en.Wedoany.com Reported - On June 22, local time, U.S. AI inference chip company Groq announced the completion of a new $650 million growth funding round. The round was led by Disruptive and Infinitum, with participation from some existing investors. The funds will be used to accelerate the expansion of Groq's AI inference cloud infrastructure.
Groq's current business focus has shifted to large-scale AI inference cloud services. The company currently operates 13 data centers across North America, Europe, the Middle East, and the Asia-Pacific region, serving over 5 million developers and thousands of AI-native enterprises, processing trillions of tokens weekly. The new funds will be used to upgrade existing data center infrastructure and deploy Groq's latest inference technologies, including the NVIDIA LPX system.
According to the company's plan, Groq aims to expand its total AI inference cloud installed capacity to 200 megawatts by the end of 2027. This target addresses the rapid growth in inference-side computing power demand. As AI applications transition from model training and experimental validation to production deployment, enterprises are seeing rising demand for inference computing power with low latency, high concurrency, and controllable costs. The inference cloud is evolving from a supporting service into a key component of AI infrastructure.
Groq's core technology foundation is the LPU inference processor architecture, primarily optimized for sequential computing tasks such as large language models. Unlike the training phase, inference services emphasize continuous operation, response speed, unit cost, service stability, and scalable scheduling capabilities. Groq's continued expansion of its cloud platform after this funding round indicates that its commercial focus is shifting from pure chip capability demonstration to sustainably deliverable inference cloud services.
The management team has also been adjusted accordingly. Alan Rice has joined Groq as Chief Operating Officer, having previously held positions at xAI and Meta in data center-related roles, with experience in large-scale infrastructure operations. Sinclair Schuller and Rakesh Malhotra will assume the roles of Chief Technology Officer and Chief Product Officer respectively starting in July, responsible for driving platform technology and enterprise-grade product development.
This funding round follows Groq's non-exclusive technology licensing agreement with NVIDIA. Groq stated that NVIDIA's next-generation LPX platform has integrated Groq's inference technology. For Groq, the parallel advancement of technology licensing and cloud business expansion means it no longer relies solely on its own chip sales, but supports business growth through inference cloud platforms, technology licensing, and data center operational capabilities.
The AI computing power market is shifting from "who can train larger models" to "who can run models stably at lower costs." Training determines the upper limit of model capabilities, while inference determines whether applications can be scaled for use. Groq's allocation of funding to global data centers and 200-megawatt inference cloud expansion reflects that the AI infrastructure competition is entering a phase of sustained operations.
Groq's immediate challenges are also clear: the 200-megawatt target requires synchronized matching of power, data center space, liquid cooling, networking, chip supply, and customer workloads. Whether the inference cloud can generate long-term revenue depends not only on computing scale but also on price competitiveness, model ecosystem, enterprise customer stickiness, and service stability. For AI application companies, what truly matters is not peak computing power, but inference capabilities that can be reliably, stably, and cost-effectively deployed over the long term.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
