Nvidia Announces Vera Rubin Space-1 Chip System For Orbital AI Data Centers 147
Nvidia unveiled its Vera Rubin Space-1 system for powering AI workloads in orbital data centers. "Space computing, the final frontier, has arrived," said CEO Jensen Huang. "As we deploy satellite constellations and explore deeper into space, intelligence must live wherever data is generated." CNBC reports: In a press release, the company said that its Vera Rubin Space-1 Module, which includes the IGX Thor and Jetson Orin, will be used on space missions led by multiple companies. The chips are specifically "engineered for size-, weight- and power-constrained environments." Partners include Axiom Space, Starcloud and Planet.
Huang said Nvidia is working with partners on a new computer for orbital data centers, but there are still engineering hurdles to overcome. "In space, there's no convection, there's just radiation," Huang said during his GTC keynote, "and so we have to figure out how to cool these systems out in space, but we've got lots of great engineers working on it."
Huang said Nvidia is working with partners on a new computer for orbital data centers, but there are still engineering hurdles to overcome. "In space, there's no convection, there's just radiation," Huang said during his GTC keynote, "and so we have to figure out how to cool these systems out in space, but we've got lots of great engineers working on it."
Intelligence ?!?! (Score:2)
Methinks they are drinking their own Kool-Aid. They seem to think their chips are actually intelligent ?!!?
Keep this guy away from anything scientific please...
Re: (Score:2)
No. He's just got a different, probably intuitive and not formalized, definition of "intelligence". And at a guess he consider "motivations similar to mine" to be a component.
Re: Intelligence ?!?! (Score:2)
Re: Intelligence ?!?! (Score:2)
This is concerning (Score:3)
It feels like they're making chips to fuel hype for a thing we all know can't work because physics.
I guess there's probably a huge circular investment with SpaceX or something though?
Re: (Score:2)
Economics, perhaps, not physics.
Re: This is concerning (Score:5, Insightful)
Even excluding launch costs do we have a feasible way to make a space data center work?
I'm sure the physics can work with math, but I'm not convinced something could be built to maintain orbit, capture energy, and then shed the heat with unlimited budget that replaces a medium sized data center.
Re: (Score:2)
As I've said before, the heat the back of the solar panel can reject at reasonable temp (near 100C) is in the same region as the front can generate (not taking into account earth visibility though, too hard for back of envelope). So cooling seems the easier problem. Capturing energy is just a question of more light weight (flexible) panels and a polar orbit for non stop sunshine.
LEO with never seen before surface area to weight ratios might be an issue.
Re: (Score:2)
Polar orbit costs more to get it there (deltaV) and it would have to be higher than LEO.
Also what about maintenance and upgrades?
Re: (Score:2)
Re: (Score:3)
The answer to the maintenance and upgrade question is simple - you don't maintain, you don't upgrade. You design the whole thing to be disposable. If an individual server breaks, you turn it off and pretend it doesn't exist anymore. Keep going until the thing is so broken you can't do anything more with it. When it stops working entirely, you abandon it as space junk or splash the whole orbital DC into the Indian Ocean where it can cause environmental havoc for centuries to come.
Re: (Score:3)
Also remember, this won't be just a handful of potential deorbiting disposable nightmares, it will be tens or hundreds of thousands if they actually executed the way they want.
Re: This is concerning (Score:4, Informative)
The math is that for each couple of GPU servers (we are talking GB300 sorts of servers), you'd need something about the size of the ISS' solar and cooling systems. Contrary to his "great engineers will solve it" sentiment, the cooling side of the equation is governed by very straightforward and incontrovertible physics, best possible solar isn't *much* better than what the ISS has already.
Others in this "let's go orbital datacenters" crowd have already pegged that you'd need tens of thousands of these setups, meaning that we'd need to launch essentially tens of thousands of ISS', and probably plan to let them all burn up after 5-6 years.
This is to compete with maybe 3 to 5 traditionally terrestial datacenters, the construction of which would probably be less than launching even a handful of those setups.
Re: This is concerning (Score:5, Funny)
Even excluding launch costs do we have a feasible way to make a space data center work?
Step 1: Place all the tech-bro AI broligarchs onto a SpaceX Starship with data terminals for each of them. Make sure live video feeds are available so we can watch them work.
Step 2: Launch to a Lagrange point.
Step 3: Allow the tech-bros access to their terminals, and flip on the live video feeds.
Step 4: Enjoy watching them try to continue to hype their own farts from space while their supplies dwindle. See how long it takes until they start to realize they are totally boned.
Step 5: Place bets on who eats who first and enjoy the show.
