$2.7 Billion in Data Center Savings: New Applied Digital Report Positions North Dakota as the Epicenter for AI Infrastructure
Rhea-AI Summary
Positive
- Potential cost savings of up to $2.7 billion over 30 years through strategic site selection and design
- Secured a major 15-year, $7 billion lease agreement with CoreWeave for 250MW
- Superior efficiency metrics with projected PUE of 1.18 and near-zero WUE
- Access to gigawatt-scale stranded power and 220+ days of free cooling annually
- 400 MW critical IT load capacity with over 1 gigawatt in load study
Negative
- AI factories require 15-30 times more power density than traditional data centers, increasing operational complexity
- Significant upfront investment required for infrastructure development
- Geographic concentration risk in North Dakota region
Insights
Applied Digital's North Dakota strategy delivers compelling cost advantages for AI infrastructure, validated by a major $7B CoreWeave deal.
Applied Digital's white paper represents a significant strategic positioning in the rapidly evolving AI infrastructure landscape. The company has identified a powerful competitive advantage by locating high-density computing facilities in North Dakota - an unconventional but increasingly logical choice driven by two critical factors: power costs and cooling efficiency.
The financial implications are substantial. The projected
Their Polaris Forge model demonstrates exceptional technical metrics, particularly the projected PUE of 1.18, which places them among the most efficient data centers globally. For context, traditional data centers typically operate at PUEs between 1.5-2.0, while even efficient hyperscale facilities struggle to consistently achieve below 1.25 in most climates.
The
Applied Digital's approach addresses one of the most pressing constraints in AI deployment today - the availability of power-efficient, scalable infrastructure. By establishing gigawatt-scale capacity in regions with stranded power assets, they're positioning themselves to capitalize on both the immediate AI compute shortage and the long-term shift toward more sustainable, cost-effective AI infrastructure.
Polaris Forge “AI Factory” model outlines how smart site selection and cooling design can cut costs by up to
DALLAS, June 17, 2025 (GLOBE NEWSWIRE) -- Applied Digital Corporation (Nasdaq: APLD) (“Applied Digital” or the “Company”), a designer, builder, and operator of next-generation digital infrastructure designed for High-Performance Computing (HPC) applications, announces the release of a new white paper, AI Factory: A Case Study for Total Cost of Ownership. The paper details how site selection and data center design decisions can dramatically reduce the long-term costs of generative AI infrastructure.
“AI factories represent a fundamentally new category of digital infrastructure,” said Wes Cummins, Chairman and CEO of Applied Digital. “Their success hinges on making the right decisions up front, especially around power and cooling. This white paper helps lay out what those decisions look like and why regions like North Dakota can outperform conventional markets on both cost and sustainability.”
Applied Digital also introduced Polaris Forge as the name for its North Dakota-based data center region. With four campuses in various stages of development, Polaris Forge is built for scale, optimized for AI workloads, driven by renewable power and supported by infrastructure that prioritizes total cost of ownership.
Key findings from the white paper include:
- AI factories require 15-30 times the power density of traditional data centers, driving the need for new power and cooling strategies.
- Site selection directly impacts cost; choosing areas with stranded power and cooler climates can reduce annual electricity costs by
$50 t o$60 million per year compared to other existing 100MW data centers, or up to$2.7 billion over a 30-year lifespan. - Liquid cooling and free cooling are essential to long-term efficiency, enabling significantly lower Power Usage Effectiveness (PUE) and Water Usage Effectiveness (WUE).
- North Dakota offers more than 220 days of free cooling annually, which contributes to lower operational costs and improved sustainability.
- Polaris Forge 01’s design, featuring a closed-loop, waterless, direct-to-chip cooling system and access to gigawatt-scale stranded power demonstrates how infrastructure purpose-built for AI can achieve a projected PUE of 1.18 and a WUE near zero.
The company’s existing Ellendale campus—now designated Polaris Forge 01—demonstrates how a purpose-built AI factory can outperform traditional markets. Applied Digital designed Polaris Forge 1 with speed and efficiency in mind, selecting a location that offers access to abundant stranded power, a favorable climate for free cooling, and a strong foundation for long-term capacity expansion. Built to support 400 MW of critical IT load, with over 1 gigawatt in load study, Polaris Forge 01 positions both the region and Applied Digital as leaders in AI infrastructure.
