India’s
artificial intelligence (AI) ambitions are triggering a race to build the infrastructure needed to power the technology. US technology giant Meta recently announced plans to lease a 168-megawatt (MW) AI-ready data centre being built by Reliance Industries in Jamnagar, Gujarat.
Earlier this year, Google began construction on its AI hub in Visakhapatnam as part of a $15 billion investment programme to build gigawatt-scale AI infrastructure in India. Global data centre operator AirTrunk, backed by Blackstone, has also announced plans to invest around ₹3 trillion in India by 2030, making it one of the country’s biggest digital infrastructure commitments.
What is a data centre?
A data centre is a specialised facility that houses thousands of computers, storage systems and networking equipment that run websites, cloud services and AI applications around the clock. Besides servers, these facilities also require heavy-duty cooling systems, backup power and advanced security to keep operations running without interruption.
Data centres have existed for decades, but AI has changed their scale and purpose. Earlier facilities mainly stored files or hosted websites. Today’s AI data centres are designed to train and run large AI models, requiring far more computing power and specialised hardware such as graphics processing units (GPUs).
Why are AI data centres so energy-intensive?
Every time someone asks ChatGPT, Gemini or Claude a question, the request is processed inside a data centre. Unlike traditional servers, AI systems rely on powerful processors that perform thousands of calculations simultaneously, making them significantly more energy-intensive.
According to the International Energy Agency (IEA), data centres consumed around 415 terawatt-hours (TWh) of electricity in 2024, about 1.5 per cent of global electricity supply. That figure is expected to nearly double to 945 TWh by 2030.
Why do AI data centres consume so much water?
The enormous amount of electricity consumed by AI servers also generates significant heat. To prevent computers and AI chips from overheating, data centres rely on advanced cooling systems, many of which use large quantities of water.
According to the Environmental and Energy Study Institute (EESI), a medium-sized data centre can consume up to 110 million gallons of water a year, equivalent to the annual water use of about 1,000 households. Larger facilities can use as much as 5 million gallons of water a day, or nearly 1.8 billion gallons annually — roughly the water needs of a town with 10,000 to 50,000 people.
The environmental impact extends beyond water use. A report by the United Nations University Institute for Water, Environment and Health (UNU-INWEH) estimates that global data centres could consume 945 TWh of electricity by 2030. The report also warns that the sector’s water footprint could equal the minimum annual domestic water needs of 1.3 billion people in Sub-Saharan Africa.
Researchers have also found that AI data centres can affect local temperatures. A study by researchers from the University of Cambridge, Nanyang Technological University and other institutions found that land surface temperatures around AI data centres increased by an average of 2 degrees Celsius after they became operational, with impacts extending up to 10 kilometres away. More than 340 million people living within that distance could be affected by the temperature rise, the study said.
How does India compare in data centre capacity?
According to Statista, India had 296 data centres as of April 2026, making it the world’s sixth-largest market by number of facilities after the US, the UK, Germany, China and France. However, the number of facilities alone does not reflect a country’s computing strength.
Instead, the industry measures capacity in megawatts (MW), which indicates how much electricity a data centre can use to power servers, storage systems, cooling equipment and AI chips. Higher capacity means the facility can support more computing workloads.
By this measure, India has expanded rapidly. According to the Ministry of Electronics and Information Technology (MeitY), installed data centre capacity increased from about 375 MW in 2020 to nearly 1.5 gigawatts (GW) in 2025. More recent estimates by Cushman & Wakefield suggest operational capacity has already crossed 1.6 GW, making India the second-largest operational data centre market in the Asia-Pacific region.
The expansion is expected to continue. According to Wood Mackenzie, India’s operational data centre capacity is projected to increase more than five-fold from 2.2 GW in 2025 to 12 GW by 2030. AI-dedicated capacity alone is expected to rise from 275 MW to 6,546 MW during the same period. The research firm also projects India’s domestic AI market to reach ₹11.7 trillion by 2032, driven by rapid digitalisation and rising AI adoption.
Can India’s power grid keep pace with AI growth?
As India builds more AI infrastructure, electricity demand is expected to rise sharply.
In a written reply to Parliament, Minister of State for Power Shripad Naik said AI data centres are projected to add 26.3 GW of electricity demand by 2031-32. The additional demand is expected to be integrated into the grid and met primarily through renewable energy. The estimate is nearly double the government’s earlier projection of 13.56 GW, announced in March this year.
To meet this growing demand, the power ministry said it is expanding transmission infrastructure in phases while adding new power generation capacity, including renewable energy projects.
The Central Electricity Authority (CEA) also expects India’s peak electricity demand to increase from 289 GW in FY27 to 388 GW by FY32, reflecting rising consumption across the economy, including from AI infrastructure.
Why are companies investing billions in India’s AI infrastructure?
The government’s higher power demand estimate comes as companies accelerate investments in AI-ready infrastructure.
Last week, HCLTech signed a memorandum of understanding with Sarvam AI to build a $1.5 billion AI data centre in Odisha. Google is developing its AI hub in Visakhapatnam, while Microsoft has committed $3 billion to expand its cloud and AI infrastructure in India.
According to research firm Wood Mackenzie, India’s growing digital economy, expanding internet user base and rising adoption of AI are turning the country into a major destination for data centre investments. The report says Maharashtra and Tamil Nadu currently account for around 65 per cent of installed IT load, while future investments are expected to expand into Andhra Pradesh, Telangana, Uttar Pradesh and Karnataka.
What lies ahead for India’s AI infrastructure?
India’s AI ambitions extend far beyond building better chatbots. The technology requires an entire ecosystem of data centres, electricity networks, cooling systems and digital infrastructure capable of operating around the clock.
While billions of dollars are flowing into the sector, the rapid expansion also raises questions about whether India can build enough clean power, secure adequate water resources and develop infrastructure fast enough to support the country’s AI ambitions. How India balances this growth with energy security and environmental sustainability could shape the next phase of its AI journey.
 "Inside India's AI boom: Data centres, energy and the infrastructure race")