Step 6: Once free of them, start to reclaim some small semblance of sanity back here on Earth. But, since we're humanity, this step is mostly optional.
Re: (Score:3)
Yes, *technically* possible but physics is why the economics is completely stupid. Given enough materials and devoting an ISS-scale power and cooling solution per server we can work, but it's monumentally stupid because ultimately it's a massive waste of resources for a highly compromised theoretical end game versus just plopping them down in terrestrial datacenters. The physics involved in the data transmission, the power, the cooling, the launching is all very stupid, possible, but stupid given the huge
Re: (Score:2)
Sorry, but you don't "know it can't work because physics".
There are good reasons to believe it's a difficult problem, and it's certainly questionable whether it's a reasonable goal. But I'd bet those chips have other uses, if "Space A!!!" doesn't take off. And I doubt that "Space AI!!" is impossible, I think it's is probably currently too expensive to be practical, however. But I'm not an engineer specializing is design of space projects, so my judgement here shouldn't be taken as "inside information".
Re: (Score:2)
Re: (Score:2)
I too, am puzzled by the claim that cooling things is "more efficient" in space. That's not my understanding at all.
I suppose if you had radiators the size of Wyoming you could radiate enough heat away but I don't see how operating a data center in space would be a financially practical solution.
It seems like a lot of work and expense for something that's going to be insanely expensive to service and maintain.
Also, maybe I missed it but where, exactly, are the financial savings in this concept? How is goin
Re: (Score:2)
I too, am puzzled by the claim that cooling things is "more efficient" in space. That's not my understanding at all.
Most people problem understand space is around 3K and computers on Earth get up to 373K. The heat differential is all they see. The method to transfer the heat on Earth is not practical in space.
It seems like a lot of work and expense for something that's going to be insanely expensive to service and maintain.
The proposal is trying to counter the fact that data centers on Earth are resource intensive. Moving it to space is the simple answer but it is an answer without any forethought that it is practical. It is a practical as sending my food into space to preserve it longer.
Re: (Score:2)
Anything to make line go up. Because that's all that matters nowadays - line goes up.
And the fact that line goes up doesn't require actually spending money or money moving around.
Re: (Score:3)
Please, explain to us how cooling in space is more efficient (without a medium like air or water surrounding you to put your heat into and get it away from your server).
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
And to answer your next question- I'll ask you a question.
Do you really actually not understand why a passive array is more efficient than blowing air over something? (Ignoring the actual amount of blowing a closed-loop datacenter needs to do, which is hundreds of watts per kilowatt of dissipated heat)
Re: (Score:2)
Please, explain to us how cooling in space is more efficient (without a medium like air or water surrounding you to put your heat into and get it away from your server).
It doesn't. The whole more efficient line of argument is rather silly. There is nothing (except economic suicide) stopping anyone from building a data center here on earth that is entirely passively cooled. You attach heat pipes from ICs and power electronics to passively cooled radiators. Convection is way better at dissipating heat than radiation alone so here on earth you would even have a density advantage over space.
Of course in the real world nobody would ever do that because the density of such a
Re: (Score:2)
Convection is more space-efficient, but requires a lot of power.
In practice, assuming you can wave away the cost of lifting your radiators (which is the real limiter right now), cooling in space is basically free, while cooling on Earth requires an ongoing cost of around 300W per kW you need to cool. It's very bad, which is why companies use evaporative cooling when they can get away with it.
So no, the argument isn't rather silly. It's an ine
Re: (Score:2)
Radiation on Earth is far less efficient than it is in space.
Convection is more space-efficient, but requires a lot of power.
This is nonsense, convection requires no power whatsoever.
In practice, assuming you can wave away the cost of lifting your radiators (which is the real limiter right now), cooling in space is basically free, while cooling on Earth requires an ongoing cost of around 300W per kW you need to cool.
This is as wrong as it is irrelevant. The energy cost of cooling in my scenario is 0. Density is what matters and there is no feasible way to achieve it in space.
So no, the argument isn't rather silly. It's an inevitability if you can follow the line on a graph.
No it is just absurd. What you are attempting to do is make an invalid comparison between present day earth based data centers where density is a key cost driver and space based data centers where anything approaching comparable density is financially infeasible.
Re: (Score:3)
And how would a huge radiator array be more efficient than just an earthly server with a heatsink & a fan like on earth?
Because the claim is not that it's possible in space, but that it's more efficient.
Re: (Score:3)
Yes "enough radiators" means something the size of the ISS for a couple of GB300 grade servers.
