SpaceXAI will use the NVIDIA Corporation (NVDA.US) Vera CPU to accelerate agent AI applications.

NVIDIA Corporation (NVDA.US) announced on Monday that SpaceX (SPCX.US) subsidiary SpaceXAI will deploy the Vera CPU to accelerate its next-generation Agentic AI applications. SpaceXAI, which was formerly known as xAI, an independent company founded by Elon Musk, is now a department under SpaceX. SpaceXAI plans to expand the infrastructure behind its AI chat Siasun Robot & Automation Grok on the NVIDIA Corporation Vera Rubin platform, while extending the optimized Vera Rubin NVL72 solution into space through its first-generation Starmind satellites. NVIDIA Corporation stated that Agentic AI applications increasingly rely on CPUs to orchestrate tools, execute code, process data, and run simulation tasks between model calls. SpaceXAI will utilize Vera to accelerate these application workloads, helping AI agents respond more quickly while ensuring that GPUs continuously receive data and maintain high utilization rates. NVIDIA Corporation pointed out that as SpaceXAI scales to exawatt-level computing power, Vera Rubin provides a universal architecture that can efficiently scale its next-generation AI factorythis is the foundation for SpaceXAI's plan to move beyond terrestrial data centers and towards orbital computing. It is reported that Vera is the first CPU specifically designed for AI agents, aimed at accelerating CPU-intensive tasks around model inferencefrom tool invocation, code execution to data processing, orchestration, and simulation. Compared to x86 CPUs, Vera can increase task completion speed by up to 1.8 times across various workloads including Agentic AI, reinforcement learning, and data processing. "Vera provides the CPU performance needed for real-time actions by AI agentsexecuting code, processing data, and coordinating complex tasks. SpaceXAI is pushing this architecture from large AI factories to the new frontier of orbital computing," said Ian Buck, NVIDIA Corporation's Vice President of Hyperscale and High-Performance Computing.