Most semiconductor fabs cost billions of dollars and take years to build. Besxar Space Industries is trying a different approach: strap a microwave-oven-sized manufacturing pod to a SpaceX rocket, let it cook for eight minutes in the vacuum of space, and bring it back down with the booster.
The Washington, D.C.-based startup completed its first test flight on July 6, 2026, during SpaceX’s Starlink 10-50 mission. Two of Besxar’s “Clipper Class Fabships” rode alongside 29 Starlink v2 Mini satellites aboard a Falcon 9 first-stage booster, making it one of the more unusual rideshare arrangements in commercial spaceflight history.
How making chips in space actually works
Besxar’s Fabships are fully autonomous manufacturing pods designed to produce ultra-pure semiconductor substrates and precursor materials. Rather than orbiting the Earth for days or weeks, they take advantage of a window that already exists in every Falcon 9 launch: the roughly eight minutes the first-stage booster spends above the Kármán line before it turns around and lands on a drone ship or back at the launch pad.
This is a critical distinction from other space manufacturing ventures, which typically focus on microgravity as the key advantage. Besxar’s bet is that the ultra-high vacuum environment matters more than weightlessness for producing the kind of pristine materials needed in advanced AI chips, quantum computing components, and defense applications.
From OpenAI to orbital fabs
Besxar was founded in 2023 by Ashley Pilipiszyn, who was an early employee at OpenAI. The company has secured a 12-mission agreement with SpaceX, first announced on October 28, 2025. It also has backing from NVIDIA’s Inception Program, which provides startups with technical resources and go-to-market support. On top of that, Besxar holds a contract with the US Department of Defense.
The initial July test flight carried terrestrial-manufactured wafers rather than attempting in-space production. The goal was simpler but essential: prove the Fabships could survive the violent forces of launch and re-entry without falling apart.
Why eight minutes might be enough
With 12 missions on the books, Besxar can fly, test, learn, redesign, and fly again on a cadence that would be impossible with dedicated orbital platforms. The approach also sidesteps the logistical overhead of ISS access, which involves lengthy approval processes and scheduling constraints. Besxar’s pods don’t need life support. They don’t need astronauts. They just need a spot on a booster that’s already going up.