Two announcements arrived this week that look unrelated. On March 16, at its annual developer conference in San Jose, NVIDIA confirmed that its next-generation Vera Rubin platform is being adapted for orbital deployment under the name Space-1 Vera Rubin, designed to run artificial intelligence data centers in orbit. Five days later, on March 21, Elon Musk launched the TERAFAB Project, a vertically integrated chip fabrication facility being built by Tesla in partnership with SpaceX and xAI to produce between 100 and 200 billion AI chips per year. The connection between them is not strategic partnership. It is shared diagnosis. Both companies looked at the same projection and reached the same conclusion: Earth cannot generate enough power or chip capacity to meet where AI demand is heading.

What NVIDIA Space-1 Actually Is

The Vera Rubin platform, named after the astronomer whose work confirmed the existence of dark matter, integrates seven chip types into a single vertically optimized system. NVIDIA claims it delivers ten times more performance per watt than its predecessor, Grace Blackwell. Space-1 Vera Rubin takes that architecture and redesigns it for orbital conditions. The stated goal is to bring accelerated computing off Earth for the first time, enabling real-time geospatial intelligence and autonomous operations in space. Partners already named include Aetherflux, Axiom Space, Kepler Communications, Planet Labs, and Starcloud.

Jensen Huang's framing at the keynote was direct: computing demand has increased by one million times over the last few years. He projected at least one trillion dollars in order revenue from Blackwell and Vera Rubin systems between 2025 and 2027, doubling an estimate he gave less than six months ago. That acceleration is what makes the orbital data center idea something other than science fiction. If inference demand keeps compounding, and if terrestrial power grids cannot keep pace, space-based solar-powered computing becomes an engineering answer rather than a marketing narrative.

What TERAFAB Is Actually Solving

TERAFAB was first confirmed at Tesla's January 2026 earnings call, where Musk told investors the company needed its own fabrication facility to avoid a chip supply constraint projected to materialize within three to four years. The facility combines logic processing, memory, and advanced packaging under one roof, targeting 100,000 wafer starts per month initially with ambitions to scale toward one million. Tesla's fifth-generation AI chip, called AI5, is among the first products it will produce, with small-batch runs expected later in 2026 and volume production in 2027. The facility is targeting 2-nanometer process technology, which is currently the most advanced node in commercial production.

The TERAFAB vision extends beyond terrestrial robotics. The announcement explicitly targets solar-powered AI satellites and orbital infrastructure, framing the chip factory as a prerequisite to a space-based energy and compute economy. That positions TERAFAB not as a competitor to TSMC, but as a supply chain foundation for a future that neither TSMC nor Samsung is currently designing toward.

The Industrial Theory Underneath Both Bets

NVIDIA's approach is to design hardware that can operate in space using partners already building orbital infrastructure. Tesla's approach is to build the factory that makes the chips those orbital systems will need. One company is designing for the environment. The other is manufacturing for the demand. Both are working backward from the same number: a compute requirement that exceeds what the current terrestrial industrial base can supply.

There is a legitimate execution question sitting under the TERAFAB announcement. Building a 2-nanometer fab from scratch is, by most accounts in the industry, one of the hardest technical challenges in existence. It took TSMC decades to reach this process node. Musk's stated planning horizon is one to two years, which is structurally incompatible with the traditional semiconductor buildout cycle. The more grounded read is that TERAFAB's March 21 launch is a project initiation, not a factory opening. Small-batch production in 2026 and volume in 2027 are still aggressive timelines, and the facility location has not been officially confirmed, though early hiring points to the North Campus of Gigafactory Texas in Austin.