Google Space AI Data Center Test Is Scheduled for October 1

Google’s first space AI data center test is scheduled for October 1, as the company prepares to send a prototype satellite into orbit under Project Suncatcher. The mission will test whether Google’s AI computing hardware can operate reliably in the physical conditions of space.

What Google Is Sending to Space

The prototype, developed with Planet, is designed to evaluate Google Tensor Processing Units (TPUs) in orbit. Google says the first mission is focused on gathering real-world data about how the hardware handles vibration, radiation, thermal extremes, and other conditions that cannot be fully reproduced on Earth.

The mission is an extension of Google’s broader investment in custom AI hardware. Saganote has covered Google’s chip designed to bring Gemini directly into hardware, as well as Google’s decision to restrict Meta’s Gemini access as compute capacity runs short. Google has also continued expanding its Gemini model lineup, including three new Gemini models and an early Gemini 4 preview.

Project Suncatcher Starts With a Test

Google describes Project Suncatcher as a long-term research effort to explore scalable machine learning infrastructure in space. The company’s future concept involves satellites equipped with multiple TPUs and linked together using high-bandwidth laser communications.

The October 1 mission is not a full-scale commercial data center in space. It is an early technology test intended to show how Google’s AI hardware performs in orbit before the company considers larger satellite-based systems.

Google says the first launch will provide data for future designs, including work on cooling AI chips in a vacuum and connecting satellites precisely enough to support high-bandwidth computing workloads.

The company’s official Project Suncatcher update says the first orbital mission is scheduled to launch aboard the Transporter-18 rideshare mission with SpaceX.