you in orbit) to (4/year)^3
A bit flip might change a 0 to a 1 or a 1 to an infinity. Even if you could just do everything three times, that triples the hardware and energy costs compared to terrestrial computing.
I don’t think you don’t understand the difference between the amout of cosmic rays, which are basically any flying particle, on Earth compared to space. Small nodes would be dealing with multiple per cpu cycle. Multiple could be 1 million a second, I am trying to figure out a way to measure. It would be something like distance from atomsphere(rate of total particles) x probably of an object the size of a transitior getting hit(rate of collions). I could probably find the bit-flip rate for an off-the-shelf space resistant chip and infere the rate for the size I need, but there are other factors. A bit will not flip on every collision, shrinking transistors exponentially increases this.
There’s various ways from better shielding to redundancy to error checking everything.
This is actually why I hope they do pursue this stupidity. We do need to figure out how to goto the next level of computing for everything we do in space. If the folly of “datacenters in space” helps them figure out reliable computing despite radiation and lack of cooling then we win
You have two options for shielding, one is shield the entire thing like we do with the ISS because people live there and radiation is bad for living cells. Two is to shield the smallest unitary entities to use the least amout of materials, like a shield per chip. This is physics not magic, radiation does not just go away. When everything add to launch weight it does not make sense. Redundancy systems also add to launch weight.
The datacenters are only a little bigger than a v3 starlink. It’s 1 rack of compute, around 125kw avg 150kw peak. The biggest part is the solar array.
If it’s one rack it’s kinda pointless? And also absurdly more expensive than putting that one rack almost anywhere else in the world. And you can’t really fix or upgrade it. And then it’s in space and more susceptible to bit flips.
I’m not sure there’s a single thing it being in space does that’s better than it not being in space.
One rack is generally going to be well more than enough for a persons inference needs.
They want to put a million of these up there long term, where are you going to find space for a million individual racks? Now it’s a large datacenter again.
I did see one thing about a company adding a cluster or something to someones house and giving them a cut of it. That decentralizes the power distribution, and space requirement, but it adds other problems like vandalism / theft as how well can you protect a thing tacked onto the side of a house worth 10s of thousands of dollars.
Edit: And you don’t fix or upgrade them, they deorbit in ~5 years, and get replaced with the next best thing. Radation protection to avoid bit flipping will be a cost issue, but they already have hardened chips that work in space, so I’m not sure how much new technology is needed, and starship can lift a shit ton of weight, so heavy shielding is possibly an extra option?
Edit2: Just to be clear, I’m not trying to say they are going to earn enough revenue to make these things profitable like they did with Starlink, I’m just talking about the technical specs of what they say they’re going to do. There’s a lot of misconceptions about what they even intend to try putting up there.
Large data centres are used for a reason, it’s way more efficient to shove all the compute in one large building than putting it in space, at all.
Its a horrifically stupid idea, with no real benefit at absurd cost.
It’s not no real benefit, these are some, which are legitimate benefits.
- Unlimited sun, without the sun having to go through the atmosphere which increases the efficiency of the solar panel.
- No opposition / red tape from communities fighting against the data centers, some communities are evening banning them.
- No theft or vandalism.
Just to restate my relevant edit you probably didn’t see above
I’m not trying to say they are going to earn enough revenue to make these things profitable like they did with Starlink, I’m just talking about the technical specs of what they say they’re going to do. There’s a lot of misconceptions about what they even intend to try putting up there.
You can get more power from solar on the ground, cheaper, because it isn’t in space and they’ll get deorbited which is worse than vandalism because the entire thing is now gone.
I’m not disagreeing and saying they won’t do it, I’m just saying it’s an incredibly stupid idea.
That just increases the land space required though which makes it harder to get it built upfront, but I agree, land based solar, even taking 3x+ the land required + battery backup is probably still cheaper given they’ll last 20 years or more.
Ground projects like that take years to complete as well, between finding a spot, permitting, building it etc and to some extent time is money in this current environment.
Weight wise, they can launch 50 AI datacenters per launch with Starships 100T capacity, but volumetric wise I don’t know how well they can fold these up and if they’ll actually reach 50, but lets say they can get 50. I honestly have no idea if they can.
That’s 7.5MW of solar panels deployed each launch which will be coming off a factory line launching multiple times a week. They did 123 starlink launches in 2025, so thats 922.5MW solar capacity launched in a year if they did that with the AI sats, but they’ll likely do way more if starship actually works.
