The Indian government has urged Microsoft to consider building future data centres closer to power generation sites rather than concentrating them near coastal cable landing stations, arguing that the shift could lower infrastructure costs and support the country’s expanding AI ecosystem.
Electronics and IT Secretary S. Krishnan made the suggestion on Monday at the launch of Microsoft’s fourth cloud region in India, located in Hyderabad. He said locating data centres closer to electricity generation could be more economical than building them primarily around cable landing hubs such as Chennai.
The remarks come as Microsoft scales up one of its largest infrastructure bets in India, with a $20.5 billion commitment to cloud and AI infrastructure.
Proposal for Conducting Cyber Crisis Drill, Tabletop Exercise (TTEx) & CCMP Readiness Exercise
Hyderabad becomes Microsoft’s fourth India cloud region
Microsoft formally launched its India South Central cloud region in Hyderabad on September 21.
The new region joins the company’s existing cloud regions in Pune, Chennai and Mumbai. Microsoft said the Hyderabad facility is designed for AI workloads and will serve as a strategic hub for India, Asia and the wider Global South.
The region became generally available on August 6 and is now part of Microsoft’s larger India expansion.
Microsoft says its data centre regions across the country are AI-enabled and equipped with infrastructure capable of supporting advanced cloud and artificial intelligence workloads.
Major Indian companies including Adani Group, Bajaj Finserv, HDFC Bank and PB Pay have already signed up to use the Hyderabad region, according to Microsoft.
Why power location matters for AI data centres
Modern AI data centres consume enormous amounts of electricity.
Thousands of high-performance processors may run continuously inside a single facility, while cooling systems and networking equipment add further energy requirements.
Krishnan’s argument is that electricity can be delivered more efficiently when data centres are built closer to large generation sources.
Long-distance electricity transmission requires grid infrastructure and can involve additional costs and losses.
Data connectivity works differently.
Fibre-optic networks can move enormous amounts of information over long distances, meaning a data centre does not necessarily need to sit directly beside a submarine cable landing station to maintain high-speed connectivity.
The government is therefore suggesting that companies reconsider the traditional preference for coastal hubs and give greater weight to where abundant electricity is available.
Krishnan said it could be easier to move data through fibre networks than to move very large quantities of electricity to coastal data centres.
Microsoft says Hyderabad region uses low-water cooling
Power is only one resource challenge facing the data centre industry.
Large facilities can also consume significant amounts of water for cooling.
Microsoft said its new Hyderabad region uses high-efficiency mechanical cooling that results in effectively zero water use for cooling operations.
That is increasingly important in India, where rapidly growing data centre clusters may compete with cities, agriculture and industry for electricity and water.
The company is also investing in connectivity infrastructure.
Microsoft is part of a consortium developing the planned India-Southeast Asia submarine cable system, known as I-2SEA.
The 3,600-km cable is expected to connect India with Malaysia and Singapore, with Indian landing points at Machilipatnam and South Chennai. It is targeted to enter service in 2029.
That combination of inland computing capacity and coastal cable infrastructure reflects the broader architecture the government is encouraging.
India’s AI push is creating a power challenge
India wants to become a major global hub for AI development and cloud computing.
But the scale of electricity needed for that ambition is becoming increasingly important.
Nearly 40% of India’s coal-fired power plants were operating with critically low fuel stocks as of September 21, according to government data reviewed by Reuters, highlighting broader pressure on power supply during periods of high demand.
AI data centres will add another large industrial electricity load.
That means decisions over where new facilities are built could influence transmission investment, land use, renewable-energy procurement and the cost of delivering AI services.
Microsoft says the wider expansion of India’s data centre sector could generate between $140 billion and $200 billion in economic output through 2030, citing a Boston Consulting Group study.
The government’s latest suggestion shows that the next phase of the data-centre race will not be driven by computing hardware alone.
Access to cheap, reliable power may become one of the most important factors determining where India’s AI infrastructure is built.
What this means for you: India’s AI expansion will require far more electricity as well as computing capacity. Building data centres closer to power sources could reduce infrastructure costs, but regulators will still need to balance energy demand, grid reliability and environmental pressures.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
