Google's $15B India AI Hub Takes Shape in Andhra Pradesh — Largest Overseas AI Build on Earth
Google’s $15B data centre hub in Andhra Pradesh is its largest AI investment outside the United States by a significant margin. Ground broke in April 2026. Construction is active across three sites totalling 601 acres on the southern Indian coastline.
Google Cloud CEO Thomas Kurian named it explicitly in October 2025 when the project was announced: “It’s the largest AI hub that we are going to be investing in anywhere in the world, outside of the United States.”
Three Sites, One Strategy
The hub spans three locations in coastal Andhra Pradesh:
| Site | Location | Area |
|---|---|---|
| Site 1 | Taluvada | 266.6 acres |
| Site 2 | Adivivaram | 160 acres |
| Site 3 | Rambilli-Achyutapuram | 174.8 acres |
The infrastructure is built for compute-heavy AI workloads: TPU and GPU clusters for deep learning, neural network training, and large-scale inference. The eastern coast location is also tied to new international subsea cable landings and fibre networks, making it a connectivity node as well as a compute one.
Construction is being handled by AdaniConnex — the data centre joint venture between Adani Enterprises and EdgeConneX — alongside Nxtra by Airtel, Bharti Airtel’s data centre subsidiary.
Why India, Why Now
India is not a new market for Google Cloud, but this build is a different order of magnitude. Previous India data centres served domestic cloud demand. This hub is sized and positioned for international AI inference load and frontier model training.
The coastal placement matters: submarine cable infrastructure connects India to both Southeast Asia and the Gulf, giving the hub latency advantages for serving AI workloads across a broad geography. The site selection in Andhra Pradesh, rather than established tech clusters like Hyderabad or Bengaluru, reflects land and power availability as primary constraints.
The project also fits a pattern visible across Big Tech. Microsoft committed $3B to India in 2024. Amazon has pledged $12.7B in AWS infrastructure by 2030. Google’s $15B commitment moves the benchmark for hyperscaler India investment.
Water and Environmental Friction
The project has faced organised criticism since announcement. Water usage is the primary concern — large-scale AI compute requires substantial cooling, and Andhra Pradesh faces recurring drought conditions. Environmental groups have challenged approvals on groundwater depletion grounds.
Google has not disclosed specific cooling technology or water consumption projections for the hub. Those disclosures typically follow construction completion rather than preceding it.
The friction is not unique to India. Data centre water use has blocked or delayed projects in Ireland, the Netherlands, and Arizona. Google’s Andhra Pradesh hub is proceeding through approvals; the environmental review process continues alongside construction.
Scale in Context
At $15B committed across three sites, the Andhra Pradesh hub is larger than Google’s individual data centre investments in any single country outside the US to date. For reference, Google’s 12-country global AI infrastructure build announced in 2025 aggregated roughly $75B total — making India’s hub 20% of the entire non-US programme.
The operational timeline has not been disclosed. TPU and GPU clusters of this scale typically take 18-36 months from ground-breaking to initial production capacity.