AM Intelligence (AMI), the AI infrastructure arm of India’s AM Group, has placed its first major order for 9,000 Nvidia Vera Rubin NVL72 rack systems to power the initial phase of its AI factory in Hyderabad, marking a substantial push into cutting-edge AI compute capabilities.
- 9,000 Nvidia Vera Rubin GPUs ordered for Hyderabad AI factory
- First phase powers 30 MW of AI compute capacity by 2027
- Part of a $8B global AI infrastructure expansion plan
What happened
AM Intelligence, the AI infrastructure venture of the AM Group founded by Greenko promoters, has announced a significant purchase of 9,000 Nvidia Vera Rubin NVL72 rack-scale systems. Scheduled for delivery in Q1 2027, this hardware will support the first 30 megawatts phase of their new AI factory in Hyderabad, India. This project is a foundational step for AMI’s ambitious plan to roll out up to 1 gigawatt of AI compute capacity internationally.
The Hyderabad AI factory aims to become one of Asia’s foremost frontier AI compute clusters. The facility will incorporate Nvidia’s latest Vera Rubin GPU technology designed to efficiently handle trillion-parameter AI models and reduce AI inference costs significantly compared to prior hardware generations. This initiative is part of AMI's larger vision to deploy AI data centers across India, the US, Finland, and Malaysia.
Why it matters
This investment highlights India’s growing role in the global AI infrastructure landscape, bolstered by the AM Group’s strategic shift from renewable energy to high-performance AI computing. Leveraging its expertise in large-scale power infrastructure, AM Intelligence is transforming energy assets into AI compute power, underscoring the convergence of clean energy and advanced technology sectors.
The deployment of Nvidia’s Vera Rubin GPUs at scale could provide a significant competitive advantage by lowering operational AI costs and enabling cutting-edge AI model training and inference. This is a critical step for India to foster a robust AI ecosystem amid increasing international demand for compute resources, supporting both domestic innovation and global AI workloads.
What to watch next
Monitor the rollout progress of AM Intelligence’s Hyderabad AI factory and subsequent compute phases to evaluate the practical impacts of Vera Rubin hardware on AI workloads and cost efficiency. The development and scaling of AMI’s planned 200 megawatts of initial compute capacity could signal the pace of India’s AI infrastructure growth.
Additionally, AM Intelligence’s expansion plans across the US, Europe, and Asia will be important to watch as they seek to create a global compute-as-a-Service platform. Innovations in cooling and power integration technology, along with future Nvidia chip generations, will also shape the long-term competitiveness of this initiative.