Nvidia Starts Vera CPU Shipments As Oracle Plans Massive 2026 Rollout

Khac Phu Nguyen May 19, 2026 1 min read This article first appeared on GuruFocus. Nvidia (NASDAQ:NVDA) has moved its Vera CPU systems into production, marking a potentially important step as agentic AI workloads begin shifting from simple responses toward systems designed to act, simulate and operate at scale. The first infrastructure systems were hand-delivered by Nvidia executive Ian Buck to early enterprise adopters including Anthropic, OpenAI, Oracle Cloud Infrastructure and SpaceXAI, showing how Nvidia is pushing its AI factory architecture deeper into the CPU layer. Warning! GuruFocus has detected 5 Warning Signs with NVDA. Is NVDA fairly valued? Test your thesis with our free DCF calculator. Buck framed Vera as part of a new CPU moment created by agentic AI, where models are moving from answering to acting and need infrastructure built to keep that work moving at scale. The first delivery went to Anthropic's SoMa offices in San Francisco, while the final delivery of the day went to SpaceXAI's offices in Palo Alto. Nvidia's team also walked Elon Musk through the system's interior, with SpaceXAI evaluating Vera for reinforcement learning workloads and agent-based simulation pipelines that support its training stack. For investors, the hyperscale angle may be the bigger signal. Oracle Cloud Infrastructure is the first cloud provider to deploy Vera at hyperscale and plans to roll out hundreds of thousands of Nvidia Vera CPUs beginning in 2026. Vera also sits inside Nvidia's broader co-design strategy alongside Rubin GPU, BlueField 4 DPU, Spectrum-X and MGX rack architecture. Beyond standalone CPU systems, Vera will serve as the host processor for Vera Rubin NVL72, where it connects through second-generation Nvidia NVLink-C2C to a pair of Rubin GPUs, possibly making it another key piece of Nvidia's expanding AI infrastructure stack.