In a new deal, Nvidia hires Groq's top engineering talent, including its founder, who built AI chips at Google
GROQ INC./Reuters
- Nvidia licensed Groq's AI inference technology and hired key engineers.
- Groq will continue to operate independently after the non-exclusive licensing agreement.
- Silicon Valley sees more deals where top AI talent and technology are acqui-hired.
Nvidia is investing more in the AI boom with new hires from and a licensing deal with AI hardware startup Groq.
Groq confirmed in a blog post on Wednesday that it has "entered into a non-exclusive licensing agreement with Nvidia for Groq's inference technology" and said the company will continue to operate independently.
The startups is known for its Language Processing Unit, which is a custom chip designed for AI inference, namely, the process by which a trained AI model makes predictions or decisions. Groq was valued at about $6.9 billion as of three months ago and raised around $750 million in its latest funding round.
"As part of this agreement, Jonathan Ross, Groq's Founder, Sunny Madra, Groq's President, and other members of the Groq team will join Nvidia to help advance and scale the licensed technology," Groq added in the blog.
A person familiar with the matter told Business Insider on Wednesday that Nvidia is not acquiring the chip startup.
Neither Nvidia nor Groq mentioned the financial terms of the agreement.
Ross and Douglas Wightman were engineers at Google who started the project that became Google's first TPU chips, before leaving to found Groq. The TPUs are custom-made to accelerate large-scale machine-learning tasks designed to handle AI workloads, and are a major rival to Nvidia's GPUs.
The deal between the two companies comes as a new type of dealmaking is on the rise in Silicon Valley. Whereas traditional startups either aim to go public or be acquired, new acqui-hire deals could leave some startup employees behind, only benefiting a small percentage of staff members with desirable AI skills and the founders.
For instance, in 2024, Google agreed to pay $2.5 billion to license Character.AI's technology but only hired its two superstar cofounders and 20% of the startup's employees. In the same year, AI developers Adept and Inflection also made similar deals with Amazon and Microsoft, respectively.
More recently, Meta's acqui-hire of Scale AI became one of the biggest bets on talent after the company agreed to invest roughly $14 billion for a 49% stake and to bring its CEO, Alexandr Wang, into the fold to lead the Meta Superintelligence Labs.
These acqui-hires don't always end well. Windsurf employees were left in limbo after the AI coding startup was nearly aquired by OpenAI for $3 billion, only for the deal to fall apart and the company to be split. Google spent billions to hire Windsurf's CEO and top engineers, while the remaining hundreds of employees were acquired by another startup, Cognition.