SpaceXAI adopts NVIDIA Vera CPU, extends Vera Rubin compute to Starmind satellites
On August 25, SpaceXAI announced adoption of NVIDIA Vera CPU and plans to use an optimized Vera Rubin NVL72 in first-gen Starmind AI satellites.
On August 25, SpaceXAI announced an expansion of its deployment on NVIDIA's new Vera Rubin platform, adopting the Vera CPU for next-generation agentic AI workloads. Vera is NVIDIA's first CPU designed specifically for AI agents, handling tool orchestration, code execution, data processing, and simulation workloads.
More notably, SpaceXAI plans to deploy an optimized Vera Rubin NVL72 in its first-generation Starmind AI satellites, extending NVIDIA's AI compute platform from terrestrial data centers into orbit. The deployment addresses orbital constraints in power, thermal management, bandwidth, reliability, and physical integration.
NVIDIA also announced the Spectrum-X Multiplane Ethernet scaling technology, which can scale a single network layer to 512,000 GPUs. An eight-plane configuration retains roughly 90% of bandwidth even if a single plane fails. ScaleIn, based on the BlueField-4 DPU and DOCA software platform, brings infrastructure services into the acceleration scope.