New FCC filing by SpaceX for constellation of million data center satellites for AI apps.

# of sats: up to 1 million!
Orbit: 500 km and 2,000 km altitude and 30°inclinations. Multiple orbital shells 50 km apart.
Cooling: radiative
Schedule: unspecified
Sat size, weight, power, cost, ..: unspecified
Launches: millions of tons of mass per year

There goes the neighborhood.
Or is this just marketing fluff and orbital hoarding?

@sundogplanets
https://www.scribd.com/document/990116826/Orbital-Data-Center-LOA-Narrative
https://fccprod.servicenowservices.com/icfs?id=ibfs_application_summary&number=SAT-LOA-20260108-00016
1/n

To keep the clueless masses intoxicated, SpaceX threw in these flowery words in the FCC filing -

"Launching a constellation of a million satellites that operate as orbital data centers is a first step towards becoming a Kardashev II-level civilization—one that can harness the Sun’s full power—while supporting AI-driven applications for billions of people today and ensuring humanity’s multi-planetary future amongst the stars."

Earth must be destroyed to save it?

https://www.scribd.com/document/990116826/Orbital-Data-Center-LOA-Narrative
2/n

The filing is short on details and long on lofty goals and promises.

And also a bit sloppy -
"Global electricity demand for data centers is projected to more than double by 2035 driven by growth in AI — reaching ~1,200-1,700 Terawatt hours and ..."

You gotta specify TW or TW-hours per month or per year.

Also, I suspect the million satellite number comes from 1 million * 100 kw = 100 GW total power.

For comparison, Starlink v3 satellites will generate 20 kW with a 60x8 m array.

3/n

A blunt assessment of "AI datacenters in space” by Prof. Matthew Buckley in Dec 2025 -

"To be blunt, the entire stupid idea is a giant middle finger to multiple fundamentals of physics, and the fact that it is apparently being taken seriously by our tech lords, mainstream journalism, and political leaders is a damning indictment of not just the ridiculous amount of money chasing bad ideas in the tech/LLM/hype sector that has eaten the American economy, ..."

https://www.physicsmatt.com/blog/2025/12/11/the-dumbest-thing-ive-seen-this-week
4/n

Another good article by a former NASA engineer/scientist on why "Datacenters in space are a terrible, horrible, no good idea."

"This is an absolutely terrible idea, and really makes zero sense whatsoever. There are multiple reasons for this, but they all amount to saying that the kind of electronics needed to make a datacenter work, particularly a datacenter deploying AI capacity in the form of GPUs and TPUs, is exactly the opposite of what works in space."

https://taranis.ie/datacenters-in-space-are-a-terrible-horrible-no-good-idea/
5/n

Datacenters in space are a terrible, horrible, no good idea.

There is a rush for AI companies to team up with space launch/satellite companies to build datacenters in space. TL;DR: It's not going to work.

Taranis

@AkaSci

Sooner or later, the unemployed, angry masses of #AiAntagonists will stop posting silly memes and start smashing #datacenters instead...

...from that perspective putting data centers in orbit is perfect protection against angry #luddite mobs...

...as a plus, cooling and power, which is the main thrust of Luddite critique (because #AI #LLM are totally useless) is taken care of in orbit...

So if you have lift capacity of $20/kg suddenly orbital data centre starts to look like a damn good investment.

@n_dimension @AkaSci
Frank Herbert's imagined "Butlerian Jihad" in Dune is looking likely.

Multiple levels of magical thinking!

@n_dimension @AkaSci
Cooling stuff in space is actually really hard. No evaporation cycles, convection or conduction. Only IR radiation on the "dark" side.
@raymaccarthy @n_dimension
Not a problem. AI will solve it 😜

@AkaSci @n_dimension
You need to adjust your clothing!

That's fierce sarcasm that's exposed!

AI might, but we have no idea how to make an AI. The LLM/Generative "AI" are statistical plagiarism machines. More useful mis-named Machine learning is simply curated databases and pattern matching, with rules created by experts.

Scrape the Internet and you end up with regurgitated "Grokipedia", a NAZIfied version of Wikipedia, and a bunch on random nonsense.

@AkaSci but hey, think of all that random bit flipping you’d get from the cosmic radiation. A whole new flavor of stochastic gradient descent! lol
@AkaSci how will they launch the sysops?
@AkaSci well, there's this simple solution, although unpopular because ICE exists: arrest that south african illegal immigrant and NEVER let him go. Problem solved.
@AkaSci As much as I think the current Data Center Boom (TM) is a massive bubble that will end badly with a LOT of bag holders, if we really really needed *that much* compute we have plenty of cold surface area that just needs "bring heavy things and folks in and out" infrastructure.
@AkaSci his hubris is insufferable
@AkaSci and all ground-based astronomy can just fsck off, right? 😡

@AkaSci Just in case anyone needs to see it: this can't work with today's or even near-future technology.

https://pivot-to-ai.com/2025/12/01/ai-data-centres-in-space-why-dcs-in-space-cant-work

#OrbitalAI #DataCentersInSpaceCantWork

AI data centres — in SPACE! Why DCs in space can’t work

Spending all the money you have and all the money you can get and all the money you can promise has a number of side effects, such as gigantic data centres full of high-power chips just to run lyin…

Pivot to AI
@AkaSci Somebody’s been hitting the ketamine hard.

@AkaSci

We are not nearly enough to have the technology to become a Kardashev I-level civilization and these fucking morons are talking about becoming a II-level one.

Are they so clueless or do they think we are?