CoreWeave Enters India With 240 MW AI Data Center Deal

CoreWeave is entering India for the first time through a partnership with AdaniConneX that will give the US-based AI cloud provider access to 240 megawatts of data center capacity at the Taloja campus in Navi Mumbai. The project will comprise three 80 MW buildings and is designed specifically for high-density artificial intelligence workloads, with the first phase expected to become available in mid-2028.

The deal brings together one of the world’s fastest-growing AI cloud providers and one of India’s major data center infrastructure platforms at a time when demand for AI computing is rapidly increasing. CoreWeave plans to deploy NVIDIA’s Vera Rubin platform at the new facilities for AI training, inference, reasoning and agentic workloads, while AdaniConneX will provide the underlying data center infrastructure. CoreWeave will initially occupy the full 240 MW deployment and has an option to add another 240 MW, potentially taking the campus allocation to 480 MW. CoreWeave

Key takeaways

CoreWeave is making its first entry into India.
It will deploy 240 MW of AI data center capacity at AdaniConneX’s Taloja campus in Navi Mumbai.
The initial project consists of three 80 MW buildings.
CoreWeave can potentially add another 240 MW, taking its total allocation to 480 MW.
The facilities are expected to use NVIDIA Vera Rubin infrastructure.
The first phase is expected to become available in mid-2028.
CoreWeave will be the sole tenant across the three initial buildings.
The company plans to establish an Indian office and hire locally.
Bloomberg reports that the project could involve multiple billions of dollars of investment over its lifetime, although CoreWeave’s announcement does not disclose a specific investment amount. Moneycontrol
The project arrives as India’s data center capacity approaches 2 GW, with Mumbai remaining the country’s dominant data center market. The Economic Times

Why CoreWeave is coming to India

CoreWeave’s India entry is fundamentally about bringing computing capacity closer to customers.

AI companies need enormous amounts of computing power not only to train large models but increasingly to run them after training. Inference, reasoning and AI agents can require substantial computing capacity when thousands or millions of users interact with models simultaneously.

That creates a different infrastructure requirement from conventional cloud computing.

AI data centers need high-density GPU clusters, extremely fast networking, advanced cooling systems and large amounts of reliable electricity.

CoreWeave has built its business around this specialised infrastructure.

Its India strategy therefore is not simply about putting conventional cloud servers in another country. The Taloja project is being designed specifically for high-density AI computing. CoreWeave says the facilities will use direct liquid cooling in GPU data halls and a non-evaporative chilled-water system. CoreWeave

That design is significant because modern AI accelerators generate much more heat than traditional enterprise computing equipment.

What the AdaniConneX deal provides

AdaniConneX is a 50:50 joint venture between Adani Enterprises and EdgeConneX focused on building and operating data center infrastructure in India.

The company has been expanding across major Indian data center markets, including Chennai, Hyderabad, Noida, Pune and Navi Mumbai.

Its Navi Mumbai platform is particularly important because the region is already India’s largest data center market.

AdaniConneX lists the Navi Mumbai campus with potential capacity of up to 1 GW of IT load, with connectivity to major internet exchange points and plans for renewable-energy integration. AdaniConneX

For CoreWeave, partnering with an established infrastructure operator can accelerate its entry into India.

Instead of building an entire data center ecosystem from scratch, CoreWeave can focus on deploying and operating its AI cloud platform on purpose-built infrastructure.

The arrangement effectively separates two layers of the AI infrastructure stack.

AdaniConneX: land, buildings, power and data center infrastructure.

CoreWeave: AI cloud platform, computing infrastructure and customer-facing cloud services.

The 240 MW deployment is large

The scale of the project is important.

The first phase will consist of three buildings, each providing 80 MW.

That gives:

80 MW × 3 buildings = 240 MW

CoreWeave will be the sole tenant across these three buildings.

It also has the opportunity to expand by another 240 MW at the same campus.

CoreWeave’s planned Navi Mumbai footprint

ADANICONNEX TALOJA CAMPUS

Initial deployment
┌──────────────┐
│   Building 1 │  80 MW
└──────────────┘

┌──────────────┐
│   Building 2 │  80 MW
└──────────────┘

┌──────────────┐
│   Building 3 │  80 MW
└──────────────┘

             │
             ▼
        240 MW initial
          capacity

             │
       Expansion option
             │
             ▼

        +240 MW possible

             │
             ▼
       Up to 480 MW
       for CoreWeave

Source: CoreWeave announcement. CoreWeave

A 240 MW initial commitment is substantial relative to India’s current data center market.

It also illustrates why AI infrastructure is changing how operators think about data center capacity.

For conventional enterprise workloads, capacity can be added incrementally.

AI clusters can require enormous blocks of power because thousands of accelerators may operate together as a single computing system.

Why Navi Mumbai matters

Navi Mumbai is not an accidental location.

Mumbai is India’s biggest data center market, benefiting from its role as the country’s financial centre and from extensive telecommunications infrastructure.

CBRE says Mumbai accounted for 52% of India’s operational data center stock at the end of H1 2026. It also identified Navi Mumbai as one of the established micro-markets attracting a large share of new development. CBRE

The wider Mumbai region also benefits from connectivity infrastructure, including submarine cable landing facilities.

That matters for AI cloud providers because data centers are not isolated warehouses full of GPUs.

They need to connect customers to computing infrastructure with high bandwidth and low latency.

Navi Mumbai therefore combines access to customers, network connectivity and a rapidly expanding data center ecosystem.

India’s data center market is approaching 2 GW

CoreWeave’s entry comes as India’s data center industry is entering a major expansion phase.

According to CBRE, India added around 130 MW of data center capacity during H1 2026, while more than 200 MW was expected to become operational during the second half of the year.

That would take total operational capacity close to 2 GW by the end of 2026. Express Computer

The investment pipeline is considerably larger.

CBRE estimates that cumulative investment commitments into India’s data center sector since 2021 had reached approximately $173 billion by the end of H1 2026, including about $38 billion during the first half of the year. CBRE

The figures represent investment commitments rather than money already spent.

That distinction matters.

A large amount of planned capacity still needs to be financed, constructed, powered and brought online.

But the size of the pipeline demonstrates how quickly India’s digital infrastructure market is expanding.