Google Tests Space-Based Data Centers for AI Computing

Google has launched an experiment to test the viability of space-based data centers for AI computing. Four Tensor Processing Units (TPUs) were sent aboard a SpaceX Falcon 9 rocket as part of Project Suncatcher, with the goal of measuring performance under orbital radiation and thermal conditions.

The refrigerator-sized satellite, MVP, was developed in collaboration with Planet and launched on October 1 from California's Vandenberg Space Force Base. Google aims to determine whether specialized AI chips can function reliably in space, a crucial step towards creating practical and economically competitive space-based AI infrastructure.

Currently, AI data centers are becoming increasingly demanding in their use of electricity, land, and water. Google estimates that solar panels in orbit could be up to eight times as productive as those on Earth, benefiting from near-continuous sunlight.