Applied Digital designs, builds, and operates specialized data centers, called AI factories, for high-performance computing and artificial intelligence. Its primary revenue comes from HPC hosting for tier-1 hyperscalers like CoreWeave, which accounts for a significant portion of its $36 billion contracted revenue. A smaller segment provides infrastructure for crypto mining, and its majority-owned subsidiary, ChronoScale, offers GPU cloud services, including a 50MW deployment with Microsoft.
Key Inputs And Sourcing
1. Tenant Fit-Out Services (Components & Labor)
Not in registry
other · North America, Global · ~71.6% of Services and other cost of revenues
Source Explicitly stated as $176.1 million out of $245.7 million in 'services and other cost of revenues' for HPC hosting business. This includes various components (electrical, cooling, networking) and labor for data center build-out. CEO mentions locking in 'key components' and 'full factory capacity output' from certain factories for a 4-year timeframe.
Confidence: high
2. Energy (Electricity/Natural Gas)
Not in registry
energy · natural-gas · North America (North Dakota, Louisiana, Alabama), Finland · ~8.1% of Services and other cost of revenues
Source Explicitly stated as $19.8 million in 'energy cost' within 'services and other cost of revenues'. Power is called the 'single largest gating factor in our industry'. APLD has a long-term power purchase agreement for a 1.2 GW natural gas facility in North Dakota.
Confidence: high
3. Data Center Electrical Components (e.g., Transformers, Switchgear)
Not in registry
component · Global · unknown
Source Implied by 'well honed supply chain' and locking in 'key components' for data center builds. Substation and grid upgrades are mentioned as significant investments. Electrical systems are crucial for data centers.
Confidence: medium
4. Data Center Cooling Systems (e.g., Liquid Cooling Infrastructure)
Not in registry
component · Global · unknown
Source APLD's facilities are 'liquid cooled'. CEO mentions cooling as a key design aspect for future-proofing.
Confidence: medium
5. Networking Equipment (Fiber, Switches)
Not in registry
component · Global · unknown
Source Fiber is mentioned as an established local supply chain component. Essential for data center operations and connectivity.
Confidence: low
Industry Publications
- Data Center Frontier (datacenterfrontier.com) — Dedicated to news and analysis of the data center industry, covering hyperscale, colocation, and enterprise data centers, power, cooling, and AI infrastructure.
- Mission Critical Magazine (missioncriticalmagazine.com) — Focuses on design, construction, and operation of mission-critical facilities, including data centers, relevant to APLD's 'AI factories'.
- DCD (Data Center Dynamics) (datacenterdynamics.com) — Provides global news, analysis, and events for the data center industry, covering technology, infrastructure, and market trends.
- Power Engineering (power-eng.com) — Covers power generation, transmission, and distribution, which is critical given APLD's focus on securing large-scale power for its campuses and partnerships like Base Electron.
- T&D World (tdworld.com) — Focuses on the transmission and distribution of electricity, directly relevant to APLD's need for grid interconnection and upgrades for its data center campuses.
Economic Data Watch
1. U.S. Census Bureau — Value of Construction Put in Place
Not in registryaccess=api
Metric/field CPICN
Cadence monthly
Why it matters Tracks overall nonresidential construction spending, a key demand driver for data center buildouts and industrial infrastructure relevant to APLD's expansion.
Signal to watch Sustained growth in nonresidential construction spending is bullish, indicating a healthy environment for data center development.
Confidence: high
2. Federal Reserve Board — Industrial Production and Capacity Utilization
Not in registry
Metric/field INDPRO
Cadence monthly
Why it matters Indicates the health of the manufacturing sector, which affects the supply and cost of materials and equipment for data center construction.
Signal to watch Increasing industrial production suggests a more robust supply chain and potentially stable or lower input costs for APLD's builds.
Confidence: high
3. U.S. Department of the Treasury — Market Yield on U.S. Treasury Securities at 10-Year Constant Maturity
Not in registry
Metric/field DGS10
Cadence daily
Why it matters APLD's financing costs for project-level debt and corporate debt are sensitive to interest rates. Lower rates support more attractive terms.
Signal to watch Declining or stable 10-year Treasury yields are bullish for APLD's ability to secure favorable financing for its campuses.
Confidence: high
4. EIA — Average Retail Price of Electricity
Not in registryaccess=api
Metric/field ELEC.PRICE.ND-ALL.M
Cadence monthly
Why it matters Power cost is a significant operational expense for data centers, especially in APLD's core North Dakota market.
Signal to watch Stable or decreasing electricity prices in key operating regions like North Dakota are bullish for APLD's operating margins.
