Groq

Groq's chip design will be incorporated into Nvidia's new AI processor, indicating a significant collaboration. Nvidia also struck a $20 billion licensing deal with Groq, which halted OpenAI's direct discussions with Groq for faster inference chips.

Nvidia licensed technology from Groq for $17 billion, integrating Groq's chips into Nvidia's new AI system to handle the 'decode' stage of inference computing. This partnership highlights Groq's specialized expertise in fast AI inference.

Groq, a chip startup, licensed its inference computing technology to Nvidia for $17 billion. Its technology will be a focus at the conference, highlighting Nvidia's efforts to strengthen its position in the inference chip market.

Nvidia acquired Groq for $17 billion, a chip startup specializing in fast and cheap inference computing. Nvidia plans to integrate Groq's technology into its CUDA platform and new server lines.

Nvidia struck a multi-billion dollar licensing deal with Groq, a market specialist in inference, and hired its top engineers to aid Nvidia's transition into the inference field.