Nvidia licenses Groq AI chip technology and hires key executives

Nvidia licenses Groq AI chip technology and hires key executives Nvidia announced a non-exclusive licensing agreement with AI chip startup Groq on...

Dec 24, 2025
4 min read
Set Technobezz as preferred source in Google News
Technobezz
Nvidia licenses Groq AI chip technology and hires key executives

Don't Miss the Good Stuff

Get tech news that matters delivered weekly. Join 50,000+ readers.

Nvidia announced a non-exclusive licensing agreement with AI chip startup Groq on Wednesday, acquiring key technology and hiring several executives including founder Jonathan Ross. The deal contradicts earlier reports of a $20 billion acquisition, with Groq confirming it will continue operating independently under new CEO Simon Edwards.

Groq founder Jonathan Ross, who helped develop Google's tensor processing unit, will join Nvidia alongside President Sunny Madra and other engineering team members. The Mountain View-based startup specializes in inference chips that accelerate responses from trained AI models, a market where Nvidia faces increasing competition from AMD and other startups.

The licensing agreement follows Groq's $750 million funding round in September that valued the company at $6.9 billion. Investors included Disruptive, Samsung, Cisco, Altimeter, and 1789 Capital, where Donald Trump Jr. serves as partner. Disruptive CEO Alex Davis, whose firm invested over $500 million in Groq since its 2016 founding, confirmed the deal came together quickly.

CNBC earlier reported Nvidia agreed to acquire Groq for $20 billion in cash, which would have marked Nvidia's largest deal ever, surpassing its $7 billion Mellanox acquisition in 2019. Neither company commented on those acquisition reports, with Groq stating in its blog post that its cloud business will continue operating separately.

Groq targets $500 million in revenue this year amid surging demand for AI inference hardware. The company's language processing units claim to run large language models ten times faster while using one-tenth the energy of traditional chips. Groq powers AI applications for over two million developers, up from 356,000 last year.

The startup employs a unique architecture using on-chip SRAM memory instead of external high-bandwidth memory chips, avoiding industry-wide memory constraints. This approach speeds chatbot interactions but limits model size capabilities. Groq's primary competitor Cerebras Systems reportedly plans to go public next year.

Nvidia CEO Jensen Huang emphasized maintaining market leadership as AI shifts from training to inference during his 2025 keynote. The company held $60.6 billion in cash and short-term investments at October's end, up from $13.3 billion in early 2023. Nvidia recently announced intentions to invest up to $100 billion in OpenAI.

Groq will notify investors about the licensing agreement later Wednesday. The deal represents Nvidia's latest move to strengthen its inference capabilities while neutralizing a potential competitor in the rapidly evolving AI hardware market.

Share this article

Help others discover this content