India's data center capacity projected to hit 3.6GW by 2030
What's the story
India's data center capacity is projected to grow to around 3-3.6 gigawatts (GW) by 2030, according to a report cited by Axis Capital. The growth will be driven mainly by hyperscalers and AI-focused cloud players. However, execution and supply-chain challenges may keep the actual operational capacity below the publicly announced targets of 6-8 GW set by industry players.
Market shift
Hyperscalers to drive demand, AI's role increasing
The report predicts that hyperscalers will account for about 90% of data center demand by 2030. This is a major jump from their current 60% share.
The change is driven by companies like Amazon Web Services, Microsoft, and Google expanding their capacity as well as the growing need for neo-cloud and large language model hosting.
Capacity analysis
Current data center landscape in India
India's installed data center capacity is between 1.3GW and 1.4GW at present.
Third-party colocation facilities take the biggest share at 60-65%, followed by hyperscaler captive or leased capacity at 25-30%.
Enterprise and edge data centers account for around 8-10% and 2-3%, respectively.
The report says a more realistic estimate of planned capacity by mid-2030 would be between 3.4GW and 3.6GW, with actual operational capacity closer to around 2.8GW due to various challenges including power availability, equipment supply, and construction timelines.
Growth hurdles
Supply-chain challenges may hinder project execution
The data center expansion pipeline is picking up pace, with an estimated capacity of 900MW to 1.2GW awarded in the last six to seven months of calendar year 2026.
However, these awards do not directly translate into operational capacity.
The report highlights that hyperscaler-grade backup generation equipment from leading global suppliers is booked about two years in advance, which could slow project execution due to supply-chain bottlenecks.
Cost implications
Rising capital intensity and cost considerations
The report also highlights a rise in capital intensity as AI-enabled facilities require more expensive infrastructure.
Traditional colocation builds cost around ₹44-46.6 crore per megawatt (MW), excluding IT hardware, while AI-enabled infrastructure costs around ₹48-54 crore per MW. This is a 10-15% premium over traditional builds.
If all currently announced projects are fully realized, India's data center capacity could reach 7-8GW by 2035.