I wonder if there could be some way to photolitograph compute circuits directly onto a radiator substrate, and accomplish a fully-passive thermal solution that way. Consider the heat-conduction problem: from dimensional analysis, the required thickness of a (conduction-only) radiator plate with a regular grid of heat sources on it shrinks superlinearly as you subdivide those heat sources (from few large sources, into many, small ones). At fixed areal power density, if the unit heat source is Q, the plate thickness d ∝ Q^{-3/2}. (This is intuitive: the asymptotic limit is a uniform, continuous heat source exactly matched to a uniform radiation heat sink; hence heat conduction is zero). So: could one contemplate an array of very tiny CPU sub-units, grided evenly over a thin Al foil—say at the milliwatt scale with millimeter-scale separation? It'd be mostly empty space (radiator area) and interconnect. It'd be thermally self-sufficient and weigh practically nothing.
Starcloud can’t put a data centre in space at $8.2M in one Starship
171–180 of 365 posts
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#172Putting a datacenter in space is one of the worst ideas I've heard in a while. Reliable energy? Possible, but difficult -- need plenty of batteries Cooling? Very difficult. Where does the heat transfer to? Latency? Highly variable. Equipment upgrades and maintenance? Impossible. Radiation shielding? Not free. Decommissioning? Potentially dangerous! Orbital maintenance? Gotta install engines on your datacenter and kee…
If you read the Starcloud whitepaper[1], it claims that massive batteries aren't needed because the satellites would be placed in a dawn-dusk sun-synchronous orbit. Except for occasional lunar eclipses, the solar panels would be in constant sunlight. The whitepaper also says that they're targeting use cases that don't require low latency or high availability. In short: AI model training and other big offline tasks. F…
> For maintenance, they plan to have a modular architecture that allows upgrading and/or replacing failed/obsolete servers. If launch costs are low enough to allow for launching a datacenter into space, they'll be low enough to allow for launching replacement modules.
This is hiding so, so much complexity behind a simple hand wavy “modular”. I have trained large models on thousands of GPUs, hardware failure happen all the time. Last example in date: an infiniband interface flapping which ultimately had to be physically replaced. What do you do if your DC is in space? Do you just jettison the entire multi million $ DGX pod that contains the faulty 300$ interface before sending a new one? Do you have an army of astronauts + Dragons to do this manually? Do we hope we have achieve super intelligence by then and have robots that can do this for us ?
Waving the “Modular” magic key word doesn’t really cut it for me.
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#173I wonder if there could be some way to photolitograph compute circuits directly onto a radiator substrate, and accomplish a fully-passive thermal solution that way. Consider the heat-conduction problem: from dimensional analysis, the required thickness of a (conduction-only) radiator plate with a regular grid of heat sources on it shrinks superlinearly as you subdivide those heat sources (from few large sources, into…
Look at the thermal shield design for the JWST. Could you have a data center that unfolds into a multi-layered plane where the outer solar collector layer faces the sun, an intermediate layer shields infrared emissions from the back side of that, and the final layer that always faces away from the sun holds (or is) a bunch of chips? Park it in an orbit where it can stay oriented this way or an L point. Free compute for the life span of the chips powered by the sun.
A lot of these is a supercomputing Dyson swarm.
Also do chips in space need casing or could the wafers be just exposed on that back layer?
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#174Earlier quoted context omitted.
The best argument I've heard for data centres in space startups is it's a excuse to do engineering work on components other space companies might want to buy (radiators, shielding, rad-hardened chips, data transfer, space batteries) which are too unsexy to attract the same level of FOMO investment...
Yes, and also just because a space data center isn’t useful today doesn’t mean it won’t be required tomorrow. When all the computing is between the ground and some nearby satellites, of course the tradeoffs won’t be worth it. But what about when we’re making multi-year journeys to Mars and we need a relay network of “space data centers” talking to each other, caching content, etc? We may as well get ahead of the prob…
> what about when we’re making multi-year journeys to Mars and we need a relay network of “space data centers” talking to each other, caching content, etc?
How would this work? Planets orbit at different speeds, so you can't build a chain of relays to another planet. I can imagine these things orbiting planets, but is that worth it compared to ground-based systems?
