Tata Consultancy Services is making one of the largest AI infrastructure commitments in India's history. HyperVault, TCS's data center subsidiary, will invest up to ₹70,000 crore — roughly $7.4 billion — to build a sprawling artificial intelligence data center campus in Hyderabad, Reuters and Bloomberg reported on September 5.
The planned facility will rise on 264 acres of secured land in the Telangana capital and is designed for a total capacity of up to 1 gigawatt, a scale that would place it among the largest AI-focused data center campuses anywhere in the world. For readers following the latest AI developments, the announcement marks India's clearest signal yet that it intends to host frontier-scale AI computing at home rather than rent it from overseas clouds.
Built for Frontier AI Workloads
According to details confirmed by TCS and Telangana state officials, the campus will cater specifically to frontier AI companies and hyperscalers, with infrastructure engineered for high-density GPU deployments used in AI training, inference, and other advanced computing workloads. Rather than committing to a single build-out, HyperVault and its partners plan to construct the campus in phases, scaling capacity in step with customer demand and evolving technology requirements.
That phased approach mirrors how AI data center capacity is being developed elsewhere: power-hungry GPU clusters are only economical when tenants are lined up, and a 1-gigawatt site could consume as much electricity as a mid-sized city at full utilization. TCS says the project is expected to create several thousand direct and indirect jobs across the region, spanning power, cooling, networking, construction engineering, and data center operations.
The company also stressed sustainability commitments, saying the campus will incorporate green energy and water-neutral design principles — a notable pledge in a state where water availability has long been a politically sensitive constraint on industrial expansion.
Telangana Officials: A 'Historic' Project
Telangana Chief Minister A. Revanth Reddy called the project "historic" for both the state and India, saying it "establishes that Hyderabad is now at the forefront of the global AI revolution with a strong focus on infrastructure."
Telangana's IT and Industries Minister D. Sridhar Babu added that the agreement was concluded at a rapid pace, attributing the speed to the state's focus on giving investors "clarity and speed of execution." The quick timeline matters: as global demand for AI-ready land, power, and cooling intensifies, regions that can permit and energize large campuses quickly are winning deals that once would have taken years to negotiate.
India Joins the Global AI Buildout Race
The HyperVault announcement lands amid an unprecedented global spending wave on AI data centers. In just the past week, Crusoe raised $3 billion at a $30 billion valuation to fund AI energy infrastructure, and August saw Anthropic sign a reported $45 billion compute deal with Nscale for Nvidia's next-generation Rubin capacity. The contrast is instructive: while US hyperscalers and AI labs lock up power in Texas, the Midwest, and the Gulf, India — the world's most populous country and one of its fastest-growing digital economies — has so far lacked homegrown capacity at frontier scale.
A 1-gigawatt campus in Hyderabad would not close that gap on its own, but it gives Indian enterprises, startups, and government AI programs a credible domestic option. It also positions TCS, India's largest IT services exporter, as an infrastructure operator rather than just a consumer of cloud capacity — a strategic pivot with implications for how India's IT sector competes globally.
Why a Gigawatt Matters
Data center capacity is measured in electricity for a reason. At modern densities, the power envelope determines how many GPU racks a campus can hold, how much cooling it needs, and ultimately how much AI training and inference work it can host. Most hyperscale campuses operating today draw tens to a few hundred megawatts; a single site designed for up to a gigawatt sits at the very top of the range, comparable to the largest AI campuses announced in the United States and the Gulf. Assuming the full build-out, Hyderabad would host one of the most powerful computing sites on the planet.
The scale also explains the structure of the deal. HyperVault's plan to invest "up to" ₹70,000 crore alongside partners, in phases tied to demand, keeps capital expenditure flexible while land, permits, and power infrastructure — the scarcest inputs — are locked in early. In today's AI market, securing grid capacity before competitors is frequently the difference between a shovel-ready campus and a slide deck.
What to Watch
Key questions remain about how the campus will be filled and powered. TCS has not publicly detailed which anchor tenants, if any, have committed, nor the mix of renewable and grid power that will feed the site. The phased construction plan means the headline 1-gigawatt figure is a ceiling, not a day-one reality.
Still, the direction of travel is unmistakable. With 264 acres secured, state government backing, and a build program aimed squarely at frontier AI tenants, HyperVault's Hyderabad campus is the strongest evidence yet that the next phase of the AI infrastructure boom will not be confined to North America and the Gulf. For India's AI ambitions, the ground — literally — has been broken.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
