Google has successfully run its first machine learning prompt on a satellite-mounted processor, bringing its space-based computing experiment closer to reality. The prototype spacecraft generated an AI response from orbit using Google’s Tensor Processing Units (TPUs).The achievement marks an early milestone for Project Suncatcher, Google’s research initiative exploring large-scale computing infrastructure in space. Engineers launched the satellite on October 1 aboard a SpaceX Falcon 9 rocket. It shared the ride with more than 100 other spacecraft.Google confirmed that its onboard systems completed initial checks after approximately three weeks of commissioning. The team has now demonstrated that an orbital TPU can handle basic machine learning workloads.First AI response from orbitDuring a ground-station communication window, engineers transmitted a prompt to Google’s Gemma 4 open model. They asked the onboard system to compose a haiku about its surroundings in space.The satellite received the prompt and generated three lines of poetry. Its subsequent transmission back to Earth confirmed that the system had successfully processed the request.Our TPUs just phoned home… from space.In early October, we sent four TPUs into orbit to test whether our AI hardware could operate in space. The initial goal was to see if our chips could turn on, process a prompt, and transmit a response back to Earth.This week, we got our… pic.twitter.com/j8NJx2nWnQ— Google (@Google) October 30, 2026The response read: “Blue marble below / Data streams in silent grace / Starlight feeds my core.” Google also tested the model with space-themed limericks. Engineers used additional prompts to retrieve telemetry information about the spacecraft’s operating conditions.The demonstration represents an important technical step. It confirms that the satellite can receive instructions from Earth, process them onboard, and return generated information.Radiation poses major challengeThe spacecraft carries four Google TPUs, although its power system allows engineers to operate only one processor at a time. Its electrical generation capacity reaches approximately 1 kilowatt.Power consumption also limits individual operating sessions. Engineers can run a TPU for roughly 20 minutes before the available energy becomes a constraint.Google’s next tests will examine how the processors withstand the harsh conditions of low Earth orbit. These include temperature fluctuations and exposure to space radiation.Radiation presents a particular challenge for conventional computing hardware. High-energy particles can trigger bit flips, changing stored values inside computer memory.Such errors can corrupt calculations without necessarily activating standard monitoring systems. Google previously examined radiation effects using particle accelerators on Earth.Those experiments suggested inference workloads could tolerate radiation better than model training. However, actual orbital testing will help researchers establish how the hardware behaves under continuous exposure.Two satellites planned for 2027Project Suncatcher eventually aims to connect multiple spacecraft into an orbital computing network. Google plans to launch two additional satellites in 2027.Those spacecraft will demonstrate higher power generation capabilities and optical communication links between satellites. Such connections could allow multiple processors to coordinate workloads across an orbital network.Google’s approach differs from other proposed space-based computing systems. Its research focuses on closely coordinated satellites capable of sharing processing tasks.Other companies, including Starcloud and Cowboy Space, are exploring orbital computing. Their initial approaches focus more heavily on running inference workloads aboard individual spacecraft.Scaling these systems will also depend on launch economics. Google’s research suggests that SpaceX’s fully reusable Starship rocket could require approximately 1,800 flights to achieve significant cost reductions.