Google to Test AI Computing in Space With Prototype Satellite

News Desk

SAN FRANCISCO:  Google is preparing to launch a prototype satellite next week as part of its research initiative, Project Sun Catcher, to explore whether space could one day host large-scale data centers dedicated to artificial intelligence (AI) computing.

The experimental mission will examine whether AI computing infrastructure can operate in orbit while taking advantage of the continuous sunlight available in space.

Google is not alone in exploring the concept. Companies including SpaceX and Star Cloud are also researching orbital data centers as a potential way to overcome some of the power constraints faced by conventional data centers on Earth.

The satellite is scheduled to fly aboard SpaceX’s Transporter-18 rideshare mission. Google is working with satellite company Planet Labs on the project.

The prototype will carry Google-developed Tensor Processing Units (TPUs), specialized chips designed for AI workloads. Researchers will assess how the processors perform under the radiation, extreme temperatures and mechanical stresses associated with spaceflight.

Cooling will be another major focus of the experiment. Since conventional air-based cooling systems cannot operate in the same way in space, Google plans to test a system combining heat pipes and radiators to manage the heat generated by the AI processors.

Google said the TPUs have already undergone testing at the University of California, but an orbital trial is needed to determine how the technology performs under actual space conditions.

The project remains at an experimental stage, with experts cautioning that high launch costs, complex engineering requirements and the challenges of manufacturing satellites at scale could keep orbital AI data centers years away from commercial deployment.

If successful, however, the research could offer a glimpse into a future in which some of the world’s growing demand for AI computing is handled not on Earth, but in orbit.