Tata Consultancy Services is planning a major new AI data center campus in Hyderabad that could eventually provide as much as one gigawatt of computing capacity, positioning the Indian city to host large-scale artificial intelligence infrastructure for global technology companies.
HyperVault, a TCS subsidiary established to develop AI-focused data centers, has secured 264 acres for the project. TCS says HyperVault and its partners could invest as much as ₹700 billion — approximately $8 billion — as the campus is developed and operated.
The project will be built in phases, with the pace of construction determined by customer demand and evolving technology requirements. If fully developed to its planned 1 GW capacity, it would be among India’s largest concentrations of computing infrastructure.
Unlike conventional enterprise data centers, the Hyderabad campus is being designed specifically for the demands of artificial intelligence. HyperVault expects its customers to include large cloud providers, or hyperscalers, as well as companies developing advanced AI models.
Training and operating those models requires large clusters of graphics processing units, or GPUs, packed much more densely than servers in traditional data centers. That creates enormous requirements for electricity and cooling. HyperVault says the Hyderabad facility will support high-density computing and direct-to-chip liquid cooling, in which cooling fluid is brought close to the processors rather than relying primarily on conventional air conditioning. The campus is also expected to incorporate renewable energy and water-neutral design principles.
The scale of the power requirement is particularly noteworthy. Data center projects of this size consequently involve much more than constructing buildings and installing servers. They require substantial power generation and transmission capacity, fiber connectivity, cooling systems, water management and specialized engineering.
For TCS, the project also represents a significant expansion beyond the company’s traditional role as one of the world’s largest providers of IT services and consulting.
HyperVault is intended to provide the underlying computing infrastructure needed by companies developing and deploying artificial intelligence, while TCS can provide cloud, engineering, AI and enterprise technology services around that infrastructure. TPG (Texas Pacific Group) is participating as a strategic partner in the HyperVault business.
Artificial intelligence is adding additonal requirements: access to enormous amounts of computing power. That means India increasingly needs the physical infrastructure — electricity, data centers, advanced cooling and high-speed networks — to accompany its engineering talent.
Hyderabad has already attracted major investments from global technology companies and has developed into one of India’s largest technology and GCC centers. The HyperVault project would add a new layer to that ecosystem by providing infrastructure intended for some of the most computing-intensive AI applications. At Amritt, we have supported an American company that provides crucial equipment to AI data centers. (We are not involved with TCS in this initiative)
The economic impact could also extend well beyond TCS. A data center campus of this scale requires suppliers and contractors in power systems, cooling equipment, networking, construction, engineering and ongoing facility operations.
There is an important qualification to the headline numbers. The 1 GW capacity and approximately $8 billion investment represent the potential full build-out, not infrastructure that is already committed for immediate construction. HyperVault says development will proceed in phases according to customer demand, a very reasonable approach in my view.
Even with that distinction, securing a 264-acre site for a purpose-built AI campus indicates the scale at which TCS is preparing for future demand. If the project develops as planned, Hyderabad would become not only a major source of technology talent for global companies but also a location where some of the physical computing infrastructure behind the AI industry is housed.
