India’s IT services sector is no longer just writing the software layer of the digital economy; it is beginning to build the physical AI infrastructure underneath it. That shift is visible in the latest earnings commentary from TCS and HCLTech, and it comes at a time when India’s data centre market is set to move from about $10 billion in 2025 to $22 billion by 2030, according to Vestian. The broader message is clear: as AI moves from experimentation to production, Indian IT firms are increasingly following the compute, power, land, and connectivity that make those workloads possible.
From software to physical AI
TCS said in its Q4 earnings call that its Hyperworld business had made “significant progress” on AI infrastructure capacity, including customer commitments, land parcel finalisation, and partnerships as it works toward a one-gigawatt buildout. The company also said demand is converging around anchor AI workloads in the 100MW to 200MW range per customer, which signals a move away from small pilots toward much larger infrastructure bets. That same call linked TCS’s HyperVault initiative with OpenAI, including an initial 100MW phase and an option to scale to 1GW in India.
HCLTech is seeing the same trend from a different angle. The company said demand is rising in robotics, AI factories, data centres, and semiconductor engineering, and described physical AI as a growing area of interest for enterprise customers. In other words, the industry is no longer talking only about cloud migration and software engineering; it is also selling the infrastructure and engineering services that sit closer to chips, edge inferencing, and automation.
Why the market is changing
This shift is happening because AI is changing what enterprises buy. The spend is moving away from traditional discretionary technology work and toward foundational infrastructure, especially data centres, storage, networking, and specialised engineering for robotics and edge devices. For Indian IT firms, that opens a new revenue pool where the work is more capital-intensive, more strategic, and potentially more durable than short-cycle software projects.
Analysts quoted in the coverage say physical AI could add 8% to 12% incremental revenue for Indian IT services firms over the next two to three years, especially from data centres, robotics integration, edge silicon engineering, and AI factory services. That estimate matters because it shows how the value chain is widening: the winners will not just be the firms that build applications, but also those that help build the underlying compute ecosystem.
India’s data centre supercycle
Vestian’s numbers underline how large the opportunity is becoming. The report says India’s data centre market was valued at $10 billion in 2025 and is expected to cross $22 billion by 2030, while installed capacity currently sits at 1.4 GW to 1.6 GW. More than 700MW is already under construction, with another 1GW to 1.2GW planned. That means the next phase of growth will be driven not only by demand, but by the ability to secure land, power, cooling, and connectivity at scale.
The NTT Global Data Centers guide attached to this story reinforces that India is not a single data centre market, but a portfolio of regional ecosystems. Mumbai accounts for roughly 53% of operational capacity, Chennai about 21% to 23%, Delhi NCR around 15% to 20%, while Bengaluru and Kolkata serve distinct enterprise and regional demand patterns. The guide also stresses that infrastructure strategy in India depends on sequencing, not just location, because power, land, connectivity, and policy conditions vary sharply across regions.
What this means for IT firms
For Indian IT services companies, this is a major business-model transition. The sector is still rooted in software and outsourcing, but the next wave of growth is likely to come from physical AI infrastructure, hyperscale partnerships, and engineering services tied to robotics and semiconductor design. Firms like Infosys and Wipro are also expanding their AI-ready delivery systems, infrastructure services, and analytics capabilities, which suggests the competitive field is already widening beyond TCS and HCLTech.
The deeper implication is that Indian IT firms are moving closer to the industrial layer of AI. That includes compute capacity, data centre planning, edge systems, and physical deployment, not just model development or application integration. As AI matures, the companies that can bridge software, infrastructure, and operations will be the ones best placed to capture the next cycle of enterprise spending.
