NVIDIA and Oracle Take Vera Rubin to Multi-Gigawatt Scale
NVIDIA and Oracle plan to expand Oracle Cloud Infrastructure (OCI) Superclusters using NVIDIA Vera Rubin NVL72 systems to multi-gigawatt scale to meet AI demand. The collaboration includes generally available Vera Rubin NVL72 and NVIDIA Vera CPUs on OCI, with plans to introduce NVIDIA HGX Rubin NVL8. OpenAI has already used OCI for training and serving GPT-6 Astra on a supercluster with over 100,000 NVIDIA Blackwell GPUs, highlighting the infrastructure's role in AI advancements.
How this was made
The 30-second read
Why it matters
The key trading takeaway is a capacity and platform availability update on OCI for training, inference, and agentic AI, which can influence expectations for near-term AI infrastructure utilization and competitive positioning of both NVIDIA and Oracle Cloud.
Market read
A major cloud deployment expansion and general availability of NVIDIA’s Rubin GPU and Vera CPU platforms on OCI, signaling continued AI infrastructure buildout for agentic workloads.
What to watch
Execution risk remains around multi-gigawatt power and cooling buildout, and customers may still prioritize existing Blackwell deployments (GB200/GB300) before fully migrating to Rubin-based systems.
Background
The article frames an expanded NVIDIA-Oracle collaboration to scale OCI Superclusters using NVIDIA Vera Rubin NVL72 systems plus Vera CPUs, and introduces additional Rubin-based server options (HGX Rubin NVL8).
Ticker impact
NVIDIA will expand OCI Superclusters to multi-gigawatt scale using NVIDIA Vera Rubin NVL72 systems and Vera CPUs, scaling GPU and CPU capacity for customers.
Likely upward bias as investors price in continued adoption of NVIDIA’s newest GPU and CPU platforms with a major cloud customer.
The article describes an expanded deployment and general availability of Vera Rubin NVL72 and Vera CPUs on OCI, which supports near-term utilization expectations and reinforces NVIDIA’s platform lock-in narrative.
Oracle says OCI Superclusters built on NVIDIA Vera Rubin NVL72 will scale to multi-gigawatt capacity, with Vera Rubin and Vera CPUs now generally available on OCI.
Likely modest positive reaction as the market interprets the expansion as incremental OCI demand and improved infrastructure efficiency.
The piece is a customer-demand-driven expansion and product availability update, but it does not provide financial terms, contract size, or explicit revenue impact.
Market effects
Reinforces the AI infrastructure capex cycle and the shift toward agentic workloads that require balanced CPU and GPU scaling.
Limited direct regional impact; primarily global hyperscale/cloud infrastructure demand.
Supports broader AI compute supply chain demand for GPUs, networking (Spectrum-X), and data-center power and cooling.
Counterpoint
Without disclosed contract value, capacity commitments, or customer names beyond OpenAI, the market may treat this as incremental marketing/roadmap detail rather than a near-term revenue catalyst.
Key entities
- technology platformNVIDIA Vera Rubin NVL72
Rack-scale GPU system referenced as the basis for multi-gigawatt OCI Superclusters.
- cloud infrastructureOracle Cloud Infrastructure (OCI) Superclusters
OCI architecture combining NVIDIA NVL72 systems with Oracle networking and data-center components.
- technology platformVera CPUs
CPU component described as generally available on OCI alongside Vera Rubin NVL72 systems.
- customerOpenAI
Cited as using OCI at scale and testing Vera CPUs for agentic sandbox workloads.
