United States-based artificial intelligence cloud firm CoreWeave has officially marked its strategic expansion into India. Through a major partnership with AdaniConneX, the firm plans to establish a massive CoreWeave data centre presence delivering 240 megawatts (MW) of critical capacity in Navi Mumbai.
Key takeaways
- CoreWeave enters the Indian digital ecosystem through an infrastructure partnership with AdaniConneX.
- The multi-phase roadmap establishes a 240 MW artificial intelligence cloud campus in Navi Mumbai.
- The purpose-built installation is dedicated to high-density computing capacity required for modern AI workloads.
- AdaniConneX will supply the domestic facility, land bank, and electrical substation connectivity.
- The move aligns with surging enterprise demand for low-latency graphics processing unit (GPU) capacity across the Asia-Pacific region.
A 240 MW Footprint in Navi Mumbai
The planned CoreWeave data centre campus will be situated in Navi Mumbai, an established digital hub for enterprise digital infrastructure in western India. Reports from industry publications, including Devdiscourse and CRN Asia, confirm that the facility will provide up to 240 MW of critical power capacity reserved specifically for accelerated cloud computing platforms and machine learning environments.
Navi Mumbai has consistently attracted large-scale enterprise capital due to its proximity to international subsea landing stations, modern civil zoning, and access to stable electrical grids. For CoreWeave, securing a dedicated 240 MW power allocation provides the scalability necessary to train massive foundational models and host compute-heavy inference tasks without the capacity bottlenecks seen in more congested hosting markets.
The CoreWeave Data Centre Partnership with AdaniConneX
To execute a deployment of this scale, CoreWeave is collaborating directly with AdaniConneX, a domestic joint venture established between Adani Enterprises and global operator EdgeConneX. The partnership pairs CoreWeave’s specialised software-defined cloud platform with AdaniConneX’s physical real estate portfolio, power procurement channels, and local project execution capabilities.
Deploying specialized AI infrastructure requires far higher power density and specialized thermal engineering compared to standard enterprise hosting. Modern graphics processing clusters generate intense thermal loads that traditional air cooling cannot support. By collaborating with AdaniConneX, CoreWeave can leverage pre-engineered industrial sites configured to accommodate liquid-to-chip cooling loops, dedicated dual-substation power feeds, and high-density rack enclosures.
AdaniConneX oversees site acquisition, shell construction, utility clearances, and regional grid integration, which significantly reduces the lead time required for CoreWeave to bring hardware clusters online. This structural division of labor lets CoreWeave focus entirely on importing, configuring, and orchestrating its specialized GPU architectures.
| Project Parameter | Details |
|---|---|
| Primary Operator | CoreWeave |
| Infrastructure Partner | AdaniConneX |
| Planned Location | Navi Mumbai, Maharashtra, India |
| Total Power Capacity | 240 Megawatts (MW) |
| Target Workloads | Artificial Intelligence Training, LLM Fine-Tuning, Inference |
| Facility Engineering | High-Density Liquid Cooling and Dedicated Grid Substation |
Technical Architecture of the CoreWeave Data Centre
The technical demands of modern deep learning have forced hyperscalers to rethink data centre architecture from the foundation up. Inside the CoreWeave data centre, systems are engineered to support rack densities exceeding 40 to 100 kilowatts (kW) per cabinet—substantially higher than the 6 to 10 kW standard found across traditional enterprise data halls.
To prevent performance throttling across thousands of interconnected nodes, the Navi Mumbai site will implement low-latency non-blocking network fabrics such as NVIDIA Quantum InfiniBand alongside ultra-high-speed RoCE (RDMA over Converged Ethernet). These networking topologies allow thousands of accelerator chips to function as a unified virtual supercomputer, sharing tensor weights across memory channels with minimal latency overhead.
Additionally, advanced power distribution units (PDUs) and uninterruptible power supply (UPS) systems are being integrated to handle the rapid electrical fluctuations typical of massive generative AI training runs, where power consumption can swing by multiple megawatts within fractions of a second during heavy matrix operations.