Confidence: medium
5. Trading Economics — Finland Industrial Production
Not in registryaccess=scrape
Metric/field Finland Industrial Production
Cadence monthly
Why it matters APLD is expanding into Finland; industrial production can be a proxy for local economic activity and potential demand for data center services, as well as supply chain health.
Signal to watch Growth in Finland's industrial production could signal a more robust economic environment supporting APLD's European expansion.
Confidence: high
Free Alt Data Watch
1. FERC / MISO — MISO Interconnection Queue
Not in registry
Metric/field queue_status:data_center_projects
Cadence irregular
Why it matters Tracks new data center projects seeking grid connections, indicating competitive landscape and future supply in APLD's key operating regions.
Signal to watch Fewer new large-scale data center projects entering the queue in APLD's regions is bullish, suggesting less competition for power and land.
Confidence: medium
2. Google Trends — Web Search Interest
Not in registry
Metric/field search_interest:AI data center buildout
Cadence weekly
Why it matters Indicates general public and industry interest in AI data center construction, reflecting underlying demand for APLD's services.
Signal to watch Sustained or increasing search interest suggests continued high demand for AI infrastructure, benefiting APLD.
Confidence: medium
3. SEC EDGAR — Form 8-K Filings
Not in registry
Metric/field filing_description:new lease agreement
Cadence event_driven
Why it matters APLD frequently announces new customer contracts, campus expansions, and financing milestones via 8-K filings, which are critical for revenue visibility and growth.
Signal to watch Timely announcements of new hyperscaler leases, expansion agreements, or favorable financing indicate strong execution and future revenue growth.
Confidence: high
4. North Dakota Public Service Commission — Utility Dockets
Not in registry
Metric/field docket_status:power_plant_approvals
Cadence irregular
Why it matters Tracks regulatory approvals for new power generation and infrastructure, which is crucial for APLD's ability to secure power for its expanding campuses, especially with partners like Base Electron.
Signal to watch Expeditious approvals for relevant power projects are bullish, ensuring APLD's long-term power supply and expansion capabilities.
Confidence: medium
5. APLD Investor Relations — Press Releases / Newsroom
Not in registry
Metric/field press_release_content:construction_milestone
Cadence event_driven
Why it matters Direct source for company-announced construction progress, operational milestones, and new partnerships, which are key indicators of execution.
Signal to watch Consistent announcements of construction milestones met on schedule and new capacity coming online are bullish, demonstrating strong execution.
Confidence: high
Paid Alt Data Watch
1. Revelio Labs — Job Postings
Not in registryaccess=file
Metric/field job_postings_count:Data Center Construction Manager
Cadence monthly
Why it matters A leading indicator of project execution and capacity ramp-ups in the data center sector, directly correlating with APLD's ability to build and deliver.
Signal to watch Increasing job postings for specialized construction and engineering roles by APLD and across the industry suggest accelerating buildouts and strong demand.
Confidence: high
2. datacenterHawk — Data Center Pricing Benchmark
Matcheddatacenterhawk_data_center_pricing_benchmark · access=file · map_only
Metric/field lease_rate_per_MW:North_Dakota_Hyperscale
Cadence quarterly
Why it matters Provides insights into market pricing for data center capacity, directly impacting APLD's revenue potential from new leases and expansions.
Signal to watch Rising lease rates for hyperscale capacity in APLD's operating regions are bullish, indicating pricing power and higher potential margins.
Confidence: high
3. Industrial Info Resources — North American Data Center Project Database
Not in registry
Metric/field project_status:APLD_campuses
Cadence monthly
Why it matters Offers detailed tracking of industrial projects, including data centers, providing independent verification of APLD's construction progress and competitive activity.
Signal to watch Positive updates on APLD's project statuses (e.g., moving from planning to construction, reaching completion) and limited delays are bullish.
Confidence: high
4. CloudPrice — GPU Cloud Pricing
Matchedcloudprice_cloud_compute_pricing_technology · access=file · map_only
Metric/field GPU_rental_rate:NVIDIA_GV300N
Cadence daily
Why it matters Monitors the pricing environment for GPU compute, which is the underlying demand driver for APLD's AI factories and ChronoScale's business.
Signal to watch Stable or increasing GPU rental rates suggest robust demand for AI compute, translating to strong demand for APLD's infrastructure.
Confidence: high
5. Similarweb — Website Traffic Analytics
Matchedsimilarweb_website_traffic_engagement · access=file · map_only
Metric/field traffic_volume:coreweave.com
Cadence monthly
Why it matters Provides a proxy for the business activity and growth of APLD's major customers like CoreWeave, indicating their demand for compute capacity.