The possibility isn't the thing, it's the sheer impossible scale is ridiculous compared to "just do it on the surface, where you have a nice convenient whole planet to sink the heat into".
Re: (Score:2)
It means a radiator about the size of the ISS' solar panels per MW, which is around 700 GB300 servers.
The possibility isn't the thing, it's the sheer impossible scale is ridiculous compared to "just do it on the surface, where you have a nice convenient whole planet to sink the heat into".
Yes, the planet is much bigger than..... space
The "nice convenient whole planet" is sitting at roughly 300K. That means it requires hundreds of watts of power to move a kilowatt of heat into that environment.
In space, it's quite literally free.
The complication is in size, but as pointed out, the size might sound large to you, but it's not larger than we already deal with.
Re: (Score:2)
even if i make very optimal assumptions, that's like 100 million dollars for just launching that 100 million dollars of cooling into space (but something closer to a billion is more likely, certainly if you also consider material/construction costs).
And then you still need to generate that MW, and get that into space, etc...
That's about the price i find for a 100MW data center on earth.
Noone is saying that it's impossible, but the cost difference between both is just insane!
Re: (Score:2)
Re: (Score:2)
The ISS power/thermal solution is good for about 100KW, not a MW.
Re: (Score:2)
I'm surrounded by fucking morons.
I said:
It means a radiator about the size of the ISS' solar panels per MW, which is around 700 GB300 servers.
Meaning, If you were to ship up a radiator the size of the solar panels on the ISS, it would be able to cool about 1MW.
The ISS can't cool anywhere near that, because its radiators are tiny.
Re: (Score:2)
The fact remains, something the size of the ISS would be needed to support ~100kw of space-hosted stuff. The radiators are only "tiny" by comparison to the big solar panels, but they are plenty big, and we can't ignore the solar footprint that would be needed.
We should not have tens of thousands of ISS magnitude things in LEO to feed the 'wouldn't space be neat?' hubris of a handful of self-congratulatory "thought leaders".
Re: (Score:2)
The fact remains, something the size of the ISS would be needed to support ~100kw of space-hosted stuff.
100kW? Poppycock.
You could cool 1MW if you had radiators the size of the ISS' solar panels. The ISS doesn't have large enough radiators to do more than 100kW of "space stuff", because it wasn't designed to, and thus has tiny radiators.
The radiators are only "tiny" by comparison to the big solar panels
No, they're tiny.
If you were to make a square of them- about 40x40 feet total.
but they are plenty big, and we can't ignore the solar footprint that would be needed.
Your sense of scale is fundamentally warped.
We can absolutely ignore the "footprint". We care only about the mass.
These things are very easily modular.
We should not have tens of thousands of ISS magnitude things in LEO to feed the 'wouldn't space be neat?' hubris of a handful of self-congratulatory "thought leaders".
Really- we should have them here on the plan
Re:This is concerning (Score:4, Informative)
That it's huge enough to radiate that energy away.
It receives about 300W/(m^2), and is able to radiate enough into space to be in an equilibrium at the current temperature.
A single nvidia h200 server requires about 10kW of power, so if we assume equal radiant cooling as earth to keep a similar temperature to earth, that's 300 square meter of just cooling radiators for just that (and somehow figure out how to prevent those radiators from heating up in the sun, so probably also 300 square meter of heat shield to shield them or so?.
Then of course you also need 10kW of solar panels that also heat up and need their cooling, and all that for just a single h200 server.
Sounds very economical and practical to shoot that into space.
But yeah, how about you stop acting all smug and start actually answering how it would work? Because a simple napking calculatiion shows this would be a space project larger than anything we've ever done before.
Our biggest space project so far, the ISS produces about 80 kW of power, so would be enough for 8 h200 servers.
But yeah, this is all very practical, and not just more hype bullshit.
Re:This is concerning (Score:4, Informative)
Re: (Score:2)
In a vacuum, what absorbs the heat?
You missed the point, didn't you?
Every 24 hours, the Earth receives 175,000 terawatts of power from the sun.
What stops the Earth from turning into a giant ball of magma?
Because things in a vacuum radiate away energy in the form of light. [wikipedia.org]
If this didn't happen, there would be no Earth. It would have vaporized long ago.
Re: (Score:2)
So rather than multiple kilowats going through your AI server, you're gonna run on a fraction of that?
Basically all power used by your chip is the power you have to dissipate, which is already a challenge on earth where we can use air to get rid of the heat.
And the claim is not that it's possible, but more efficient in space.
Re: This is concerning (Score:2)
Then get back to us when you come up with a scheme to remove that waste heat that works in the vacuum of space, and isn't super expensive to implement and maintain.
