AMI Commits $20 Billion to 300 MW AI Data Center Capacity Push

Hyderabad-based AI infrastructure company AM Intelligence (AMI) has secured two additional binding orders for 20,000 NVIDIA Rubin GPUs, which will be installed as NVIDIA Vera Rubin NVL72 rack-scale systems across India and Malaysia.

Latest Orders Double Down on Rubin Platform

The orders, announced on October 5, 2026, from Hyderabad, represent the company's second and third firm commitments and are aimed at bringing large-scale compute capacity closer to the cloud providers, enterprises and AI developers it serves.

The new commitments arrive roughly six weeks after AMI placed its first order on August 25, 2026, covering 9,000 Rubin GPUs for its inaugural AI Factory in Hyderabad. The follow-on orders account for roughly 70 MW of capacity, will be delivered in Q2 2027, and will be split between India and Malaysia.

Committed Book Approaches 100 MW

With all three orders combined, AMI has now committed to approximately 29,000 Rubin GPUs and close to 100 MW of frontier compute capacity. The company states that this volume of firm orders places it among the earliest and largest committed Rubin GPU buyers outside North America. Anil Chalamalasetty, Chairman of AM Intelligence, said the two new facilities will follow the same architecture as the Hyderabad AI Factory.

Each will pair the NVIDIA Vera Rubin platform with high-throughput RDMA over Converged Ethernet (RoCE) networking and advanced storage systems. The sites are also designed for high rack power density with liquid cooling throughout, a configuration the company says allows the same physical footprint to house successive generations of AI silicon.

NVIDIA's managing director for Asia South, Vishal Dhupar, said AM Intelligence is bringing the next generation of AI factories to India and Malaysia, pairing accelerated computing with high-performance networking and liquid-cooled infrastructure to give developers and enterprises a foundation for more ambitious AI work.

Beyond the First 100 MW

The committed 100 MW is only the starting point of AMI's expansion plans. Over the next 15 months, the company intends to bring a further 300 MW of capacity to market, backed by capital expenditure exceeding USD 20 billion. This comes on top of the USD 6 billion already committed to the initial 100 MW tranche.

A Wider Compute-as-a-Service Pipeline

AMI has assembled a global pipeline of roughly 400 MW of Compute-as-a-Service capacity spanning India, Malaysia, Europe and the United States. The three orders placed so far constitute the first committed portion of that pipeline, while the remainder remains under development, targeted for delivery in later phases through 2027 and into early 2028.

The pipeline underpins AMI's stated goal of delivering 1 GW of Compute-as-a-Service capacity for global AI workloads, itself part of a broader programme covering 5 GW of powered AI data centres across India, the United States and Europe.

Electron-to-Token Economics

AMI positions itself as a strategic builder of long-term assets, arguing that its power solutions, capex advantage and access to next-generation AI infrastructure yield what it calls best-in-class electron-to-token economics.

The approach integrates energy infrastructure with the compute needed to train and run AI models, enabling AI training and inference at scale. The company plans to serve hyperscalers, neo cloud providers, sovereign AI initiatives, frontier labs, AI natives and local developer communities through this collaborative offering.

Group Background and Greenko Ties

AM Intelligence is constructing a full-stack AI infrastructure ecosystem that covers energy, data-centre infrastructure, hardware and customised AI models. With backing from Greenko's owned renewable generation and storage assets, the company is developing gigawatt-scale, liquid-cooled compute capacity intended for global hyperscalers, frontier AI labs, enterprises and India's sovereign AI initiatives, while also broadening chipset access for Indian developers working on domestic and global AI solutions.

AMI is promoted by Greenko's founders and operates platforms across energy, molecules and AI. Its energy business spans a 50 GW energy infrastructure platform being built across more than 20 states, covering solar, wind, hydro and energy storage, alongside what the group describes as the world's first interconnected energy storage platform, targeting 200 GWh of storage capacity in India.

On the molecules side, AMI is developing a low-carbon molecule platform, including green ammonia, with a target of 5 MTPA of capacity across multiple Indian locations. Its 1 MTPA green ammonia project now under construction is expected to rank among the world's largest RFNBO-compliant green ammonia facilities.