In the heart of North Dakota, a quiet revolution in artificial intelligence infrastructure is unfolding. Applied Digital Corporation, a Dallas-based data center operator, has secured a landmark $5 billion lease agreement with an unnamed U.S.-based investment-grade hyperscaler for its Polaris Forge 2 campus in Harwood, near Fargo. This 15-year deal, announced on October 22, 2025, commits 200 megawatts of critical IT capacity, with the hyperscaler holding the right of first refusal on an additional 800 MW. The agreement underscores the escalating demand for high-performance computing (HPC) tailored to AI workloads, positioning North Dakota as an unexpected epicenter for the next wave of digital infrastructure.

Drawing from recent announcements, the Polaris Forge 2 project spans 900 acres and is designed to support massive GPU clusters essential for generative AI and machine learning. Construction is set to begin in September 2025, with initial capacity online by 2026 and full operations by early 2027. According to a press release from GlobeNewswire, this lease brings Applied Digital’s total leased capacity in North Dakota to 600 MW across two campuses, involving deals with two major hyperscalers.

The Rise of Regional AI Hubs

North Dakota’s appeal lies in its abundant energy resources, particularly stranded natural gas, and supportive infrastructure like the newly approved pipeline to Fargo. As noted in a post on X by investor Kevin O’Leary, the state is ‘perfectly positioned for AI infrastructure’ due to these assets. This deal isn’t isolated; it’s part of a broader trend where regional players are capitalizing on hyperscalers’ insatiable need for power and cooling, amid global data center demand projected to consume up to 420 terawatt-hours by 2030, per X discussions from users like Gen Z Investor.

Applied Digital’s strategy emphasizes cost efficiency. A June 2025 white paper released by the company, as reported in GlobeNewswire, highlights how site selection in North Dakota could save up to $2.7 billion in data center costs through optimized design and cooling. The Polaris Forge 2 campus boasts a projected power usage effectiveness (PUE) of 1.18, making it one of the most efficient facilities for AI factories.

Power and Water Challenges in AI Expansion

The project’s scale raises questions about resource demands. An article in GovTech details how Polaris Forge 2 is planned with careful consideration of water and power usage, securing energy through Cass County Electric Cooperative. Yet, as AI infrastructure booms, concerns over energy grids and water scarcity are mounting. Posts on X from users like Next 100 Baggers emphasize bottlenecks in power, cooling, and fiber optics, noting that hyperscalers could spend $7.9 trillion on data centers by 2030.

Applied Digital’s CEO, Wes Cummins, stated in the announcement, ‘We believe Polaris Forge 2 represents the next stage in Applied Digital’s rapid growth and our position as a leader in delivering high-performance AI infrastructure,’ as quoted in Data Center Dynamics. This sentiment echoes the company’s August 2025 plans for a $3 billion, 280-MW campus expansion, reported by CRN.

Economic Ripple Effects on North Dakota

The lease is poised to inject significant economic stimulus into the region. With projected revenues of $5 billion over 15 years, Applied Digital is not only bolstering its operations but also creating jobs and attracting investment. X posts from analysts like Shanu Mathew highlight how the deal elevates North Dakota’s total leased capacity to 600 MW, potentially transforming the local economy. As per Proactive Investors, this positions Applied Digital as a key player in the AI infrastructure race.

Beyond economics, the deal signals a ‘stealth land grab’ for compute resources. Industry insiders on X, such as Sam Badawi, discuss a ‘datacenter crunch’ where cloud providers like Amazon AWS and Oracle OCI are reserving long-term energy contracts. Applied Digital’s move could lock up 20-30% of U.S. data center capacity by 2026, per speculative analyses in various X threads, though exact figures remain estimates.

Strategic Implications for Hyperscalers

The unnamed hyperscaler’s involvement—described as investment-grade—suggests major players like Microsoft or Google are securing footholds in underutilized regions. This aligns with broader trends, as evidenced by Nebius AI’s $700 million raise for U.S. expansion, mentioned in X posts by Beth Kindig. Applied Digital’s focus on AI-optimized designs, including GPU clusters, addresses the frenzy around chips from Nvidia, without direct partnerships but through pragmatic infrastructure plays.

Critically, the lease includes provisions for expansion, with the hyperscaler able to scale to 1 GW total. As reported in StockTitan, initial phases are slated for 2026-2027, aligning with surging AI demands. This could alleviate supply chain bottlenecks, but it also intensifies competition for resources in critical sectors.

Future Horizons in AI Infrastructure

Looking ahead, Applied Digital’s North Dakota push may inspire similar developments elsewhere. X user George Koutalidis notes pipelines like Nebius’s 300 MW New Jersey campus, indicating a national scramble for AI capacity. The company’s report on $2.7 billion in savings, via Applied Digital’s IR site, positions North Dakota as an ‘epicenter’ for cost-effective AI builds.

However, challenges loom. Environmental impacts from high energy use are under scrutiny, with posts on X warning of global consumption spikes. Applied Digital’s efficient designs mitigate some risks, but as the industry scales, regulatory hurdles could emerge. This $5 billion lease isn’t just a deal—it’s a blueprint for how regional operators are reshaping the AI landscape, one megawatt at a time.

Beyond the Horizon: Industry Shifts

As AI evolves, deals like this highlight the shift from chip-centric narratives to infrastructure fundamentals. Insights from Quiver Quant project substantial revenue for Applied Digital, boosting its Nasdaq-listed stock by 7% post-announcement, as per CryptoRank. For industry insiders, this underscores the value of under-the-radar plays in a hyped market.

Ultimately, North Dakota’s emergence as an AI hub reflects broader geopolitical and economic dynamics, where access to power trumps proximity to tech hubs. With hyperscalers eyeing similar leases, the next few years could see a rebalancing of U.S. compute capacity, driven by pragmatic bets like Applied Digital’s.