Re: (Score:2)
Why do you think the surface of the moon is cold, and the surface of the Earth is warm?
Person you ignorantly replied to is correct- the only thing you need to worry about is your internal cooling loop. After that, it's just making sure you have a big enough radiator.
Re: (Score:2)
Re: (Score:2)
You do know the surface of the moon temp has like a 4/500F degree range depending if its in sunlight or not?
200F. 500F would have made it impossible to walk upon.
It's not like space is cold ergo cooling is easy is it.
Space is very cold.
You obviously do not aim your radiator at the sun.
How big is a big enough radiator and how are you going to get it into space and what is it radiating into?
How in the fuck do you think the ISS stays cool?
It radiates into space.
Cooling in space is very easy, and requires a fraction of the power that cooling on Earth does.
The technical requirement is literally as simple as "The radiator cannot face the sun".
The one valid point you make here, is the one I already mentioned above- getting it up there.
And that comes down to launch costs.
Re: (Score:2)
How in the fuck do you think the ISS stays cool?
With massive radiators to cool the passive heat coming from humans and equipment. How massive would the radiators have to be to cool off a data center: orders of magnitude larger than the data center.
Re: (Score:2)
To put it into perspective, the ISS solar panels are ~2500m^2. This is not a difficult engineering problem.
Re: This is concerning (Score:2)
Cool, cool maybe they can use Time Lord technology.
Like I said, get back to me when you hsve a cost effective design that works.
Re: (Score:2)
So the radiator has to be simultaneously both bigger than the space data centre and also smaller than the space data centre so it can stay in the space data centre's shade away from the sun.
Engage your brain.
Like actually fucking engage it.
Why would it need to stay in the shade?
What if high-reflectivity materials worked... And what if the radiator panel could be.... perpendicular to the solar panel...
You're not a stupid person- why the fuck are you acting like you are?
Cool, cool maybe they can use Time Lord technology.
And yet, I think the writers of that show were more knowledgeable in basic scientific concepts than you are.
Like I said, get back to me when you hsve a cost effective design that works.
The mass of said radiator would be ~33T, or half of an F9H launch.
The cost effectiveness is indeed the problem- bu
Re: (Score:2)
The surface of the moon reaches 120-130 C.
Oh, you meant the non-sunlit part? Tough to run a datacentre in space without sunlight.
Re: (Score:2)
Do you think that "make radiator not face sun" is more difficult than "make solar panel face sun?"
This is a simple engineering problem. The requirements are not more extreme than the ISS.
You're shutting off your brain because you have a visceral dislike of the people who want to do this. Making yourself stupid to spare yourself cognitive dissonance is not a fucking virtue.
Re: (Score:2)
a big enough radiator.
Yes, that's the whole problem, "big enough" is just massively big, and you end up with a *whole* lot of overhead cluttering LEO and a whole lot of waste launching it...
And the *only* rationalization offered so far has been "oh, this one company that makes a part for natural gas power plants is backordered". There is zero chance that the *easiest* solution for that is tens of thousands of ISS sized datacenters.
Re: (Score:2)
Mass-wise, it's about half of a F9H launch payload per MW.
Size-wise, it's about 1 ISS total-solar-panel area per MW.
And the *only* rationalization offered so far has been "oh, this one company that makes a part for natural gas power plants is backordered". There is zero chance that the *easiest* solution for that is tens of thousands of ISS sized datacenters.
No, they're merely extrapolating out.
If launch costs lower enough, it quite simply becomes cheaper to put them in space. This is ignoring the ongoing efficiency improvement.
Closed-loop cooling on Earth requires a lot of power. Hundreds of kilowatts per megawatt.
You can cheat by using evaporative cooling, but then you're goin
Re: (Score:2)
However, when the person says something like:
Then get back to us when you come up with a scheme to remove that waste heat that works in the vacuum of space, and isn't super expensive to implement and maintain.
Which is cocky ignorance, then the desire to kindly educate that person is gone. That person is an active threat to this fucking world's grip on reality, which slips further every fucking day.
A perfect blackbody with perfect emissivity radiates 1kW/m^2 at 100C.
Nothing is a perfect blackbody, though. Realistically, you can expect 300-500W, depending on how good you are at getting the heat to the radiator.
Re: (Score:2)
I assume they'd have some kind of nuclear reactor to get the 1.21 GW necessary to start the process, but once the chips get hot enough, they should be able to power themselves.
Re: (Score:2)
It's more efficient to run a datacenter in space, in terms of cooling and power.
