TCS’s HyperVault puts India’s AI-compute race on a 1 GW footing

The development

HyperVault, a subsidiary of Tata Consultancy Services, has secured 264 acres in Hyderabad to develop a large-scale artificial-intelligence data-centre campus with capacity of up to 1 GW. The company and its partners are expected to invest up to ₹70,000 crore to build and manage the infrastructure. The campus will be developed in phases according to customer demand and technology requirements and is designed to serve frontier AI companies, hyperscalers and global enterprises with high-density, liquid-cooled GPU infrastructure for AI training, inference and advanced computing workloads.

Why it matters

The strategic bottleneck in AI is increasingly compute access. Training and serving advanced models require large power blocks, high-density GPU systems, liquid cooling, high-capacity networks, reliable operations and access to suitable land and energy. A 1-GW campus would therefore be more than another commercial data centre: it would add a substantial physical node to India’s emerging AI infrastructure base. The project also has potential spillovers into power, cooling, networking, construction, engineering and operations, helping develop the supply chain around compute rather than treating AI as a software-only sector.

The Techadyant view

The HyperVault project is important because it brings together a large Indian technology company, a named infrastructure vehicle, secured land, a defined capacity target and a substantial investment envelope. It may give Hyderabad a stronger position in the competition to host frontier-AI workloads. Yet the announcement also exposes the policy choices ahead. If the campus primarily serves global hyperscalers, India gains jobs, services and infrastructure but may not automatically gain sovereign access to compute. The strategic value will be higher if Indian start-ups, public research, universities and national missions can obtain predictable access, while domestic firms participate in power, cooling, networking, operations and hardware supply.

What to watch

The first test is financial and physical execution: project approvals, power procurement, construction milestones, customer commitments and the commissioning of the first halls. Watch for the actual fraction of the 1-GW target contracted and installed, the ratio of green power to grid and backup supply, water-use performance, GPU availability and pricing, and access arrangements for Indian AI companies and research institutions. The June 2028 target is an important date, but phased delivery and usable compute—not the full-build headline—will determine the project’s strategic effect.