SpaceX launches Google AI chips into orbit in push toward space-based data centers

HAWTHORNE, Calif. — SpaceX carried Google's custom artificial intelligence processors into low-Earth orbit aboard Planet Labs satellites, according to a report published Oct. 2, 2026, by CNBC Technology. The deployment marks an operational step toward establishing space-based data center infrastructure, utilizing commercial launch capability and earth-imaging satellite platforms to test compute workloads beyond the terrestrial grid.

Strategic Context

The October payload integration brings together three distinct segments of the commercial technology stack: heavy-lift launch vehicles provided by SpaceX, earth observation hardware built by Planet Labs, and proprietary silicon developed by Google. Google's tensor processing units (TPUs), designed specifically for heavy machine learning and inference workloads, were integrated directly into the satellite payloads ahead of the launch.

By placing high-performance AI processors into orbit, technology operators are testing the viability of running compute clusters in space. Terrestrial data centers face mounting constraints regarding grid power availability, liquid cooling infrastructure, and real estate acquisition. Space-based architecture introduces alternative constraints, including radiation hardening, thermal management in a vacuum, and reliance on solar generation arrays.

Forward Outlook

Enterprise technology operators will monitor the operational performance of the orbital hardware to determine whether space-based data processing can scale beyond experimental payloads. As satellite operators seek higher-margin revenue streams beyond traditional earth observation and broadband relay, on-board AI compute represents a potential shift in how raw sensor data is processed before downlink transmission.