AM Intelligence to deploy one of Asia’s 1st NVIDIA Vera Rubin GPU clusters at Hyderabad AI Factory

AM Intelligence orders 9,000 NVIDIA Rubin GPUs for Hyderabad AI factory

AM Intelligence, the AI infrastructure platform promoted by Greenko's founders, said on Tuesday it has placed a binding order for 9,000 NVIDIA Rubin GPUs for deployment at its first AI factory in Hyderabad, positioning the site among Asia's earliest Vera Rubin frontier compute clusters.

The GPUs will be installed as NVIDIA Vera Rubin NVL72 rack-scale systems and are scheduled for delivery in the first quarter of 2027, the Hyderabad-based company said. The facility will initially provide 30 megawatts of capacity and form the first tranche of AMI's planned 1 gigawatt Compute-as-a-Service buildout across India, the United States, Finland and Malaysia.

AMI said it plans to bring 200 MW of capacity to market in the near term, backed by capital expenditure exceeding $8 billion. The company is also developing 5 GW of powered AI data centres across India, the United States and Europe, designed for dense, liquid-cooled AI computing.

AMI chairman Anil Chalamalasetty said the order marks AMI's shift from "transforming electrons into value" toward converting power infrastructure into AI compute at scale. "For over two decades, we have focused on transforming electrons into value — addressing critical and emerging societal needs. Today, the electron-to-token opportunity allows us to take this capability further, converting power infrastructure into frontier AI compute at scale," he was quoted as saying. "Bringing the latest-generation NVIDIA Vera Rubin to India marks the next step in this journey and establishes AMI at the forefront of the emerging AI electron-to-token economy," he added.

The Hyderabad AI factory will combine NVIDIA's Vera Rubin architecture with RoCE networking and advanced storage, according to AMI. The systems will support some of the largest and most sophisticated AI workloads, including trillion-parameter models and agentic AI applications. NVIDIA's Vera Rubin NVL72 systems include the NVFP4 low-precision format and next-generation HBM4 memory, technologies intended to improve inference efficiency.

AMI said the facility has been engineered to deliver about 450 exaFLOPS of NVFP4 inference compute, which it expects will lower token-serving costs compared with earlier systems and expand access to advanced AI infrastructure for Indian developers.

The facilities being developed by AMI are designed to accommodate successive generations of AI silicon, the release said, positioning them for long-term use.

Perspectives on the announcement focal on the company's strategic shift toward AI infrastructure, with the chairman emphasizing the transition from power delivery to AI-driven value creation. The two reports carried the same core details, with one adding the precise announcement date and the other offering more technical specifics.