NVIDIA Groq 3 LPX in full production, hits 3,431 tokens/s on Gemma 4 31B

NVIDIA announced Groq 3 LPX, its interactive AI inference accelerator for the Vera Rubin platform, is now in full production. In an Artificial Analysis benchmark, it delivered 3,431 output tokens per second on Gemma 4 31B with 100K context, 4x faster than the nearest alternative. Nebius is the first AI cloud to adopt it.