India data centre capacity to jump to 12 GW by 2030 as AI fuels demand: Wood Mackenzie

India's operational data centre capacity is projected to increase more than five-fold to 12 GW by 2030 from 2.2 GW in 2025, driven by rapid adoption of artificial intelligence, hyperscale cloud investments and the country's expanding digital economy, according to a Wood Mackenzie report released on Monday.

AI-dedicated capacity is expected to rise multifold to 6,546 MW by 2030 from 275 MW in 2025, while data centre electricity demand is forecast to grow to 191 terawatt-hours (TWh) by 2040 from 10 TWh in 2025, the report said.

"India's data centre market becomes a structural investment thesis," said Souhardya Pal, Research Associate at Wood Mackenzie. "The convergence of hyperscale capital, AI workload growth and a decade of policy support have created the conditions for India to rival any market in Asia-Pacific. The question for developers and investors is no longer whether to enter India, but where and how."


The report said India's digital economy, valued at Rs 32 lakh crore in 2025 and accounting for about 12 per cent of GDP, together with more than 1.03 billion active internet users and around 22 billion monthly UPI transactions, is driving demand for data centre infrastructure. India's domestic AI market is projected to reach Rs 11.7 lakh crore by 2032.

Wood Mackenzie said Maharashtra and Tamil Nadu currently account for about 65 per cent of installed IT load, but future investment is expected to expand into Andhra Pradesh, Telangana, Uttar Pradesh and Karnataka. These emerging markets are attracting significant commitments from global hyperscalers, including Amazon Web Services and Google, alongside domestic operators such as AdaniConnex, which has announced a development pipeline of 2.6 GW.

The report said access to reliable, cost-competitive electricity has overtaken land and capital as the industry's primary constraint, with captive power generation and long-term renewable power purchase agreements emerging as preferred strategies.

"Land and capital are no longer the limiting factors for data centre developers in India. What determines site selection and delivery timelines now is access to firm, round-the-clock power at the node level," said Rashika Gupta, Vice President of Research at Wood Mackenzie. "Developers who secure their power strategy early through captive generation or long-term renewable PPAs, will lock in a structural cost and sustainability advantage for the life of their assets."

The report also identified water availability as an emerging investment risk as AI-driven computing increases cooling requirements, particularly in water-stressed states such as Tamil Nadu and Karnataka. It said developers adopting closed-loop cooling systems and zero liquid discharge technologies could gain an advantage as regulatory requirements evolve.

According to Wood Mackenzie, data centre electricity demand is forecast to grow 20-fold by 2040 from 10 TWh in 2025 and would account for 7 per cent of the total electricity demand.

The report also identifies water availability as one of the sector's most underappreciated investment risks. As AI workloads increase rack densities and cooling requirements, water access is becoming an increasingly important site selection criterion, particularly in water-stressed markets such as Tamil Nadu and Karnataka.

Wood Mackenzie notes that developers adopting closed-loop cooling systems and zero liquid discharge technologies are already reducing freshwater consumption and positioning themselves ahead of potential future regulation.

"India's combination of strong digital demand, supportive policy, expanding renewable energy markets and accelerating AI adoption is creating one of the world's fastest-growing data centre markets. However, long-term success will increasingly depend on developers' ability to secure reliable power, manage water resources and select locations that balance growth potential with infrastructure resilience," the report added.