India’s Race to Build AI Ready Data Powerhouses

India is attracting massive global investment in AI‑ready data centres as it produces 20% of the world’s data but has limited capacity. Tech giants like Google, OpenAI, Reliance, AWS and TCS are building large GPU‑driven facilities requiring huge power and advanced cooling. States are competing for these projects, which boost high‑skill jobs and support India’s growing data‑localisation needs.

India is rapidly becoming a focal point for global tech firms seeking to expand their artificial intelligence infrastructure. As AI adoption surges worldwide, the country’s fast-growing data output and underdeveloped data‑centre ecosystem are creating a compelling opportunity for massive investment.

Although India generates close to one‑fifth of the world’s data, it hosts only a sliver of total global data‑centre capacity. This imbalance is pushing technology giants to bet big on large‑scale AI‑optimised facilities across the country.

Over the past year, several major announcements have underscored this momentum. Google unveiled plans for a $15‑billion AI data‑centre campus in Visakhapatnam, expected to employ more than 100,000 people during the build‑out phase. OpenAI has reportedly held discussions on establishing a 1‑GW facility in India, while Reliance Industries and its partners are planning an $11‑billion AI data‑capacity project in Andhra Pradesh. Jio is also expanding its own infrastructure in Jamnagar. Other commitments include AWS’s $8.3‑billion cloud investment in Maharashtra and TCS’s roadmap to spend roughly $6.5 billion over the next several years to build 1 GW of data‑centre capacity.

Established Indian operators—such as Yotta, CtrlS and Sify Technologies—are also scaling up aggressively, particularly in major tech hubs including Mumbai, Hyderabad and Bengaluru.

Industry experts say this wave of investment is being driven primarily by the soaring need for GPU‑powered compute. Vishnu Subramanian of E2E Networks explained that GPU demand for AI workloads is rising so quickly that the industry is nowhere near a plateau.

What Makes an AI Data Centre Different?

AI‑focused data centres differ substantially from traditional facilities. They require high‑performance chips like GPUs and TPUs, ultra‑fast networking and enormous storage resources to train and run AI models. They also consume significantly more power and generate far more heat, necessitating advanced cooling and specialised engineering talent.

According to Deloitte India’s S. Anjani Kumar, even the supporting areas of an AI facility—such as substations, heavy‑duty power equipment and sophisticated cooling—require substantial capital outlays. The main computing area must then support dense clusters of GPUs, advanced networking hardware and orchestration software.

Such facilities often need multiple gigawatts of power when fully operational. Gartner’s Naresh Singh noted that the engineering involved spans everything from internal compute clusters to external connectivity requirements. These projects also create opportunities for numerous service providers, including local players, across engineering, operations and ancillary technology services.

Government support is another essential factor, with states providing land, incentives and dedicated data‑centre zones to attract investment. But one challenge overshadows the rest: power availability. AI‑first data centres require huge, stable energy supplies, driving additional demand for transformers, transmission lines and backup systems. Developers often integrate microgrids, generators and batteries to ensure reliability.

Why Older Data Centres Can’t Just Be Upgraded

Despite India already hosting large cloud‑computing estates from global hyperscalers, converting traditional centres into AI‑ready sites is far from simple. Yotta Data Services co‑founder Sunil Gupta explained that conventional racks usually draw around 6–8 kW of power. With GPUs, that jumps to at least 50 kW—and can hit 200 kW.

This dramatic increase breaks existing cooling and power-delivery designs. AI‑grade facilities often require chip‑level liquid cooling rather than standard liquid or air‑based systems. Yotta has managed some retrofitting—for example, its Navi Mumbai NM1 centre now operates at nearly double its original 30‑MW capacity, with half its floors fully populated with GPUs—but Gupta acknowledged that such upgrades are possible only when a building’s footprint and infrastructure allow for it.

The cost difference is also stark. Traditional CPU‑based facilities typically require $10–12 million per MW. For AI‑driven sites, just the GPUs can cost $35–50 million per MW.

Gupta noted that Yotta’s NM1 facility currently hosts India’s largest GPU cluster, with E2E Networks following with about 1,000 GPUs at L&T’s Chennai centre. While major cloud providers have likely deployed some GPUs, overall AI compute capacity in India remains in an early stage.

Why States Are Competing for AI Infrastructure

Deloitte estimates that India may need an additional 45–50 million square feet of data‑centre space and 40–45 TWh of power by 2030 to support projected growth. Total capacity could reach 6–10 GW by the end of the decade.

These multi‑billion‑dollar projects bring significant foreign investment and generate thousands of construction‑phase jobs. Yet once operational, workforce requirements fall sharply. A 50‑MW or larger facility may need only around 50 employees to run day‑to‑day operations.

However, experts argue that the high‑skill jobs created—particularly in cooling systems, power engineering, networking and GPU operations—will have a lasting impact on India’s talent pool. Each direct role tends to create additional employment in the surrounding ecosystem, including maintenance, logistics and technical services.

With the Digital Personal Data Protection (DPDP) rules pushing companies toward more on‑shore data storage, demand for new data‑centre capacity is poised to accelerate. States are actively courting investments with incentives and dedicated policies, while global companies view India as a strategic long‑term market.

As AI reshapes digital economies worldwide, India is positioning itself not only as a major consumer of data but as a future hub for the infrastructure that powers next‑generation computing.