Groq Raises $350 Million to Accelerate Neocloud Pivot

Groq built its business around its proprietary Language Processing Unit, or LPU, designed to run artificial-intelligence inference workloads with lower latency. The startup is now broadening its focus from developing chips to operating a “neocloud” platform that supplies GPUs and computing capacity, positioning it against hyperscale cloud providers and a growing field of specialist AI infrastructure companies as demand for model inference accelerates.

Groq raised $350 million in a funding round led by Disruptive, with Nvidia participating, valuing the company at $3.5 billion. The capital will support its expansion into GPU and AI compute services. Groq currently operates 13 data centers worldwide and plans to increase its power capacity to more than 200 megawatts by 2027, a key measure of how much computing infrastructure it can deploy.