AM Intelligence Orders 9,000 NVIDIA Rubin GPUs for Hyderabad AI Factory

AM Intelligence (AMI), an AI infrastructure company promoted by the founders of Greenko Group, has placed a binding order for about 9,000 NVIDIA Rubin GPUs for its first AI factory in Hyderabad, with deliveries scheduled for the first quarter of 2027.

The GPUs will be deployed as NVIDIA Vera Rubin NVL72 rack-scale systems across 30 MW of capacity at the Hyderabad facility. AMI said the deployment will make the facility one of the first frontier AI compute clusters in Asia based on the Vera Rubin platform.

The Hyderabad AI factory is the first tranche of AMI’s planned 1 GW of Compute-as-a-Service (CaaS) capacity across India, the US, Finland and Malaysia. The company plans to bring 200 MW of capacity to market in the near term, with capital expenditure of more than $8 billion.

“For over two decades, we have focused on transforming electrons into value - addressing critical and emerging societal needs,” Anil Chalamalasetty, Chairman of AM Intelligence, said. “Today, the electron-to-token opportunity allows us to take this capability further, converting power infrastructure into frontier AI compute at scale.”

The Hyderabad facility is being designed with high-throughput RDMA over Converged Ethernet (RoCE) networking and advanced storage systems. AMI said the infrastructure will support AI workloads including trillion-parameter models and agentic AI applications.

NVIDIA’s Vera Rubin NVL72 platform uses NVFP4, a low-precision computing format for AI workloads, along with HBM4 memory. NVIDIA says the architecture can reduce AI inference token-serving costs by up to 10 times compared with its Grace Blackwell generation.

AMI said its Hyderabad AI factory is being engineered to deliver about 450 exaFLOPS of NVFP4 inference compute. It is developing 5 GW of powered AI data centres across India, the US and Europe. The facilities are being designed for high-density AI computing and will use liquid cooling to handle the power and heat requirements of AI chips.

The company also plans to provide 1 GW of CaaS capacity for global AI workloads, with an initial 200 MW expected to come online in the near term.

AMI said its infrastructure will serve hyperscalers, frontier AI labs, enterprises, sovereign AI initiatives, AI-native companies and developers.

The company is building its AI infrastructure alongside its existing energy platform. AMI said it is developing a 50 GW energy infrastructure platform across more than 20 states in India, covering solar, wind, hydro and energy storage.