Indian AI Infrastructure Company AM Intelligence Orders 9,000 NVIDIA Vera Rubin Systems
India's AM Intelligence orders 9,000 Vera Rubin
Indian AI infrastructure company AM Intelligence has ordered 9,000 NVIDIA Vera Rubin systems, preparing to become one of the first companies in Asia to adopt this advanced computing platform.
The order consists of Vera Rubin NVL72 rack-level systems, scheduled for delivery in the first quarter of 2027, to be deployed at a 30 MW AI factory in Hyderabad.
This factory will be one of the first in Asia to use the Vera Rubin cutting-edge AI computing cluster, supporting trillion-parameter models and agent-based AI applications.
This is the first batch of AM Intelligence's planned 1 GW computing-as-a-service capacity, with a near-term target of 200 MW and capital expenditures exceeding $8 billion, covering India, the U.S., Finland, and Malaysia.
The company was founded by the founder of Greenko, emphasizing the transition from energy to AI computing.
From a market mechanism perspective, large-scale pre-orders lock in next-generation GPU supply, with funds flowing from energy infrastructure to AI computing clusters. Event-driven orders are increasing attention on AI computing layout in Asia, benefiting NVIDIA and local Indian AI developers, while putting pressure on other buyers facing tight computing supply.
Source: Public Information
ABAB AI Insight
As an AI infrastructure platform associated with Greenko, AM Intelligence extends its renewable energy background to large-scale AI computing. The order of 9,000 Vera Rubin systems continues its path of transitioning from power assets to computing as a service, similar to other energy companies entering the AI data center space.
In terms of capital pathways, by binding with NVIDIA's latest architecture, it aims to secure an early mover position in Asia, motivated by reducing inference costs and serving ultra-large-scale clients and local models, with resources flowing from power infrastructure to cutting-edge GPU clusters, forming a closed loop from energy to token economy.
Comparing with other early Vera Rubin deployment plans globally and India's sovereign AI initiatives, the current order is in the phase of locking supply from planning to actual execution, with the industry position shifting from traditional data center operators to builders of cutting-edge AI factories.
The structural judgment indicates a technological substitution, where the mechanism lies in the next-generation low-precision and HBM4 architecture significantly reducing token costs, allowing early deployers to gain inference efficiency advantages, leading to a shift in AI computing competition from quantity to architectural generations.
ABAB News · Cognitive Laws
Early locking of next-generation architecture determines computing advantages
Energy companies transforming into AI factories is a natural extension
Large-scale orders signal early mover advantage under scarce supply