Groq

Mountain View-based AI inference company that designed the LPU chip and runs the GroqCloud service. After licensing its technology to Nvidia in December 2025, it raised $650 million and a $350 million Series A in 2026.

Industry: AI inference chips and cloud | Location: Mountain View, United States | Founded: 2016 | Total Funding: USD 2.8B

About Groq

Groq is a US AI infrastructure company founded in 2016 by former Google Tensor Processing Unit engineers, including Jonathan Ross, together with Douglas Wightman, a former Google X engineer who served as its first chief executive. It designed the LPU, or language processing unit, a chip built specifically for AI inference, the processing that happens after a user prompts a model, as opposed to the heavier work of training. Most profiles place its headquarters in Mountain View, California. In 2017 Social Capital's Chamath Palihapitiya invested USD 10 million as seed funding, and in April 2021 Groq raised a USD 300 million Series C led by Tiger Global Management and D1 Capital Partners, reaching a valuation above USD 1 billion. A USD 750 million round in September 2025 was reported at a post-money valuation of USD 6.9 billion. Groq launched GroqCloud in February 2024, a platform that lets developers and enterprises run inference workloads. The company says it operates 13 data centres across North America, Europe, the Middle East and Asia-Pacific and serves more than five million developers and thousands of AI-native companies. In December 2025 Nvidia struck what both sides described as a non-exclusive licensing agreement for Groq's inference technology rather than an acquisition. Press reports, citing investor Alex Davis, put the value at about USD 20 billion, though terms were not disclosed. Ross, president Sunny Madra and part of the engineering team joined Nvidia. Groq remained independent, with Simon Edwards taking over as chief executive, and GroqCloud was reported to be excluded from the agreement and to continue operating. In 2026 Groq announced USD 650 million in new growth capital, led by existing investors Disruptive and Infinitum, after Reuters first reported the raise in May. The money is for upgrading its data centres with newer hardware, including Nvidia's LPX system, and for reaching 200 megawatts of compute capacity by the end of 2027, a shift that outlets describe as a move from chip maker to AI inference cloud operator. Groq has a Gulf footprint. It deployed a cluster in Saudi Arabia in December 2024, and in May 2025 it was named an official inference provider for HUMAIN, the Saudi AI company led by Tareq Amin. Its Dammam data centre, part of a USD 1.5 billion commitment announced at LEAP 2025, has served traffic since February 2025 and supported day-zero access to OpenAI's open-weight gpt-oss models on GroqCloud.

Leadership

Related News & Coverage