Vizag Anchors India's Gigawatt-Scale AI Infrastructure

Google officially broke ground on an AI hub in Visakhapatnam on April 28, 2026, according to a Google Cloud press release. The project is being developed in partnership with AdaniConneX and Nxtra by Airtel and the ceremony included Andhra Pradesh Chief Minister N. Chandrababu Naidu and Union Minister Ashwini Vaishnaw, per the press release. Multiple Indian outlets, including NDTV and The Economic Times, report that Google will invest $15 billion in the campus, described by coverage as one of the largest foreign direct investments in India's history. NDTV and The Economic Times also report statements by Jeet Adani, Director at Adani Group, that the Adani Group has committed $100 billion to build an integrated platform across energy generation, transmission, digital networks and data centers to support AI growth.

Google's announcement and press materials state the AI-ready campus will total approximately 600 acres across Tharluwada, Adavivaram, and Rambilli and target 1 gigawatt (GW) of capacity, according to The Economic Times. The Google release and ministerial remarks at the ceremony reference associated digital infrastructure, including subsea cable landings, that will improve connectivity for the region.

Public reporting frames this project as a vertically integrated approach-co‑locating large-scale compute capacity with grid and network infrastructure-to reduce marginal energy and connectivity costs for AI workloads. Observers writing about comparable megaprojects note that co-investment by hyperscalers and local infrastructure partners can accelerate data‑centre scale-up by aligning land, power, and fiber availability.

For practitioners and infrastructure planners, the combination of a 1 GW target and major hyperscaler capital matters because power availability and network latency shape which workloads are economical to run at scale. Industry coverage emphasizes energy provision and efficiency as gating factors; several outlets quote Jeet Adani stressing that "AI may be written in code, but it runs on electricity," highlighting the public narrative linking energy strategy with compute affordability.

Observers should track three indicators over the next 12-24 months:

• •grid-scale power agreements and on-site renewable capacity announcements tied to the campus
• •subsea and terrestrial fiber build schedules that affect international and intra‑India latency
• •phased technical specifications for the campus (PUE targets, modular data‑centre designs, and timeline for reaching incremental MW/GW milestones). Reporting by NDTV and Google's materials provide the initial project numbers, but formal project timelines, procurement contracts, and environmental clearances will be the concrete milestones that confirm delivery versus aspiration

Data‑center operators, cloud partners, and enterprise ML teams will watch how localized hyperscaler investments change cost structures for large-scale training and inference in India. Industry observers note that when compute, power, and connectivity are addressed jointly, it lowers the barrier for deploying latency‑sensitive and data‑sovereign AI applications domestically.

Google's press release and the contemporaneous coverage include quotes and project-level metrics, but they do not publish full long-term timelines, vendor lists for hardware procurement, or detailed power-purchase agreements in the initial announcements. Multiple news outlets cite Jeet Adani's commitments and project-scale figures; however, detailed contractual milestones remain to be disclosed by the project partners.