Comment by repple
9 hours ago
Their goal of moving compute to space combined with their capacity to launch tons of payload will make this look like a tiny blip.
9 hours ago
Their goal of moving compute to space combined with their capacity to launch tons of payload will make this look like a tiny blip.
What is the benefit of "moving compute to space"?
It's hard for an uprising of poor people to shut it off. It's the ideal place to run your CEO / President simulations.
I say this tongue in cheek, but in all seriousness, I can't really think of any other benefit, and I no longer have a lot of faith in the good sense of some of the people involved.
Elon makes a relatively good case in the Dwarkesh podcast. I recall it like this:
1) Energy infra is going to be seriously limited on the production side well, well below demand
2) energy engineering solar for space requires less materials than for gravity-based solar (!)
3) you cut out distribution network needs when you just launch stuff all per-pod in space
4) SpaceX thinks it can create a scalable vertically integrated production facility to turn raw materials into space datacenter pods, with the exception of chips.
As a business bet, this is predicated on 10,000x inference demand growth - if we have that, and SpaceX can get the integrated production rolling, and get Starship launching, then these will be actively utilized at scale.
Whether you are bullish on the whole plan should, I think come down to your take on those priors: 10kx growth, ability to manage supply chain and production, Starship outlook, and silicon access.
I'm not bearish on this after listening to the podcast; it has a very Elon-like returns distribution - if they're wrong on a lot of this, they'll probably have some moderately price-competitive datacenter facilities in space and a lot of built organizational knowhow while Brooklyn journalists dunk on them for spending all that effort to just replicate what we have on Earth. If they're right about most of this, they'll have an unreplicable head start, both due to years of experience, and due to the cheap launch they gambled on ten years ago, they'll have a nearly insurmountable moat.
11 replies →
Not having to deal with having to defend in court why polluting an area where you built your datacenter and fucking it up for the residents there is actually better for all man kind.
> What is the benefit of "moving compute to space"?
I’ll bite. It’s cheaper and quicker to permit a launch than permit, zone and interconnect a datacenter. And solar panels in space don’t need glass cladding, which makes them cheaper to make and lift.
The downside is launch cost. But there is a breakeven between these factors that seems to have most of its error bars within Starship’s target. (By my math, around $35/kg.) So if Starship works, and all indications seem to show that it will, eventually, then that puts space-based data centers at cost parity with terrestrial ones within a decade. Which was, well, unexpected when I ran the numbers.
(The surprising finding when you run the numbers is launching the chips and solar panels isn’t the limiter, it’s launching the radiators. Which opens up whole new questions about at what scale it makes sense to stop sending those up the well.)
> It’s cheaper and quicker to permit a launch than permit, zone and interconnect a datacenter
There's plenty of empty land sufficiently far from cities and not being used for anything else and that shouldn't have permitting or zoning problems.
For interconnect do that via satellite.
The capacity of a single datacenter would require thousands of launches to get the equipment into space. I don’t believe for a second that this would be easier in any way. Cooling and bandwidth are also completely unsolved for compute on a useful scale.
What about maintenance? I’d naively assume that’s the killer.
That xAI fails faster, hopefully.
Here's an explainer video: https://youtu.be/Jth4yATniS4?si=zIzZXwLk6RTgyeCQ&t=37