Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

xAI joins SpaceX

spacex.com

Post reply on HN
Print view View original

321–330 of 1001 posts

Re: xAI joins SpaceX

#321

Post by iandanforth »

qaq wrote: post #241
SimianSci wrote: post #113

The Goalpost shift continues, If elon were working for me, I would have fired him for having never delivered on any of his projects. Hyperloop > Neuralink > Self-Driving Cars > Robotaxi fleets > Personal Robots > Orbital Datacenters > [Insert next]

hmm Tesla shipped millions of cars SpaceX launches 90% of space payloads, Starlink is working well. Thats hard to categorize as never delivered on any of his projects

The crucial thing is that Tesla's valuation has the hype projects baked in. The fact that it never delivered self driving or a robotaxi fleet and is now being saved solely by an import ban on Chinese EVs means that any success he had with Tesla is now an illusion.

Re: xAI joins SpaceX

#324

Post by fooker »

rybosworld wrote: post #70

> The basic math is that launching a million tons per year of satellites generating 100 kW of compute power per ton would add 100 gigawatts of AI compute capacity annually, with no ongoing operational or maintenance needs. Ultimately, there is a path to launching 1 TW/year from Earth. > My estimate is that within 2 to 3 years, the lowest cost way to generate AI compute will be in space. This is so obviously false. Fo…

> in what fantasy world

It is already more expensive to performance maintenance on SOCs than it is to replace them. Remember, these machines are not for serving a database, there are practically no storage needs (and storage is the component that fails most often.)

Given that, the main challenge is cooling, I assume that will be figured out before yeeting 100 billion $ of computers into space. Plenty of smart people work at these companies.

Re: xAI joins SpaceX

#325

Post by bakies »

padjo wrote: post #73
kemotep wrote: post #56

If Musk and SpaceX are serious about putting 1 million datacenter satellites into space, then they are not serious about Mars. You cannot simultaneously build and launch 10’s of thousands of Starships to deliver 1 million tons of equipment and supplies bound to Mars while also committing to launching 10’s of thousands of Starships to orbit full of satellites. They would need to quadruple their launch rate, and half o…

Why would you put data centres in space?

Seemingly to prop up the value of your companies

Re: xAI joins SpaceX

#326

Post by lovidico »

reilly3000 wrote: post #269

I suppose one of the ADR’s read something like “…who cares about bitflips, man. Isn’t AI all about probability?” Knowing the insane level of hardening that goes into putting microcontrollers into space, how to the expect to use some 3nm process chip to stand a chance?

A trend at the moment is to just hope for the best in cubesats and other small satellites in LEO. If you’re below the radiation belt it’s apparently tenable. I worked somewhere designing satellite hardware for LEO and we simply opted to use consumer ARM hardware with a special OS with core level redundancy / consensus to manage bit flips. Obviously some problems will present for AI there… but there are arguably bigger problems with AI data centres like the fact that they offer almost no benefit with respects to the costs of putting and maintaining stuff in space!

Re: xAI joins SpaceX

#327

Post by smileson2 »

Buttons840 wrote: post #50

SpaceX is too big to fail. It's important for national security. I wonder if Elon wants to tangle all his businesses into SpaceX so they are all kept afloat by SpaceX's importance.

national security is pretty felixaeble

Re: xAI joins SpaceX

#328

Post by Ifkaluva »

brokensegue wrote: post #139
paxys wrote: post #125

Just a neat bit of financial engineering. You can tell because Elon picked SpaceX instead of Tesla – which would have actually made sense at some level (Optimus Robots + AI). But Tesla is public and so he'd need to follow laws and reporting requirements.

does he need spacex/xai to prop up tesla or the other way around?

I guess the difference is Tesla is a public company, so requires more paperwork. SpaceX isn't public yet, but will be soon, meaning it will have a cash infusion.

Re: xAI joins SpaceX

#329

Post by camdenreslink »

wongarsu wrote: post #117
rybosworld wrote: post #70

> The basic math is that launching a million tons per year of satellites generating 100 kW of compute power per ton would add 100 gigawatts of AI compute capacity annually, with no ongoing operational or maintenance needs. Ultimately, there is a path to launching 1 TW/year from Earth. > My estimate is that within 2 to 3 years, the lowest cost way to generate AI compute will be in space. This is so obviously false. Fo…

You operate them like Microsoft's submerged data center project: you don't do maintenance, whatever fails fails. You start with enough redundancy in critical components like power and networking and accept that compute resources will slowly decrease as nodes fail No operational needs is obviously ... simplified. You still need to manage downlink capacity, station keeping, collision avoidance, etc. But for a large con…

But you could just run a “zero maintenance” data center on Earth and not pay to blast it into orbit.

Re: xAI joins SpaceX

#330

Post by bastawhiz »

moeadham wrote: post #276
gok wrote: post #45

> it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power We currently make around 1 TW of photovoltaic cells per year, globally. The proposal here is to launch that much to space every 9 hours, complete with attached computers, continuously, from the moon. edit: Also, this would capture a very trivial…

In fairness, solar cells can be about 5x more efficient in space (irradiance, uptime).

And how much of that power would be spent on high speed communications with Earth that aren't, you know, a megabit or two per second
Post reply on HN