SpaceX launches Google's first AI chips into space in the initial orbital test of Alphabet's Suncatcher project

SpaceX put Google-designed artificial intelligence chips into orbit on Thursday, October 1, aboard a Falcon 9 rocket that lifted off from Vandenberg base in Santa Barbara County, California. The launch window opened at 11:15 a.m. Pacific time, and the company confirmed the successful deployment of the entire payload at around 1 p.m. local time. The mission, named Transporter-18, carried several Planet Labs satellites, among them a solar-powered prototype fitted with Google tensor processing units (TPUs).

The launch marks the first in-orbit test of Alphabet's Project Suncatcher, an initiative billed as a 'moonshot' and first unveiled in November 2025. The goal is to develop reliable, solar-powered AI computing infrastructure capable of operating continuously in space. Travis Beals, senior director of Project Suncatcher, wrote in a blog post after the launch that the team had 'confirmed contact with the satellite and that it is operating as expected', describing the milestone as the first step in a long-term research effort to explore whether space could one day host large-scale machine learning infrastructure.

The rationale behind the project rests on a physical advantage: in low Earth orbit, satellites can access almost constant sunlight, generating up to eight times more solar power than on Earth, according to Alphabet. Longer term, the company envisions linking multiple satellite constellations together to handle larger AI workloads directly in orbit. Google had already tested its TPUs running AI workloads at a University of California, Davis facility, but it still does not know how its chips will behave in the harsh conditions of low Earth orbit, which include extreme temperature swings and radiation exposure.

The launch also highlights the close business ties between Alphabet and SpaceX, even as their respective artificial intelligence divisions compete with each other. Alphabet is a significant investor in SpaceX, which went public in June in a landmark offering, and its stake in SpaceX is currently valued at more than $82 billion.

SpaceX, for its part, is pursuing a parallel and in some ways rival strategy: the company led by Elon Musk has announced its own plans to build and launch orbital data centers, made up of satellite swarms developed at its facilities in Redmond, Washington, equipped with graphics processing units (GPUs) and solar panels it plans to manufacture together with Tesla, Musk's carmaker. Earlier this year, Musk said data centers in space would be the cheapest way to train AI, and that this 'will be true within two years, maybe three at most'. In the same vein, SpaceX chief operating officer Gwynne Shotwell said at an event in September that the company will deploy 'supercomputing in space' in 2027.

The article nonetheless notes skepticism from industry experts, who see the idea of space data centers as a still-distant bet — if it proves viable at all — in part because rocket launches remain limited in capacity and expensive. Added to that are the technical challenges of operating hardware in space: the need for cooling systems, chips able to withstand extreme temperatures, radiation shielding, and the risk that accumulating orbital debris poses to the viability of this kind of infrastructure.

The Transporter-18 flight was also SpaceX's second major launch of the day: that same Thursday it began ferrying astronauts to the International Space Station from Florida's Space Coast on behalf of NASA, starting a six-month mission in orbit. Taken together, the day reflects the extent to which SpaceX has become central both to institutional space programs and to Big Tech's most speculative bets on moving AI computing off Earth, at a time when the growth of terrestrial data centers is running increasingly into constraints on land, power and public acceptance.