space is cold!
No, space is empty. If you want to be pedantic, space is very hot actually
The Physics of Data Centers in Space
Submitted 1 day ago by JensSpahnpasta@feddit.org to technology@lemmy.world
https://blog.koehntopp.info/2026/02/25/physics-of-data-centers-in-space/
Comments
FilthyShrooms@lemmy.world 1 day ago
turboSnail@piefed.europe.pub 20 hours ago
How do you measure the temperature of particles that aren’t there?
Also, the hot argument refers to intense solar radiation, which is available only on one side of the satellite. The other side doesn’t receive sunshine, so it will loose heat.
FilthyShrooms@lemmy.world 18 hours ago
That’s the thing, there are particles there, just very few. Temperature is measuring how fast the particles move, and they move pretty fast in space. We feel heat based on how much energy the particles transfer to us, but because there’s so few particles, it would feel cold.
This is the same reason why getting rid of heat is so hard in space. The best ways of cooling here on earth involve giving those particles the energy we don’t want, but because there’s so few in space, conventional cooling is nearly impossible.
gandalf_der_12te@discuss.tchncs.de 1 day ago
This also reminds me of something else: Yes, you can have all kinds of problems in space missions. Not just bit flips.
I have been thinking about planning a mars mission. And one of the things that you need to guarantee is the safety of the crew in untested environments. Like on the surface of mars, you have to guarantee that the crew can survive at least until the next ship can arrive to help them. And launch windows to mars appear only once every two earth years (once every martian year), so crew has to survive that long.
You also have to survive when there’s a power outage for any reason. Could be that the coworker stumbled over a cable and pulled out a power line’s connection. Could be that there’s 3-months-long dust storms (which would block solar panels). (they appear each martian year, sometimes more sometimes less severe, i think typical duration is less than 50 days, but just to be sure make that 100 days). Could be that there’s a problem with your electronics and they simply won’t work anymore and you can’t figure out why.
So imagine you’re standing in your apartment. I recommend you live inside the Starship for the first few years because it’s already a habutat since you flew in it for 6 months to get to mars in the first place.
Power outage. How do you get out of the door? There’s obviously a pressure door between your apartment and the outside world. And the power line’s outside. So if the power line disconnects because a rover messed up and pulled it out, there’s a power outage. And if the door is electrically operable, you can’t get out of the door to repair it. That is why i say that every door has to be manually operable. Like, pressure doors should have manual operation mode. You need to be able to operate the pressure door with your bare hands. I already thought of some designs that can actually deliver that, but i’m too lazy to do a drawing now.
HertzDentalBar@lemmy.blahaj.zone 2 hours ago
SuspciousCarrot78@lemmy.world 1 day ago
Wouldn’t the more logical first approximation be to bury them underground, and then progress towards (perhaps) placing them in or near the ocean (obviously, within sealed containers, yadda yadda, salt corrosion, yadda yadda, inhospitable environ yadda yadda makes Poseidon angry).
I like the “yeet them into the sea” idea conceptually because (1) yeet them into the sea (2) in theory, you could power them via tidal/wave/OTEC (3) water cooling.
Seems…too obvious. There’s probably a good reason (or bad ones - $$$) why this hasn’t been tried yet. But I bet those reasons are eminently more solvable that “send em into space”
Cocodapuf@lemmy.world 20 hours ago
Yeah, I mean take advantage of geothermal heating/cooling. It does seem obvious. The only actual advantage to space is the 100% solar availability, but that’s actually not a huge advantage in the grand scheme of things.
lepinkainen@lemmy.world 1 day ago
People don’t read enough hard sci-fi, in a realistic space battle you use missiles and other projectiles because they don’t heat up your ship
If you have any kind of laser crap, you need to dissipate the heat and that requires surface area
And that surface area is an easy target for your enemy’s kinetics… take out the cooling system and everyone is cooked alive. Nice.
EndlessNightmare@reddthat.com 1 day ago
This seems monumentally stupid.
jaschen306@sh.itjust.works 1 day ago
Don’t you need an atmosphere to transfer heat? Otherwise the dissipation is super slow.
MonkderVierte@lemmy.zip 21 hours ago
A swarm of ideally flat slab-like satellites, vould actually work. Sun side for solar, back side a sirfaxe optimized for infrared heat radiation.
But still, even with 10x10 meter and x1000 pieces, you only get the equivalent to a small crunching center. Ignoring the problems of strong radiation and replacing parts.
