Nvidia Confirms Vera Rubin’s Full Production and OpenAI Deployment
Nvidia has confirmed that its Vera Rubin platform is in full production, with systems shipping to leading tech companies like OpenAI, CoreWeave, Google Cloud, Microsoft Azure, Meta, and Dell. According to Bloomberg, OpenAI plans to adopt Vera Rubin at scale during Q3.
Key Highlights:
- CoreWeave’s NVL72 Racks: Reporting a tenfold increase in token output compared to the previous generation.
- Vera CPU Performance: Outperforming AMD’s Turin chip on Python workloads, making it nearly twice as fast.
- Nvidia’s Cooling Innovation: The NVL72 system uses liquid cooling to eliminate internal cabling, enhancing component density.
Quote from Ian Buck:
"…CoreWeave is already measuring a tenfold increase in token output from the new NVL72 racks, while the Vera CPU outpaces AMD on Python workloads."
Context and Stock Performance:
This production milestone comes at an interesting time for Nvidia’s stock. While the company’s shares have shown modest growth this year, the broader chip index has experienced a significant surge. Intel, ARM, and AMD have all seen their stocks double or more.
Analysts predict Nvidia’s revenue will climb by 82% to approximately $393 billion for the current fiscal year, but its stock price hasn’t mirrored the same momentum as the previous bull market.