I am unaware of any current examples which would support that conclusion. The problem with cooling in space is the lack of a medium of heat transfer. Space is mostly vacuum, and vacuum makes for good insulation. Cooling computers on Earth is dumping heat either to air or water.
Even if a datacenter could be built in space, the next problem would be communication. Retrieving the data to build the model on Earth requires a large fiber optic network. Communicating with a space data center would be via radio. A
Re: (Score:2)
I am unaware of any current examples which would support that conclusion.
It's basic physics.
The problem with cooling in space is the lack of a medium of heat transfer.
Dumping heat to the atmosphere is quite expensive. Here on Earth, we cheat by letting that atmosphere phase change our energy that we have stored in water.
If we do not do that (dry coolers), then it requires a lot of power to get that heat dumped into the atmosphere, as your primarily limited by the temperature differential of the outside air vs. the inside air- its Carnot efficiency. And it's quite bad.
For a closed loop system, space is quite simply more efficient, because the tempera
Re: (Score:2)
Re: (Score:2)
It's basic physics.
And you missed the basic part of it. Heat transfer at the most basic level is transferring the movement of some molecules to other molecules. Space is a vacuum with very few molecules. Radiation is the only way to transfer the heat and radiation is not very effective.
Dumping heat to the atmosphere is quite expensive.
In what world is it "expensive" again? An air cooler for a CPU in this case is quite cheap.
Here on Earth, we cheat by letting that atmosphere phase change our energy that we have stored in water.
That is the biggest bunch of bullocks I have heard. There is no phase change in water coolers for CPUs. The liquid which is mostly water gets hotter or
Re: (Score:2)
And you missed the basic part of it. Heat transfer at the most basic level is transferring the movement of some molecules to other molecules. Space is a vacuum with very few molecules. Radiation is the only way to transfer the heat and radiation is not very effective.
Wrong.
Stop it, you have no idea what the fuck you're talking about.
It requires about 1 ISS-square-footage-of-solar-panels for a radiator to radiate 1MW of heat in space.
This is an actual physicist explaining why space data centers are bad idea [youtube.com]
You didn't watch that video, did you? lol.
Hossenfelder explains it quite simply. So simply, that even your dumb ass should have understood it.
"You need a large radiator to cool efficiency."
Re: (Score:2)
Wrong.
Space is a vacuum. This is a fact. Radiation is the only way to transfer heat in this situation. Please point out where this is wrong.
It requires about 1 ISS-square-footage-of-solar-panels for a radiator to radiate 1MW of heat in space.
That is a lie [nasa.gov]. "The EATCS is designed to provide 35 kW of heat rejection per loop for a total capability of 70 kW." Maybe in your world 1MW == 70kW but not in mine.
Also the size of the radiator panels is not "1 ISS-square-footage-of-solar-panels".
Heat collected by the EATCS ammonia loops is radiated to space by two sets of rotating radiator wings—each composed of three separate radiator ORUs. Each radiator ORU is composed of eight panels, squib units, squib unit firmware controller, Integrated Motor Controller Assemblies (IMCAs), instrumentation, and QDs. Each Radiator ORU measures 76.4 feet (23.3 meters) by 11.2 feet (3.4 meters) and weighs 2,475 pounds (1,122.64 kilograms)
Total size computed: 2 sets X 3 ORUs X 76.4 ft x 11.2 ft = 5134 sq ft. You are just wrong.
Total capacity of the radiator is 7
Re: (Score:2)
Space is a vacuum. This is a fact. Radiation is the only way to transfer heat in this situation. Please point out where this is wrong.
That isn't the part that was wrong.
and radiation is not very effective
That is the part that was fundamentally wrong.
That is a lie [nasa.gov]. "The EATCS is designed to provide 35 kW of heat rejection per loop for a total capability of 70 kW." Maybe in your world 1MW == 70kW but not in mine.
So, you're illiterate.
I did not say 1-ISS-Worth-Of-Radiator-Square-Footage can dissipate 1MW.
I said 1-ISS-Worth-Of-Solar-Panel-Square-Footage can dissipate 1MW.
It requires ~2800m^2 to dissipate 1MW in cynical conditions. Reality is better.
ISS has ~2500m^2 of solar panels.
The point of saying this, is to demonstrate to you that the size we're talking here isn't obscenely huge.
ISS has smaller radiators than that, because it
Re: (Score:2)
That isn't the part that was wrong.
But you said that part was wrong.
I said 1-ISS-Worth-Of-Solar-Panel-Square-Footage can dissipate 1MW. . . . ISS has ~2500m^2 of solar panels.
