Monday’s announcement turns a new page for Groq: the $350M round backs a cloud operator, not a chipmaker. The company’s own silicon, the LPU line built to challenge Nvidia, is no longer the headline.
Disruptive led the round with Nvidia planning to participate, and the deal values Groq at $3.5B. That is down from the $6.9B it commanded last September, before Nvidia hired founder Jonathan Ross and other key staff under a $20B licensing arrangement. Groq’s spokesperson casts the new figure not as a discount but as a clean valuation for the company that exists after that deal.
The pivot began in June with a $650M raise, and Groq now aims to grow from 54 megawatts of capacity to more than 200 megawatts in 2027. It operates 13 data centers across North America, Europe, the Middle East, and Asia Pacific, serving more than 6 million developers, enterprises, and AI-native companies.
The fresh capital targets customers who want medium and larger clusters of Nvidia accelerated computing for both training and inference. Groq’s chairman, Alex Davis, calls inference the coming centerpiece of AI infrastructure, a prediction the market is still testing: neoclouds face heavy hardware bills and rapid depreciation, with profits far from guaranteed.
The same supplier relationship now defines much of the neocloud sector, where Nvidia sells the hardware and also buys equity in the companies running it. Groq’s financials remain private for now.
