AI infrastructure provider Groq has raised $350 million in a Series A funding round at a $3.5 billion valuation, led by investment firm Disruptive with expected participation from Nvidia subject to customary closing conditions. The financing accelerates the company's structural pivot from developing custom inference silicon toward operating an enterprise-grade inference cloud powered by Nvidia accelerated computing systems.
The round follows a $650 million capital raise completed in June 2026 and builds on a strategic shift that began late last year. In December, Nvidia signed a non-exclusive technology licensing agreement with Groq to access its inference architecture while allowing Groq to maintain independent operations and expand its cloud platform, GroqCloud.
Expanding Compute Footprint and Power Capacity
Groq originally focused on designing proprietary Language Processing Units (LPUs) to execute LLM inference workloads with deterministic low latency. Following changes to its core engineering group, the firm repositioned itself as an inference cloud operator ("neocloud"), deploying clusters of Nvidia GPUs to meet growing enterprise demand for reliable model serving.
The company currently manages 13 data center sites across North America, Europe, the Middle East, and the Asia-Pacific region, representing 54 megawatts of operational power capacity. According to company disclosures, the fresh capital will fund infrastructure expansion targeting more than 200 megawatts of active compute capacity during 2027.
Target Workloads and Developer Reach
Groq reports an active developer base exceeding 6 million users across enterprise software vendors, research teams, and AI-native startups. Management stated that the capital will support customers requesting medium- and large-scale Nvidia compute clusters for high-throughput inference deployments and hybrid training workflows.
The strategic transition reflects broader market dynamics in AI infrastructure, where dedicated inference providers are increasingly competing on capacity orchestration, power allocation, and multi-region deployment rather than silicon design alone.
