this post was submitted on 01 Jul 2026
1065 points (98.4% liked)
Technology
86036 readers
5241 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
LOL LOL you CANNOT cool something like that in space. The entire concept is flawed.
You can, they just have to be smaller rather than a massive single orbital data centrer like this proposal. https://www.youtube.com/watch?v=FlQYU3m1e80
Still not a great idea because of the economics, but the same can be said for the data center build out on earth too, so why would they let that stop them?
For scale. roughly a two server datacenter needs to have solar and radiator about as big as the ISS.
Which is possible, but insane. However insane plain old datacenter is, just tons more insane.
That simply isn’t true, the video I linked explains everything clearly, for a 20kw satellite the cooling area is needed very modest.
Still not a great idea and I am not advocating for it, but people need to stop fighting bullshit with bullshit and start fighting it with truth.
For reference, a GB300 server is now at about 8.5kw for a single server. A fully populated NVL8 server is about 15kw. Looking online, looks like the ISS is about up to 90kw, so I guess I'm off and the actual number is something like 6 to 10 of these servers per ISS scale facility.
I would still argue this is a crazy overhead for what they would now consider meager capacity, with, luckily for nVidia, a pretty hard deadline where the very expensive equipment burns up without potential for extended lifetime use.
For a single satellite, we'd need a football sized array for heat dissipation. The dissipation capacity isn't equal across the entire array. And you need some way to move the heat from the centre out towards the edges.
And aside from that, 100k satellites is the limit of objects we can put into low earth orbit before we start risking cascade collisions that break everything into small bits and make getting anything into orbit impossible. We're currently at 14k objects. Space X is proposing ONE MILLION satellites. And they'll each need huge heat dissipation arrays.
That simply isn't true, the video I linked explains everything clearly, for a 20kw satellite the cooling area is needed very modest.
Still not a great idea and I am not advocating for it, but people need to stop fighting bullshit with bullshit and start fighting it with truth.
I know who Scott Manley is. I'm subbed to his channel and saw that video when it came out.
That being said, one of the top comments in that video is someone who, claims to be, a spacecraft thermal engineer. And they bring up a few good points. But the one I'm most interested in is the loss of efficiency in heat dissipation the further the heat is pushed along the array. Which means you can't treat the entire surface area of the dissipation array at equal performance, so you need an even bigger array.
And btw, a 20kw satellite is peanuts for AI workloads. Which is the reason they're suggesting putting up a million of these. And that right there is, IMO, the biggest issue. We're already at 14k satellites (most of those are Star Link). And 100k satellites is the current figure we expect collisions between satellites to start becoming unavoidable, with the possibility of an out of control cascade of collision becoming a major concern from there upwards.
I think Kyle Hill did a better job at being objective on the problem:
https://www.youtube.com/live/4mx9Rp-SMNk
A 20KW data center is utterly feeble, barely worth the name, by ground-based data center standards, which easily go into tens of megawatts.
It's a single rack rather than it being a whole data center. The whole constellation of satellites is the data center.
Again, I'm not advocating for this as a good idea, just saying that cooling is not the reason it is a bad one.