Meta and Reliance partner to build India's first 168 MW AI data center in Jamnagar

Meta has signed its first built-to-suit AI data center agreement in India, partnering with Reliance Industries to build a 168-megawatt facility in Jamnagar, Gujarat. The announcement deepens a relationship that began with Meta's $5.7 billion Jio Platforms investment in 2020 and extends to open-source AI model collaboration.

What's new

Key terms of the partnership:

• Scale: 168 MW of AI-optimized compute capacity in Jamnagar, with options to expand beyond the initial phase.
• Structure: Meta will lease capacity from the Reliance-owned facility. Reliance builds and operates; Meta anchors the demand.
• Sustainability: Meta contracted nearly 1 GW of clean energy from CleanMax (837 MW) and Fourth Partner Energy (88 MW) across multiple Indian states. The facility will be cooled using desalinated seawater. Meta has committed to covering the entire cost of the energy and water required to support its operations there.
• Connectivity: The site ties into Project Waterworth, described as the world's longest subsea cable system, linking India to global AI infrastructure.
• Timeline: The facility is expected to be operational within approximately two years.

Mark Zuckerberg stated: "This world-class facility in Jamnagar will help us scale our AI infrastructure globally while deepening our long-term investment in India's economy."

Reliance chairman Mukesh Ambani added: "Building India's first built-to-suit AI data centre for a global technology leader of Meta's scale demonstrates India's readiness to be at the forefront of the global AI revolution."

Context

Meta's relationship with Reliance goes back to a $5.7 billion investment in Jio Platforms in 2020, one of the largest FDI deals in Indian history at the time. The companies have since collaborated on digital services and enterprise AI. This data center deal is the infrastructure layer of that partnership, giving Meta a significant physical AI footprint in the world's most populous country.

India has become a priority geography for global AI investment. Microsoft, Google, and Amazon have each announced multi-billion-dollar India commitments in the past 18 months, targeting both the developer workforce and the domestic enterprise market. Meta's Jamnagar facility plants its flag in that race.

The 168 MW figure is meaningful: a single NVIDIA GB200 NVL72 rack system consumes roughly 120 kW, meaning the facility could house approximately 1,400 such racks at full draw — enough for substantial training and inference capacity at scale.

Why it matters

For Meta AI, this is a statement of infrastructure intent. The company runs Llama as an open-weight model family and increasingly competes with OpenAI and Anthropic for developer and enterprise workloads. More owned compute — even via a lease — reduces dependence on hyperscaler pricing and gives Meta flexibility to run large training runs on its own schedule.

For India's AI ecosystem, the deal validates the country as a credible site for frontier AI infrastructure. The renewable energy and desalinated water cooling arrangements suggest Meta expects the facility to run at very high utilization. India's emergence as a strategic AI infrastructure destination, alongside its large developer workforce, makes this partnership significant beyond the two companies involved.