Applied Digital Signs $7B Lease Agreement with CoreWeave for AI Capacity

DALLAS, Texas — Applied Digital announced on June 3 that it has signed two long-term lease agreements with CoreWeave, promising approximately $7 billion in revenue over approximately 15 years. The agreements will support CoreWeave’s artificial intelligence (AI) and high-performance computing (HPC) infrastructure at Applied Digital’s new data center campus in Ellendale, North Dakota.

Shares of Applied Digital surged about 48% on Monday following the announcement, while CoreWeave’s stock increased by about 8%. The leases will provide 250 megawatts (MW) of critical IT load to meet the rapidly growing demand for AI capabilities. CoreWeave retains an option for an additional 150 MW.

Wes Cummins, Chief Executive Officer of Applied Digital, stated that these agreements are a vital part of the company’s strategic expansion into advanced compute infrastructure. The leases guarantee an annual revenue stream of approximately $467 million for Applied Digital, which will bolster its long-term financial stability.

The first data center, expected to be operational by the fourth quarter of 2025, will provide 100 MW of power. A second building housing 150 MW is projected for completion in mid-2026, with a potential third facility under consideration for 2027. The Ellendale campus is designed to accommodate up to 1 gigawatt of IT load, positioning it as a significant player in the evolving AI infrastructure landscape.

Applied Digital’s design incorporates rapid deployment capabilities and efficient access to low-cost energy, key components in fulfilling the dynamic requirements of AI and HPC applications. The agreements mark a substantial step forward in establishing Applied Digital as a key provider of essential data center infrastructure.

In addition to scaling up its operations, Applied Digital aims to further consolidate its position in the delivery of cutting-edge compute solutions. The company envisions the Ellendale facility not just as a development project, but as a launchpad for the future of AI infrastructure.