Google Cloud and NVIDIA announced on October 21 that A5X bare-metal instances powered by NVIDIA Vera Rubin NVL72 GPUs will be generally available next month, targeting large-scale reinforcement learning and agentic inference workloads.
Single-site deployments scale to 80,000 Rubin GPUs and multisite configurations reach 960,000 GPUs, with the infrastructure delivering up to 10 times lower inference cost per token and up to 10 times higher token throughput per megawatt versus the prior generation. Ineffable Intelligence, founded by AlphaGo creator David Silver, was the first customer to access A5X and is using it for reinforcement learning research.