AI data centre manufacturing in India is entering a new phase as AM Intelligence (AMI), backed by the promoters of Greenko, placed two additional orders for 20,000 NVIDIA Rubin GPUs for deployment across India and Malaysia.
The new orders add about 70 MW of capacity and take AMI’s committed Rubin portfolio to around 29,000 GPUs and nearly 100 MW.
The company also plans another 300 MW of capacity over the next 15 months, requiring more than USD 20 billion in capital expenditure.
The larger question for Indian manufacturing is whether this AI-factory buildout can create demand for locally produced servers, liquid-cooling systems, power equipment and other high-density data-center hardware.
AMI’s announcement is significant because AI factories require considerably more than GPUs. Its planned NVIDIA Vera Rubin NVL72 systems will use high rack-power densities, RoCE networking, advanced storage, and liquid cooling.
The infrastructure therefore creates a wider equipment requirement around every GPU deployment.
India already has a substantial data-center base. Government data shows national data-center capacity increased from about 375 MW in 2020 to around 1,500 MW in 2025, while 38,231 GPUs had been onboarded through the government-backed AI compute framework.
For manufacturers, the opportunity is consequently moving beyond conventional data-center construction towards equipment capable of supporting much higher compute density.
Also Read: Karnataka Manufacturing Industry 2026: Policy & Key Sectors
Some pieces of this manufacturing ecosystem are already developing domestically. Schneider Electric has identified India as a manufacturing location for AI data-center infrastructure and said its local operations can serve around 90 percent of a data center’s product requirements.
Its Bengaluru facility also manufactures Motivair liquid-cooling solutions, an increasingly important technology as AI rack densities rise.
Netweb Technologies provides another example. Its collaboration with Vertiv covers rack-scale GPU platforms, liquid cooling, UPS systems, busways and power infrastructure, with the companies stating that the solutions are engineered and manufactured in India.
AMI’s expansion could strengthen demand across these categories as operators build larger AI facilities.
The immediate manufacturing opportunity may not be the advanced GPU itself but the infrastructure surrounding it.
High-density AI facilities need specialized cooling distribution, chillers, power conversion, UPS systems, busways, racks, networking equipment and high-performance computing platforms.
This could widen the addressable market for Indian electrical-equipment, electronics and data-centre manufacturers. It also creates an opportunity for domestic suppliers to move higher up the value chain from conventional infrastructure towards AI-specific systems.
AMI’s planned additional 300 MW makes the opportunity more significant. If the company executes this expansion alongside its existing pipeline, demand for locally manufactured equipment could grow with each new AI factory.
For India, therefore, AMI’s NVIDIA Rubin order is more than a compute-capacity announcement. It is an early test of whether the country can build a domestic manufacturing ecosystem around AI factories, even while the core accelerators remain globally sourced.
We use cookies to ensure you get the best experience on our website. Read more...
