aiqbench takeChips & Infra
aiqbench analysis
This is a capacity announcement: Oracle is committing to multi-gigawatt AI cloud deployments built on Nvidia's Vera Rubin systems. Its practical significance depends on delivery timelines, power availability, and whether the claimed gains in GPU utilization hold up in production. Readers tracking AI compute supply should watch for concrete deployment dates and pricing rather than the scale figures alone.
4
Sources covering36Kr, HPCwire, Oracle Blogs, VentureBeat
Nvidia and Oracle Expand AI Infrastructure Partnership, with Vera Rubin System Deployments Set to Reach Multi-Gigawatt Scale
36Kr1 min read
From 36Kr
SummaryOn October 27 local time, Nvidia and Oracle announced they will expand Oracle Cloud Infrastructure (OCI) Supercluster featuring Nvidia Vera Rubin NVL72 systems to multi-gigawatt scale, and will officially offer Vera Rubin NVL72 systems and Vera CPU cloud services. Oracle also plans to launch HGX Rubin NVL8 systems equipped with eight Rubin GPUs to support frontier model training, inference, and agentic AI applications. The two companies said the partnership aims to improve GPU utilization and provide scalable computing power for AI workloads.