AM Intelligence said it has placed two additional binding orders for about 20,000 NVIDIA Rubin GPUs, expanding planned AI factory capacity in India and Malaysia and lifting its total committed Rubin capacity to roughly 29,000 GPUs and close to 100MW.
The company, an AI infrastructure platform set up by the promoters of Greenko, said the new capacity is expected to be delivered in Q2 2027. It follows an earlier commitment announced on August 25, 2026, for 9,000 Rubin GPUs at its first AI factory in Hyderabad.
AM Intelligence said the follow-on orders represent about 70MW of additional capacity to be deployed across Malaysia and India. The company framed the expansion as an effort to provide cloud providers, enterprises and AI developers with access to large-scale computing closer to where AI systems are built and deployed.
“Both facilities will be built to the same architecture standard as the Hyderabad AI Factory, integrating the NVIDIA Vera Rubin platform with high-throughput RDMA over Converged Ethernet (RoCE) networking for fast, efficient data movement, alongside advanced storage systems,” said Anil Chalamalasetty, chairman of AM Intelligence. “Each site is engineered for high rack power density and liquid cooled throughout, allowing the infrastructure to accommodate successive generations of AI silicon within the same physical footprint.”
NVIDIA said the projects would combine accelerated computing, high-performance networking and liquid-cooled infrastructure. “AM Intelligence is bringing the next generation of AI factories to India and Malaysia with NVIDIA Vera Rubin,” said Vishal Dhupar, managing director, Asia South, NVIDIA. “By combining accelerated computing with high-performance networking and liquid-cooled infrastructure, AMI is creating the foundation for developers and enterprises to build more ambitious AI.”
The expansion comes as hyperscalers and new AI infrastructure entrants compete to secure next-generation GPUs amid tightening supply and rising demand for training and inference capacity, while also navigating power availability and data centre build timelines.
AM Intelligence said it has developed a pipeline of about 400MW of “Compute-as-a-Service” capacity across locations including India, the United States, Europe and Malaysia, with additional phases under development for delivery in 2027 and early 2028. It also said it intends to bring 1GW of compute-as-a-service capacity to market, within a broader programme of 5GW of powered AI data centres across India, the United States and Europe.
The company said it plans to bring an additional 300MW of capacity to market in the next 15 months, involving capital expenditure of more than $20+ billion in addition to the $6 billion it said was already committed for the initial 100MW.
