Google has set a concrete timeline for its ambitious space-based datacentre plan, aiming to launch its first two prototype satellites by early 2027. This marks the first major milestone in its “Project Suncatcher” initiative, which the company hopes will lead to a “scalable space-based AI.”
This plan is Google’s answer to the rising demand for AI, which is fueling a $3 trillion spending boom on terrestrial datacenters. The company’s research suggests that by the mid-2030s, falling launch costs could make orbital datacenters, powered by hyper-efficient solar panels, economically comparable to their Earth-bound counterparts.
These prototypes will be crucial for testing solutions to the “significant engineering challenges” Google has acknowledged. Engineers must prove they can manage the intense heat of AI processors in a vacuum, ensure the reliability of the systems in a high-radiation environment, and establish high-bandwidth optical links to the ground.
The 2027 launch will be a closely watched event. Competitors, including Elon Musk’s Starlink and an Nvidia-backed startup named Starcloud, are also aggressively pursuing this new frontier. Starcloud is even planning to launch its first AI chips into space later this month.
This “moonshot” research, as Google calls it, represents a new phase in the development of AI infrastructure. The success or failure of these early prototypes will determine if the future of AI computation truly lies 400 miles above the Earth’s surface.