Google’s Project Suncatcher took off on a SpaceX Falcon9 rocket from Vandenberg Space Force Base in California on Thursday, kicking off a new era for the company.
According to Google, Project Suncatcher is a long-term research program exploring the feasibility of scaling AI infrastructure in space. Their first step is testing whether Google's machine learning processors survive the trip to low earth orbit.
The idea is to build AI compute hardware into satellites, then launch the satellites into orbit potentially reducing data center energy demands. They say, low earth orbit provides “near-constant sunlight,” and nearly “eight times more solar power than on earth.” Eventually they want to link solar-powered satellites into constellations.
But, Dr. Aaron Boley, Professor of Astronomy at the University of British Columbia, and co-director of the Outer Space Institute says, the cost of launching data center infrastructure into space creates new environmental problems and compounds old ones. "Every ton you put into orbit requires ten to thirty tons of fuel," he said. "That's actually dumping, or has the potential to dump, a tremendous amount of exhaust into the upper atmosphere."
Google isn't alone in its orbital compute ambitions. This summer, SpaceX and Nvidia partnered to develop Starmind AI1, a proposed fleet of one million satellites. On September 22, the Center for Space Environmentalism submitted a public statement to the FCC imploring the Commission not to "treat 4-tonne orbital data centers as routine telecommunications satellites."
The Indian orbital data center company TakeMe2Space, and the micro-gravity life sciences company Varda Space Industries, sent hardware into low earth orbit with Google yesterday.
The three companies participated in SpaceX's SmallSat RideShare Program — indicating growing commercial interest in space computing and manufacturing.
