Reliance, Brookfield and Digital Realty Commit $11 Billion to AI Data Centre

A consortium led by Reliance Industries said on November 26 it would invest $11 billion over five years to build roughly one gigawatt of AI native data centre capacity on a 400 acre campus in Visakhapatnam, Andhra Pradesh. The project, a joint venture with Brookfield and Digital Realty, is designed to serve demand for large scale compute needed by generative AI applications as global cloud providers expand regional hubs.

The announced scale and price tag underscore the capital intensity of next generation data centres. At the headline level the plan equates to about $11 million of investment for each megawatt of capacity, and roughly $27.5 million invested per acre on the 400 acre site. Those figures reflect costs associated with power delivery, advanced cooling systems, redundant infrastructure, and real estate in an emerging market context.

Reuters reporting has highlighted projections that India’s data centre capacity will expand rapidly in coming years, and the Visakhapatnam campus is part of a broader wave of private investment aimed at securing local AI infrastructure and jobs. For investors and cloud customers, onshore facilities shorten latency, address data residency requirements and make it easier to deploy large scale models regionally. For India, they offer a chance to capture higher value parts of the digital economy and to anchor engineering and operations jobs outside existing tech hubs.

The financial and strategic logic is clearest in the marketplace. Generative AI models are both compute intensive and latency sensitive, encouraging enterprises and hyperscalers to seek nearby, high density infrastructure. The Reliance led venture positions Indian capacity to compete for workloads that otherwise would be routed to U.S or European data centres. It also signals that institutional investors and large owners of digital real estate see sustained demand growth rather than a short lived spike.

There are, however, practical and policy challenges. Delivering one gigawatt of usable compute requires secure grid connections, often backed up by on site generation and substantial cooling capacity. Those systems raise questions about energy sourcing and water use in regions hosting large campuses. The project will test how quickly local utilities, regulators and state governments can coordinate land allocation, grid upgrades and approvals while meeting environmental and renewable energy commitments.

Longer term the investment is a marker of structural change in global cloud and AI infrastructure. Physical capacity is becoming a competitive battleground as much as software and models. Companies that control real estate and connectivity can influence where compute heavy industries locate. For India the stakes include not only immediate job creation and local procurement but also the opportunity to integrate into supply chains for chips, power equipment and data centre services.

The Visakhapatnam announcement adds a major private sector bet to a landscape where Google and other large cloud providers are already expanding regional AI capacity. How effectively India converts such investments into sustained jobs, exports and technological capability will shape the country’s role in the next phase of the global AI economy.