Nvidia to Acquire AI Chip Startup Groq for $20 Billion in its Largest Deal Ever
Nvidia agrees to buy AI accelerator chip startup Groq for $20 billion in cash, its largest acquisition ever. The move targets a key rival founded by the creators of Google's TPU, reshaping the AI semiconductor landscape.
The AI chip wars just escalated dramatically. Nvidia has agreed to acquire nine-year-old startup Groq for a massive $20 billion, absorbing a company founded by the very engineers who created Google's competing TPU. It’s Nvidia's largest purchase ever, a move that both strengthens its market dominance and removes a potent future rival from the board.
A Record-Breaking Deal
The $20 billion all-cash transaction was confirmed by Alex Davis, CEO of Disruptive, which led Groq's latest financing round. This figure shatters Nvidia's previous acquisition record, the nearly $7 billion purchase of Israeli chip designer Mellanox in 2019. With a war chest of $60.6 billion in cash and short-term investments as of late October, Nvidia is well-equipped to fund the deal. However, Davis noted that Groq's nascent cloud business is not part of the transaction.
Who is Groq, the TPU Challenger?
Founded in 2016, Groq designs high-performance AI accelerator chips meant to compete with Nvidia'sGPUs. Its key founder and CEO, Jonathan Ross, was one of the creators of Google's custom tensor processing unit (TPU). This pedigree has made Groq a closely watched player. Just three months ago in September, the company raised $750 million at a $6.9 billion valuation from investors including Blackrock and Samsung. It was targeting $500 million in revenue this year.
Authors
Related Articles
Samsung Electronics' union branch revealed that 84% of survey respondents oppose the government's roughly $290 billion (₩400 trillion) semiconductor megaproject in Gwangju. Government, management, the union and shareholders are all reading the same national project in sharply different ways.
Nvidia shipped roughly a billion RISC-V cores in 2024, then announced it would run CUDA on the open standard. We break down how royalty-free instruction sets and open software stacks are trying to route around CUDA's lock-in. Part 2 of the Semiconductor Sovereignty series.
US AI-chip export controls split into three layers in the first half of 2026 — January easing, a May crackdown on circumvention, and a pending bill. Nvidia erased China from its guidance and still posted a record $81.6 billion quarter. A look at the export policy that both shields and cages it.
AMD's MI325X matches or beats Nvidia on memory and bandwidth — yet Nvidia's 86-92% share holds. The real moat is CUDA, 20 years in the making. Part 1 of 4.
Samsung Electronics' union branch revealed that 84% of survey respondents oppose the government's roughly $290 billion (₩400 trillion) semiconductor megaproject in Gwangju. Government, management, the union and shareholders are all reading the same national project in sharply different ways.
Nvidia shipped roughly a billion RISC-V cores in 2024, then announced it would run CUDA on the open standard. We break down how royalty-free instruction sets and open software stacks are trying to route around CUDA's lock-in. Part 2 of the Semiconductor Sovereignty series.
US AI-chip export controls split into three layers in the first half of 2026 — January easing, a May crackdown on circumvention, and a pending bill. Nvidia erased China from its guidance and still posted a record $81.6 billion quarter. A look at the export policy that both shields and cages it.
AMD's MI325X matches or beats Nvidia on memory and bandwidth — yet Nvidia's 86-92% share holds. The real moat is CUDA, 20 years in the making. Part 1 of 4.
Thoughts
Share your thoughts on this article
Sign in to join the conversation