What does an AI firm do after a type of not-acqui-hire offers, the place a rival pays buyers a hefty IP “licensing” charge whereas poaching its essential expertise? For AI chipmaker Groq, the reply seems to be increase more cash from buyers — who have been mentioned to have profited handsomely after a take care of Nvidia in December — rent extra expertise, and pivot.
On Monday, Groq announced a brand new $650 million funding spherical, confirming earlier reports. The increase comes roughly six months after Nvidia signed a non-exclusive licensing settlement for Groq’s know-how and employed away founder and CEO Jonathan Ross, president Sunny Madra, and different staff. Groq didn’t disclose its new valuation. It was final valued at $6.9 billion following a $750 million round in September.
Ross, who got here from Google, was identified within the AI chip world for serving to create Google’s AI chip, the Tensor Processing Unit. He teamed up with one other Google engineer, Doug Wightman, to launch Groq a decade in the past. Wightman stayed on after the Nvidia deal and have become CEO.
Groq created a chip it referred to as a language processing unit (LPU), used for inference, and offered it as a part of a cloud service or an on-premises {hardware} cluster.
With Nvidia now proudly owning the IP for LPUs, the GPU large introduced its personal {hardware} cluster, the Nvidia Groq 3 LPX inference {hardware} system, at its GTC occasion in March.
In response, Groq has pivoted to its neocloud enterprise, it mentioned. That enterprise had been run by Madra after Groq acquired his AI information analytics firm Definitive Intelligence, in 2024. It has grown to 13 information facilities throughout North America, Europe, the Center East, and APAC and is serving over 5 million builders and hundreds of AI corporations, processing trillions of tokens every week, the corporate says.
Groq has additionally been hiring substitute execs. It added Alan Rice as COO, beforehand at xAI and Meta, after a profession within the U.S. Navy.
It additionally added an entrepreneurial duo, Sinclair Schuller, who joins as CTO, and Rakesh Malhotra as CPO. They beforehand labored collectively at Apprenda, an enterprise cloud software program firm based by Schuller; they then co-founded Nuvalence, a software-engineering agency acquired by EY in 2024. Malhotra beforehand spent a few decade engaged on Microsoft’s cloud merchandise.
Whether or not Groq can succeed after nearly promoting itself is dependent upon how aggressive its inference cloud can stay, now that the important thing {hardware} IP is shared with Nvidia. Actually, it has a shot. Inference-related tech is an space experiencing great demand (and VC investment). Nevertheless it’s additionally seeing growing innovation and competitors.
Nonetheless, others appear to have survived these kinds of offers. Scale AI’s CEO Jason Droege informed Forbes that enterprise has rebounded after Meta did a $14.3 billion not-acqui-hire a few yr in the past, and that the corporate is on monitor to do $1 billion in income.
Within the big-money sport of AI, something appears attainable.
Whenever you buy by way of hyperlinks in our articles, we may earn a small commission. This doesn’t have an effect on our editorial independence.
