Science & Technology6 August 20265 min read
Data centre capacity rises from 375 MW to 1,575 MW since 2020, Lok Sabha told
Installed data centre capacity in India has grown from 375 MW in 2020 to about 1,575 MW, driven by demand from artificial intelligence and high-performance computing, Minister of State for Electronics and IT Jitin Prasada told the Lok Sabha on 05.08.2026. He said data centres are an integral part of the Digital India Infrastructure ecosystem.
The main hubs are Mumbai, Navi Mumbai, Chennai, Hyderabad, Bengaluru, Delhi NCR and Gujarat. Andhra Pradesh, Madhya Pradesh, Chhattisgarh and West Bengal are emerging as new investment destinations.
The sector is mainly developed, owned and operated by private companies, with domestic and global investment. Firms weigh demand projections, operating costs, power availability, skilled manpower and long-term viability when choosing sites. The Government of India provides no subsidy for setting up data centres.
What the IndiaAI Mission has delivered
The government acknowledged the risks of limited domestic capability in semiconductor manufacturing, computing infrastructure, foundational models and advanced research. It said it is using a multi-pronged approach under the IndiaAI and India Semiconductor Missions to build technological sovereignty across the AI value chain.
The IndiaAI Mission was approved on 7 March 2024 with an outlay of ₹10,371.92 crore over five years. Its progress, as listed in the reply:
- Foundation models: 506 applications were received under the Call for Proposals of 30 January 2025, and 20 proposals (12 Large Multimodal Models and 8 Small Language Models) were selected, with intellectual property retained by applicants. Launched outputs include Sarvam AI's 30 billion and 105 billion parameter models, Gnani.AI's speech-to-speech model, BharatGen's Param2-17B, Patram-7B and Shrutam-2, and Avataar AI's video generation model.
- Compute: 15 Compute Service Providers have been empanelled over four rounds, 237 projects approved for subsidised compute and 93.18 lakh GPU hours sanctioned. A Purchase Order has been issued for a High-Performance AI Compute System of about 1.1 EFLOPS at the NIC Data Centre, Shastri Park, Delhi.
- Applications: 11 national hackathons or innovation challenges were launched, 62 AI prototypes developed and 20 AI solutions deployed in public-sector institutions.
- Centres of Excellence: 58 are approved across States and UTs, of which 22 have been approved and initiated across 13 States/UTs.
- Safe and Trusted AI: 13 projects on bias mitigation, machine unlearning, privacy-preserving AI, algorithm auditing and explainability have been selected.
The Department of Science & Technology also implements the National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS), approved with an outlay of ₹3,660 crore. It has set up 25 Technology Innovation Hubs at academic institutions, covering areas such as AI and machine learning, robotics, cybersecurity and quantum technologies. It also funds fellowships and skilling and supports deep-tech startups.
Chips and dependence on imported GPUs
The Semicon India Programme, launched in January 2022 for semiconductor fabs, display fabs, packaging and Indian-designed chips, has approved 12 manufacturing projects with committed investment of ₹1.64 lakh crore. Three have begun commercial production. The Union Cabinet approved Semicon 2.0 on 15 July 2026 with an outlay of ₹1,27,500 crore, to accelerate chip design, expand fabs and advanced packaging, and support materials, gases and equipment.
The government said India's compute ecosystem still relies on globally sourced GPUs procured through empanelled providers, in line with international AI hardware supply chains. The Shastri Park system is among the steps to reduce this dependence over time.
India joined the Pax Silica coalition at the India AI Impact Summit 2026, with the United States and other partners, to secure the global silicon supply chain. The Stanford Global AI Vibrancy Report (November 2025) ranks India 3rd globally in AI competitiveness and ecosystem vibrancy, and second-largest contributor to GitHub AI projects.
Water and environmental rules
In a written Rajya Sabha reply, Minister of State for Environment, Forest and Climate Change Kirti Vardhan Singh said AI data centres do not per se need Environmental Clearance under the EIA Notification, 2006. Clearance is needed only where a data centre forms part of a Building and Construction Project under Item 8(a) (built-up area above 20,000 sq. m) or a Township and Area Development Project under Item 8(b) (area of 50 ha or more and/or built-up area of 1,50,000 sq. m or more). Prior clearance then comes from the State Environment Impact Assessment Authority (SEIAA). Appraisal considers freshwater availability, especially in water-stressed regions, water balance, greywater generation, recycling and reuse, and prescribes safeguards.
Water and greywater use and discharge are regulated under the Water (Prevention and Control of Pollution) Act, 1974 and the Air (Prevention and Control of Pollution) Act, 1981, with compliance monitored by State Pollution Control Boards or Pollution Control Committees. Groundwater extraction follows Ministry of Jal Shakti guidelines (Notification S.O. 3289(E) of 24.09.2020, amended 29.03.2023).
The Bureau of Energy Efficiency's District Cooling Guidelines, 2023 recommend treated water as cooling tower make-up water where feasible, energy-efficient cooling, alternative water sources and conservation measures.
Water needs depend on the cooling technology used. The minister said the industry is adopting direct-to-chip liquid cooling, adiabatic cooling and immersion cooling, and moving towards closed-loop liquid cooling with high-density racks, which together improve Water Use Efficiency (WUE) and cut power and water use. The 'Resilience, Innovation & Efficiency' Working Group of the India AI Impact Summit examines energy efficiency, water management and environmental resilience in digital and AI infrastructure.