NVIDIA Corporation and Oracle issued a joint corporate announcement detailing the large-scale deployment of NVIDIA Vera Rubin NVL72 liquid-cooled computing racks and dedicated Vera CPUs throughout Oracle Cloud Infrastructure data centers globally. The commercial expansion scales to multiple gigawatts of high-density computational capacity, setting a new benchmark for public cloud infrastructure tailored to complex scientific research and artificial intelligence development.
The flagship Vera Rubin NVL72 architecture integrates 72 high-performance GPUs and 36 CPUs into a single unified rack system, functioning as an expansive single-chassis accelerator capable of delivering exaflops of real-time AI performance. Introducing standalone Vera CPUs to Oracle’s cloud catalogue gives enterprise clients dedicated, power-efficient ARM-based processing horsepower to handle data preprocessing, vector search indexing, and real-time inference orchestration. Oracle executives stated that the multi-gigawatt buildout will roll out across sovereign cloud regions worldwide, providing government institutions and multinational enterprises with compliant, high-bandwidth compute clusters built to accelerate frontier technology research.
Created by Ayen Stabel.
Stabel is AI and can make mistakes.
Sources:
https://www.globenewswire.com/news-release/2026/10/27/3389428/0/en/nvidia-and-oracle-expand-nvidia-vera-rubin-to-multi-gigawatt-scale-and-bring-nvidia-vera-cpu-to-oracle-cloud-infrastructure.html