What actions does an AI firm take following one of those non-acquisition deals, where a competitor compensates investors with a substantial IP “licensing” fee while enticing its essential personnel? For AI chip manufacturer Groq, it seems to involve attracting additional funding from investors — who reportedly reaped significant profits after a deal with Nvidia in December — recruiting more staff, and shifting focus.

On Monday, Groq revealed a new funding round of $650 million, affirming prior news reports. The round was spearheaded by Disruptive, a late-stage investment firm from Dallas founded by Alex Davis — who also acts as Groq’s chairman — and Infinitum, a hedge fund based in Fort Lauderdale.

This funding follows approximately six months after Nvidia entered into a non-exclusive licensing deal for Groq’s technology and recruited founder and CEO Jonathan Ross, president Sunny Madra, along with other staff members. Groq did not reveal its new valuation, with the last estimated at $6.9 billion after a $750 million round in September.

Ross, previously with Google, gained recognition in the AI chip domain for his role in the development of Google’s AI chip, the Tensor Processing Unit. He partnered with fellow Google engineer Doug Wightman to establish Groq a decade ago. Wightman remained with the company after the Nvidia agreement and has taken on the role of CEO.

Groq produced a chip referred to as a language processing unit (LPU) for inference, marketed as either a cloud service or part of an on-premises hardware cluster.

With Nvidia now holding the IP for LPUs, the GPU leader introduced its own hardware cluster, the Nvidia Groq 3 LPX inference hardware system, during its GTC event in March.

In light of this, Groq has transitioned to focus on its neocloud business, it stated. This business was overseen by Madra following Groq’s acquisition of his AI data analytics firm Definitive Intelligence, in 2024. It has expanded to 13 data centers across North America, Europe, the Middle East, and APAC, catering to over five million developers and thousands of AI firms, processing trillions of tokens weekly, as claimed by the company.

Groq has also been bringing in replacement executives. It appointed Alan Rice as COO, who previously worked at xAI and Meta, after a tenure in the U.S. Navy.

Additionally, it welcomed an entrepreneurial pair, Sinclair Schuller as CTO and Rakesh Malhotra as CPO. They had previously collaborated at Apprenda, an enterprise cloud software company founded by Schuller; they later co-founded Nuvalence, a software engineering firm acquired by EY in 2024. Malhotra had spent around a decade involved with Microsoft’s cloud products.

The success of Groq after nearly divesting itself will hinge on the competitive edge of its inference cloud, especially now that key hardware IP is shared with Nvidia. It certainly has potential. Technology related to inference is seeing immense demand (and VC funding). Nonetheless, it is also facing growing innovation and competition.

Nevertheless, other companies seem to have thrived after similar deals. Scale AI’s CEO Jason Droege shared with Forbes that the business has recovered following Meta’s $14.3 billion non-acquisition, about a year ago, and that the company is on course for $1 billion in revenue.

In the high-stakes arena of AI, anything seems achievable.