India’s growing demand for AI computing, cloud services, and digital infrastructure is pushing data centre operators to expand capacity and rethink where they build.
CtrlS Datacentres is targeting 4.4 GW of capacity by 2030, with investments spanning established technology hubs and smaller cities
It recently announced a ₹10,000 crore investment in Bihar to set up a 150 MW AI factory in Bodh Gaya, while expanding its presence in the state with a ₹400 crore data centre in Patna. CtrlS’s Patna data centre currently has 12 MW of IT capacity, with scope to scale up as demand increases.
In an exclusive interview with AIM, Rahul Dhar, President of Global Data Centre Operations, Enterprise Colo Delivery, Programme Management and Energy R&D at CtrlS Datacentres, says the strategy is to bring infrastructure closer to customers, particularly content-based businesses that need to serve users in smaller cities.
Dhar says the new facility would support demand for digital infrastructure in Bihar and eastern India.
The company also plans to establish more than 20 edge data centres across tier-2 and tier-3 cities over the coming years with improved fibre connectivity. It currently operates 19 data centres across 10 cities, including Lucknow, Noida, Kolkata, and Gandhinagar, and operates a containerised data centre in Bhubaneswar.
Nikhil Kamath, Co-founder of Zerodha, and entrepreneur Sreeram Reddy Vanga, recently invested a combined ₹250 crore in CtrlS Datacenters. Canadian Pension Plan Investment Board also invested around ₹7,000 crore in the company earlier this year.
Dhar attributes investor interest in CtrlS to its focus on hyperscale campuses, advance planning for high-capacity grid power, and modular designs.
Hyperscalers Drive Demand
Hyperscalers constitute the largest source of demand for CtrlS, although the company also serves nearly 40 Indian public and private sector banks. Startups and neocloud providers are another growing customer segment.
CtrlS is building greenfield data centre campuses for hyperscalers in Hyderabad and Mumbai. It is also in discussions with neocloud companies seeking capacity, and expects to deliver against its existing commitments over the next 12–24 months.
But it also faces the possibility of competition from Big Tech. This year, the central government introduced a 21-year tax holiday for foreign companies providing global cloud services using India-based data centres.
However, Dhar says it won’t affect CtrlS’s existing demand, which is driven by domestic customers as well as global technology companies. “Even though there is a tax holiday, and there might be an influx of foreign investors, the Indian colocation providers have a good order book in hand,” he says.
Power Availability Remains a Challenge
“India needs to shorten the period for forecasting data centre power requirements and speed up approvals for power transmission lines and fibre networks to keep pace with the industry’s expansion,” says Dhar.
Utilities and government departments typically seek power consumption on 7–8 year horizons. However, the growing power requirements of AI workloads and increasingly dense GPU racks are making long-term forecasts harder to maintain.
“The projection that the typical governments or the various power or utility departments are asking us can be reduced to three to four years,” Dhar says.
Data centres are growing from facilities with capacities of around 5–8 megawatts to campuses ranging from 10–100 MW. As newer chips enter the market, operators must continually reassess their power requirements and the capacity of existing facilities.
“We are bulk consumers of power, so we need better planning or better awareness among various government departments on how much power we need, at what time and at which location,” he says.
Utilities need advance visibility into demand to plan transmission lines, substations, and power connections to data centre sites. Dhar also calls for faster right-of-way approvals to lay fibre networks and transmission lines, particularly when operators acquire land on city outskirts.
CtrlS is providing inputs to government bodies and industry associations, including the Ministry of Electronics and Information Technology, NASSCOM, and the Confederation of Indian Industry, on India’s recently proposed data centre policy. Several states, including Telangana, Uttar Pradesh, Karnataka, Tamil Nadu, and Maharashtra, have already introduced data centre policies.
Dhar says the Centre’s proposed policy should address infrastructure planning and approval timelines.
Sustainability
Data centres face growing scrutiny over their water and electricity consumption. According to Mordor Intelligence estimates, India consumed about 150 billion litres of water in 2025 to cool about 1.5 GW of data centre capacity. This figure is expected to more than double by 2030.
CtrlS uses closed-loop chillers to limit water loss and is reusing treated wastewater at one of its Mumbai facilities to reduce dependence on fresh water.
The company also owns and operates a nearly 100 MW captive solar farm in Nagpur. The power is fed into the grid to offset consumption at its Mumbai data centres, with similar initiatives planned for other locations, including Hyderabad.
CtrlS also operates its facilities at a power usage effectiveness (PUE) of around 1.4. PUE measures a data centre’s total energy consumption against the energy used by its IT equipment, with lower values generally indicating better energy efficiency.
Changing Data Centre Design
As GPU deployments grow, operators must adapt to higher power densities and new cooling requirements. Dhar says CtrlS is prepared to support customers deploying liquid-cooled infrastructure, including GPU providers and cloud companies.
“All our data centres are rated tier-4 data centres,” he says, referring to the highest classification of data centres with zero single points of failure and close to 100% uptime. He adds that the company uses external certification and has introduced an Integrated Command and Control Centre in Hyderabad to monitor infrastructure centrally.
The centre supports predictive analysis and preventive maintenance. CtrlS is also applying AI to infrastructure monitoring, including CCTV systems, to improve responses to operational issues.
As AI workloads increase infrastructure requirements, Dhar says data centre operators will need to keep adapting their facilities to changing chip technologies, power densities, and cooling systems.


