Google and SpaceX are planning orbital data centers — first prototypes by 2027
Google and SpaceX are in active talks to build data centers in orbit, reports WSJ, with two prototype satellites planned for launch as early as 2027 under a project called Suncatcher. The partnership would also involve Planet Labs, the satellite imaging company. For Google, the pitch is simple: AI models are consuming electricity faster than grid operators can keep up, and space offers virtually unlimited solar power with no land costs, no cooling water bills, and no local planning disputes.
The appeal — and the catch
The logic behind orbital computing is not as far-fetched as it sounds. Google's own Trillium TPUs — the chips that power its AI workloads — have already passed radiation testing in a particle accelerator simulating low-Earth orbit conditions. Cooling via radiators is viable. Inter-satellite data links at 10 Tbps are achievable with off-the-shelf technology. Data Center Dynamics has detailed Google's vision of 1km-wide arrays of 81-satellite compute clusters.
The catch is cost. Getting hardware into orbit currently runs $1,500–$2,900 per kilogram. Google's internal models put the break-even point at $200 per kilogram — a figure that only becomes realistic if SpaceX's Starship rocket flies around 180 times a year and drives prices down roughly 90% from today's rates. Google's own projection puts that target at 2035. Until then, orbital data centers are roughly three times more expensive than their terrestrial equivalents.
The IPO angle
For SpaceX, the timing is not accidental. The company is targeting an IPO in June 2026 at a reported valuation of $1.75 trillion, and orbital AI infrastructure is a central plank of its investor story — sitting alongside Starlink and a reported AI division merger. Any prototype failure or cost overrun before the listing would put direct pressure on that narrative.
SpaceX has already flagged the risk itself: its IPO filing warns investors that orbital AI infrastructure may never become commercially viable. Regulatory exposure is also real — millions of planned satellites raise Kessler syndrome concerns (the risk that a cascade of debris collisions makes entire orbital bands unusable), which the FCC is beginning to treat as a policy problem, not just a theoretical one.
Google invested $900 million in SpaceX back in 2015, so the relationship has deep roots. But this bet is far larger. The 2027 prototypes will be the first real test of whether the physics and the finances can be made to meet — and the answer will land well before Starship hits its 180-launches-a-year target.
