Nvidia acquires chip startup Groq for approximately $20 billion in a licensing deal that eliminates a potential competitor while expanding the company's artificial intelligence infrastructure. The transaction, structured to avoid regulatory scrutiny, marks Nvidia's largest deal and follows a similar pattern established three months earlier with Enfabrica.
What Happened: Asset Acquisition
Nvidia is acquiring substantially all of Groq's assets except its cloud computing business, though the companies characterize the arrangement as a "non-exclusive licensing agreement."
Groq CEO Jonathan Ross, a former Google engineer who helped develop the search giant's Tensor Processing Unit, will join Nvidia along with president Sunny Madra and other senior executives.
The deal closed three months after Groq raised $750 million at a $6.9 billion valuation from investors including BlackRock, Samsung, Cisco and 1789 Capital, where Donald Trump Jr. serves as a partner. The startup will continue operating independently under CFO Simon Edwards as its new chief executive.
The transaction mirrors Nvidia's September deal with Enfabrica, where the company paid over $900 million to hire the startup's CEO and employees while licensing its technology.
Both deals use licensing structures rather than outright acquisitions, potentially avoiding antitrust scrutiny that blocked Nvidia's $40 billion bid for Arm Holdings in 2022.
Also Read: AI Threats Push Governments Toward Blockchain Infrastructure In 2026, Experts Warn
Why It Matters: Market Dominance
Groq's Language Processing Unit uses on-chip SRAM rather than external DRAM, enabling what the company claims is up to 10 times better energy efficiency. This architecture excels at real-time inference but limits model size.
The timing follows Google's recent unveiling of its seventh-generation TPU, codenamed Ironwood, and release of Gemini 3, trained entirely on TPUs, which topped benchmark rankings. Nvidia responded on X: "We're delighted by Google's success… NVIDIA is a generation ahead of the industry—it's the only platform that runs every AI model."
The deal reinforces narratives driving decentralized AI computing projects like io.net, which position themselves as alternatives to centralized infrastructure.
"People can put their own supply onto a network, whether that's data centers or yourself with your laptop, contributing your available GPU power, and getting fairly compensated for it using tokenomics," Jack Collier, io.net's chief growth officer, told BeInCrypto.
However, Nvidia's acquisition of Groq's low-latency technology extends its technical lead, making competitive performance harder for alternatives to achieve.
Read Next: Dogecoin After The Election Surge: What Went Wrong With The $1 Thesis