WanderingThoughts@europe.pub 1 day ago
You would need to launch so ridiculously much into orbit. The ISS is rounded up 200 kW of solar power. 0.2 MW. Say you want the equivalent of a gigawatt data center. 1000 MW. Yeah, that’s about 5000 ISS sizes objects. It needs a bunch more cooling, and a data center doesn’t need habital zones, so a pretty barebones ISS. Launch about 3 roughly ISS sized objects per day and you’re done in 1.7 years. Somehow I don’t really see that as a realistic plan.
GreyEyedGhost@piefed.ca 1 day ago
The current solar panel system of the ISS weights about 8 tonnes, the Falcon Heavy can deliver 63 tonnes to LEO. That’s about 715 launches of the Falcon Heavy, assuming space solar panel W/kg hasn’t improved since then, that Starship never becomes commercially viable, and doesn’t include batteries, cooling, or the working components. This still isn’t in the range of feasible for a data center, but could be an option for microgravity industry. The value of a more successful or precise silicon crystal production method, for instance, may make it worthwhile.
phoenixz@lemmy.ca 1 day ago
And yet another lost of reasons why these billionaires are fucking removed. Yes, im using the horrible R word, but “idiots” just doesn’t suffice.
These people are investing billions of dollars on unproven technology and now want to put trillions more in pushing those unproven technologies to space which is just removed levels of stupid. But the investments will happen, a ginormous amount of pollution will be made and it’s all nothing for humanity and a huge amount of money for a select few billionaires.
Fuck this time line, fuck removed billionaires
We need a wealth cap. Nobody should be able to do shit like this with impunity just because they have “enough money to do so”.
Give me a 1M wealth cap world wide and I’ll give you a world where everyone is happy at relatively equal levels of wealth
gandalf_der_12te@discuss.tchncs.de 1 day ago
Like, very interesting article. It also mentions bit flips.
I think the major protection against bit flips is to use larger structure sizes. Like, today’s desktop computers use structure sizes around 2 nm or sometimes even less than that. Which means you only need to shift a few electrons around to cause a bit flip. If you use larger wires inside the processor, they store more electrons so it takes more energy to flip them around. So there’s a much much smaller number of radiation particles that have a minimum of that energy, so lower risk of bit flips.
Does anybody have actual numbers on the structure sizes needed to effectively prevent bit flips? I mean, outer space missions like NASA flights already need this today.
gandalf_der_12te@discuss.tchncs.de 1 day ago
You know how much it costs to transport stuff into space? Well, currently, around $1k/kg of material. Let that sink in.
eleitl@lemmy.zip 19 hours ago
More interestingly, how much energy you need for launch your photovoltaics power plant. You never EROEI over lifetime.
Cocodapuf@lemmy.world 20 hours ago
Well, that’s not the strongest argument at the moment, the launch is a one time cost, and that cost is in the process of dropping dramatically. But don’t worry, there are a plethora of other reasons this is a bad idea.
gandalf_der_12te@discuss.tchncs.de 19 hours ago
The price is dropping but not that rapidly. IIRC the price dropped like 95% over the last 25 years … That’s a 12% reduction annually on average. Compare that to computer chips where prices dropped 60% annually IIRC and solar panels where it dropped 30% annually.
teft@piefed.social 1 day ago
I’ve had to explain these points a few times to my friends who love the muskrat. They think he’s a genius for wanting orbital data centers.
Xaphanos@lemmy.world 1 day ago
All without explaining heat radiation, total power budget, physical maintenance, etc.
goferking0@lemmy.sdf.org 1 day ago
Data transmission/delays and damage from debris not just radiation
Gsus4@mander.xyz 1 day ago
Every single time…
AA5B@lemmy.world 1 day ago
SpaceX has done some great things to revolutionize space access, and already launches to orbit 80% by mass of all the worlds space “stuff”
Now they’re getting close with StarShip/Superheavy, designed to carry 10+ toned that every year. It will become much cheaper than anything before it, partly based on that economy of scale.
But where’s the market for that scale? How the heck are they launching 800+% of the worlds satellites every year? Business as usual will not support that. SpaceX needs to create new markets to drive that business, perhaps mars colonies or datacenters in space.
Gsus4@mander.xyz 1 day ago
Yeah, like when he undermines/debases mass transit to sell electric cars and shitty tunnel machines…for…Mars colonies…yea…there is a definitely plan in there, sort of…certainly not a grift.