You can’t build an almost 1GW solar array (edit and datacenter) on land that quickly. (edit2: Oops I didn’t do the 3x+ for the 1GW solar i mentioned above, it would need to be 4-5GW on land so it can overproduce to store enough in the batteries for overnight)
The downside of course is it’s only going to last 5-10 years. That’s a lot of costs to try to recoup in that time frame.
What about the latency hit getting data back to earth?
LEO isn’t that far away. StarLink has quite good latency.
Orbital data centers are a good idea if one wants to get massive golden parachutes that siphon money from all the investor cash that the entirely u realistic promises brings in. They are a grift that will most likely benefit Musk in the same way all his other pie int he sky bullshit does.
Being technically terrible hasn’t really stopped that from happening with his other bullshit projects.
Orbital data centers are the new “hyperloop”
Space being “cold” doesn’t matter since vacuum INSULATES.
it’s not even cold…! The matter that DOES exist in it is very hot plasma but it’s just really thinly spread out.
This is what I’ve been tearing my hair out over any time this comes up. If you put a computer in space it will heat up until it achieves incandescence. Which is bad for the performance.
This kills the
crabserver.and sheesh you sure ain’t KIDDING! The amount of wattage in terms of thermal energy they’d have to radiate off would in fact LITERALLY require incandescence @.@;
for anyone else who stumbles across this comment section, i’m referring to “black body radiation” - it is specifically the thermal energy leaving something via radiating, and if it is radiating out at a high enough rate, that makes it glow. Even if we were not literally seeing the thermal radiation with our eyes, it is still technically incandescent in the infrared end of the spectrum!
(unless of course you’re defining incandescence as specifically only when the radiation reaches all the way up into the visible wavelengths and frequencies… which, if so, that’s quite fair really.)
I don’t have a good grasp on what technology exists for space, but I would assume that radiators of some sort would be possible. Not in the conventional way that they ineract with a medium to release heat, but instead that the radiators emit heated particles - kind of in the same way that water evaporates without boiling. With that being said, I have no idea what efficiency that would operate under, and I have no idea if such a radiator would be used up fast. It sounds like a terrible idea, but I don’t posess the facts.
Some people don’t believe in space travel because there is no air to push against in the way that jet engines work. But they fail to understand that space travel operates under other ways to generate force. I just don’t want to end up in the same sort of argument as them, believing that it’s not possible to cool down machines in space just because there is no medium for conventional cooling.
For an intro to real proven methods you can look up how the cooling for the ISS works. It’s quite interesting, but takes a lot of space, for relatively little power.
So is the plan is to use a heat-pump to cool the computers while getting some waste-product as hot as possible, and then eject the waste product? Or perhaps rather than ejecting it, the heat could be put into a large surface-area heat-sink thing to just radiate the energy black-body-style…
I think ideas like this are fair and reasonable if we have to have a data centre in space (for example, if we wanted people to live indefinitely on a space-colony or something like that). But its pretty clear that no plan will ever be anywhere near as good as what we can do on the planet’s surface.
Building and operating these things on the Earth’s surface is already expensive and resource intensive. And doing it in space is going to be a lot worse.
I agree. I just saw that the comments were only considering conventional heat dispertion to a surrounding medium and that it wouldn’t work in space. I feel it would be counterproductive to base the arguments on that narrow idea, but much more productive to realize that there are alternatives, but the current alternatives are much worse.
But apparently still somehow good for The Economy™!
They want to put them where normal people cannot damage them.
It’s so expensive. Putting data centers in bunkers or caves is more reasonable. (There actually are data centers like this).
Proton’s primary servers are stuffed in a cold war bunker in the Alps, for example.
And yet the IPO of SpaceX was justified with the presumed future success of its space-based data center program.
Wait is that really a thing
Yes. He unironically promised to build data centers in space. So basically, his company’s stock went up, because he falsely advertised something that is practically impossible to achieve at this time.
Hasn’t he said full whee drive automatically driving your car would be availbe a decade ago and it’s still not here
The one I heard was a fleet of Tesla self-driving taxis across the US by 2025.
Instead we have news reports of Teslas on autopilot killing people while the driver is distracted.
There are entire websites dedicated to tracking his “promises” and lies.
Basically, if you believe a word that guy says at this point, you deserve to lose all your money.
Yes. That’s not even the most flagrant one
It’s comical how finance is the most ‘vibes’ based of any discipline, yet they try so hard to LARP as hard math/science experts. Much projection for such massive insecurity, it seems.
Shits gonna be like 99% radiator
One consumer grade graphics card surrounded by the largest heatsink ever produced.
