Overview
- The company reported record fiscal Q1 revenue and on May 20–21 projected nearly $20 billion in standalone CPU revenue for the current fiscal year.
- Nvidia publicly unveiled the Vera CPU architecture and the Vera Rubin platform as purpose‑built hardware for agentic AI workloads, citing 88 custom Arm cores and high memory bandwidth.
- Initial Vera production systems were delivered to major AI customers and OEMs, and Dell began shipping Vera in PowerEdge servers as Nvidia moves to scale production toward broader H2 2026 deliveries.
- U.S. export controls complicate near‑term China sales: Nvidia excluded China data‑center compute from its immediate guidance even though management counts China in its longer‑term $200 billion CPU market assumptions.
- Analysts predict Vera could challenge Intel and AMD on performance and unit shipments, a shift that could change data‑center supplier roles and affect how cloud operators buy and run multi‑agent AI systems.