Signal to watch Sustained growth in website traffic for APLD's key customers is bullish, suggesting increasing demand for their services and thus for APLD's infrastructure.
Confidence: medium
Search Keywords Brand Product
- AI factories
- HPC hosting
- data center hosting
- ChronoScale
- AI infrastructure
- high-performance computing
- data center development
- hyperscale data centers
- GPU cloud
- power purchase agreement
- digital infrastructure
Search Keywords Event Phrases
- APLD earnings
- Applied Digital Finland expansion
- CoreWeave lease
- Base Electron PPA
- What They Do (Plain English & Analogies)
- Applied Digital (APLD) designs, builds, and operates massive data centers, which they call 'AI factories,' specifically tailored for the intense power and cooling needs of artificial intelligence (AI) and high-performance computing (HPC) workloads. Think of them as specialized landlords for the AI industry: they provide the land, the buildings, the power, and the cooling infrastructure, and then lease this 'AI factory' space to large tech companies (hyperscalers) that run powerful AI models. They also have a smaller business hosting cryptocurrency mining and own a significant stake in ChronoScale, a company that provides GPU cloud computing services, effectively reaching all the way to the actual AI compute hardware.
- Very Brief History
- Applied Digital Corporation was formerly known as Applied Blockchain, Inc. and officially changed its name to Applied Digital Corporation in November 2022 to reflect its pivot towards high-performance computing and AI. The company began signing significant leases for its AI factory campuses, notably securing its first major lease with CoreWeave in May 2025 for its Polaris Forge 1 campus in North Dakota, rapidly expanding its footprint in the AI infrastructure market.
- "Street Stereotype"
- Applied Digital is generally perceived on the street as a key 'AI factory' landlord, offering rare early-stage exposure to long-duration, contracted AI compute demand. Investors see it as a scarce execution play in AI infrastructure, validated by significant long-term leases with top-tier AI customers like CoreWeave and strategic financing structures. The market focuses on its ability to build out large-scale AI data centers rapidly and efficiently.
- Subsidiaries On Linked In*
- ChronoScale — Applied Digital holds an approximately 96% ownership stake.; LinkedIn: chronoscale-inc
- Customer Sectors & Example Clients
- Applied Digital's primary customer sectors are high-performance computing (HPC) and artificial intelligence (AI) industries. Their top client is CoreWeave, an AI hyperscaler. Through its subsidiary ChronoScale, they also serve clients like Microsoft for AI compute deployments. They target other 'tier-1 investment-grade hyperscalers' for their large-scale AI factory campuses.
- New Customers / Segments They'Re Targeting
- Applied Digital is actively targeting additional tier-1 investment-grade hyperscalers for new and expanded campuses, particularly in North America. They are also selectively pursuing opportunities in new international markets, with Finland being their first step into Europe, where they are marketing a gigawatt-scale site to leading global hyperscale customers. Furthermore, their ChronoScale subsidiary is expanding its GPU cloud computing business, securing capacity at smaller, more distributed sites (50-100 MW) that may have been overlooked by larger-scale developers, targeting the growing 'neocloud' segment.
- Sales Geographies And Expansion Plans
- Currently, Applied Digital primarily operates in North America, with significant development and operational activities in North Dakota (Ellendale, Harwood, Center), Louisiana, and Alabama. The company has recently taken its first step outside the United States by securing an agreement for up to 1 gigawatt of potential power capacity in Finland, with initial power availability anticipated in 2028. While their near-term focus remains on domestic projects, Finland represents a measured long-term expansion into attractive European markets.
- How Key Themes May Help/Hurt
- Applied Digital is strongly positioned to benefit from the 'AI '25: Data Center Buildout' theme. The accelerating global construction of hyperscale data centers, driven by AI, directly increases demand for APLD's 'AI factories.' Development constraints like limited powered land, lengthy permitting processes, and skilled labor shortages favor established operators like APLD, who have already secured sites, power, and community support, leading to stronger pricing power and project backlogs. The trend towards larger and more complex facilities with higher rack densities and advanced cooling requirements also increases the value of APLD's specialized design and construction capabilities. Conversely, the company could be hurt if an AI capital expenditure slowdown occurs, reducing demand for new facilities. While APLD actively mitigates power availability and permitting delays through strategic partnerships and established campuses, broader industry challenges could still create headwinds. Overbuilding in the data center market could also lead to pricing pressure, though APLD's focus on long-term, premium-priced contracts with tier-1 clients aims to counter this.
