Close Menu
    Zukunftsforschung

    Tech-Milliardäre wollen Rechenzentren im Weltraum errichten. Die Mathematik könnte schnell hässlich werden.


    Tech-Milliardäre wollen Rechenzentren im Weltraum errichten. Die Mathematik könnte schnell hässlich werden.

    https://www.businessinsider.com/elon-musk-jeff-bezos-space-data-centers-scientists-ask-why-2026-4?utm_source=reddit&utm_medium=social&utm_campaign=insider-futurology-sub-post

    View 11 Comments

    11 Kommentare

    1. businessinsider on

      **From Business Insider’s Ellen Thomas:** 
      From the coming SpaceX IPO excitement to Nvidia’s foray into space computing, data centers in space have become the tech industry’s new favorite talking point.

      The space economy could become a $1.8 trillion opportunity by 2035, according to consulting firm McKinsey and Company. At the same time, while some major companies are ready to spend billions launching GPU-loaded satellites into orbit, not everyone on this planet is buying in yet.

      Tech leaders like SpaceX CEO Elon Musk, Amazon founder Jeff Bezos, and Google CEO Sundar Pichai are framing data centers in space as a cost-effective solution to AI’s insatiable appetite for resources, such as land and power.

      Proponents of data centers in space make a simple case. Space data centers can get free power from the sun and don’t have to pay for real estate — and best of all, both resources are unlimited.

      Critics wary of data centers in space say the endeavor is far more technically complicated, cumbersome, and expensive than it might appear. OpenAI CEO Sam Altman has also called the idea „ridiculous.“

      …
      Scientists cited a litany of obstacles that space-based data centers face.

      Matthew Buckley, a theoretical physicist and associate professor at Rutgers University’s department of physics and astronomy, estimated that every data center launched into space would require 450 football fields of solar panels to power it.

      That alone could cost $10 billion per data center, he estimated, on top of an additional $10 billion to launch it into orbit — and that’s not counting the cost of maintaining the data center in space.

      [Read more about the pushback on data centers in space. ](https://www.businessinsider.com/elon-musk-jeff-bezos-space-data-centers-scientists-ask-why-2026-4?utm_source=reddit&utm_medium=social&utm_campaign=insider-futurology-sub-post)

    2. EnchantedTaquito8252 on

      Why do people want data centers in space? Is it just because people incorrectly assume space is cold? 

    3. Narf234 on

      I don’t see what the issue is. The opponents are saying that it’s technically difficult and expensive. Big deal? If companies want to use their investor’s money to try something why not let them?

    4. WiglyWorm on

      I’d like to be able to see the dentist.

      Tax these rich assholes already.

    5. jesterOC on

      Let them try, it if works (it won’t) great no more data centers here sucking up all our resources.

    6. parkway_parkway on

      „Buckley, the theoretical physicist, estimated that every data center launched into space would require 450 football fields of solar panels to power it.

      That alone could cost $10 billion per data center, he estimated, on top of an additional $10 billion to launch it into orbit — and that’s not counting the cost of maintaining the data center in space.“

      Elon wants to build a swarm of small sats which are similar in size and power to the Starlink satellites and then have them all communicate by optical link.

      That way the difficulties of putting them up are no greater than for Starlink which they are already doing.

      And yes $20b is nothing compared to the $600b they’re spending on AI build out this year.

      As for maintainance you just don’t bother. The chips have a lifespan of 3-5 years and so after that you let them burn up in the atmosphere and if something goes wrong with one of the swarm you just down it early.

      These critics really have no clue what they’re talking about.

    7. xxAkirhaxx on

      There’s something else going on with the whole ‚data centers in space‘. AT best AT BEST…we make data centers up there, and in order to get past the heat problem, because no air, so heat has no where to go, we expel the heat onto x (x being a variable, not the nazi platform). Then we eject x into space. The result would be, in orbit we would have x all over the place. In the amounts we’re talking about, we’d probably eventually reverse global warming, but not climate change, because climate would begin to change, just the other way instead because x would begin to accumulate and stop sun light from reaching the planet. This would drive down the efficiency of solar power, cool the planet far too low and kill off millions of species. It would be a massive, massive issue.

      And that’s the thing, this happens no matter what x is. The only way it wouldn’t happen is if x allowed sun light to pass through it. And even then you’d have to worry about the side affects of whatever x is permanently being in orbit in space, where our most precious data centers would be.

      Like what the fuck? Want a better idea? Build the data centers in deserts, use the solar power from the sun to cool them, and if water becomes a problem, pump it from the fucking ocean, it doesn’t have to be drinkable so desalination becomes a much easier process. It only needs to dissipate heat enough, not at the same rate as fresh water.

    8. garry4321 on

      This is such a fucking stupid concept. It would overheat so quickly. Heat is an issue in space. Oh, one $20 hardware component broke? Whelp, better jump into a fucking spaceship and spend millions to go replace it

    9. misterguyyy on

      ~~Tech Billionaires~~ The owners of SpaceX and Blue Origin ~~want to put data centers in space~~ are trying to milk the AI hype for another grift.

      ~~The math could get ugly fast~~ Capitalism allocates resources most efficiently, and by that I mean that people with hoarded wealth get to blow millions on crackhead bus rant tier ideas and still end up rich. Therefore we’re taking those ideas seriously.

    10. KidKilobyte on

      Arm-chair physicists on Reddit explaining why it won’t work. Thinks billionaires don’t do their homework before investing. Maybe they’ll burn thru a couple of million and decide it’s not worth it. But could be Starship will dramatically lower launch costs and make this practical. We have been studying putting Solar Power in space for decades. With Starship those numbers would work. But why send the power back to Earth by microwave, risking unknown problems, when you can use it in situ with the fastest growing consumer of electricity.

      If they actually start doing this, rest assured it will work. It won’t be a case of we won’t know until we get there. It’s all about the money and numbers and how much Americans hate living next to data centers.

    11. _ii_ on

      No body is launching a data center to space. Laser inter-connected compute racks in space is more like it. You’ll need large solar panels and radiators but not football field size like the article claims. The author of the article simply doesn’t understand how distributed systems work.

      Some companies, including Google, have been working on the “space data center” concept for a while. It’s not new, it didn’t originate from Musk, and the idea has merits.

    Leave A Reply