India’s installed data-centre capacity is projected to grow nearly threefold to 5–6 gigawatts (GW) by 2030, from around 1.65 GW in 2025, as artificial-intelligence workloads, rising digital-data consumption and tougher data-localisation requirements accelerate demand for large computing infrastructure.
According to Redseer Strategy Consultants, the expansion would require India to commission several times its currently installed capacity within the next five years, while also securing major additions to electricity supply, land and network infrastructure.
“Installed capacity stood at around 1.65 GW in calendar year 2025 and is expected to reach 5–6 GW by CY2030,” Redseer said in its India Data Centres report. The projected expansion represents a compound annual growth rate of about 27–32% over the decade from 2020.
Why India’s data-centre demand is rising
Data centres are facilities that house servers, storage systems and networking equipment used to run websites, cloud platforms, digital payments, streaming services, enterprise software and artificial-intelligence applications. As more consumer, business and government services shift online, the amount of information that needs to be stored, processed and moved securely is rising sharply.
One of the key drivers is India’s push to store sensitive data within the country. Rules and requirements linked to the Digital Personal Data Protection Act, along with sector-specific directions from regulators such as the Reserve Bank of India (RBI) and the Insurance Regulatory and Development Authority of India (IRDAI), have increased the need for domestic data-storage and processing capacity.
The report also identified rapid growth in data consumption and AI workloads as major demand drivers. Artificial intelligence requires enormous computing power, especially for training large models and processing high volumes of data. The growing use of cloud services, online commerce, digital payments, streaming and enterprise digitisation is adding to the pressure on existing data-centre capacity.
India’s cost advantage attracts investment
India is becoming an attractive location for data-centre investment because of the comparatively low cost of building facilities. Redseer estimates that a data centre can be set up in India at approximately $8 million per megawatt (MW) of capacity.
This cost advantage, combined with growing domestic demand, data-residency requirements and access to a large digital user base, could encourage global cloud companies, colocation providers, telecom firms and domestic infrastructure players to expand their India footprint. However, the report indicates that availability of reliable electricity will be the key condition for turning announced projects into operational data centres.
Power requirement could rise to 7–9 GW
The expected 5–6 GW of installed IT load by 2030 will require considerably more electricity from the grid. Redseer estimates that this scale of data-centre capacity would translate into approximately 7–9 GW of grid draw, reflecting the additional power needed for cooling systems, backup infrastructure and other facility operations.
The report estimates that India will still need to secure, contract and energise around 5 GW of incremental grid capacity to support the projected expansion. At present, around 2.7 GW of grid capacity has been secured or is available for the sector.
This means data-centre growth will depend not only on demand from technology companies, but also on the ability of developers and state utilities to provide uninterrupted power, strengthen transmission networks and bring renewable-energy options online. Power availability, pricing and reliability are especially important because data centres must operate continuously and cannot tolerate long outages.
Mumbai and Chennai dominate
India’s data-centre industry is currently concentrated in a few established locations. Mumbai alone accounts for around 50% of installed capacity, while Mumbai and Chennai together account for about 66%, Redseer said.
Mumbai has emerged as a key hub because of its strong connectivity through submarine cable landing stations, proximity to financial-services firms and concentration of large enterprises. Chennai also benefits from international cable connectivity, established digital infrastructure and access to enterprise and cloud demand.
However, the report expects this geography to change as AI-related computing needs rise and developers seek locations with easier access to land and electricity.
AI could shift growth to new locations
Redseer expects latency-sensitive inference workloads and regulated data workloads to remain close to established hubs such as Mumbai and Chennai. Inference refers to the process in which an already trained AI model is used to respond to real-time requests—for example, a chatbot answering a query, a recommendation engine serving content, or a fraud-detection system reviewing a transaction.
Such workloads require low delay, or low latency, because they need to respond quickly to users, businesses or regulated systems. This makes established metro data-centre clusters important for customer-facing and compliance-sensitive operations.
However, AI training and batch-processing workloads may increasingly move to locations that offer larger parcels of land, cheaper and more reliable power, and better scope to build very large computing campuses. Training involves processing massive datasets to build or improve AI models, while batch workloads are computing tasks that can be run over a longer period and do not require instant response times.
Hyderabad, Andhra coast, Gujarat and Rajasthan emerge
The report identifies Hyderabad and the Andhra coast as locations likely to record significant relative gains as India’s data-centre market expands. These locations could benefit from access to land, improving connectivity and the ability to develop new infrastructure clusters.
Gujarat and Rajasthan are also emerging as potential destinations for AI training and batch-processing workloads, according to Redseer. Their appeal is likely to depend on the availability of large land parcels, renewable-energy capacity, transmission connectivity and competitive power costs.
The shift could create a more distributed data-centre map in India, reducing dependence on Mumbai and Chennai and allowing new regional hubs to specialise in power-intensive workloads such as AI model training.
India’s data-centre capacity is set for a major expansion, but the scale of the opportunity also highlights the infrastructure challenge. Reaching 5–6 GW by 2030 would require rapid project execution, financing, reliable grid connections, new transmission infrastructure and a growing share of clean energy.
The report indicates that demand is no longer the main concern. AI, cloud use, digital growth and data-localisation rules are driving demand. The real challenge is whether India can provide enough power, land, fibre connectivity and quick approvals to build large-scale data centres.