JupiteX Get the app
Science & Technology19 Nov 2026 · about 6 min

India's Data Centre Boom Faces Growing Power And Cooling Challenges As AI Demand Surges - The 420.in

The brief

A data centre is a large building filled with computer servers. Those servers store, process and transmit digital information. Data centres matter because many ordinary online activities rely on that computing infrastructure, even when users do not see it. The article points to online banking, digital payments, video streaming and business software as examples. These services use data centres to handle information and support digital activity. As companies expand cloud services and artificial intelligence operations, facilities also need to support more powerful computing systems. India had nearly 1.8 gigawatts of operational data centre capacity by the end of the first half of 2026, according to Cushman & Wakefield. Another 3.9 GW was in development. These figures show the scale of infrastructure being built to serve digital demand, and why dependable power and cooling are important: facilities need to keep servers running safely and continuously.

01

What is a data centre, and what everyday digital services depend on it?

A data centre is a large building filled with computer servers. Those servers store, process and transmit digital information. Data centres matter because many ordinary online activities rely on that computing infrastructure, even when users do not see it.

The article points to online banking, digital payments, video streaming and business software as examples. These services use data centres to handle information and support digital activity. As companies expand cloud services and artificial intelligence operations, facilities also need to support more powerful computing systems.

India had nearly 1.8 gigawatts of operational data centre capacity by the end of the first half of 2026, according to Cushman & Wakefield. Another 3.9 GW was in development. These figures show the scale of infrastructure being built to serve digital demand, and why dependable power and cooling are important: facilities need to keep servers running safely and continuously.

02

How much data centre capacity does India have now, how much is planned, and what does the 17 GW electricity-demand projection mean?

By the end of the first half of 2026, India had nearly 1.8 gigawatts of operational data centre capacity, according to Cushman & Wakefield. Its assessment put the development pipeline at about 3.9 GW. These figures describe data centre capacity, not the amount of electricity the industry uses over a period.

A separate government-cited estimate concerns future electricity demand. In August 2026, the Ministry of Electronics and Information Technology cited Central Electricity Authority projections that data centre electricity demand could reach approximately 17 GW by 2031–32. That is a forecast of power demand, not a count of operating or planned computing facilities.

The distinction matters as India plans infrastructure for expansion. Crisil Ratings estimated installed data centre capacity could rise fivefold, to about 7.5 GW by March 2030, and said growth depends on timely power connections, approvals and upgrades for AI workloads. More facilities mean greater pressure on electricity infrastructure.

03

Why do AI workloads require so much electricity and produce so much heat?

Artificial intelligence workloads use powerful computing systems to process enormous amounts of information. The article explains that advanced AI processors generate substantial heat because they perform enormous numbers of calculations. More demanding workloads therefore add pressure to both electricity supply and cooling systems.

The challenge grows when more powerful processors are installed closer together. Servers already produce heat while operating, and tightly grouped, high-performance equipment can make it harder to keep temperatures safe. Traditional air-conditioning is common, but conventional air cooling may be less effective for these increasingly demanding AI systems.

That is why the industry is considering approaches such as liquid cooling, which transfers heat away from equipment through specially designed systems. This is not simply a comfort issue: if temperatures rise too high, equipment can malfunction, slow down or shut down. India’s expansion plans will depend in part on reliable electricity and cooling that can handle growing AI workloads.

04

What can happen to servers and digital services if a data centre loses power or cannot stay cool?

Data centres support digital services that people and businesses use every day, including banking, payments, video streaming and business software. They must operate continuously. The article warns that even brief power interruptions can disrupt essential digital services, so reliable electricity connections are central to the industry’s expansion.

Heat creates a separate risk. Servers generate heat as they operate, and cooling systems keep temperatures within safe limits. If temperatures rise too high, equipment can malfunction, slow down or shut down. AI processors intensify the challenge because they generate substantial heat while performing large numbers of calculations.

These risks help explain why building facilities is only part of the task. Operators also need dependable power and effective cooling. India’s development pipeline is growing, but Crisil Ratings says expansion depends partly on timely electricity connections and upgrades for demanding AI applications. Without reliable infrastructure, a facility may be built yet struggle to keep services running safely and continuously.

05

How do liquid-cooling methods compare with air cooling, and how can their water use differ?

Traditional data centres commonly use air-conditioning to keep servers at safe temperatures. But conventional air cooling may be less effective as powerful AI processors are placed closer together and generate substantial heat. Liquid cooling is an alternative that transfers heat away from equipment using specially designed systems.

Two approaches described in the article work differently. Direct-to-chip cooling circulates coolant through components positioned near processors. Immersion cooling places specialised computing equipment in electrically non-conductive liquid. Both are intended to manage high temperatures, but their costs and infrastructure requirements differ. The article does not rank either method as universally better.

Water use also varies by design. Some cooling systems require substantial water, while closed-loop arrangements recirculate coolant. That difference can affect operating costs and environmental sustainability, so the choice of technology matters as data centres expand. The article does not quantify water use or provide a direct comparison of efficiency between air and liquid systems.

06

Why do data centres need transmission lines, substations and backup systems—not just enough electricity generation?

A data centre needs power delivered to its site reliably, not merely electricity generated somewhere on the grid. The article identifies timely electricity connections as a condition for growth and notes a pipeline of projects in electricity transmission and distribution. It does not describe transmission lines, substations or backup systems in detail.

As general background, transmission lines carry electricity over distance, while substations help connect parts of the grid and adjust voltage for delivery. A site connection links the facility to that network. Backup systems can provide a temporary source of power if the usual supply is interrupted, helping a facility manage an outage; these particular mechanisms are not specified in the article.

The underlying concern is clear in the source: data centres must operate continuously, and even brief interruptions can disrupt essential digital services. India’s expansion will depend on timely connections and power infrastructure that can serve new facilities. More generation alone does not ensure that each site receives dependable electricity.

07

What does a gigawatt measure, and how is power capacity different from the amount of electricity used over time?

A gigawatt, or GW, measures power: the rate at which energy can be supplied or used at a moment. One GW equals one billion watts. This is general measurement background; the article uses GW to report data centre capacity and projected electricity demand.

Electricity consumed over time is measured as energy, commonly in kilowatt-hours or gigawatt-hours. For example, a power level maintained for a period results in a quantity of energy used during that period. So a GW figure does not, by itself, say how many units of electricity a facility consumes in a day or year.

That distinction helps read the article’s numbers accurately. Cushman & Wakefield reported nearly 1.8 GW of operational data centre capacity and about 3.9 GW in development. Separately, a government-cited projection put data centre electricity demand at approximately 17 GW by 2031–32. The article explicitly describes 17 GW as projected future power demand, not current operational computing capacity.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web