NVIDIA CorporationNvidia announces full-scale production of Groq 3 LPX, a new AI accelerator.
Nvidia announced that its Groq 3 LPX artificial intelligence inference accelerator is now in full-scale production. The announcement was made at the annual Hot Chips event, where Nvidia said the Groq 3 LPX, an extension of the next-gen Vera Rubin platform, can deliver 3,400 output tokens per second running the Gemma 4 31B open-source agentic model, according to benchmarking from Artificial Analysis. Nvidia also said Nebius is the first AI cloud to adopt Groq 3 LPX. The company entered a non-exclusive licensing agreement with Groq for its inference technology in December 2025, reportedly acquiring assets from Groq for $20 billion.
NVIDIA CorporationNvidia announces full-scale production of Groq 3 LPX, a new AI accelerator.
Nebius Group N.V.Nebius is the first AI cloud to adopt Groq 3 LPX, boosting its product offerings.