SpaceXAI Adopts NVIDIA Vera CPU to Accelerate Agentic AI at Massive Scale
nvidia ·
NVIDIA announced that SpaceXAI will deploy its new Vera CPUs to accelerate next-generation agentic AI workloads, including orchestration, code execution, data processing, and simulation between model calls. The Vera CPU is described as the first CPU built for AI agents, featuring 88 NVIDIA-designed Olympus cores, Spatial Multithreading, and LPDDR5X memory with up to 1.2TB/s of bandwidth, enabling up to 1.8x faster task completion than x86 CPUs. SpaceXAI plans to expand its AI infrastructure behind Grok using the NVIDIA Vera Rubin platform as it scales toward gigawatts of computing capacity. In addition, SpaceXAI intends to extend NVIDIA accelerated computing into orbit with a first-generation Starmind AI satellite based on an optimized Vera Rubin NVL72 system. NVIDIA executives highlighted how Vera helps AI agents act in real time while keeping GPUs fully utilized, and SpaceXAI's president noted the CPU provides performance and memory bandwidth for large-scale orchestration and data processing. The integrated Vera Rubin architecture combines compute, networking, and software to optimize AI factory performance and energy efficiency while lowering cost per token. NVIDIA and SpaceXAI are adapting the platform for the unique constraints of orbital computing, such as power, thermal management, and reliability.
nvidia