Groq raises $350M Series A at $3.5B valuation to scale AI inference

Groq's $350 million Series A reflects accelerating demand for dedicated AI inference infrastructure. The company currently operates 13 data centers across four major geographic regions and serves a large customer base of six million developers, Fortune 500 enterprises, and AI-native companies. The funding round's timing and structure—with NVIDIA as a planned participant—underscores the strategic importance of the inference workload in the broader AI stack. Alex Davis, Executive Chairman and CEO of lead investor Disruptive, frames inference as the emerging priority layer of AI infrastructure, suggesting that as foundation models mature and deployment scales, the compute needed for inference (generating answers from trained models) is expected to exceed that required for training. Groq's stated plan to expand power capacity from 54 to 200+ megawatts in 2027 indicates confidence in rising customer demand and positions the company to serve customers seeking access to large clusters of NVIDIA accelerated computing.