In addition to its infrastructure development, Applied Digital continues to play a key role in supporting local economies and workforces through its Community and Economic Development Initiatives. The company’s approach to integrating community priorities, such as workforce housing and job creation, into its data center expansion strategy has helped it scale responsibly while maintaining strong local support.
“With Polaris Forge, we’re building something that’s efficient, scalable and community-focused,” added Cummins. “We believe AI infrastructure can thrive in places like North Dakota—not in spite of their location, but because of it.”
Applied Digital recently signed a 250MW 15-year lease agreement, worth
To learn more about how site selection and design can impact long-term AI infrastructure costs, and why Applied Digital is betting big on regions like North Dakota, download the full white paper.
About Applied Digital
Applied Digital (Nasdaq: APLD) develops, builds and operates next-generation data centers and cloud infrastructure. Different by design, the company’s purpose-built facilities are engineered to unleash the power of accelerated compute and deliver secure, scalable and sustainable digital hosting, along with turnkey CSaaS and GPU-as-a-Service solutions. Backed by deep hyperscale expertise and a robust pipeline of available power, Applied Digital accommodates AI Factories and beyond to support the world’s most exacting AI/ML, blockchain and high-performance computing (HPC) workloads.
CAUTION ABOUT FORWARD-LOOKING STATEMENTS
This press release contains “forward-looking statements” as defined in the Private Securities Litigation Reform Act of 1995 regarding, among other things, future operating and financial performance, product development, market position, business strategy and objectives and future financing plans. These statements use words, and variations of words, such as “will,” “continue,” “build,” “future,” “increase,” “drive,” “believe,” “look,” “ahead,” “confident,” “deliver,” “outlook,” “expect,” “project” and “predict.” Other examples of forward-looking statements may include, but are not limited to, (i) statements that reflect perspectives and expectations regarding the data center campus development, (ii) statements about the high-performance computing (HPC) industry, (iii) statements of plans and objectives, including an evolving business model, or estimates or predictions of actions by suppliers, (iv) statements of future economic performance, and (v) statements of assumptions underlying other statements and statements about the Company or its business. You are cautioned not to rely on these forward-looking statements. These statements are based on current expectations of future events and thus are inherently subject to uncertainty. If underlying assumptions prove inaccurate or known or unknown risks or uncertainties materialize, actual results could vary materially from our expectations and projections. These risks, uncertainties, and other factors include: our ability to complete construction of the Ellendale HPC data centers; changes to AI and HPC infrastructure needs and their impact on future plans; risks associated with the leasing business, including those associated with counterparties; costs related to the HPC operations and strategy; our ability to raise additional capital to fund ongoing and future data center construction and operations; our ability to obtain financing of the lease agreements on acceptable financing terms, or at all; the inability to comply with regulations, developments and changes in regulations; cash flow and access to capital; availability of financing to continue to grow our business; decline in demand for our products and services; maintenance of third party relationships; and conditions in the debt and equity capital markets. A further list and description of these risks, uncertainties and other factors can be found in the Company’s most recently filed Annual Report on Form 10-K and Quarterly Report on Form 10-Q, including in the sections captioned “Forward-Looking Statements” and “Risk Factors,” and in the Company’s subsequent filings with the Securities and Exchange Commission. Copies of these filings are available online at www.sec.gov, on the Company’s website (www.applieddigital.com) under “Investors,” or on request from the Company. Information in this release is as of the dates and time periods indicated herein, and the Company does not undertake to update any of the information contained in these materials, except as required by law.
Media Contact: Jaymie Scotto & Associates jsa_applied@jsa.net
FAQ
What cost savings does Applied Digital's Polaris Forge data center project in North Dakota offer?
What is the size of Applied Digital's (APLD) recent lease agreement with CoreWeave?
What are the efficiency metrics for Applied Digital's Polaris Forge 01 facility?
Why did Applied Digital choose North Dakota for their AI data centers?
What is the power capacity of Applied Digital's Polaris Forge 01 facility?
