Nvidia Launches Space-1 Vera Rubin Module for Orbital AI Computing

Nvidia has launched a suite of space AI computing platforms, headlined by the space-hardened Space-1 Vera Rubin Module, designed to enable artificial intelligence applications from orbit and on Earth. The move positions the chipmaker at the forefront of a rapidly expanding space economy that McKinsey & Co. projects will reach $1.8 trillion by 2035. CEO Jensen Huang acknowledged that current economics for space-based AI data centers remain poor but said they will improve over time, pointing to solar power in space as a potential solution to the power bottlenecks that constrain Earth-based data centers.

Several companies are already leveraging Nvidia's accelerated computing platforms for space applications. Axiom Space is using them for next-generation missions, Kepler Communications for orbital data relay, and Planet Labs for its satellite imagery business. Most notably, startup Starcloud launched a satellite in November equipped with a space-hardened Nvidia H100 chip, marking the first time an advanced AI chip was sent into space. Nvidia is further deepening its commitment to orbital computing by recruiting an Orbital Datacenter System Architect to build dedicated products for AI operations in orbit. The combination of hardware launches, existing partnerships, and new hiring signals Nvidia's intent to establish a lasting infrastructure presence in space as satellite networks and orbital data demands grow.