DALLAS, April 27, 2026, 09:01 CDT
Applied Digital Corp. has secured a 15-year lease worth around $7.5 billion with an undisclosed U.S. hyperscaler, locking in 300 megawatts of critical IT load at its Delta Forge 1 campus. With this deal, the company’s total contracted lease revenue now tops $23 billion. Details came out in an April 23 filing and press release.
Timing is key here: building out AI data centers has become a contest not just of software, but of power and cooling muscle. The hyperscalers — those massive cloud firms with bottomless appetite for compute and storage — need locations that can handle racks packed with hot, power-hungry chips. Reuters points out that Amazon, Google (part of Alphabet), and Microsoft are all pouring money into securing that critical capacity.
Applied Digital doesn’t see instant revenue from the lease. The timeline puts Delta Forge 1’s first phase in mid-2027, so the company now faces the challenge: convert those locked-in contracts into operational capacity, and do it on time.
Applied Digital has landed a second U.S. investment-grade hyperscaler as a tenant, this time at one of its three AI Factory campuses. That makes three hyperscale customers in total for the company. An “investment grade” label signals solid credit—a make-or-break factor for data-center operators, who often rack up debt long before new facilities begin to generate consistent revenue. GlobeNewswire
The stock surged over 12% at the open following Thursday’s announcement. By Monday, shares were last seen at $35.09, just 0.3% higher, putting the Dallas-based firm’s market capitalization around $9.9 billion.
Chairman and CEO Wes Cummins said, “We remain focused on delivering operational AI capacity at scale.” Execution is still front and center, he noted, pointing to the importance of getting new capacity up and running on time. GlobeNewswire
According to the company, Delta Forge 1 covers over 500 acres in the southern U.S. and targets large-scale AI tasks. The site is set up for both training and inference—the latter refers to putting trained AI models to work, generating outputs like answers or images for users.
Investors have zeroed in on the lease pipeline following Applied Digital’s latest numbers. The company logged fiscal third-quarter revenue of $126.6 million, a 139% jump from last year. Still, net loss attributable to common stockholders grew to $100.9 million. Adjusted EBITDA landed at $44.1 million. As of quarter-end, Applied Digital held $2.1 billion in cash, cash equivalents and restricted cash, set against $2.7 billion in debt.
In a Saturday piece, Motley Fool’s Harsh Chauhan highlighted Applied Digital’s ongoing AI data-center builds as fuel for a possible uptick in growth once these come online. Chauhan noted the company is working on 400 MW in data-center capacity for CoreWeave, plus 200 MW more for a different hyperscaler. Of that, 100 MW for CoreWeave is already done.
The new lease ups the company’s funding requirements. Applied Digital is planning to secure as much as $300 million through a senior secured bridge facility to keep Building 3 at Polaris Forge 1 moving forward. Another senior secured revolving credit facility—also up to $300 million—is in the works, earmarked for development, working capital, and handling transaction costs.
The environment is still fiercely competitive. Applied Digital hasn’t disclosed the customer behind its latest deal, which arrives as big cloud players scramble for data-center capacity outfitted to handle AI—power, cooling, the works. The same rush is drawing in specialized AI infrastructure firms and some heavyweight data-center landlords as well.
Still, plenty could go wrong. Applied Digital has flagged a host of potential pitfalls: construction setbacks, customer lease choices, swings in AI and high-performance computing demand, power issues, equipment breakdowns, regulatory hurdles, tight cash flow, and funding access—all could alter the picture. Plus, those credit facilities? They’re not finalized yet.
Applied Digital locks in a fresh long-term revenue stream linked to AI infrastructure with this deal. The bigger unknown: can the company actually turn contracted megawatts into live, operating data centers quickly enough to back up the market’s renewed interest?
