Nvidia has signed a major technology agreement with Groq, one of its key competitors in the field of artificial intelligence chips. Under the agreement, Nvidia has obtained a license to use Groq’s proprietary processor technologies. At the same time, Groq’s founder and CEO, Jonathan Ross, along with the startup’s senior leadership and core engineering team, have transferred to Nvidia.
Although the financial details of the deal have not been officially disclosed, some sources claim that Nvidia paid approximately $20 billion for these assets. If confirmed, this would represent the largest acquisition or technological investment in Nvidia’s history. Groq as a startup will not be fully shut down; instead, it will continue to operate independently under new leadership and will keep providing its cloud services.
What makes Groq particularly attractive to Nvidia is its alternative chip architecture known as the “LPU” (Language Processing Unit). The startup claims that these chips can run AI models up to 10 times faster while consuming 10 times less energy than Nvidia’s current processors. This innovation could allow Nvidia to further strengthen its dominance in the AI inference market.
