The race for artificial intelligence dominance is no longer just about who has the best algorithm; it’s about who has the power, the cooling, and the physical space to run it. Building the infrastructure to support next-generation AI is one of the most capital-intensive and logistically challenging endeavors in modern tech history.
Applied Digital (NASDAQ: APLD) proved that it is firmly in the driver’s seat of this AI infrastructure boom.
In a blockbuster announcement, the data center developer and operator revealed a substantial 15-year, take-or-pay lease agreement at its fourth AI Factory campus, Polaris Forge 3. The deal, signed with the same U.S.-based, high investment-grade hyperscaler that recently partnered with them at their Delta Forge 1 site, pushes Applied Digital past 1 Gigawatt (GW) of total contracted capacity.
The Numbers Behind Polaris Forge 3
Data center agreements of this magnitude are rare, and the financial metrics attached to this lease are staggering:
- The 15-year lease represents $7.5 billion in guaranteed, take-or-pay contracted revenue.
- If the hyperscaler exercises all renewal options, the value of this single campus agreement could increase to $18.2 billion.
- Located in a northern state, Polaris Forge 3 spans over 600 acres. It is designed to deliver 300 MW of critical IT load, backed by a whopping 430 MW of grid-connected utility power.
The Big Picture: $73 Billion in Cumulative Potential Revenue
While the Polaris Forge 3 deal is significant on its own, it is part of Applied Digital’s rapid scaling effort. By repeating its disciplined AI Factory model across four distinct campuses (Polaris Forge 1, 2, 3, and Delta Forge 1), the company has built a powerhouse portfolio in record time.
With this latest agreement, Applied Digital’s total contracted baseline revenue across its footprint climbs to $31 billion. If all options and renewals are exercised across its campuses, that total potential revenue jumps to $73 billion.
Additionally, approximately 65% of this total contracted revenue is backed by top-tier, U.S.-based investment-grade hyperscalers. For investors and industry onlookers, this provides Applied Digital with an incredible level of long-term revenue visibility and financial stability.
Why Hyperscalers Keep Coming Back: The AI Factory Model
In the tech world, reliability is everything. The fact that this lease is with the same blue-chip hyperscaler that signed with Applied Digital just a month prior at Delta Forge 1 is a resounding vote of confidence.
As Applied Digital Chairman and CEO Wes Cummins pointed out, this repeat business is a direct result of execution:
“Polaris Forge 3 is a direct extension of what we’ve proven works: a disciplined, repeatable AI Factory model that delivers large-scale capacity to the world’s most demanding compute customers. This second 300 MW lease…reflects the confidence we’ve built through disciplined execution and our ability to consistently advance large-scale AI infrastructure projects. We’ve earned a seat at the table with blue-chip customers, and we intend to keep that position through flawless execution and long-term operational reliability.”
Polaris Forge 3 is engineered from the ground up specifically for high-density AI workloads. The site will feature Applied Digital’s proprietary waterless cooling technology, advanced liquid-cooling architecture, and ultra-high-density power delivery systems.
What’s Next for Applied Digital?
Initial operations at Polaris Forge 3 are expected to kick off in August 2027, but the company isn’t hitting the brakes anytime soon.
While locking down 1.2 GW of contracted capacity over the past 11 months is a historic feat, Cummins teased that the company is actively marketing more than 1.7 GW of additional grid-connected utility power across its newly expanded footprint of AI Factory sites.
As tech giants continue to pour billions into AI development, the bottleneck will inevitably remain power and infrastructure. With over 1 GW already locked down and another 1.7 GW in the pipeline, Applied Digital has cemented itself as the backbone of the AI revolution.
To read the full press release, click here.



