India's Hyperscale Data Centre Megaprojects

Imagine a single company committing more capital to one category of infrastructure than most countries' entire annual government budgets, that's roughly the scale several Indian conglomerates have now committed to AI data centre infrastructure specifically, treating it as a genuinely new core business line rather than a peripheral technology investment.

Reliance Industries has announced a $110 billion commitment over seven years toward AI and data centre infrastructure, with over 120 megawatts of capacity expected to go live at its Jamnagar, Gujarat campus by late 2026, while Adani Enterprises partnered with Google on a roughly $15 billion AI data centre campus in Visakhapatnam, Andhra Pradesh, planned across 2026 to 2030, together the two conglomerates' commitments total roughly $210 billion toward building out India's AI data centre backbone.

Global specialists are moving at genuinely large scale too, NTT is investing over $2 billion to double its Indian capacity by 2026, and Yotta Data Services is building what it describes as one of Asia's largest AI computing hubs using Nvidia's latest Blackwell Ultra chips, in a project also costing more than $2 billion, showing this isn't purely an Indian-conglomerate story, global data centre and cloud specialists are racing for the same Indian AI compute opportunity.

The sheer scale of capital committed here, hundreds of billions of dollars across just a handful of announced projects, is itself the clearest signal that major Indian and global players believe AI compute demand in India is going to be genuinely durable, not a speculative bet, exactly the kind of capital commitment companies don't make lightly given how expensive and hard to reverse hyperscale data centre construction actually is.