Groq raises $350M to expand its Nvidia-powered neocloud

Groq has secured 350 million dollars in a new funding round led by investment firm Disruptive, with planned participation from Nvidia. The fresh capital values the startup at 3.5 billion dollars, down from its 6.9 billion dollar valuation last September. This adjustment follows a 20 billion dollar licensing deal in which Nvidia hired Groq founder and former CEO Jonathan Ross alongside other key talent. Consequently, Groq has shifted its strategy from designing its own Language Processing Units to operating as a neocloud provider that hosts Nvidia systems.

The new investment follows a 650 million dollar funding round in June, which initiated Groq's pivot. The company plans to use the capital to scale its power capacity from 54 megawatts to more than 200 megawatts by 2027. Currently, Groq operates 13 data centers across North America, Europe, the Middle East, and the Asia Pacific region. These facilities serve more than 6 million developers, enterprises, and AI-native companies.

For AI practitioners and developers, Groq's evolution provides another alternative for accessing high-demand Nvidia GPUs. The company is targeting users who need medium to large clusters of accelerated computing for both training and inference workloads. However, Groq enters a highly competitive and capital-intensive neocloud market alongside players like CoreWeave, Lambda, and Nebius. While demand for inference is soaring, these specialized clouds face significant financial hurdles, including rapid hardware depreciation and massive capital expenditures.