What kind of nonsense unit conversion is this? That's like saying my Air Cooler can dissipate "1-Coca-Cola-Cans sq ft" worth of 1 BILLION candles of heat. The fact of the matter is it takes 5,134 sq ft of radiator to dissipate 70kW of heat. That is orders of magnitude lower than a simple CPU air cooler. For a small datacenter to dissipate 10MW of heat would require 733,428 sq ft of panels.
Re: (Score:2)
But you said that part was wrong.
Actually, when you asked me "what part of that is wrong", you re-stated what I said was wrong excluding the part that was actually wrong
What kind of nonsense unit conversion is this? That's like saying my Air Cooler can dissipate "1-Coca-Cola-Cans sq ft" worth of 1 BILLION candles of heat. The fact of the matter is it takes 5,134 sq ft of radiator to dissipate 70kW of heat. That is orders of magnitude lower than a simple CPU air cooler. For a small datacenter to dissipate 10MW of heat would require 733,428 sq ft of panels.
It wasn't a conversion- it was to give you a sense of the size.
For 1MW, it requires about as many square meters of radiator, as the ISS has in solar panels- demonstrating that it is well within our capability of putting shit that large up there.
As for your math?
Completely wrong.
It requires about 28500m^w of panels, or 306,000 square feet.
The EATCS, which is old, requi
Re: (Score:2)
For 1MW, it requires about as many square meters of radiator, as the ISS has in solar panels- demonstrating that it is well within our capability of putting shit that large up there.
Which is irrelevant to total actual size needed to dissipate that heat. By "within or capability", I do not recall any space station with 73,342 sq ft of radiators to dissipate 1MW of heat.
It requires about 28500m^w of panels, or 306,000 square feet.
The ISS has 27,000 square feet of solar panels. So you lied again.
The EATCS, which is old, requires about 3000m^2/MW. Modern systems can do about 2800m^2/MW
And what is the capacity of the your system? Oh it doesn't exist, does it?
In total, the PVR can reject up to 14 kW of heat into deep space. The PVR weighs 1,633 pounds (740.7 kilograms) and when deployed measures 10.24 feet (3.12 meters) by 44.62 feet (13.6 meters)."
*Sigh* That was for the PREVIOUS radiator system. You can't read can you?
Let me know if you need any more help with the math.
The problem is you were caught in a lie and now you are trying to fudge numbers by gaslighting. Bottom line:
Re: (Score:2)
Which is irrelevant to total actual size needed to dissipate that heat. By "within or capability", I do not recall any space station with 73,342 sq ft of radiators to dissipate 1MW of heat.
It doesn't require 73,342 sq ft of radiator to dissipiate 1MW of heat.
The ISS has 27,000 square feet of solar panels. So you lied again.
We were talking about space it would take to cool 10MW. What in the fuck are you talking about?
And what is the capacity of the your system? Oh it doesn't exist, does it?
What? Do you think satellites don't have radiators?
ok, dude.
Here's the thing- you are way too fucking stupid to have this conversation.
Every single post I make is correcting 10 fucking things you got wrong or misunderstood, and you just can't be deterred- you just keep throwing spaghetti at the wall with your fucking potato-ass IQ.
Re: (Score:2)
It doesn't require 73,342 sq ft of radiator to dissipiate 1MW of heat.
Please show me a system that can dissipate 1MW of heat. It doesn't exist. By the current designs of the current ISS radiator it will be 73,000 sq ft.
We were talking about space it would take to cool 10MW. What in the fuck are you talking about?
You said the ISS has "306,000 square feet." of panels. It does not . You lied again.
What? Do you think satellites don't have radiators?
Again, please show me a satellite that can dissipate 1MW, 10MW of heat. It doesn't currently exist, does it?
Here's the thing- you are way too fucking stupid to have this conversation
Logical Fallacy: Ad hominem attack.
Every single post I make is correcting 10 fucking things you got wrong or misunderstood, and you just can't be deterred- you just keep throwing spaghetti at the wall with your fucking potato-ass IQ.
Again on the ISS right now it requires 5,134 sq ft of radiator panels to dissipate a max of 70kW of heat according to NASA. Instead of
Re: (Score:2)
Please show me a system that can dissipate 1MW of heat. It doesn't exist. By the current designs of the current ISS radiator it will be 73,000 sq ft.
No. It would be 32291.73ft^2 (3030m^2)
Again,
"In total, the PVR can reject up to 14 kW of heat into deep space. The PVR weighs 1,633 pounds (740.7 kilograms) and when deployed measures 10.24 feet (3.12 meters) by 44.62 feet (13.6 meters)."
Let me know if you need any additional help with the math.
