- Nvidia (NVDA) has agreed to acquire Groq’s assets for $20 billion in cash, marking its largest deal ever and surpassing the $7 billion Mellanox acquisition in 2019, while officially framing it as a non-exclusive licensing agreement for Groq’s inference technology per CNBC reporting.
- Groq founder and CEO Jonathan Ross, president Sunny Madra, and other senior leaders will join Nvidia to advance the technology, with Groq continuing as an independent company under new CEO Simon Edwards and its GroqCloud service uninterrupted.
- The transaction strengthens Nvidia’s AI inference capabilities by integrating Groq’s low-latency processors, amid Groq’s recent $750 million funding round at a $6.9 billion valuation and targeted $500 million revenue this year.

Nvidia (NVDA) has reached a significant agreement involving Groq, a specialist in high-performance artificial intelligence accelerator chips focused on inference tasks. According to a CNBC report citing Alex Davis, CEO of Disruptive – which led Groq’s latest financing round – the transaction values Groq’s assets at $20 billion in cash, marking Nvidia’s largest deal to date and surpassing its 2019 purchase of Mellanox for close to $7 billion.
Groq, founded in 2016 by Jonathan Ross – a key contributor to Google’s tensor processing unit – and other former Google engineers, has developed technology aimed at accelerating inference for large language models. The company targeted $500 million in revenue this year, reflecting strong demand for efficient AI processing chips. Just three months prior, Groq raised $750 million in a round led by Disruptive, achieving a valuation of about $6.9 billion, with participation from investors including Blackrock (BLK), Neuberger Berman, Samsung, Cisco (CSCO), Altimeter, and 1789 Capital.
In its official statement, Groq described the arrangement as a ‘non-exclusive licensing agreement’ for its inference technology, with founder and CEO Jonathan Ross, president Sunny Madra, and other senior leaders joining Nvidia to advance the licensed capabilities. Groq will remain an independent company under new CEO Simon Edwards, the former finance chief, and its GroqCloud service will continue operations uninterrupted. Nvidia CEO Jensen Huang, in an internal email obtained by CNBC, emphasized that the company is licensing Groq’s intellectual property and integrating its low-latency processors into Nvidia’s AI factory architecture to support broader inference and real-time workloads, while explicitly noting that Nvidia is not acquiring Groq as a full company.
This structure aligns with recent industry patterns where major technology firms, including Meta (META), Google (GOOG), and Microsoft (MSFT), have used licensing deals combined with talent acquisition to secure AI expertise and technology amid regulatory scrutiny of traditional mergers. Nvidia previously employed a similar approach in September, spending over $900 million to license technology and hire key personnel from Enfabrica.
Nvidia’s financial position supports such moves, with $61 billion in cash and short-term investments as of late October, compared to $13.3 billion in early 2023. The company has also increased investments across the AI ecosystem, including in Crusoe, Cohere, CoreWeave (CRWV), a planned up to $100 billion commitment to OpenAI involving deployment of Nvidia products, and a $5 billion investment in Intel (INTC).
The agreement enhances Nvidia’s position in AI inference, a growing segment distinct from its dominance in model training via graphics processing units. Competitors in specialized inference chips include Cerebras Systems, which recently withdrew its IPO filing after raising over $1 billion but expressed intent to pursue a public offering soon. Groq’s approach, rooted in expertise from Google’s TPU development, provides Nvidia with additional tools to address diverse AI workloads as demand evolves.
WallStreetPit does not provide investment advice. All rights reserved.
- Bulenox: Get 45% to 91% OFF ... Use Discount Code: UNO
- Risk Our Money Not Yours | Get 50% to 90% OFF ... Use Discount Code: MMBVBKSM
Disclaimer: This page contains affiliate links. If you choose to make a purchase after clicking a link, we may receive a commission at no additional cost to you. Thank you for your support!
Leave a Reply