Nvidia Expands Vera Rubin Platform with Dedicated Inference Hardware
Nvidia announced major updates to its Vera Rubin platform at GTC 2026, including custom CPU racks, dedicated inference chips, a new storage architecture, an inference operating system, open model alliances, and agent security software.
Why it matters
Nvidia's expansion of the Vera Rubin platform signals the company's continued push to dominate the AI hardware and software market, now with a stronger focus on inference.
Key Points
- 1Nvidia introduced the Vera Rubin platform with custom CPU racks
- 2The platform now includes dedicated inference chips, the Groq 3 LPX
- 3Nvidia also unveiled a new storage architecture, inference OS, and security software
Details
At its annual GTC conference, Nvidia expanded the Vera Rubin platform it had introduced earlier at CES. The updates include custom CPU racks, the addition of dedicated inference hardware in the form of the Groq 3 LPX chips, a new storage architecture, an inference-focused operating system, open model alliances, and agent-based security software. This marks the first time Nvidia has incorporated dedicated inference silicon into its platform, which has traditionally focused on training and accelerating large language models and other AI workloads. The new inference capabilities, along with the other platform enhancements, aim to provide a more comprehensive AI infrastructure solution for enterprises and researchers.
No comments yet
Be the first to comment