Nvidia and Oracle expand Vera Rubin to multi-gigawatt scale

Nvidia and Oracle are taking their AI partnership to a scale usually reserved for power plants. The two companies announced on October 27, 2026 that Oracle Cloud Infrastructure (OCI) Superclusters will grow to multi-gigawatt capacity using Nvidia’s Vera Rubin platform. The announcement goes beyond a promise. Vera Rubin NVL72 racks and Nvidia’s Vera CPUs are now available on OCI, so customers can rent the hardware today rather than read about it in a roadmap. What Oracle is actually getting Vera Rubin is the successor to Nvidia’s Grace Blackwell generation. The flagship configuration is the NVL72 rack. It packs dozens of GPUs into a single cabinet and stitches them together with NVLink 6, Nvidia’s latest chip-to-chip interconnect. Memory is the other headline spec. Each Vera Rubin GPU uses HBM4 memory supporting up to 288 GB and 22 TB/s of bandwidth. Oracle will wrap these racks in advanced Ethernet networking and Nvidia’s BlueField-4 DPUs. A DPU, or data processing unit, handles networking, storage and security chores so the GPUs can stay focused on AI work. The deal also brings HGX Rubin NVL8 systems into the mix. These smaller eight-GPU configurations are aimed at frontier model ...