SpaceX (SPCX) is about to carry a very interesting payload into orbit: AI chips from one of its biggest rivals.
According to a widely shared post on X, Alphabet (GOOG) is launching Google’s TPUs (the custom chips it builds to run AI) this week on a SpaceX Falcon 9 to test AI data centers in orbit.
Elon Musk’s response? “The amount of compute in space will obviously round up to 100% of all compute.”
(No hedging there!)
For SpaceX investors, though, the bigger deal is who’s buying the ride.
I wrote earlier this week about Musk’s aggressive timeline for the Moon and Mars. But this feels like one of those weeks where decades happen – because orbital compute is officially longer a SpaceX-only bet. The competition is paying to get in on it.
SpaceX gets paid either way
Think about the setup here. SpaceX has its own AI business (hello, Grok). Google is one of its fiercest rivals in AI. And when Google wanted to test its chips in orbit, it went to SpaceX for the launch.
If Coca-Cola had to ship every can on Pepsi’s trucks, you’d want to own the trucking company.
Same idea here. Whether Grok or Gemini wins AI in space, the chips have to get up there somehow.
(Why space at all? The short version is power. In the right orbit, solar panels get near-constant sunlight, and you don’t have to wait in line for a spot on a crowded electric grid.)
And Google’s already ramping spending…
As you can see, CapEx is rapidly ramping. Even a sliver of it heading to orbit would matter to whoever’s doing the launching.
Let’s test that 100 GW
It gets bigger. SpaceX’s CFO says the company is on track to deploy 100 GW of orbital compute every year by the end of the decade.
Meanwhile, the FCC accepted SpaceX’s application to deploy up to 100,000 Gen3 Starlink satellites for filing on September 18. And Musk added a key detail:
“Each sat is 250kW and will have a SpaceX-designed Nvidia Vera Rubin NVL72 computer.”
So here’s some quick envelope math: 100,000 satellites × 250 kW each = 25 GW.
That’s the entire proposed Gen3 constellation…and it’s only a quarter of what the CFO says SpaceX will deploy every year. Even as payloads likely get bigger…it’s still a lot of rockets. And SpaceX owns the rockets.
NVIDIA’s along for the ride, too
Notice what Musk says goes inside every one of those satellites: a computer built on chips from NVIDIA (NVDA).
Nvidia has already ridden the data center boom on the ground to huge revenue growth…
If Musk is even partly right, orbit becomes another market for those chips. I’d call that a nice bonus for Nvidia shareholders.
Here’s the catch
Of course, one Falcon 9 test launch is a tiny contract. Orbital compute could stay experimental for years, and SpaceX’s revenue may not show much of this until well after the decade’s end.
And SpaceX is going to need all the help it can get to deliver on analysts’ growth expectations and maintain its extraordinarily rich valuation.
As for me, I’m comfortable sitting on the sidelines for now. There’s a lot of execution risk at play here, but the opportunity, frankly, is enormous.
And if you want to model that opportunity more precisely, you can build with TIKR.com’s proprietary modeling tool, which requires just three inputs, to get a full financial model for what SpaceX stock could be worth in three years.
It’s simple, straightforward, and robust – using institutional-quality data – and you can build it for free.
Disclaimer:
Please note that the articles on TIKR are not intended to serve as investment or financial advice from TIKR or our content team, nor are they recommendations to buy or sell any stocks. We create our content based on TIKR Terminal’s investment data and analysts’ estimates. Our analysis might not include recent company news or important updates. TIKR has no position in any of the stocks mentioned. Thank you for reading, and happy investing!
