Jonathan Ross, the founder and chief executive of Groq, has left the company. Nvidia paid to hire him away in December alongside Sunny Madra and other staff, then secured a non-exclusive licensing deal for the chipmaker’s intellectual property.
On Monday, Groq confirmed it has raised $650 million to replace the departed leadership and restructure its strategy. The company did not reveal its current valuation. Its last figure was $6.9 billion, recorded in September after a $750 million funding round.
Ross joined forces with Doug Wightman ten years ago to build Groq. Both engineers came from Google, where Ross helped design the Tensor Processing Unit. Wightman remained after the Nvidia agreement and assumed the role of chief executive.
The firm previously sold language processing units, or LPU, for inference work. These chips ran either in the cloud or within on-premises hardware clusters.
Nvidia now controls the LPU IP. At its GTC event in March, the GPU giant launched its own inference system, the Nvidia Groq 3 LPX hardware cluster.
Groq has since shifted focus to its neocloud business. Madra oversaw this unit after Groq bought his data analytics firm, Definitive Intelligence, in 2024. The operation has expanded to 13 data centres across North America, Europe, the Middle East and APAC. It currently serves more than five million developers and thousands of AI companies, processing trillions of tokens weekly.
The company is also appointing new executives. Alan Rice has joined as chief operating officer. He previously worked at xAI and Meta, following a career in the US Navy. Sinclair Schuller and Rakesh Malhotra have been hired as chief technology officer and chief product officer respectively. They previously co-founded Nuvalence, a software engineering firm acquired by EY in 2024. Schuller started Apprenda, an enterprise cloud software company, before that. Malhotra spent roughly ten years developing Microsoft cloud products.
Groq’s future success depends on keeping its inference cloud competitive while Nvidia holds the key hardware technology. The inference sector faces high demand and significant venture capital investment. However, innovation and competition are rising in that area.
Other firms have survived similar situations. Jason Droege, the chief executive of Scale AI, told Forbes that business has recovered after Meta executed a $14.3 billion not-acqui-hire deal last year. Scale AI is on track to generate $1 billion in revenue.
In the high-stakes world of artificial intelligence, any outcome remains possible.
What it means
Investors in Groq have profited from the Nvidia deal. They now see a path to return by funding a pivot away from hardware sales. The company is betting on its cloud infrastructure and new management to compete without owning the underlying chip IP.