“That ought to do the trick!” 😆
Reagan approves
Well that’s investors fault for believing him isn’t it. The biggest thing ever put up in space is the international space station we don’t even have the capability to do that anymore. How did they think he was going to build a data centre in orbit? Which of course is completely ignoring all the other technical reasons wouldn’t work.
It’s like Tesla: Everyone knows it’s a scam. Whenever there’s a panic or a dip then Musk’s army of sock puppets and wash traders will coordinate to purchase more stock and drive the value back up. That’s how TSLA has remained so ridiculously over valued for decades.
It’s an exponential money printer that’s going to break the entire world eventually.
Self landing rockets were practically impossible not that long ago. Self driving cars were practically impossible not that long ago.
There’s nothing fundamentally impossible about orbital data centres. The main factor against it is the $/kg of payload into space. That’s one of the many issues SpaceX is working to solve, and there’s nothing to suggest they won’t get there.
The limiting factor for space data centers is absolutely not the cost to build them. The AI industry has already proven that they will burn the GDP of a small country just to have an LLM that shits out slightly better text that has coin flip level odds of being true.
The limiting factor here is one of thermodynamics. A travel mug for coffee works because it has a near vacuum between the inner lining and the outer shell. Vacuums are fantastic insulation because there’s no atoms there to transfer heat away. Very useful if you want hot coffee for a few hours. Space is a big vacuum, and data centers are giant heat generators. You’re basically putting a computer in a perfect insulation medium. It’s really stupid.
Came here to say this, thanks.
Self landing rockets were practically impossible not that long ago
Technically, they still are.
Self driving cars were practically impossible not that long ago.
There have been implementations of a sell-driving vehicle since the 1980s, and we’re still far away from “true” full self diving.
Both of these examples demonstrate the adage of “the first 90% of the work takes 10% of the effort, and the last 10% work takes 90% of the effort”.
The main factor against it is the $/kg of payload into space.
Oh my lord no. Although, technically yes but not for the reason you think.
The number one issue is heat dissipation. To radiate the heat from one DC satellite (at the power levels needed to run AI workloads) would need a football sized dissipation array. Even if Space X can invent some magical new physics and cut that down to a quarter of that size (hint: they can’t), we’re still talking about an order of magnitude increase in payload per satellite.
Next on the list is volume. We’re currently at around 14k man-made objects in low earth orbit. As it is, satellites (including the ISS) have to perform collision avoidance maneuvers every so often. The calculated limit of satellites we can put up to low earth orbit before orbital collision maneuvers start to become unmanageable is 100k. Basically after that amount we enter into a state where several corrections for each satellite are made regularly, and a single collision at the 100k limit would result in a cascading series of collisions that will render low earth orbit impossible to use. Basically after that anything you put up will get shredded by the insane amounts of debris.
Space X wants to put up a MILLION massive satellites that will require extremely large structures to dissipate the heat from the very power hungry AI chips.
They fully know the impossibility, and when challenged about the over crowding issue during an interview, an engineer brushed it off as “it’s not a problem”. People who speak that way about science and engineering issues are not serious people.
That’s one of the many issues SpaceX is working to solve, and there’s nothing to suggest they won’t get there.
There are countless engineering and physics reasons why they won’t. Stop sniffing Elon’s farts. They’re not good for your brain.
Edit: all of this is to say: space datacentres are the dumbest idea yet to come out of that idiots face hole. And he’s said a lot of really really dumb things.
I suspect Musk is trying to ensure his satellites are up before anyone else’s, so when the inevitable legislation is enacted to control who launches what, he already “owns” the lion’s share of “orbital real estate”.
They’re not going to put up these satellites, because they won’t be close to usable or affordable. Either the workloads will be miniscule, or the cost to put them into orbit is prohibitive.
The whole pitch was a cool sounding “space age” solution to a problem with AI datacentres that everyone is aware of. It was just a snakeoil salesman’s promise just so he could con investors out of money for his sweet 1.7 trillion.
So like you said, I’m correct. The only issue with thermal radiation is $/kg of payload. Again, this is an issue that they’re working to solve via methods like reusable rockets.
So like you said, I’m correct. The only issue with thermal radiation is $/kg of payload.
That’s not what you said. You implied the only limiting factor is a reasonable payload that can be resolved by the current incremental improvements to rocket tech.
What I said is that even with massive improvements to rocket tech, it will still be a near impossibility to get as many AI (or even regular datacentre) satellites into space.
