NVIDIA and Oracle plan multi-gigawatt Vera Rubin expansion on Oracle Cloud
NVIDIA and Oracle announced plans to expand Oracle Cloud Infrastructure Superclusters with Vera Rubin NVL72 systems to multi-gigawatt scale. The companies also made Vera Rubin NVL72 and Vera CPUs generally available on OCI.
A forecast as of publication. An announcement and on-sale availability require separate confirmation.
NVIDIA and Oracle plan to expand Oracle Cloud Infrastructure (OCI) Superclusters with NVIDIA Vera Rubin NVL72 systems to multi-gigawatt scale. The infrastructure is intended for training and serving frontier AI models and for agentic applications.
The companies announced the general availability of Vera Rubin NVL72 and Vera CPUs on OCI. Customers can scale GPU and CPU resources independently to match their workloads.
OCI Supercluster components
Each NVL72 rack connects 72 GPUs through NVIDIA NVLink. The clusters also use Oracle Acceleron networking, NVIDIA Spectrum-X Ethernet, Spectrum-6 switches and BlueField-4 DPUs; OCI’s network can scale across hundreds of thousands of GPUs.
Oracle plans to introduce NVIDIA HGX Rubin NVL8, an x86-based server configuration with eight Rubin GPUs, two 64-core Intel Xeon 6 CPUs, 4TB of DDR5 memory and 6.4 Tbps of RDMA networking.
OpenAI runs GPT-6 Astra on OCI
OpenAI trained and served GPT-6 Astra in Abilene on an OCI Supercluster with more than 100,000 NVIDIA Blackwell GPUs. Sreekanth Pothanis, an OpenAI engineering leader, said Vera Rubin on OCI provides the scale and performance needed for increasingly demanding workloads.
Vera CPUs target agent workloads
NVIDIA Vera CPUs on OCI are designed for operations AI agents perform between model calls, including running tools, executing code and retrieving data. Oracle said high single-thread performance and memory bandwidth can reduce GPU idle time during those operations.
OCI Kubernetes Engine allows CPU and GPU resources to scale independently, while Oracle Acceleron provides hardware-based isolation for workloads running untrusted AI-generated code. OpenAI said it is bringing Vera into production for agentic sandbox workloads.
What we know
- NVIDIA and Oracle plan to scale OCI Superclusters with Vera Rubin NVL72 systems to multi-gigawatt capacity.
- Vera Rubin NVL72 and Vera CPUs are generally available on Oracle Cloud Infrastructure.
- The HGX Rubin NVL8 configuration will include eight GPUs, two 64-core Xeon 6 CPUs and 4TB of DDR5 memory.
- OpenAI uses an OCI Supercluster with more than 100,000 Blackwell GPUs for GPT-6 Astra.
- Vera CPUs target tool execution, code execution and data retrieval for AI agents.
Trust: Primary source
- The story cites a primary statement. It establishes what its author said, but does not guarantee every claim is true.
- Sources: NVIDIA Newsroom.
COMMUNITY
Discussion
Sign in to join the discussion.
Loading comments…