Meeting Surging AI Compute Needs Across India
India is experiencing an unprecedented surge in demand for compute capacity as local software firms, global capability centers (GCCs), research institutions, and enterprises develop domain-specific generative models. The government’s broader digital initiatives have further spurred demand for local compute sovereignty, encouraging multinational operators to invest in domestic hyperscale facilities.
Traditional multi-tenant hosting environments in the region have historically catered to web hosting, database storage, and conventional SaaS workloads. By introducing a dedicated CoreWeave data centre to Maharashtra, local engineers and global enterprises gain access to infrastructure designed natively for deep learning frameworks like PyTorch, JAX, and TensorFlow.
Rather than routing high-value workloads to overseas clusters in North America or Europe—which introduces latency penalties and complicates cross-border data transfer compliance—enterprises operating in South Asia can train and fine-tune models directly within the country. This local presence reduces network transit costs while keeping critical training data within Indian borders.
Liquid Cooling and High-Density Power Infrastructure
Managing heat dissipation is one of the most pressing engineering hurdles in high-performance cloud campuses. The CoreWeave data centre in Navi Mumbai will feature direct-to-chip liquid cooling systems and advanced secondary fluid loops designed to maintain optimal semiconductor junction temperatures under continuous operational load.
Liquid cooling provides significant efficiency advantages in tropical and sub-tropical climates like western India. Closed-loop liquid distribution units (CDUs) extract heat directly from processors far more efficiently than ambient air handlers, lowering the facility’s Power Usage Effectiveness (PUE) ratio. This efficiency reduces overall electrical draw, saving operational expenditure while minimizing strain on local municipal utility grids.
Navi Mumbai’s Rise as a Hyperscale Digital Hub
The choice of Navi Mumbai reinforces the city’s position as a focal point for institutional data centre investments in India. The industrial corridors of Airoli, Mahape, and Dronagiri host a growing ecosystem of carrier-neutral campuses, international fiber routes, and robust utility networks.
Because multiple undersea fiber-optic cable systems terminate near Mumbai, the district functions as the primary digital gateway connecting India to Europe, the Middle East, and Southeast Asia. For CoreWeave, anchoring its Indian footprint in this corridor ensures access to redundant, multi-terabit carrier transit routes, ensuring dependable connectivity for international customers connecting into the Navi Mumbai cluster.
Industry Impact and Next Steps for the Maharashtra Facility
Following project confirmations reported by regional industry trackers including Indian Startup News and Konsulteer, focus turns to physical site readiness, regulatory milestones, and electrical infrastructure staging in Navi Mumbai. CoreWeave’s entrance introduces direct competition to incumbent hyperscalers and domestic cloud providers currently expanding their own accelerated hardware footprints.
As the multi-stage buildout moves forward, the 240 MW CoreWeave data centre campus will play a major role in shaping accelerated computing availability throughout the region. Enterprise developers, research laboratories, and artificial intelligence startups looking to train and deploy complex machine learning architectures will soon benefit from high-density, specialized computing power operating directly within western India.
Frequently Asked Questions
What is the capacity of the planned CoreWeave data centre in India?
The planned facility in Navi Mumbai is designed to provide 240 megawatts (MW) of power capacity for artificial intelligence and cloud workloads.
Who is CoreWeave partnering with for its entry into India?
CoreWeave is partnering with AdaniConneX, the data centre joint venture between Adani Enterprises and EdgeConneX.
Where is the new AI data centre campus located?
The project is located in Navi Mumbai, Maharashtra, which is one of India's primary hubs for digital infrastructure.
Sources and References
- Indian Startup News: AI cloud firm CoreWeave enters India, plans 240 MW data centre in Navi Mumbai with AdaniConneX
- Konsulteer: CoreWeave Enters India With 240 MW AI Data Centre Campus in Navi Mumbai
- Devdiscourse: CoreWeave enters India with 240 MW AdaniConneX data center deal in Navi Mumbai
- CRN Asia: CoreWeave brings 240 MW AI cloud capacity to India with AdaniConneX



