Groq secures $650M, rebuilds team after Nvidia’s $20B failed deal
▼ Summary
– Groq raised $650 million roughly six months after Nvidia signed a non-exclusive licensing deal for Groq’s IP and hired away its founder and CEO, president, and other employees.
– After Nvidia acquired the IP for Groq’s language processing unit (LPU) and announced its own hardware system, Groq pivoted to its neocloud business.
– Groq’s neocloud now operates 13 data centers globally, serves over five million developers, and processes trillions of tokens weekly.
– Groq hired new executives, including a COO from xAI and Meta, and a CTO and CPO who previously founded enterprise cloud companies.
– Groq’s success depends on how competitive its inference cloud remains after sharing key hardware IP with Nvidia, though the inference market sees high demand and competition.
What happens when a major AI chip company nearly sells itself, only to have a rival pay off investors, license its core technology, and poach its top talent? For Groq, the answer has been to raise fresh capital, rebuild its executive team, and pivot hard into the cloud business.
On Monday, Groq announced a $650 million funding round, confirming earlier speculation. This infusion comes roughly six months after Nvidia signed a non-exclusive licensing deal for Groq’s technology and hired away founder and former CEO Jonathan Ross, president Sunny Madra, and several other key employees. Groq did not disclose its new valuation, but it was last pegged at $6.9 billion following a $750 million round last September.
Ross, a former Google engineer who helped create the Tensor Processing Unit, co-founded Groq a decade ago with fellow Google engineer Doug Wightman. After the Nvidia deal, Wightman remained at the company and stepped into the CEO role.
Groq originally built a chip called a language processing unit (LPU) designed for inference, selling it either as a cloud service or as an on-premises hardware cluster. Now that Nvidia controls the LPU intellectual property, the GPU giant unveiled its own hardware system, the Nvidia Groq 3 LPX inference platform, at its GTC event in March.
In response, Groq has shifted focus to its neocloud business, a division previously led by Madra after Groq acquired his AI data analytics firm Definitive Intelligence in 2024. According to the company, that business now spans 13 data centers across North America, Europe, the Middle East, and Asia-Pacific, serving over five million developers and thousands of AI companies while processing trillions of tokens each week.
Groq has also been rebuilding its leadership team. It brought on Alan Rice as COO, a veteran from xAI and Meta with a prior career in the U. S. Navy. The company also added an entrepreneurial pair: Sinclair Schuller as CTO and Rakesh Malhotra as CPO. They previously worked together at Apprenda, an enterprise cloud software company Schuller founded, and later co-founded Nuvalence, a software engineering firm acquired by EY in 2024. Malhotra spent about a decade working on Microsoft’s cloud products.
Whether Groq can thrive after nearly being absorbed depends on how competitive its inference cloud remains now that its core hardware IP is shared with Nvidia. The company certainly has a shot. Inference technology is seeing explosive demand and heavy VC investment, but it is also attracting rapid innovation and fierce competition.
Other firms have survived similar shakeups. Scale AI CEO Jason Droege told Forbes that business rebounded after Meta executed a $14.3 billion non-acquihire about a year ago, and the company is now on track to hit $1 billion in revenue.
In the high-stakes world of AI, almost anything is possible.
(Source: TechCrunch)