Thread

Google’s Project Suncatcher to put ML infrastructure in space

233 points · 553 comments · xnx

zactato · 1790267254 · focus · HN ↗
How are they solving the heat dissipation issues?
1. GMoromisato · 1790270501 · focus · HN ↗
  Please stop with this argument. This is one of the easiest problems to solve. Heat dissipation requires a radiator, which is just a dumb hunk of metal. All you need is a cheap way to launch mass to orbit, which is exactly what SpaceX (and Chinese companies) are doing.
  The only valid argument against data centers in space is the economic cost. If the per megatoken price for space datacenters is higher than terrestrial, then this won't work.
  But the price for terrestrial datacenters keeps going up and technology keeps dropping the price of space-based.
  1. 0cf8612b2e1e · 1790272676 · focus · HN ↗
    That is not a fake argument, but real physics. Yes, you can design out X kw can be dissipated by this much radiators, but that adds an enormous quantity of mass, more than the solar panels that feed it.
    If you are trying to generate a profit, every extra kg puts you further in the red. It still costs over $1000/kg of mass to get into orbit.
    GMoromisato · 1790275041 · focus · HN ↗
    So is it a physics problem or an economic problem? Sounds to me like you are acknowledging that it is just an economic problem. If it cost $1/kg to get to orbit then this wouldn't be a problem, right? Just make a bigger radiator.
    But if it's an economic argument, then you need to do the actual math. How big is the radiator? How low could the price to orbit go? How much can you charge per million tokens in 2030?
    The reason this is a fake argument is because the validity depends on the math, and nobody advancing the "you can't cool stuff in space" argument is actually doing the math.
    sobellian · 1790289584 · focus · HN ↗
    Okay any argument about why space is uniquely challenging is going to revolve around physics. Sure it's not literally physically impossible, but we need to explain to people why this is different from shipping the GPUs to Ohio.
    If you want math then <a href="https://andrewmccalip.com/space-datacenters" rel="nofollow">https://andrewmccalip.com/space-datacenters exists. The numbers are grim for orbital DC. Even if you drag the launch cost slider all the way to $1/kg (by the way this is literally sci-fi, per ChatGPT air freight of semiconductors from Taiwan to Ohio costs $9/kg and ocean/train freight costs a bit under $1/kg for a reasonable shipment so good luck with $1/kg to LEO this century) it is still more than twice as expensive as terrestrial DCs.
    GMoromisato · 1790293404 · focus · HN ↗
    But that calculator shows most of the cost is in the satellite. At $8 per watt (or about $2 million per satellite) the cost of orbital compute matches terrestrial.
    That sounds absolutely possible. But in any event, we're now arguing a different thing.
    The only thing I'm arguing is that cooling is a solved problem. I don't know if SpaceX will ever get the price down so that it is economical. But I'm convinced that there is no universe in which they hit their foreheads and say, "Oh no, we forgot about cooling!"
    sobellian · 1790295771 · focus · HN ↗
    Note that this calculator is actually quite optimistic for orbital wrt. many things including cooling and effect on launch, as:
    > No additional mass for liquid cooling loop infrastructure; likely needed but not included
    > Thermal: only solar array area used as radiator; no dedicated radiator mass assumed
    In hardware and mfg. solvable vs. solved is a big difference. And I too believe that SpaceX's engineers know about radiator panels. But the more cynical interpretation is that whatever the SpaceX engineers think about the technical merits, they are not being asked for that. They are just being asked for a pretext that justifies the xAI acquisition. Elon is also discussing lunar satellite factories that launch the satellites via railgun. Now, is this physically impossible? No, that isn't physically impossible either and I will seriously defend the physical possibility of this. It's not going to happen though.
    And you could spend all the engineering costs on building some seriously efficient terrestrial DCs, but somehow all these analyses start with "assume that launch and satellite technology advances manyfold and terrestrial DCs stagnate or become less efficient, then if you squint the two numbers get kinda close."
    Ajedi32 · 1790344932 · focus · HN ↗
    Certainly launch has to advance a lot for this to make sense. SpaceX is betting big on Starship getting close to $100/kg. But does satellite tech have to advance at all to make this work? SpaceX has already successfully deployed one megaconstellation into orbit and it's extremely profitable. Isn't this just another Starlink with bigger solar panels and radiators?
    sobellian · 1790347442 · focus · HN ↗
    Yes it has to advance a lot. In a DC all your supporting equipment has returns to scale and can be repaired if it breaks. If a GPU breaks, a sysadmin walks over to the offending rack and swaps the card. In space all that equipment serves just a few cards (this is more like orbital server racks) and it has to work in space (so instead of using an ~infinite heat sink like the Earth, you use radiators etc). We still don't exactly know what effect radiation in LEO would have on stock GPUs - IIRC the experiments to determine this started after Elon went all in on orbital DCs. If anything breaks, you have a flying brick.
