Google said in a post on Oct. 30 that its Project Suncatcher prototype satellite has returned its first AI response from orbit. The company described it as the first AI response from a data center class TPU in orbit. TPUs are the chips Google is testing for AI computing in space.
According to the post by Travis Beals, Google’s senior director for Paradigms of Intelligence, the satellite launched on Oct. 1. Google describes Project Suncatcher as a long-term research moonshot to explore powering AI computing in space.
Beals wrote that the team spent the weeks after launch completing a commissioning checklist. That work included powering up the host and TPU trays and running early machine learning tests. Google said the team reached what it called its “Hello World” milestone this week.
During a ground-station pass, the team sent a prompt up to the Gemma 4 open model on board the satellite. The prompt asked the model to generate a haiku about being in space. Google said the poem was sent back down as the satellite passed Svalbard, Norway, and confirmed the test succeeded.
After that first output, the team tested space-themed limericks. It also sent telemetry prompts, which asked the model to report the technical status of the satellite’s payload systems.
Google said it will next run a series of tests to evaluate how the TPUs perform under thermal stress and radiation exposure in orbit. Beals wrote that verifying the chips can reliably run workloads in space paves the way for the next milestone, twin interconnected satellites in 2027.
The post did not say how many chips are on the prototype, what orbit it is in, or how much power the system uses. It also did not give results from the planned thermal and radiation tests. The account of the test comes from Google’s own announcement.