Google is preparing to take artificial intelligence somewhere it has never operated quite like this before — into orbit.
The company’s Project Suncatcher is preparing for its first in-orbit experiment, with a prototype satellite carrying Google Tensor Processing Units, or TPUs, scheduled to launch into low-Earth orbit.
The mission isn’t an operational data center in space. Instead, Google wants to answer a fundamental question: Can powerful AI computing hardware reliably operate in the harsh environment of space?
Getting there is only the first challenge. During launch, components can experience enormous vibration and acceleration. Once in orbit, the chips will encounter radiation and extreme thermal conditions that electronics on Earth don’t normally face.
Cooling may be an even bigger challenge. AI processors generate substantial heat, but space has no air to carry that heat away. Google is experimenting with heat pipes and radiators designed to transfer heat from the processors and release it into space.
So why put AI computing in orbit at all?
One potential advantage is solar energy. Google says satellites in low-Earth orbit could receive near-continuous sunlight and potentially generate substantially more solar power than comparable systems on Earth.
The company’s longer-term vision goes considerably further. Future satellites could carry multiple AI processors and communicate with neighboring satellites through high-bandwidth laser connections, potentially creating clusters of computing power in orbit.
But that future remains experimental. This first mission is about discovering what works, what fails, and what engineers would need to change before orbital AI computing could ever operate at a meaningful scale.
If the experiment succeeds, however, the places where humanity builds some of its most powerful computers may someday extend far beyond Earth.
Have you heard?