Begone. Adults are talkig.
Re: (Score:2)
Sigh. Current radiator system 70,000 W of heat dissipation using 5134 sq ft.
Heat collected by the EATCS ammonia loops is radiated to space by two sets of rotating radiator wings—each composed of three separate radiator ORUs. Each radiator ORU is composed of eight panels, squib units, squib unit firmware controller, Integrated Motor Controller Assemblies (IMCAs), instrumentation, and QDs. Each Radiator ORU measures 76.4 feet (23.3 meters) by 11.2 feet (3.4 meters) and weighs 2,475 pounds (1,122.64 kilograms)
Total size computed: 2 sets X 3 ORUs X 76.4 ft x 11.2 ft = 5134 sq ft. 70,000 W / 5134 sq ft = 13.63 W/sq ft. To size 1 MW = 1,000,000 W / (13.63 W/sq ft) = 73,000 sq ft.
"In total, the PVR can reject up to 14 kW of heat into deep space. The PVR weighs 1,633 pounds (740.7 kilograms) and when deployed measures 10.24 feet (3.12 meters) by 44.62 feet (13.6 meters)."
Again, that was PREVIOUS radiator system. The current system is not the previous system. Why must you lie so much?
Re: (Score:2)
You're talking about the room temp radiators.
I used the power radiators for a reason- they emit twice as much W/m^2.
A space datacenter will not be trying to cool room temperature heat loads. Stefan-Boltzmann scales at the 4th power of temperature.
Re: (Score:2)
> For a closed loop system, space is quite simply more efficient, because the temperature differential is larger, period.
Wikipedia cites a printed book and mentions 100 to 350 watts per square meter for radiative cooling in space.
As a low ballpark, one cubic inch of automotive radiator has about 24 to 30 square inches of surface area, and even more approximations say you need about 1 cubic inch per horsepower which is roughly 250 watts of thermal dissipation to the air.
So car radiator is, on the low end,
Re: (Score:2)
Wikipedia cites a printed book and mentions 100 to 350 watts per square meter for radiative cooling in space.
Correct.
As a low ballpark, one cubic inch of automotive radiator has about 24 to 30 square inches of surface area, and even more approximations say you need about 1 cubic inch per horsepower which is roughly 250 watts of thermal dissipation to the air.
Your automotive radiator requires power, in the form of moving air.
Remember, the word above given was efficiency.
So car radiator is, on the low end, about 1500 times better at rejecting heat than a radiator on a spacecraft. And that's with the temperature difference advantage.
Anything with a fan will be "better". It will not be more efficient.
At MW scales, the cost to move the air is not insignificant. Particularly because nobody has figured out how to put a datacenter on wheels and roll it down an infinitely long hill yet.
Re: (Score:2)
It's more efficient to run a datacenter in space, in terms of cooling and power.
I know little about orbital mechanics, so I'm curious; how do they ensure that their waste heat radiators are always on the dark side of the craft, in the shadows where they can't do the opposite of what they're intended to do?
BTW, it seems to me that "shadow" part would imply a very large craft, unless you're going to deploy big reflector panels which always face the Sun. And at that point, wouldn't you have a kind of solar sail whose effect on your trajectory or attitude you'd have to counter? I don't kno
Re: (Score:2)
I know little about orbital mechanics, so I'm curious; how do they ensure that their waste heat radiators are always on the dark side of the craft, in the shadows where they can't do the opposite of what they're intended to do?
You don't.
You make sure they're perpendicular to the sun, like the ISS, at least as much as reasonably can be, and you coat them in high-reflectivity materials, like the ISS
The more oblique the angle, and the higher the surface albedo, the lower the insolation of the panel. Realistically, you can get it down to a few tenths of a percent.
They missed the obvious solution! (Score:3, Funny)
Huang: stop with the LSD, NOW (Score:2)
"Space computing, the final frontier, has arrived" Too bad we don't get to see him jumping around a stage in a Star Trek uniform.
I miss the Monkey Boy Dance from Steve Ballmer: https://www.youtube.com/watch?... [youtube.com].
Elmo cannot hold a candle to him: https://www.youtube.com/watch?... [youtube.com].
And we should give la Presidenta honorable....well, nothing about him is honorable....let's just leave it at mention: https://www.youtube.com/watch?... [youtube.com].
These guys need to form a dance troupe behind Huang when he goes into orbit on hi
Bullshit (Score:2)
Re: (Score:2)
Re: (Score:2)
Nah. Oxygen++ would turn all of the space computers into Rust. :)
I don't think he is talking about satellite level (Score:2)
They don't make satellites.
