Applied Digital Expands AI Data Center Capacity by 75 MW at Polaris Forge 1 Amidst Surging Demand

Applied Digital has announced the operational readiness of an additional 75 megawatts (MW) of AI infrastructure at its Polaris Forge 1 campus in Ellendale, North Dakota. This expansion completes the second phase of Building 2, bringing its total critical IT load to 150 MW and the overall campus capacity to 250 MW. The entire Polaris Forge 1 campus is fully leased and is contracted to deliver 400 MW upon full buildout. This development is significant for the AI industry as it directly addresses the escalating need for specialized computing power. The continuous growth of AI models and applications, particularly in areas like large language models and advanced machine learning, requires substantial and dedicated infrastructure. Applied Digital's ability to rapidly bring such large-scale capacity online helps alleviate the ongoing compute scarcity that many organizations face. For practitioners, this means improved access to the necessary hardware for training complex models, running intensive simulations, and deploying AI solutions at scale. The expansion also highlights the strategic importance of geographical distribution for AI infrastructure, with Ellendale, North Dakota, becoming a key hub for high-performance computing. This expansion fits within a broader trend of aggressive investment in AI infrastructure across the tech landscape. Companies like Anthropic are making massive, long-term commitments to secure cloud computing, specialized hardware, and services, with agreements spanning hundreds of billions of dollars over the next decade with major players like Amazon, Google, and Broadcom. Similarly, the US economy is seeing significant business investment driven by AI infrastructure, contributing to overall economic growth. The demand for AI-specific data centers is so high that even regions like Japan are seeing multi-billion dollar projects aimed at bolstering domestic AI capacity. This collective push underscores the industry's recognition that robust, scalable infrastructure is the bedrock upon which the future of AI will be built. In practice, this means that organizations should continue to evaluate their AI infrastructure strategies, considering both hyperscale cloud providers and specialized data center operators like Applied Digital. The availability of more dedicated AI capacity can lead to better performance, reduced latency, and potentially more cost-effective solutions for specific workloads. Practitioners should monitor the geographical distribution of these new capacities, as proximity to data centers can impact performance and compliance. Furthermore, the emphasis on “purpose-built” infrastructure suggests a growing need for specialized cooling, power, and networking solutions tailored for AI, which developers and architects should factor into their design choices. The ongoing expansion by companies like Applied Digital indicates a sustained commitment to meeting AI's infrastructure demands, offering a more diverse and robust ecosystem for AI development and deployment.