India's AI Powerhouse Buys 9,000 NVIDIA Rubin Systems, First Compute Capacity Secured by U.S. Firms

Indian artificial intelligence infrastructure company AM Intelligence has procured 9,000 Vera Rubin systems from NVIDIA (NVDA-US), aiming to become one of the first in Asia to deploy this advanced computing platform at scale.

The Hyderabad-based company announced on Tuesday (25th) that rack-level integrated server systems are expected to go online next year in southern India, serving major cloud service providers, AI labs, and institutions aiming to develop indigenous Indian AI models.

High-end NVIDIA computing clusters remain in tight supply. Massive demand for AI model training from data centers and companies like Microsoft (MSFT-US), Google under Alphabet (GOOGL-US), and Amazon (AMZN-US) has pushed order lead times to several years. Japan also plans to procure NVIDIA’s next-generation Rubin chips to build domestic AI models for robotics, with the first related data center scheduled to launch by June 2028.

AM Intelligence belongs to the group that owns renewable energy firm Greenko Group Plc. The company has signed a binding purchase agreement under an $8 billion initiative, part of its broader plan to build 1 gigawatt of computing capacity. The Hyderabad facility aims to run trillion-parameter AI models and next-generation agent-based AI applications.

Mahesh Kolli, founder and president of Greenko Group, said in a video interview that the group will offer computing power at competitive prices to clients in India, the U.S., Finland, and Malaysia. He noted that the first batch of computing capacity has already been acquired by a U.S. client, though he declined to name the buyer due to a non-disclosure agreement (NDA).

He emphasized that clients are “urgently seeking” compute capacity and added that India faces minimal operational constraints, as U.S. enterprises can access AI compute with only about 300 milliseconds of latency via undersea cables.

Kolli highlighted AM Intelligence’s advantage in accessing low-cost renewable energy. “In global token economics, energy pricing plays a critical role,” he said, adding, “We are one of the lowest-cost AI computing infrastructure providers globally.”

The group plans to fund its proposed expansion through debt and equity financing. Kolli stated, “The time to convert capital expenditure into cash flow for this business is very short. We can partially self-fund through our balance sheet and have prior experience raising capital in bond markets.”

Anil Chalamalasetty, chairman of AM Group, said, “The shift from electrons to tokens enables us to transform large-scale power infrastructure into cutting-edge AI compute capacity.”