India Places a Massive AI Hardware Order-Why Nvidia's Vera Rubin Is Key

HYDERABAD — Indian artificial intelligence (AI) infrastructure firm AM Intelligence has placed a binding order for 9,000 advanced Nvidia Vera Rubin graphics processors, positioning itself among the earliest adopters of the next-generation architecture in Asia.

Announced on Tuesday, August 25, 2026, the hardware acquisition will power the initial 30-megawatt phase of an advanced AI factory in Hyderabad, slated to come online in early 2027. Backed by the founders of renewable energy giant Greenko Group, the multi-billion-dollar deployment forms part of a broader global strategy to tackle surging worldwide demand for high-end AI compute capacity.

Scaling Frontier AI Infrastructure in Hyderabad

The upcoming Hyderabad facility will utilize Nvidia’s Vera Rubin NVL72 rack-scale systems, high-throughput networking, and advanced storage infrastructure to manage trillion-parameter models and agentic AI workloads. According to company disclosures, the cluster is engineered to deliver approximately 450 exaflops of low-precision NVFP4 inference compute.

The deployment is part of an expansive $8 billion capital expenditure plan by AM Intelligence to roll out a 1-gigawatt Compute-as-a-Service platform spanning India, the United States, Finland, and Malaysia. By leveraging Greenko’s low-cost renewable generation and energy-storage assets, the firm aims to address soaring power demands and reduce token-generation costs for global cloud providers, sovereign AI initiatives, and domestic model developers.

According to official company statements, corporate disclosures, and industry announcements:

• Order Specification: Binding procurement of 9,000 Nvidia Vera Rubin GPUs deployed via Vera Rubin NVL72 rack-scale systems.

• Deployment Location: Hyderabad, India, marking one of Asia's earliest frontier Rubin clusters.

• Operational Timeline: Initial 30-MW facility scheduled to launch in the first quarter of 2027.

• Financial Scope: Initial phase anchors a broader $8 billion investment targeting 200 MW of near-term capacity within a 1-gigawatt framework.

Official Sources Section

• Nvidia Vera Rubin Platform Official Portal: Official technical specifications, rack-scale architecture details, and agentic AI deployment frameworks.

• AM Intelligence Corporate Disclosures: Official project announcements, capital expenditure outlines, and strategic renewable integration briefs.

• Bloomberg Technology Reporting Archives: Primary market interviews and institutional tracking of international hardware allocations.

Quote Section

"For over two decades, we have focused on transforming electrons into value. Today, the electron-to-token opportunity allows us to take this capability further, converting power infrastructure into frontier AI compute at scale,"

— stated Anil Chalamalasetty, Chairman of AM Intelligence.

Why It Matters

For global technology enterprises and domestic startups, access to cutting-edge computing clusters remains severely constrained by multi-year industry backlogs. By combining next-generation Nvidia silicon with captive, low-cost renewable energy, the initiative establishes India as a competitive hub for heavy AI model training and agentic reasoning workloads. Furthermore, international connectivity via undersea cables enables foreign tech companies to utilize Indian data center capacity with minimal latency.

Key Facts at a Glance

• Company: AM Intelligence (AM Group / Greenko-backed).

• Hardware: 9,000 Nvidia Vera Rubin GPUs (NVL72 rack systems).

• Target Performance: ~450 exaflops of NVFP4 inference compute.

• Global Reach: Capacity pre-contracted by international buyers alongside domestic sovereign AI initiatives.

FAQ Section

What hardware did AM Intelligence order for its India data center?

The company placed a binding order for 9,000 Nvidia Vera Rubin graphics processors integrated into Vera Rubin NVL72 systems.

When will the Hyderabad AI facility become operational?

The initial 30-megawatt cluster is scheduled to come online in the first quarter of 2027.

Who are the target customers for this computing capacity?

The facility targets hyperscalers, major cloud service providers, AI research labs, neo-cloud providers, and sovereign AI initiatives in India and abroad.

How is the data center powered sustainably?

The project draws on Greenko Group’s extensive renewable energy and energy-storage assets to supply low-cost, high-density power required for liquid-cooled AI hardware.

Source: Nvidia, AM Intelligence / Bloomberg, Financial Express