Bezos und Musk wetteifern darum, Rechenzentren ins All zu bringen
https://www.wsj.com/tech/bezos-and-musk-race-to-bring-data-centers-to-space-faa486ee?st=6wPhmY
Bezos und Musk wetteifern darum, Rechenzentren ins All zu bringen
https://www.wsj.com/tech/bezos-and-musk-race-to-bring-data-centers-to-space-faa486ee?st=6wPhmY
21 Kommentare
If only we had any problems left down here which these billionaires could help with all their tech and money. But alas …
We should tax these losers back into being millionaires and use that money fix the problems they’ve created with their greed.
It’s going to be a nightmare shuttling all of the coal and natural gas those data centers need into space.
May the space debris find these data centers swift and with zeal.
I’m beginning to think that these guys are kinda dumb and got so wealthy by sheer chance.
What problems does bringing data centers to space solve?
How are they going to handle cooling?
I’d love to see their plans for radiation/particle shielding, hardware failures like disk drives, cooling, power, cables getting jostled during launch, etc.
„We’re going to build this (expensive boondoggle with no use case)!“ Stonks go up.
A year or so later:
„We’re cancelling the program to build this (expensive boondoggle with no use case)!“ Stonks go up.
The one thing space isn’t really good at is transferring heat. The one thing data centers are really good at is generating heat. Let’s race to see which one can drain the taxpayers dry first.
“Race” – Blue Origin have done impressive stuff recently, but they’ve only launched two rockets into orbit in their entire history. Isn’t it a bit early to discuss their plans to launch megatons?
Great! I hope they both go there! Bye bye!!!
All I read about is how all these data centers are going to require gigawatts of energy and billions of gallons of water, neither of which are abundant in space. So I get space is really cold, so that takes care of cooling and the water problem, but where are they going to get the power?
I am surprised how many comments mention cooling / dissipation of heat as the obvious problem that will certainly stop this project.
It won’t. While cooling things in space is somewhat more complicated than on Earth, relatively cheap heat radiators will do the job just fine, plus you can reduce the problem somewhat by shielding the computers from the Sun using the very same solar panels that produce the necessary energy.
Guys and gals, I get your billionaire hate, but don’t let it cross into wishful thinking. These are two functional space corporations that have built their own launch vehicles. And they won’t be stymied by the heat rejection problem. Other problems, maybe, but not this one.
The entire notion is absurd.
I really don’t understand why every insane thing they say is being covered liked it’s actually feasible (or sensible).
Actual science question though… how do you cool things in space when there’s no air to dump heat into? Isn’t a vacuum the perfect insulator?
This is a scam to try to hide the ball on the ridiculous data center spend with a shiny object. The ISS has a power capacity of 120kW and generally runs in the 90kW range. The Thermal Control System can manage that much along with the astronaut waste heat (not trivial!). The ISS as the largest orbital structure ever, has an estimated total cost of $150 billion and a mass of 450 metric tons.
A data center is in the 100+ MW range, some of them are over 300.
Enough said.
So, building a giant thing that generates huge amounts of heat and putting into a vacuum where there’s nothing to take any of that heat out of the system is a recipe for… ? Things to get very hot?
Yeah, because data centers require no maintenance after you fire them up, just set it and forget it. /s
Good to know that PC parts are going to be unaffordable and then destroyed in space in the name of slop generation.
How will they cool the servers?
I read the article, and I get that they’re saying this is still the experimental stage and that it may take 20 years to build out to the point it makes sense … BUT … I don’t think their end goal is actually for these satellites to host AI workloads in space. Yes, the solar power would be a much more efficient way of addressing the high power needs of AI workloads, but it still doesn’t make much sense when you figure it’d be orders of magnitude cheaper to build out better on-planet power infrastructure over the next 20 years. Additionally, wouldn’t putting AI workloads in space introduce a huge amount of latency when the data needs to be transferred to space as opposed to along an existing fiber optic cable? Plus there’s the obvious problem of maintenance. There is something else driving this push of data centers into space, imo. Perhaps it’s the AI processing capabilities required for other satellites, which could be more easily served by servers in space (lower latency)? Perhaps its actually other projects these large tech companies are working on that run better in a zero gravity environment free of „noise“ and interference, like a quantum computer. I don’t know. I just know it doesn’t make much sense for a satellite to render stupid AI video memes.
It would be more popular and practical to launch a Chuck E Cheese into space