India’s AI data centre boom spreads from Mumbai to new hubs

India’s emergence as a multi-city AI data centre hub is crucial for understanding where the compute behind next‑generation models will actually live. With 1.6 GW running and roughly double that in the pipeline, India is no longer just a source of software talent; it’s becoming a physical home for AI workloads across cloud, enterprise and public sector.

Strategically, the geographic diversification beyond Mumbai into Hyderabad, Chennai, Pune and Vizag helps alleviate land and power constraints while spreading economic benefits. It also creates a more resilient regional footprint for hyperscalers and model labs that want to hedge geopolitical and regulatory risk. The flip side is that India now has to solve hard infrastructure problems around high‑density power, liquid cooling for GPU clusters, and grid reliability at scale.

For the race to AGI, this buildout matters because timelines are increasingly bottlenecked by power and datacenter capacity rather than pure algorithmic progress. Countries that can stand up dense, AI‑optimized facilities quickly will have disproportionate influence over where frontier training and inference happens. India is positioning itself to be one of those places.