NVIDIA: SpaceXAI to Deploy Vera CPU Accelerated Agent-Based AI, and Expand Vera Rubin Platform to On-Board Computing

BlockBeats News, August 24th

On August 24th, Nvidia officially announced that SpaceXAI will deploy the Nvidia Vera CPU to accelerate its next-generation AI applications. Vera is the first CPU specifically built for AI agents, with 88 Nvidia-designed Olympus cores, using spatial multithreading technology and LPDDR5X high-bandwidth memory with a bandwidth of 1.2TB/s. In agent-based AI, reinforcement learning, and data processing scenarios, the task completion speed can be increased by up to 1.8 times compared to an x86 CPU.

SpaceXAI President Mike Nicolls stated that Vera's CPU performance and memory bandwidth are sufficient to support large-scale task orchestration, code execution, and data processing, allowing the GPU to remain highly efficient, thereby achieving more effective output per watt of computation.

SpaceXAI also plans to expand the Grok-backed AI infrastructure based on the NVIDIA Vera Rubin platform, moving towards a gigawatt-scale computing capacity. Vera Rubin integrates Nvidia's accelerated computing, NVLink interconnect, Spectrum-X network, and BlueField data processing technologies, aiming to reduce the cost per token and enhance the overall efficiency of AI factories. Of particular note, SpaceXAI will send an optimized version of the Vera Rubin NVL72 system into space, with its first-generation Starmind AI satellite adopting this rack-level architecture, marking the first iconic deployment of Nvidia accelerated computing from ground data centers to orbital computation.