Item 1 of two A NVIDIA emblem seems on this illustration taken August 25, 2025. REUTERS/Dado Ruvic/Illustration
The deal follows a well-known sample lately the place the world’s greatest know-how companies pay massive sums in offers with promising startups to take their know-how and expertise however cease wanting formally buying the goal.
Sign up here.
Nvidia has agreed to a “non-exclusive” license to Groq’s know-how, Groq mentioned. It mentioned its founder Jonathan Ross, who helped Google begin its AI chip program, in addition to Groq President Sunny Madra and different members of its engineering group, will be a part of Nvidia.
An individual shut to Nvidia confirmed the licensing settlement.
Groq didn’t disclose monetary particulars of the deal. CNBC reported that Nvidia had agreed to purchase Groq for $20 billion in money, however neither Nvidia nor Groq commented on the report. Groq mentioned in its weblog publish that it’s going to proceed to function as an impartial firm with Simon Edwards as CEO and that its cloud enterprise will proceed working.
“Antitrust would seem to be the primary risk here, though structuring the deal as a non-exclusive license may keep the fiction of competition alive (even as Groq’s leadership and, we would presume, technical talent move over to Nvidia),” Bernstein analyst Stacy Rasgon wrote in a notice to purchasers on Wednesday after Groq’s announcement. And Nvidia CEO Jensen Huang’s “relationship with the Trump administration appears among the strongest of the key US tech companies.”
Groq is one in all quite a lot of upstarts that don’t use exterior high-bandwidth reminiscence chips, liberating them from the reminiscence crunch affecting the worldwide chip business. The method, which makes use of a type of on-chip reminiscence referred to as SRAM, helps velocity up interactions with chatbots and different AI fashions but in addition limits the scale of the mannequin that may be served.
Nvidia’s Huang spent a lot of his greatest keynote speech of 2025 arguing that Nvidia would have the option to keep its lead as AI markets shift from coaching to inference.
Reporting by Stephen Nellis in San Francisco; Additional reporting by Harshita Mary Varghese in Bengaluru; Editing by Shailesh Kuber, Matthew Lewis and William Mallard
Our Standards: The Thomson Reuters Trust Principles.