Their problem is probably more making lightweight boiling refrigerant based cooling blocks which can keep the their system cool in a vacuum, including PCB, connectors, cables, etc. Dumping the heat from refrigerant to space is not their problem.
Re: (Score:2)
Have they even considered how much radiation hardening they need - even though they will still operate within the Van Allen belt, the increased radiation is a serious concern.
Re: (Score:2)
A vacuum is the perfect insulator. How do they cool in space? heat pipes are fine but you have to radiate heat somewhere and it's all in a vacuum.
Re: (Score:2)
Radiators.
I don't know if it will be better to use heatpumps or just use more area. With heatpumps, you could probably get away using front for PV and back for radiators. Without heatpumps you'd probably have a bunch of the sun side area be mirrored, with more radiator area than PV.
Re: (Score:2)
You can ONLY radiate IR light - you can't "radiate" like you do on earth which is mostly done by convention not by giving off IR light and that takes AIR to do. you need to dump a ton of watts somewhere and light is all you've got to work with -- last I looked, my CPU/GPU isn't lighting up like a light bulb.
Re: (Score:2)
You DON'T want to boil off coolant in space, because resupplying it is quite difficult. Since you need radiation hardened chips anyway, what you need to do is use designs that work at a higher temperature, so cooling is a smaller problem. (Basically you're going to need to depend on radiation based cooling at the system level. The efficiency of that goes up, IIRC, as the fourth power of the difference between the emitter temperature and the background.)
OTOH, these were probably designed to meet the specs
I keep hearing space datacenters (Score:2)
What I am not hearing is what problems this actually solves..
Cooling - nope you can only radiate, so that is actually harder, maybe you can get the waste heat really low and turn most of it into power generation - but that is expensive.
Power - Nobody wants to put that much radio active isotope in orbit for obvious reasons, solar works really well in space but that is a lot rather expense material and lift cost.
Space - well there is a lot of space in space... but there is a lot of space on the ground too tha
Re: (Score:2)
What I am not hearing is what problems this actually solves..
It keeps suckers from losing interest in propping up comically inflated stocks.
None - except raising share prices of NVidia (Score:2)
Datacentres in space are such an obviously bad idea that only an idiot or a grifter would even suggest them and Huang is no idiot.
it does makes sense... for Musk (Score:2)
Launch cost - irrelevant xAI now part of SpaceX
latency comparable or reduces compared to other AI datacenters - up to Starlink/Datacenter then down with no other connection needed
Same setup works for moon and Mars
Re: (Score:2)
Power: We have plenty of power now, and as was stated, the panel sizes are still huge for the power needed, even in space
Launch cost: Doesn't matter who 'owns' it, the launches are still expensive, regardless of how you move the shell game of accounting
Latency is still at a disadvantage compared to a boring old fiber, regardless of your fondness for Starlink
There is *zero* reason to be concerned about a common solution for Moon and Mars at this point. Even if there were, landing on those rocks provides a m
Re: (Score:2)
Vera Rubin (Score:2)
If I were Vera Rubin's relatives I would be pissed that her name is being used for something so utterly unintelligent.
Re: (Score:2)
What's going to kill space data centers (Score:2)
Is cosmic rays, they flip bits, cause resets.
And high computation applications are especially sensitive.
Re: (Score:2)
Too many problems (Score:2)
There are good reasons why satellites might need one or two powerful chips, but there are no good reasons to build orbiting data centers. It makes far more sense to build more small data centers on the planet if your goal is to avoid outages. A satellite is located where any major power can easily shoot at it, so all you've done is reduce the pool of potential bad actors to those who have the greatest motive while increasing the odds of Kessler syndrome.
Data centers already have the problem that once you've
Re: (Score:2)
Though think of the opportunity for nVidia, if they get everyone sold on the 'your GPU will burn up on re-entry in 5 years', then they never have to worry about a client delaying a refresh...
Orbital data centers (Score:2)
There is no air in space and air is how cooling works. So if you don't have air you have to use enorm
Venture Scam (Score:2)
Why do people keep talking about putting something HOT in space. Outer space is NOT COLD. The main requirement for computing (after power) is cooling. Infrared radiation is the usual form of cooling, and that ain't fast. They have to know this. So either someone is Dunning Krugering all over the place, or it's another scam to get money from people who are ignorant of the reality.
not radiation hardened? (Score:2)
This Is Peak AI Bullshit, Right Here (Score:2)
The bubble gonna pop any second now. NVIDIA wasting resources on "space chips" when supply chains are currently so restricted is the kind of hype that happens when the bubble is hyped out.