Business Overview: Applied Digital Corporation (Nasdaq: APLD)

---

Executive Summary

Applied Digital Corporation designs, builds, and operates large-scale, purpose-built data center infrastructure, with a business increasingly centered on high-performance computing (HPC) and AI cloud infrastructure, alongside its legacy Bitcoin mining hosting operations. Applied Digital's core strategy is to develop energy-efficient, next-generation data centers — securing low-cost power and land — that can be leased or converted to serve either cryptocurrency mining hosting customers or, increasingly, GPU-based AI/HPC compute customers as demand for AI infrastructure has surged.

---

1. Core Business Model & How They Work

Applied Digital operates a data center infrastructure development and hosting model:

[ Site Selection & Power Procurement ] ➡️ [ Data Center Design & Construction ] ➡️ [ Hosting Capacity for Bitcoin Mining Customers or AI/HPC GPU Cloud Customers ] ➡️ [ Recurring Hosting/Cloud Revenue ] ➡️ [ Reinvest in Additional Capacity Buildout ]

Key Operational Drivers

1. Strategic Pivot to AI/HPC Infrastructure: Applied Digital has increasingly repositioned its data center development pipeline toward high-performance computing and AI cloud hosting (GPU infrastructure), a segment experiencing explosive demand growth as enterprises and AI labs seek data center capacity for model training and inference.
2. Low-Cost Power Sourcing: Because electricity is the dominant variable cost in both Bitcoin mining hosting and power-hungry AI/HPC infrastructure, securing access to low-cost, reliable power (often in regions with abundant or stranded energy resources) is a central competitive lever.
3. Long-Term Hosting/Lease Contracts: The AI/HPC infrastructure business model relies on securing longer-duration hosting or colocation contracts with enterprise and AI customers, offering more revenue visibility than the more volatile, Bitcoin-price-linked mining hosting business.
4. Capital-Intensive Buildout: Developing large-scale data center campuses requires substantial upfront capital for land, power infrastructure, and specialized cooling/electrical systems, financed through a mix of equity, debt, and project financing.

---

2. Business Segments

• Data Center Hosting (Bitcoin Mining) — legacy business hosting third-party Bitcoin mining hardware, providing power and facilities infrastructure in exchange for hosting fees.
• Cloud Services / AI & HPC Infrastructure — newer, faster-growing segment providing GPU-based cloud computing infrastructure and data center capacity for AI training/inference and other high-performance computing workloads.

---

3. Competitive Landscape

Competitors by Domain

AI / HPC Data Center Infrastructure

• Key Competitors: Large hyperscale cloud providers (Amazon Web Services, Microsoft Azure, Google Cloud) building their own AI infrastructure, alongside specialized neocloud/GPU infrastructure providers such as CoreWeave, Crusoe Energy, and other emerging AI data center developers, compete for the same enterprise and AI lab customer base and for scarce power/land resources.
• Dynamics: Demand for AI/HPC data center capacity has significantly outpaced available supply in recent years, benefiting well-positioned developers with access to power and land, but the space has also attracted enormous capital investment from both hyperscalers and specialized competitors, intensifying the race for power interconnection queues and construction capacity.

Bitcoin Mining Hosting

• Key Competitors: Core Scientific, Marathon Digital, and other large-scale hosting and self-mining operators compete for Bitcoin mining hosting customers and low-cost power sites.
• Dynamics: This legacy segment faces the same Bitcoin-price and network-difficulty cyclicality affecting the broader mining industry, which is part of the strategic rationale for Applied Digital's pivot toward the less commodity-price-exposed AI/HPC hosting business.

---

4. Strategic Strengths & Moats vs. Strategic Risks

Competitive Strengths (The Moat)

• Early-mover power site development: Securing power interconnection agreements and land for data center development can take years, and sites already in the development pipeline represent a real, time-based advantage over new entrants starting from scratch.
• Dual-use infrastructure flexibility: Facilities capable of serving either Bitcoin mining hosting or AI/HPC customers provide some flexibility to shift capacity toward the more attractive demand segment over time.
• First-mover positioning in the AI infrastructure supercycle: Establishing hosting relationships and long-term contracts with AI/HPC customers early in the current infrastructure buildout cycle can create durable, multi-year revenue visibility.

Strategic Risks & Vulnerabilities

1. Intense Capital Competition for Power and Sites: Hyperscalers and well-funded competitors are aggressively competing for the same scarce power interconnection capacity and suitable data center land, potentially compressing the advantage of any single developer over time.
2. Execution and Construction Risk: Large-scale data center buildouts carry significant execution risk around construction timelines, cost overruns, and power delivery delays.
3. Customer Concentration Risk: Revenue in the AI/HPC segment can be concentrated among a small number of large hosting/cloud customers, creating dependency risk if any single relationship is not renewed or scaled as expected.
   • Mitigation Strategy: Diversifying the customer base across multiple AI/HPC and enterprise hosting relationships as the segment scales.
4. Capital Structure / Financing Risk: The capital-intensive nature of data center development requires continued access to debt and equity financing, exposing the company to financing market conditions and potential dilution.

---

5. Financial Overview & Performance Matrix

---

6. Summary Conclusion

Applied Digital has repositioned itself from a Bitcoin mining hosting infrastructure provider into a participant in the broader AI and high-performance computing data center buildout, leveraging early data center site development and power procurement work to pursue the substantially larger and more secular AI infrastructure demand opportunity.

The company's long-term success depends on its ability to convert its data center development pipeline into signed, long-duration AI/HPC hosting contracts fast enough to justify its capital-intensive buildout, while competing against hyperscalers and well-capitalized specialized neocloud providers for the same scarce power and land resources central to the entire industry's growth.