CoreWeave commits 240 MW of GPU compute capacity to India, marking a major geographic expansion for the AI cloud provider in Asia.
CoreWeave has announced plans to develop 240 MW of AI data center capacity in India through a collaboration with AdaniConneX at its Taloja campus in Navi Mumbai. The move capitalizes on accelerating global demand for AI computing infrastructure and India's emergence as a booming market for AI development, cloud computing, and data center investment.
The initial deployment will consist of three 80 MW buildings, with CoreWeave serving as the sole tenant. The campus has the potential to double capacity by another 240 MW, giving the company a possible total expansion opportunity of 480 MW. The first phase is expected to become operational in mid-2028, with additional capacity coming online in phases.
A key component of the project is CoreWeave's plan to deploy the NVIDIA Vera Rubin Platform, targeting agentic AI and inference workloads. This next-generation infrastructure will give customers access to advanced AI computing without requiring massive upfront infrastructure investments.
CoreWeave has been aggressively building a global footprint across North America, Europe, and the Asia-Pacific region, with its entry into Indonesia in August representing another step in that strategy. The company has approximately 4.2 GW of contracted power. The India project fits into a broader strategy of securing power supply, building specialized data centers, deploying advanced NVIDIA hardware, and providing AI infrastructure close to customers—a model that can create competitive advantage if the company maintains high utilization and attractive pricing.
In the competitive landscape, Nebius Group recently acquired Inferize to strengthen production inference capabilities, aiming to cut cold-start delays, reduce idle GPU time, and improve capacity utilization. Nebius has committed to investing £1.7 billion in U.K. AI infrastructure with three advanced NVIDIA-powered deployments, expects connected power capacity to reach 800 MW to 1 GW by year-end 2026, and recently partnered with Palantir to provide AI compute and inference infrastructure to Palantir's commercial customers.
Microsoft has continued expanding its partnerships across AI, cloud, and enterprise technology, broadening its decade-long partnership with Databricks into the 2030s and expanding ties with Mistral to bring its models across Microsoft Foundry, Copilot Studio, and Azure. Microsoft also partnered with 3M on AI data-center infrastructure with Azure deploying 3M's Expanded Beam Optical technology, partnered with NASA on AI tools for scientific data discovery, and struck an AI content agreement with Nine Entertainment allowing Copilot to reference its journalism. Microsoft's relationship with OpenAI continues under an IP arrangement extending to 2032.
CoreWeave shares have gained 23.5 percent year to date compared with the Internet Software industry's rise of 7.6 percent. On a price-to-book basis, CoreWeave trades at 8.08X, higher than the Internet Software industry's 5.44X. The Zacks Consensus Estimate for CoreWeave's earnings for the current year has been revised downward over the past 60 days, and the stock currently carries a Zacks Rank of 3 (Hold).