3 Main Long-Term Bull Details
- Massive Contracted Revenue & Hyperscaler Validation: Applied Digital boasts approximately $36 billion in contracted revenue across five campuses, primarily from long-term (15-year take-or-pay) leases with tier-1 investment-grade hyperscalers like CoreWeave. This provides exceptional revenue visibility and validates their platform's appeal to leading AI customers.
- Differentiated, Future-Proofed AI Factory Design: The company's proprietary data center design is built for a 30+ year horizon, emphasizing flexibility and adaptability to evolving compute architectures, network configurations, and electrical/mechanical requirements. This long-term asset strategy is expected to generate outsized returns and support continuous technology refreshes within the same facilities.
- Strategic Power Access & Scarcity Value of Campuses: APLD's proactive approach to securing power, exemplified by its 1.2 gigawatt power purchase agreement with Base Electron in North Dakota, and its focus on developing established, community-supported campuses, creates significant barriers to entry. This scarcity value is expected to drive premium pricing for expansions and renewals, enhancing the long-term terminal value of their assets.
3 Main Long-Term Bear Details
- High Customer Concentration Risk: A significant portion of Applied Digital's contracted revenue, approximately 30% of the $36 billion, is tied to CoreWeave. While CoreWeave is a top-tier AI customer, this concentration exposes APLD to risks associated with a single client's financial health or strategic shifts.
- Thin Profitability & High Leverage: Despite strong revenue growth, the company reported an adjusted net loss in the recent quarter, and its GAAP net loss was substantial due to non-cash items like stock-based compensation and warrant revaluations. The company also carries significant debt, approximately $6.4 billion, which, while largely project-level and long-dated, represents considerable leverage for a company of its market capitalization.
- Execution and Timing Risks in Rapid Expansion: Applied Digital is undertaking a rapid and ambitious expansion, expecting to place over 600 megawatts into service in the next 12 months. This aggressive build-out across multiple campuses simultaneously carries inherent execution and timing risks, including potential supply chain constraints, construction delays, and the challenge of consistently bringing contracted capacity online on schedule and within budget.
- Competitors And Differentiation
- Applied Digital competes with other data center developers and operators, particularly those focused on high-performance computing and AI workloads, as well as hyperscalers that build their own facilities. Their differentiation strategy centers on several key aspects: a proprietary and proven data center design built for a 30+ year lifespan, offering substantial adaptability to evolving compute architectures and cooling requirements; a well-honed supply chain and deep energy partnerships; a demonstrated ability to deliver on lease commitments; and a focus on securing dedicated power resources (like the 1.2 GW PPA with Base Electron) to mitigate industry-wide power constraints. They also emphasize their 'franchise model' of building large, sustainable AI factories and securing durable, high-quality, long-term (15+ year take-or-pay) contracts with tier-1 investment-grade hyperscalers. The company also highlights the increasing scarcity value of its established, powered, and community-supported campuses, which creates significant barriers to entry for competitors.
- Recent Performance & What The Market'S Focused On
- Applied Digital reported strong fiscal first-quarter 2027 results, with total revenue surging 322% year-over-year to $341.9 million and adjusted EBITDA reaching $64.4 million, up significantly from the prior year. However, the company also reported a net loss from continuing operations of $221 million, largely influenced by non-cash stock-based compensation and warrant valuation changes. The market is currently focused on APLD's ability to continue executing its ambitious build-out plan, particularly bringing over 600 megawatts online in the next 12 months. Key areas of investor attention include securing favorable financing for upcoming campuses (Polaris Forge 3, Delta Forge 1 & 2) with investment-grade hyperscalers, the potential for materially higher pricing on new and expansion leases, and the growth trajectory of its ChronoScale subsidiary, which is targeting $1 billion in annual recurring revenue by 2027. The strategic value and scarcity of APLD's established, powered campuses are also central to market discussions.
- Revenue Segments And Estimated Mix
- HPC Hosting Business — Mix: ~76.8%; Source: Q1 FY27 transcript; Trend: Generated $262.6 million of revenue, including $183.5 million from tenant fit-out services and $65.8 million of base rent.
- Data Center Hosting Business (Bitcoin mining) — Mix: ~11.1%; Source: Q1 FY27 transcript; Trend: Generated $37.8 million in revenue, in line with the prior year quarter, providing steady high-margin revenue with minimal ongoing capital investment.
- ChronoScale — Mix: ~12.1%; Source: Q1 FY27 transcript; Trend: Generated $41.5 million in revenue, including $23 million of GPU hardware sales, and is on track to reach $1 billion of ARR in 2027.
- Product Brands
- {"brands":[]}