The Data Centre Boom Is Creating A New Management Challenge For India
Share
Every time someone asks a chatbot a question, runs a fraud check on a UPI payment or streams a cricket match, a server somewhere does the work. Artificial intelligence may be the headline, but the data centre is the backbone holding it up. For India, that backbone is becoming one of the decade’s biggest business opportunities, and one of its most demanding management problems.
An opportunity arriving at the right time
The numbers are striking. India’s data centre capacity was about 1.5 GW in 2026, and most forecasts expect it to reach 4 to 6.5 GW by 2030. Morgan Stanley’s more bullish estimate goes as high as 10.5 GW by FY31. Vestian expects the market to grow from $10 billion in 2025 to $22 billion by 2030, backed by close to $30 billion of investment. Separately, AWS, Microsoft and Google have together announced commitments worth about $67.5 billion.
The bigger prize lies in what this infrastructure enables. NITI Aayog’s AI for Viksit Bharat report estimates that faster AI adoption could add $500 to 600 billion to India’s GDP by 2035, over and above its current growth path. Financial services and manufacturing are expected to gain the most, with AI potentially contributing up to a quarter of each sector’s output by 2035.
Talent is flowing back too. Talent firm Xpheno counted about 15,100 Indian professionals returning from abroad in 2025, up from 9,800 in 2024.
New jobs, new businesses
Square Yards estimates that India’s data centre pipeline could create about 4.33 lakh jobs by 2030. Most of these are in construction, power, cooling, security and facility services, with around 27,000 direct roles. The Union government expects close to one lakh engineering jobs from the sector over the same period.
Further up the chain, the IndiaAI Mission has sanctioned about 93 lakh GPU hours of subsidised compute to 237 projects, giving startups and researchers access to capacity they could not otherwise afford. NITI Aayog also cautions that AI will displace clerical, routine and low-skill roles, so the gains will not be evenly spread.
Table 1. India’s data centre opportunity at a glance
Rethinking skills and the classroom
This needs a different workforce. Colleges will need courses in data centre design, power and thermal engineering, cloud operations and sustainability management. Business schools will need to teach managers to make decisions about compute budgets and energy procurement.
The challenges worth solving
Power. CEEW estimates that data centre electricity demand could rise from about 13 TWh in 2024 (0.8% of national demand) to 57 TWh by 2030 (2.6%). TERI puts the 2030 share as high as 6%.
Carbon. India’s grid emits over 600 grams of CO₂ per kWh, according to an Allianz Trade study. Applying that figure to CEEW’s 57 TWh projection gives about 34 million tonnes of CO₂ a year if the demand is met entirely from the grid. TERI estimates that greening data centres could cut their emissions by up to 88%.
Water. Data centres used about 150 billion litres of water in 2024, a figure Systemiq and CEEW expect to reach roughly 358 billion litres by 2030. A single 100 MW facility using evaporative cooling can consume around 8 lakh litres a day. More than half of India’s facilities sit in water-stressed regions.
Table 2. Resource footprint of India’s data centres
These pressures call for energy-efficient processors, liquid and immersion cooling, recyclable coolants, treated wastewater for cooling and long-term solar and wind contracts.
Disruption in education. Courses that prepare students for the clerical and routine roles NITI Aayog expects AI to displace will need rework, and their students and teachers will feel it first. The answer is affordable reskilling, closer industry-academia ties and curricula reviewed every year rather than every decade.
The “lower end” criticism. Critics argue India is settling for the infrastructure layer while others build frontier language models and agentic automation. The concern is fair, but India is doing both. Alongside its shared compute pool, the ₹10,371 crore IndiaAI Mission has identified 20 indigenous sovereign AI model proposals for support. Infrastructure and model-building can grow together, if policy keeps pushing that way.
Figure 1. Challenges of the data centre boom and the responses to them
A management opportunity
Building data centres is an engineering task. Making the boom work for India is a management one: balancing growth with sustainability, speed with skill-building, and infrastructure with innovation. The figures show both the size of the opportunity and the scale of the resources it demands. Managers, educators and policymakers who plan for both will decide how much of that opportunity India captures.
Views are personal
The author, Dr Vivek N, is Chairman – Centre for Ai in Business and Professor – Operations, Great Lakes, Chennai
