Could Buying Applied Digital Today Set You Up for Life?

Applied Digital is repositioning from bitcoin hosting to purpose-built AI infrastructure and has secured multiyear revenue visibility via large hyperscaler deals: CoreWeave expanded its Polaris Forge 1 lease from 250MW to 400MW, lifting contract value from roughly $7bn to $11bn over 15 years and expected to generate about $500m of annual NOI, while Polaris Forge 1 is slated to expand beyond 1GW (2028–2030) as transmission capacity arrives. The company is also constructing Polaris Forge 2 (initial 300MW, scalable to 1GW) and has pre-leased 200MW to an unnamed U.S. hyperscaler for $5bn over 15 years, positioning it to capture a slice of an estimated ~$350bn hyperscaler AI data-center buildout in 2025. These agreements give Applied Digital significant runway, but the thesis hinges on capital-intensive expansion, execution of large-scale builds and concentration risk from a small number of hyperscaler customers.

Analysis

Applied Digital is repositioning from bitcoin-hosted data centers to purpose-built AI infrastructure to capture demand where the current bottleneck is facility power density and advanced liquid-cooling rather than GPUs; the company says it has assembled land, power, construction crews, supply-chain contracts and engineering talent to build these "AI factories." With public hyperscalers estimated to invest about $350 billion in AI data centers in 2025, Applied Digital's infrastructure focus directly targets a large, addressable buildout opportunity.

CoreWeave expanded its Polaris Forge 1 lease from 250MW to the full 400MW, increasing contract value from roughly $7 billion to $11 billion over 15 years, and management expects that lease to contribute nearly $500 million in annual net operating income. Applied Digital plans Polaris Forge 1 capacity to exceed 1GW between 2028 and 2030 as transmission capacity arrives; Polaris Forge 2 is being built initially at 300MW and scalable to 1GW, with 200MW pre-leased to an unnamed U.S. hyperscaler for $5 billion over 15 years.

These agreements provide multiyear revenue visibility and material runway, but the thesis is contingent on capital-intensive expansions, timely delivery of additional transmission and construction, and customer concentration with a small number of large hyperscaler contracts. Execution risk, power-delivery schedules and the company's ability to convert contracted value into realized cash flow will be the primary drivers of upside or downside from current expectations.

Key Decisions for Investors

• Consider a measured long-term exposure to APLD to play AI data-center demand given multiyear contracts (CoreWeave expansion to 400MW and ~$11bn over 15 years, plus expected ~$500m annual NOI), but size positions conservatively because of high capex and customer concentration
• Monitor near-term catalysts before adding: CoreWeave NOI realization, progress on Polaris Forge 1 transmission and the 2028-2030 expansion timeline, and construction/lease progress at Polaris Forge 2 (200MW pre-lease for $5bn over 15 years)
• Use staged-buying or position-trimming tied to execution milestones (completed capacity coming online and recognized revenue) rather than binary conviction today
• If risk-averse, wait for broader customer diversification beyond the few large hyperscalers or for clearer cash-flow conversion from the announced contracts