Nvidia Corporation Licenses Groq Tech to Accelerate AI Inference Speed

Nvidia Corporation is developing a new inference computing platform that integrates technology from the startup Groq Inc. to increase the speed and efficiency of AI responses. The project targets performance bottlenecks for high-profile clients such as OpenAI Inc., which had expressed dissatisfaction with the speed of existing Nvidia hardware for ChatGPT tasks, including software development.

To protect its market position, Nvidia entered into a $20 billion licensing agreement with Groq Inc. This move effectively terminated OpenAI Inc.'s independent negotiations with Groq Inc. and other hardware startups, including Cerebras Systems.

Nvidia Corporation also invested in OpenAI Inc. to provide capital for chip procurement. Reports on the scale of this financial commitment vary, with one source citing up to $100 billion and another stating the arrangement was restructured into a $30 billion investment. Nvidia plans to officially unveil the new platform at its GTC developer conference in San Jose.