Nvidia announces space compute modules, including Space-1 Vera Rubin

Will be used by Aetherflux, Axiom Space, Kepler Comms, Planet, Sophia Space & Starcloud

Nvidia has developed a space-specific module of its Vera Rubin GPU-CPU platform.

The company said that Nvidia compute hardware will operate in orbit from Aetherflux, Axiom Space, Kepler Communications, Planet, Sophia Space, and Starcloud.

Last November, Starcloud sent an Nvidia H100 GPU to space on a test satellite - the first time an Nvidia GPU has gone to orbit. Nvidia said that the new module delivers up to 25x more AI compute for space-based inferencing, compared to the H100.

Nvidia said that the new module will target orbital data centers, advanced geospatial intelligence processing, and autonomous space operations. For less intensive workloads, the company said that Jetson Orin was capable for real-time processing of vision, navigation, and sensor data. It also offers IGX Thor, based on Blackwell, which is targeted at the Edge.

“Space computing, the final frontier, has arrived,” Jensen Huang, founder and CEO of Nvidia, said. "As we deploy satellite constellations and explore deeper into space, intelligence must live wherever data is generated.

“AI processing across space and ground systems enables real-time sensing, decision-making and autonomy, transforming orbital data centers into instruments of discovery and spacecraft into self-navigating systems. With our partners, we’re extending Nvidia beyond our planet — boldly taking intelligence where it’s never gone before.”

Baiju Bhatt, founder and CEO of Aetherflux, said: “At Aetherflux, we’re pioneering a new paradigm for power and compute in space. Nvidia Vera Rubin delivers high-performance, energy-efficient AI at the Edge in orbit, powered by solar energy. This enables autonomous operations and mission-critical services, and unlocks scalable, space-based AI infrastructure beyond Earth.”

Sophia Space said that it planned to use Jetson Orin for its modular platforms, which it offers to satellite operators. Kepler Communications will also use Jetson.

IGX Thor and Jetson Orin are both currently available, while the Vera Rubin Space Module will be available at "a later date."

The announcement comes after Nvidia posted a job listing earlier this month for an orbital data center system architect. "This is an opportunity to join the leader in AI systems at the inception of a completely new industry," the listing said.

Rival Google plans to launch a number of TPUs into space, already having tested them with a particle accelerator to simulate low-earth orbit levels of radiation. The company has partnered with Planet for a small deployment, but envisions eventually sending gigawatts into space.

SpaceX CEO Elon Musk also seeks to deploy a million-satellite orbital AI data center megaconstellation.

Musk has suggested that the satellites will use chips developed by Tesla, another Musk business, but timelines remain unclear.

Space data centers have a number of critics, including OpenAI CEO Sam Altman, short seller Jim Chanos, AWS CEO Matt Garman, and analysts at Gartner.