Skip to main content

Applied Digital completes phase II Ready for Service at Polaris Forge 1 fully energizing first 100 MW building

Applied Digital reached a milestone by completing the Ready for Service status for the second 50 Megawatt (MW) phase at its Polaris Forge 1 Artificial Intelligence (AI) Factory Campus in Ellendale, North Dakota. This completion brings Building 1 to full capacity with an operational 100 MW critical IT load.

The achievement marks the fulfillment of Applied Digital's delivery commitments for the first of three contracted buildings at the facility. Polaris Forge 1 is part of a broader 400 MW deployment project secured through long-term leases with CoreWeave.

The infrastructure comprises high-density, production-scale systems intended to provide AI compute capacity aligned with current demand. The company's approach integrates precise construction methodologies and design models to meet deployment timelines.

Applied Digital executed the initial 50 MW phase in October and then proceeded with the second phase on schedule to reach the full 100 MW operational status. Additionally, the company entered into a lease agreement valued at approximately $5 billion with a U.S.-based investment-grade hyperscaler for its Polaris Forge 2 campus, contributing to a total contracted revenue near $16 billion across both campuses.

“Bringing this first 100 MW building to full critical load on schedule reflects the engineering discipline and execution reliability our customers expect,” said Todd Gale, chief development officer of Applied Digital. “These AI Factory builds are complex, high-density systems, and delivering them at this speed and quality demonstrates the strength of our design, the rigor of our construction model, and our commitment to meeting the real deployment timelines of large-scale AI workloads.”

Applied Digital continues to develop the Polaris Forge campuses, focusing on the delivery of AI infrastructure that meets requirements for speed, precision, and sustainability relevant to modern AI computational workloads.