India’s data center growth is expanding beyond traditional hubs like Mumbai to secondary cities, driven by AI’s massive energy needs. Government projections estimate electricity demand for these facilities will reach 26.3 GW by FY 2031-32. This shift is creating new revenue opportunities for network operators providing high-speed optical links, though investors may monitor the heavy capital expenditure and execution risks involved in such large-scale infrastructure projects.
India’s digital infrastructure is undergoing a significant geographic transition. For years, the country’s data center capacity was concentrated in coastal hubs like Mumbai and Chennai, which acted as the primary entry points for global subsea internet cables. However, the rise of artificial intelligence (AI) is forcing a change in strategy. Companies are now scouting for land and power capacity in inland secondary cities rather than relying on proximity to legacy financial centers.
The Surge in Power Requirements
The scale of this transition is directly linked to the massive electricity requirements of modern AI systems. Official government projections indicate that power demand from data centers is expected to climb to 26.3 GW by the 2031-32 fiscal year. This estimate represents a significant upward revision from earlier projections, reflecting the rapid pace at which AI-specific facilities are being planned and deployed across the country.
A clear example of this corridor-centric growth is the recent investment by global technology giants like Google, which committed $15 billion to develop an AI hub in Visakhapatnam. Projects of this size aim to build dedicated international gateways, allowing companies to avoid the congestion seen in traditional port cities while placing infrastructure closer to reliable power sources.
Impact on Connectivity and Network Operators
This geographic expansion is redefining the business for network operators. Unlike traditional internet traffic, AI infrastructure requires two specialized layers of connectivity. First, it needs ultra-fast, high-capacity links to support the training of massive AI models across multiple physical locations. Second, it requires complex connectivity for the inference operations that follow. This creates a rising demand for so-called wave services—high-capacity optical links that can transmit vast amounts of data with minimal delay.
For investors, this shift offers a nuanced look at the infrastructure sector. Companies that specialize in providing fiber-optic connectivity and high-speed data transmission are likely to see sustained demand for these specialized services as AI clusters multiply across the country.
Risks and Investor Monitorables
While the demand outlook appears promising, building this infrastructure carries real risks. Developing large-scale data centers in secondary cities requires massive capital investment, which can lead to high debt levels if the projects face delays or cost overruns. Execution risk remains a critical monitorable; securing reliable power supply in new locations and ensuring timely completion of connectivity networks will dictate the profitability of these projects.
Investors should track three key factors in the coming quarters: the actual commissioning timelines for these new inland projects, the ability of power infrastructure companies to meet the 26 GW demand target, and the revenue growth for telecom and optical fiber players specifically from AI-related wave service contracts.