The other issue is that the thermal dissipation problem is not solved for such a large amount of heat in space. It’s quite hard to dump large amounts of heat in space, and it needs to be done rapidly with computing. And the larger you make your dissipators the more you run into “how do I move that heat from the source and out towards the edge of the dissipators?” Because you need to utilize ALL of the dissipator if you want to keep your server parts cool. But moving that heat around a massive array is not trivial. If you’re using a fluid and moving it with pumps, then now your adding even more heat with the pumps.
And then there’s the issue with long term investment. These server components are going to be obsolete in a few years (and nevermind failures). And IIRC, they have plans to regularly de-orbit these things every number of years, which means even more launches at a regular basis to keep the swarm numbers constant.
And none of that matters in the face of the low earth orbit crowding issue that IS a massive problem.
Completely different level of difficulty.
And be pedantic self-driving cars are not a solved problem. If you try driving a Tesla in autopilot and don’t intervene when it makes mistakes you’ll be in a wall within about 2 minutes.
Building data centres in space isn’t just a technical problem, although it’s a major technical problem, it’s also an economic problem. Obviously yes it’s physically possible to do it but there’s no way of building a data centre in space for anything close to the price of just doing it on earth and with no other obvious advantage to it being in space there’s no reason to do it.
It’s the same reason we don’t build a transatlantic railroad tunnel. Obviously it’s technically possible it’s just a very long tunnel, but it would be hella expensive.
So like I said, the only problem is payload cost……
Free unlimited power is the benefit……
Right so go build a space elevator because little old Elon is not going to fix that no matter what he tells you. Even if you could get starship working you need something with 100 times the lift capacity.
Elon Musks companies have done so many things that people like you said he couldn’t do lol. You’re so blinded by irrational hate for someone who doesn’t even know you exist.
Other people have already explained the topic in sufficient detail, so I’ll just leave a quote from a former NASA engineer and a link to their article.
Taking the NVIDIA H200 as a reference, the per-GPU-device power requirements are on the order of 0.7kW per chip. These won’t work on their own, and power conversion isn’t 100% efficient, so in practice 1kW per GPU might be a better baseline. A huge, ISS-sized, array could therefore power roughly 200 GPUs. This sounds like a lot, but lets keep some perspective: OpenAI’s upcoming Norway datacenter is intending to house 100,000 GPUs, probably each more power hungry than the H200. To equal this capacity, you’d need to launch 500 ISS-sized satellites. In contrast, a single server rack (as sold by NVIDIA preconfigured) will house 72 GPUs, so each monster satellite is only equivalent to roughly three racks.
Source: Datacenters in space are a terrible, horrible, no good idea.
It all comes down to $/kg of payload. There’s infinite space up there.
How is this a rational argument? With infinite money we could put the Empire State building on Venus but it would be really fucking stupid.
It’s rational because the payload cost is coming down and down and down with SpaceX. Reusable rockets was a massive step.
Once it’s affordable, why wouldn’t they do it?
The “good idea” isn’t the data centers but the stock pumping. You propose something insanely difficult and expensive (also hopelessly impractical and stupid in this case) and because it is so difficult and expensive you claim you can monopolize the market if you succeed which is the ultimate dream of every capitalist but you just need some insane amount of investment to get there. Then when the money runs out you go back and ask for more and exploit sunk cost fallacy. All the while valuations increasing. It is an amazing way for already rich scammers to get much, much richer than could happen in a sane economy and slurp up huge amounts of capital that otherwise could have gone into more productive endeavors.
Obviously in any well regulated economic system this shit would be subject to some proper oversight to protect the interests of the majority, particularly all the people whose pensions and livelihoods are at risk when this all goes to shit.
In our time, squeezing everyone for their worth is all there is left for these people to get rich. They’ve run out of ways to squeeze and so are going for the last hurrah.
isnt thats what all the tech conferences are for, to hype the investors up. how long can they keep up with the peddling on those conferences? everytime a new con is our hub , its almost always a different usage of AI, so they need to come up with new “ideas” each time.
deleted by creator
- slow clap *
Does that moniker apply to the ones who support or criticize it?
For the cost of one orbital data centre, you could probably build 10 terrestrial data centres, bribe literally everyone involved in the construction to pretend that they built it in space, buy an island, fake your death, and spend the rest of your life off-grid.
Yea honestly, orbital data centers are the dumbest shit I’ve heard during this bubble, and a huge indication of peak bubble hype.
The bubble is like this at this point:
Are there actually people who think this is a good idea that are not in a position to make money off it?
