$SPCX Plans $NVDA-Powered AI Compute in Orbit

SpaceXAI plans to base its first-generation Starmind AI satellite on an optimized Nvidia Vera Rubin NVL72 system, extending the same architecture used in terrestrial AI factories into space.
On Earth, SpaceXAI will deploy Nvidia Vera CPUs for agentic workloads and expand the infrastructure behind Grok on Vera Rubin as it scales toward gigawatts of compute capacity. Vera is designed to handle CPU-heavy work around AI inference, including tool orchestration, code execution, data processing, and simulations, while keeping GPUs focused on model workloads.
The chip features 88 Nvidia-designed Olympus cores and up to 1.2 TB/s of memory bandwidth, with Nvidia saying it can deliver up to 1.8x faster task completion than x86 CPUs across certain agentic AI and data-processing workloads.
Nvidia and SpaceXAI are also working to adapt Vera Rubin for the different power, thermal, bandwidth, and reliability requirements of orbital computing.