Mumbai [India]: Tata Consultancy Services Ltd. (TCS) announced plans to establish a wholly owned subsidiary focused on building world-class AI infrastructure, starting with a 1 gigawatt (GW) AI data centre in India. The project represents a $6.5 billion investment, which equals India’s total current installed data-centre capacity, marking a transformative leap for the country’s digital ecosystem.
TCS bets big on AI infrastructure
Speaking during the company’s September quarter earnings call, TCS CEO Krithi Krithivasan said the project will be implemented in phases over 5–7 years.
“We have set a target of 1 GW. We will be doing it in phases. We expect to do 1 GW in 5–7 years,” Krithivasan stated.
The project underscores TCS’ pivot from a traditional outsourcing-driven IT model to a capital-intensive AI compute business, aiming to strengthen India’s position in the global AI ecosystem.
According to TCS’ internal estimates, every 150 MW of AI compute power will require $1 billion of investment, bringing the total to $6.5 billion for the planned capacity. The company will fund the expansion through a mix of equity and debt, also bringing on board financial investors.
Clients and revenue outlook
TCS expects clients for the AI data centre to include AI startups, hyperscalers, deeptech firms, and government entities. The facility will primarily serve as a passive data centre, providing backup and failover capacity for active AI workloads across industries.
“Revenue from this investment will start accruing in 18–24 months, once the first phase goes live,” TCS CFO Samir Sekhsaria confirmed.
Although the company did not disclose the exact location of the data centre, it confirmed that the entire AI and compute infrastructure will be hosted within India, supporting the country’s push for data sovereignty and secure AI development.
Strategic shift for the Tata Group
The initiative marks a major milestone for the Tata Group, aligning with its broader goal of building a vertically integrated digital and AI ecosystem—from semiconductor and cloud infrastructure to enterprise software and consumer applications.
TCS’ Chief Operating Officer Aarthi Subramanian said the move will enable the firm to provide end-to-end enterprise AI solutions, including training large language models (LLMs), running generative AI workloads, and hosting India-based secure AI clouds.
“This will give us coverage across the entire AI stack—from infra layers to apps,” Subramanian said.
To oversee the initiative, TCS has formed a new AI and Services Transformation Unit led by company veteran Amit Kapur, reporting directly to the COO.
India’s expanding data-centre landscape
India’s data-centre market, valued at $4–8 billion, is expected to grow at a 15–20% CAGR through the decade, driven by AI, cloud, and digital infrastructure demand.
According to JLL, India’s installed data-centre capacity surpassed 1 GW in 2024, up from 350 MW in 2019. ICRA projects this to double to 2.5 GW by FY28, requiring an investment of nearly ₹90,000 crore ($10 billion). Colliers anticipates even faster growth to 3.5–4.5 GW by 2030, at a total investment of $25–30 billion.
Krithivasan added that the demand for AI compute power is rising exponentially.
“Currently, India has about 1.2 GW capacity, and demand is expected to grow tenfold in the next five to six years,” he said.
Market response
Following the announcement, TCS shares rose 1.09% to ₹3,060.20 on the NSE, outperforming the broader market as the Nifty 50 index gained 0.54% to close at 25,181.80 points.
The move positions TCS at the forefront of India’s AI revolution, setting the stage for a new era of digital infrastructure leadership that could redefine the country’s role in the global AI supply chain.