NVIDIA extends Vera Rubin for agentic AI inference

NVIDIA has announced new AI infrastructure and partner adoption updates for the Vera Rubin platform, including Groq 3 LPX in full production and related networking and infrastructure technologies.

• NVIDIA Groq 3 LPX is now in full production and is described as extending Vera Rubin NVL72 for fast token generation in agentic systems; in an Artificial Analysis benchmark on Gemma 4 31B, it reached 3,400 output tokens per second for 100,000-token long-context use cases.
• Nebius is the first AI cloud to adopt Groq 3 LPX; CoreWeave has deployed Spectrum-X Multiplane in production; SpaceXAI plans to build future AI architecture around Vera Rubin and deploy Vera CPUs for agentic AI workloads. The article also introduces NVIDIA Scale-In and NVLink Fusion as part of NVIDIA’s broader AI factory infrastructure stack.