Status: Announced / Pre-development and Planning | Location: Jamnagar, Gujarat, India
The proposed Reliance-NVIDIA AI Data Centre Campus in Jamnagar, Gujarat, is poised to become the most ambitious and largest data center project globally, with a staggering planned capacity of up to 3 Gigawatts (GW). This venture, estimated to cost between $20 billion and $30 billion, is not just about scale; it is a strategic national initiative aimed at democratizing access to Artificial Intelligence throughout India, aligning with Mukesh Ambani's vision to offer the "lowest AI inferencing costs in the world."
The collaboration between Reliance, which owns the world's largest data company (Jio), and NVIDIA, the world leader in AI technology, is centered on creating a next-generation AI Supercomputing Infrastructure. NVIDIA will supply its most advanced chips, including the Blackwell AI processors, which are essential for the complex, large-scale computations required to train Large Language Models (LLMs) customized for India’s diverse languages and generative AI applications. This advanced hardware requires specialized cooling, meaning the facility will heavily rely on Direct Liquid Cooling (DLC) systems to handle the immense heat generated by the dense GPU clusters, a necessary evolution from traditional air-cooling.
The choice of Jamnagar is strategic, leveraging the existing site of Reliance’s massive industrial complex. Crucially, the campus is planned to be powered entirely by green energy, drawing from a massive solar, wind, and green hydrogen power complex being built nearby. While maintaining a continuous 3 GW supply remains a monumental challenge for any project, this commitment underscores the scale and long-term sustainability goals embedded in the development.
By building this high-performance computing backbone, Reliance aims to replicate the market disruption achieved by Jio in the telecom sector, providing the foundational infrastructure for scientists, developers, startups, and 450 million Jio customers. The project serves as a clear statement of India’s ambition to move beyond merely consuming digital services to actively creating AI solutions made in India, for India, rapidly accelerating the nation's journey toward becoming a global AI powerhouse.
| Field | Value |
|---|---|
| Project Name | Reliance-NVIDIA AI Data Centre Campus |
| Location | Jamnagar, Gujarat, India |
| Status | Announced / Pre-development and Planning |
| Commissioning | Phased, starting estimated 2027-2028 |
| Total IT Load | Not applicable (Initial phase) |
| Total Capacity | Up to 3 Gigawatts (GW) (Total planned IT load capacity) |
| Tier Level | Tier IV (Expected for world's largest, mission-critical AI compute) |
| Project Type | AI Factory/Hyperscale Compute Hub |
| Header | Details |
|---|---|
| City Name | Jamnagar |
| Population | around 699,000 for the city in 2025 |
| Urban Agglomeration | 877,000 for the metro area |
| City GDP | (US$7.87 billion) for 2022-23 |
| Per Capita Income | $8.8k (USD) GDP per capita |
| City Tier | Tier 2 city, though some sources list it as a developing Tier 3 city |
| Key Strengths | Site of Reliance's massive existing industrial and oil refining complex; large, available land parcels; dedicated power infrastructure; proximity to major green energy projects. |
| Header | Details |
|---|---|
| Developer / Operator | Reliance Industries Limited |
| Strategic Infrastructure Partner | NVIDIA (Advanced Chip Supplier/Technology Partner) |
| Construction Contractor | Not disclosed publicly (Likely RIL's infrastructure divisions) |
| MEP Engineering | Focus on Immersion Cooling and advanced thermal management for AI clusters |
| Network Connectivity | Ultra-high-speed domestic fiber backbone (via Jio) |
| Power Infrastructure | Powered by a dedicated Green Energy Complex encompassing solar, wind, and green hydrogen (Crucial for 3 GW scale) |
| Header | Details |
|---|---|
| Power Capacity | Planned 3,000 MW (3 GW) |
| UPS Redundancy | 2N (Highest level of redundancy required for such a facility) |
| Cooling System | Direct Liquid Cooling (DLC) and potentially Immersion Cooling (Essential for NVIDIA Blackwell GPU clusters) |
| Connectivity | Leveraging Jio's extensive 4G/5G/Fiber network for low-cost, high-volume data transfer |
| PUE Target | Targeting ultra-low PUE, heavily influenced by green energy and advanced cooling |
| Energy Mix | 100% Renewable Energy commitment from dedicated nearby power generation facilities |
| Header | Details |
|---|---|
| Announcement | September/October 2024 (Partnership confirmed at NVIDIA AI Summit) |
| Construction Start | Expected 2025 |
| Phase 1 Go-Live | Est. 2027-2028 |
| Full Buildout | Est. 2030+ (To reach 3 GW capacity) |
| Header | Details |
|---|---|
| Total Investment | Est. $20 Billion to $30 Billion USD (Total estimated cost) |
| Funding | Reliance Industries Limited corporate reserves and capital |
Subscribe to our exclusive newsletter and get the latest insights on data center trends, market forecasts, and infrastructure innovations delivered straight to your inbox.