[taffydavid]

Why are we considering data centers in space? What problem does it solve? I don't understand this concept at all.

Surely putting server racks in orbit where you can't do basic maintenance without sending someone up there is a terrible idea?

It just always puts me in mind of the story of an IT guy having to travel for two days to go to a premises and turn on a server that three separate people ensure him was already on.

[martheen]

The only people considering data centers in space are those too out of touch to remember that their vacuum flask retain heat for hours. At any usable capacity we'd be talking about massive field of dissipators, which can't be possibly profitable considering there are far more lucrative payload compared to dumb panels. Some would pretend there's a just around the corner tech to solve the heat waste, but if such tech really exists one of the criticisms of terrestrial data center (wasting stupid amount of water) would've been solved anyway making space data centers even less viable.

And there's also bitflips. Acceptable for astronauts carrying normal laptop to watch movies, but in a data center? Solving it through redundancy easily cut the usable capacity to half, and there's not enough shielding with any competition for the payload slot.

The usual criticism of Mars colonization is maintaining a base in Antarctic is already a heroic effort despite being relatively easier. Google and Microsoft mostly abandoned their "data center underwater in nearby shore" plans even though in those the tech would've only sat for few minutes going down or alternatively waiting for the problematic module being floated to the surface.

[taffydavid]

Great, so I'm not just some idiot who doesn't get it. Why then is anyone entertaining this IPO when it's clearly based on such a dumb plan?