Nvidia and Oracle announce expanded partnership

Nvidia and Oracle plan to expand Oracle Cloud Infrastructure superclusters to multi-gigawatt scale using Nvidia Vera Rubin NVL72 systems, with Vera Rubin NVL72 and the Vera CPU now generally available on OCI. The expanded setup covers both GPU-heavy frontier training and inference and CPU-heavy agent work, with CPU and GPU resources scalable independently to cut idle capacity. Why it matters: it gives customers a single infrastructure route for both kinds of AI workloads instead of sizing them together.

• 36氪 AI2026-10-27 23:02