Oracle and NVIDIA plan multi-gigawatt Vera Rubin expansion

NVIDIA and Oracle plan to expand Oracle Cloud Infrastructure (OCI) Superclusters with NVIDIA Vera Rubin NVL72 systems to multi-gigawatt scale for AI training, inference and agentic applications. The companies also announced general availability of Vera Rubin NVL72 and NVIDIA Vera CPUs on OCI, allowing customers to scale GPU and CPU resources independently to match workloads and reduce idle capacity.

OCI Superclusters combine rack-scale NVIDIA systems with Oracle Acceleron and NVIDIA Spectrum-X networking, alongside gigawatt-class power and cooling. Each rack connects 72 GPUs through NVLink, while OCI’s network scales across hundreds of thousands of GPUs. Oracle also plans to introduce HGX Rubin NVL8 systems, offering eight Rubin GPUs in an x86-based server configuration for training, fine-tuning and inference.

Vera CPUs target the work agents perform between model calls, including executing tools, testing code and retrieving data. OCI Kubernetes Engine supports independent scaling of CPU and GPU resources, while Oracle Acceleron provides hardware-based isolation for untrusted, AI-generated code. OpenAI said it is bringing Vera CPUs into production on OCI for agentic sandbox workloads. It has also trained and served GPT-6 Astra on an OCI Supercluster in Abilene with more than 100,000 NVIDIA Blackwell GPUs.