„Die Server in den Orbit zu bringen ist eine dumme Idee“: Könnten Rechenzentren im Weltraum helfen, eine KI-Energiekrise zu vermeiden? Experten sind hin- und hergerissen. Googles Vorschlag, eine weltraumgestützte KI-Infrastruktur zu erforschen, wirft grundlegende Fragen zu Energie, Physik und Machbarkeit auf – und ob die Erde keine Optionen mehr hat
https://www.livescience.com/technology/artificial-intelligence/putting-the-servers-in-orbit-is-a-stupid-idea-could-data-centers-in-space-help-avoid-an-ai-energy-crisis-experts-are-torn
3 Kommentare
Earth would have so many options if its resources weren’t being used for extraneous bullshit
Immensely awful idea unless you’re also mining the resources to build them in space.
Weight of 42U server: ~1000 kg.
Cost of sending it to space: $1 million USD
Cost of building solar on earth to power it: ~$7500
So, it’s still about a hundred times cheaper to just leave it on the ground, only counting the launch costs, not even accounting for additional tech costs, maintenance, R&D, etc.
It’s just more space junk
>*“They found that regions in low Earth orbit (LEO) at altitudes around 300 miles (500 kilometers), where most satellites of megaconstellations like SpaceX’s Starlink reside, could see a collision in as little as 2.8 days. For comparison, the team ran an identical simulation with numbers of satellites and space debris in orbit from 2018. At that time, it would have taken 128 days for the first collision to occur…“*
[https://www.space.com/space-exploration/satellites/crash-clock-reveals-how-soon-satellite-collisions-would-occur-after-a-severe-solar-storm-and-its-pretty-scary](https://www.space.com/space-exploration/satellites/crash-clock-reveals-how-soon-satellite-collisions-would-occur-after-a-severe-solar-storm-and-its-pretty-scary)