    This is why that calculator, even under extremely optimistic assumptions for orbital, and even if you assume launch is zero, still cannot make it competitive with terrestrial.
    Ajedi32 · 1790354768 · focus · HN ↗
    Doesn't Starlink already handle all that though, just with networking racks instead of GPU racks? What's so unique about GPUs compared to other computer equipment, besides the power budget?
    sobellian · 1790356251 · focus · HN ↗
    Commercial GPUs already fail a lot and that's before you put them in LEO. Starlink is doing something that can't be accomplished anywhere else than space. If you could somehow have ground stations that can give wireless internet to every corner of the Earth you would much rather do that. Again, this is all physically possible, but the question is whether it's wise, whether it's saving money or if it's just a giant Rube Goldberg machine. If you assume Starlink satellite $/W then even if launch is zero, it costs twice as much as terrestrial using assumptions charitable to orbital.
    Ajedi32 · 1790357018 · focus · HN ↗
    Fair point about the failure rates. I wonder what causes that, given that there are no moving parts... if it's temperature fluctuations, maybe you could just run the GPU at 100% all the time to avoid that, given that the marginal cost of power for a satellite already in space is $0?
    > If you assume Starlink satellite $/W
    Why would you assume that, given that high power output is not a design goal of Starlink?
    sobellian · 1790357506 · focus · HN ↗
    The calculator I've cited elsewhere uses <a href="https://chatgpt.com/share/69391474-4b24-8005-bb93-ebd4340c65d8" rel="nofollow">https://chatgpt.com/share/69391474-4b24-8005-bb93-ebd4340c65.... The various starlink versions have approximately the same cost per watt across very different power levels. Much of the cost is going to be ~proportional to power - solar panels, heat rejection, power electronics, etc. Listen, if they could make it cheaper to deliver the same amount of power they would. Starlink is approximately all of SpaceX's business.
    Ajedi32 · 1790358636 · focus · HN ↗
    ChatGPT decided to use a constant $800/kg to estimate cost, and you're using that as evidence that cost/W doesn't decrease with scale?
    Anyway, my point is that Starlink is not primarily designed to harvest as much solar power as possible the way data center sats are. I'm sure they're not making it inefficient on purpose, but what you're suggesting is like using the cost/W of a solar powered traffic camera to estimate the cost/W of a solar farm.
    sobellian · 1790359507 · focus · HN ↗
    It's reasonable to 0th order to assume a ball of heat management and phased array antennae has a roughly constant cost per kg, yes. And as I said starlink absolutely wants to use as much power as they can, as communications from space are heavily limited by power.
    Ajedi32 · 1790359958 · focus · HN ↗
    FWIW, SpaceX is estimating their AI sats will have over double the W/kg of what that ChatGPT summary is estimating for Starlink. <a href="https://www.spacex.com/spacexai/starmind" rel="nofollow">https://www.spacex.com/spacexai/starmind Cost is not disclosed.
    sobellian · 1790361698 · focus · HN ↗
    Okay so even assuming all of SpaceX's numbers, which is extremely optimistic, even if we're an angel's advocate and assume that they spend all this money on engineering and optimize their launch costs and their satellite costs and they don't hit any hiccups, and DCs on Earth stay pretty much the same and don't get any added efficiencies. Even then over every time period orbital is still more expensive. And it gets particularly punishing the longer you extend the analysis period as the satellites need to get replaced while the DCs don't spontaneously combust all their infra.
    Ajedi32 · 1790609280 · focus · HN ↗
    I don't think those numbers are optimistic, I suspect they modeled it in CAD and that's what the calculation spat out; it's probably quite accurate.
    Cost is a different matter of course, but it seems incongruous to me that after being confronted with the fact that your power density assumptions were off by a factor of 2x you would continue to insist ChatGPT's weird assumptions about costs scaling linearly with mass are accurate. As was already pointed up up thread, even if all your calculator's assumptions are correct they can still turn a profit if they get the per satellite cost below $2 million.
    sobellian · 1790837162 · focus · HN ↗
    Upthread comment and therefore the premise of your $8/W target is wrong. As I explained, I plugged in all the announced engineering parameters for Starmind, $8/W, and $76/kg. It ties terrestrial over a 5yr period (the slider default) but gets crushed over a 10yr period. Naturally a longer time window is prudent absent some world-ending calamity in 5 years that renders the need for compute irrelevant. And even the 5yr tie is very sensitive to these parameters and requires everything to go right.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.