Yes. All the retail investors in SpaceX that will be left to hold the bag.
That’s a funny way to spell "our 401k"s.
If it helps, you’re VTSAX 401k right now would only lose about 0.1% value if SpaceX dropped to $0. By June 2027 when all insider shares are unlocked it could go as high as 1.5%, but again only if it went to $0.
I hate Musk, I think SpaceX as an IPO was trash, and I disagree with the rule changes for index inclusion. But big picture, our 401k will be fine.
My aunt and uncle are Elon fans (yes, even after his fucking Nazi salute) and have said orbital data centers are a good idea.
They have money, and my uncle is usually pretty smart, so I’m not sure what the fuck is happening.
I don’t know you, and I certainly don’t know your aunt and uncle, but I would generally assume in Nazi situations that “even after” probably means “particularly after” even if they don’t admit it.
I know what’s happening! Your aunt and uncle are racist dumbasses.
as much as underwater datacenter, extremely high cost despite the ocean being able to cooldown datacenters fast. also these large LLM have never overcome one major flaw, There is no profit generation in the industry.
There is no profit generation in the industry.
That’s because we’re in relatively early R&D and are trying to speedrun things in a way we generally don’t with most tech. To compare to tech you understand and are familiar with, where we’re at currently would be equivalent to like the internet in the late 80s/early 90s, except we’ve all seen how that went and everyone wants to be the Amazon, Microsoft or Google of the AI market when it moves from high R&D and little to no profit to becoming a mainstream part of everyday life.
That transition point will probably be when bipedal robots that can do most tasks as well as a human while running a local model get cheap enough to be sold as industrial equipment. It’s why Asian labs (especially Chinese ones) keep showing off bipedal robots doing tasks that require either significant agility or fine motor skills involving predicting where body parts are (like doing needlepoint without being able to see it’s hands or doing dance routines).
Orbital politicians, and billionaires are a hella cool idea though
Which planet would they be orbiting?
Earth, we’re planning on doing an orbital bombardment of politicians, we just have to get enough of them up there.
And not tell them there’s no re-entry vehicle
Planet Hell
It’s literally just taking the piss out of idiotic investors. Data Centers, AI, space, new frontier, new markets. It checks all the boxes to get idiots excited to dump money into your tech company so people keep talking about it because talking about it is what gets results. Hopefully nobody is dumb enough to actually try it, it’s an absolute scam.
I mean, space construction companies are happy to try it.
If Bezos keeps giving them money, who are they to say no? It may not even all be a waste in the long run, as it could be a “testing ground” for future scientific missions.
I see it as a win-win.
Either Elom and his fellow dipshits waste a ton of cash chasing literal pie in the sky,
Or
A viable commercial presence in space is validated and terrestrial power needs eased.
/S
Can you all not even tell this or sarcasm?
Nah, you’re not jaded enough yet, it’ll be used to facilitate a wealth transfer. Huge wealth management firms will tie up venture capital in financing SpaceX and BO and smaller space construction firms in a boondoggle and when it fails to materialize a profit they will pocket the proceeds and the losses will be covered by taxes and retirement accounts.
They do this again and again. It’s the kleptocracy playbook, they’re just getting bigger and bolder and more in your face with it.
My bad. Edited.
I see it as a win-win.
Except when they put up about 70x new satellites then what we currently have up there, and cause a cascade of collisions that results in nothing being able to enter orbit (including humans travelling to space).
We’re at 14k satellites.
100k is the cascade failure limit.
Space X is proposing one million satellites.
Edit: my “70x” figure is very fuzzy just-woke-up early morning math. No guarantees.
Edit 2: dang! One million divided by 14k is 71.428! I almost nailed it.
No, you don’t get it. (Good job on the math though, btw)
Every AI gets to have one, huge Moonraker sized LLM space station. Grok Station 1, Gemini Geosynchronous, Claude Comet. Whatever. But all eggs in one huge solar-powered basket each.
Staffed by 1 or 2 humans each. Required. By law. Armed people. Because, of course they should have guns in space.
Then just see what happens. I bet it would be amazing.
It’s such a hilariously dumb idea I hope the tech bros sink millions into it.
Lets make a data center we can’t maintain, upgrade or access for any practical reason. Waiting for the suggestions to put them in geostationary orbits so that way their latency is even higher but going to struggle staying powered when in Earth’s shadow. Or get put in the Earth-Sun L1 so they always have solar power but now have to have significant more radiators on top of even MORE latency beyond beyond the moon’s orbit.
