How do you merge AI’s relentless compute demands with strategic site selection? In this fireside chat, we sit down with Applied Digital’s CEO Wes Cummins to discuss why next-generation AI workloads require a new class of purpose-built data centers. Using Applied’s Ellendale AI Factory—a 363,000-square-foot, three-story HPC facility supporting a 100 MW IT load with advanced liquid cooling and North Dakota’s free-cooling climate—as a case study, Wes will walk through the design and construction from the ground up. Drawing on global projections of 171–219 GW of data center capacity needed by 2030, Wes will illustrate how these AI factories unlock efficiency and economic advantages compared to legacy hubs. Finally, he will explore how leveraging stranded renewable power and the region’s climate positions North Dakota as the premier AI data center hub.
Key Talking Points

• Global AI Compute Demand
• Demand is projected to reach 171–219 GW by 2030, driven by generative AI workloads requiring high power density and advanced cooling.
• Reclassification of Data Centers
• AI-specific workloads need purpose-built infrastructure with specialized cooling, high-density racks, and high-bandwidth fabrics, distinct from legacy facilities.
• Ellendale AI Factory Case Study
• A 363,000-square-foot, three-story HPC facility supporting a 100 MW IT load, featuring advanced liquid cooling and leveraging North Dakota’s free-cooling climate.
• Site Selection Strategy
• Applied Digital situates data centers near stranded renewable power sources and in cold climates to maximize efficiency, sustainability, and speed-to-power.

North Dakota Advantages
Abundant untapped energy and minimal risk of brownouts ensure reliable, cost-effective operations, positioning the state as a premier AI hub