Nvidia has begun delivering computing systems powered by its new Vera Rubin architecture to major artificial intelligence developers. The silicon giant confirmed the hardware is already entering active use at data centers operated by several major tech firms. However, the exact performance benchmarks of these production chips remain under wraps.
During a briefing at the corporate headquarters in Santa Clara, Ian Buck, the vice president of the company, confirmed the manufacturing milestones.
We are absolutely in full production.
He indicated that the technology is already in the hands of early buyers, stating that the systems are actively being stood up at all of their major customers.
According to details first reported by Bloomberg News, OpenAI plans to deploy these new systems at scale by the 3rd quarter. Other major tech entities, including CoreWeave, Google Cloud, Microsoft Azure, Meta, and Dell Technologies, are also working to integrate the architecture into their existing infrastructure. This rapid deployment schedule aligns with the aggressive timeline previously outlined by CEO Jensen Huang.
The Rubin architecture succeeds the Hopper and Blackwell generations as the primary computing engine for advanced training workloads. This rollout comes at a time when investors are closely monitoring supply chain timelines to watch for potential manufacturing bottlenecks. Showing a smooth production schedule helps calm market anxieties while helping Nvidia maintain its lead over competitors like AMD and various in house custom silicon projects.