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#175Earlier quoted context omitted.
I've had actual, real-life deployments in datacentres where we just left dead hardware in the racks until we needed the space, and we rarely did. Typically we'd visit a couple of times a year, because it was cheap to do so, but it'd have totally viable to let failures accumulate over a much longer time horizon. Failure rates tend to follow a bathtub curve, so if you burn-in the hardware before launch, you'd expect lo…
It would be interesting to see if the failure rate across time holds true after a rocket launch and time spent in space. My guess is that it wouldn’t, but that’s just a guess.
I think it's likely the overall rate would be higher, and you might find you need more aggressive burn-in, but even then you'd need an extremely high failure rate before it's more efficient to replace components than writing them off.
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#176I wonder what the implication on data protection / privacy laws and the like would be. Would it be as simple as there'd be no laws, or is the location of the users still relevant?
Laws definitely still apply.
https://en.wikipedia.org/wiki/Outer_Space_Treaty
> Article VI of the Outer Space Treaty deals with international responsibility, stating that "the activities of non-governmental entities in outer space, including the Moon and other celestial bodies, shall require authorization and continuing supervision by the appropriate State Party to the Treaty" and that States Party shall bear international responsibility for national space activities whether carried out by governmental or non-governmental entities.
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#177I wonder if there could be some way to photolitograph compute circuits directly onto a radiator substrate, and accomplish a fully-passive thermal solution that way. Consider the heat-conduction problem: from dimensional analysis, the required thickness of a (conduction-only) radiator plate with a regular grid of heat sources on it shrinks superlinearly as you subdivide those heat sources (from few large sources, into…
Look at the thermal shield design for the JWST. Could you have a data center that unfolds into a multi-layered plane where the outer solar collector layer faces the sun, an intermediate layer shields infrared emissions from the back side of that, and the final layer that always faces away from the sun holds (or is) a bunch of chips? Park it in an orbit where it can stay oriented this way or an L point. Free compute f…
I understand that the multi-layer insulation idea doesn't accomplish much unless you're trying to reach deep cryogenic temperatures passively (as infrared telescopes do!) It's a difficult structural design which would only cut your heat budget by a small constant factor. Remember that much of that heat on the solar side is making its way over to the cold side by way of electricity—the compute units are a heat "source" of similar magnitude as the solar input itself.
edit: I think the optimal packing could be a simple rolled-up scroll, that unfurls in space into a ribbon. A very lazy design where the ribbon has no orientation control, randomly furls and knots; and only half of it is (randomly) facing the sun at any given time. And the compute units are designed work under those conditions—as they are to be robust against peers randomly disappearing to micrometeorites, to space radiation, and so forth.
Because, you could make up for everything in quantity. A small 3x5 meter cylinder of rolled-up foil stores—at the mm-thickness scale, 10's of gigawatts of compute; at the micron scale, 10's of terawatts. Of course that end is far-future sci-fi stuff!
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#178Earlier quoted context omitted.
If mass is going to be as cheap as is needed for this to work anyway, there's no reason you can't just use people like in a normal datacenter.
Space is very bad for the human body, you wouldn't be able to leave the humans there waiting for something to happen like you do on earth, they'd need to be sent from earth every time. Also, making something suitable for humans means having lots of empty space where the human can walk around (or float around, rather, since we're talking about space).
Underwater welder, though being replaced by drone operator, is still a trade despite the health risks. Do you think nobody on this whole planet would take a space datacenter job on a 3 month rotation?
I agree that it may be best to avoid needing the space and facilities for a human being in the satellite. Fire and forget. Launch it further into space instead of back to earth for a decommission. People can salvage the materials later.
Re: Starcloud can’t put a data centre in space at $8.2M in one Starship
#179Why do they want to put a data center in space in the first place? Free cooling? Doesn't make much sense to me. As the article points out the radiators need to me massive. Access to solar energy? Solar is more efficient in space, I'll give them that, but does that really outweigh the whole hassle to put the panels in space in the first place? Physical isolation and security? Against manipulation maybe, but not agains…
Not every datacenter use case is latency sensitive. Backup storage or GPU compute, for example.