Applied Digital (APLD) Q2 2024: Cloud Contracted ACV Climbs to $398M Amid GPU Supply Delays
Applied Digital’s cloud service annual contract value surged to $398 million, reflecting robust AI-driven demand, but GPU cluster deployments lagged due to networking supply chain bottlenecks. Management reaffirmed surging enterprise and VC-backed customer interest in high-density compute, while highlighting supply and grid constraints as near-term gating factors. With data center capacity approaching 500 megawatts and a pipeline of 400 megawatts in development, APLD is positioning for outsized growth as AI infrastructure demand outpaces supply.
Summary
- AI Compute Demand Outpaces Supply: Cloud contract value reached $398 million as enterprise and startup customers accelerate GPU adoption.
- Operational Bottlenecks Persist: Networking component shortages delayed cluster revenue ramp despite strong customer pipeline.
- Power and Land Pipeline Expands: 400 megawatts in development positions APLD to capture future AI infrastructure scarcity.
Business Overview
Applied Digital (APLD) operates as a digital infrastructure provider, specializing in data center hosting and high-performance computing (HPC) services for blockchain mining and AI workloads. Revenue streams include multi-year hosting agreements for Bitcoin miners and cloud services contracts for GPU compute capacity, with major facilities in North Dakota and Texas. The business is structured around three segments: data center hosting (providing space and power for customer-owned hardware), cloud services (offering GPU clusters for AI clients), and HPC data centers (purpose-built for high-density compute).
Performance Analysis
Revenue growth was driven by the full ramp of the Ellendale and Jamestown hosting facilities, as well as initial contributions from the Garden City site and the first cloud service contract. The company’s total hosting capacity reached 280 megawatts across North Dakota, while the Texas-based Garden City facility began energization, contributing 132 megawatts this quarter and aiming for full 200 megawatts by April. Cloud services continued to scale, with annual contract value now at $398 million, reflecting broadening demand from both VC-backed and enterprise customers.
Profitability remains challenged by supply chain delays and pre-revenue asset amortization—notably, $3.7 million in GPU amortization and $1.5 million in lease expense for idle capacity weighed on results. Adjusted EBITDA turned positive year-over-year, but net losses persisted due to higher energy, depreciation, and personnel costs as new facilities ramped. The company’s cash position was bolstered by $81.8 million in customer prepayments and proceeds from an ATM equity offering, supporting ongoing buildout and equipment financing.
- Hosting Uptime Consistency: Jamestown and Ellendale operated at full capacity for the fifth consecutive quarter, underpinning stable baseline revenue.
- Cloud Cluster Ramp Delayed: Only four 1,024-GPU clusters were operational by quarter-end, with another four expected to ship imminently, trailing original deployment schedules.
- Prepayments Fuel CapEx: Customer deposits and upfront payments remain a key funding lever, helping offset capital intensity of new site construction and GPU procurement.
Despite near-term margin pressure, management expects a sharp cash flow inflection in March as prepayment credits burn off and new capacity is monetized.
Executive Commentary
"Since our last earnings announcement, we have added an additional cloud customer, which brings our total annual contract value of cloud service contracts at full capacity to approximately $398 million. We tailor our agreements to our customers so that as they raise money, we can exercise options embedded in the contract to deploy GPUs and ramp up hosting capacity over time."
Wes Cummings, Chairman and CEO
"A significant headwind we faced during the fiscal second quarter of 2024 was amortization and occupancy charges for leases of computing equipment and data center space that have been accessed by the company but are not yet supporting revenue. We expect this impact to decrease in the future quarters as we resolve supply chain delays and are able to stand up full computing clusters that support revenue."
David Wrench, Chief Financial Officer
Strategic Positioning
1. Cloud and HPC Demand Acceleration
AI and HPC compute demand is accelerating across both VC-backed and large enterprise customers, with APLD’s contracted cloud services ACV now at $398 million. The company is seeing new enterprise entrants—defined as $50 billion-plus market cap public companies—beginning to contract for large GPU clusters, expanding beyond the initial startup-heavy client mix.
2. Data Center Capacity and Power Scarcity
APLD’s ability to secure and develop power-rich sites is emerging as a key strategic moat. The company is developing 400 megawatts of new capacity in North Dakota and Utah, with all sites adjacent to abundant, often renewable, energy sources. Management forecasts intensifying industry-wide power constraints, positioning their land and power pipeline as a future pricing lever.
3. Supply Chain and Deployment Discipline
Management is deliberately pacing GPU deliveries and installations to avoid carrying underutilized assets, only accepting hardware when all networking components are available. This approach minimizes idle amortization drag but also limits near-term revenue ramp, as seen with the delayed cluster activations tied to InfiniBand transceiver shortages.
4. Flexible Commercial Models and Prepayment Structure
Contracts are structured to include significant upfront prepayments and options for customers to scale deployments as they secure funding. This model supports APLD’s capital needs and provides downside protection, especially as most hosting contracts are fixed price and multi-year, insulating the business from crypto volatility and AI pricing swings.
5. Strategic Sales Build-Out for Enterprise Penetration
For the first time, APLD is investing in a dedicated sales team to target enterprise customers, signaling a shift from organic inbound demand to proactive market development. Management sees this as a prudent move to capture emerging demand as Fortune 500 companies finalize AI strategies and seek GPU capacity at scale.
Key Considerations
APLD’s quarter highlights the tension between overwhelming end-market demand and the realities of supply chain, grid, and capital constraints. The company’s approach to risk management and capital allocation will be central to its ability to capture the AI infrastructure cycle while avoiding overextension.
Key Considerations:
- Cluster Commissioning Pace: Revenue realization is tightly linked to the timely delivery and installation of networking components for GPU clusters.
- Power and Land as Strategic Assets: Scarcity of available power for new data centers is driving up the value of APLD’s pipeline, with potential for pricing power if supply tightens further.
- Customer Mix Evolution: The arrival of large enterprise clients could transform revenue stability and contract duration, reducing reliance on less predictable startup customers.
- Capital Structure Flexibility: Use of prepayments, project-level financing, and equipment leases enables rapid buildout but introduces balance sheet complexity and future refinancing risk.
- Crypto Hosting Resilience: Hosting contracts remain locked in for several years, with rising Bitcoin prices driving incremental inbound demand, though new capacity is fully contracted.
Risks
APLD faces execution risk on both supply and demand fronts: Delays in critical networking components could persist, further deferring cluster revenue and exacerbating asset amortization drag. Capital intensity and reliance on upfront customer payments and project financing expose the company to funding risk if demand or credit markets soften. AI infrastructure demand is robust, but competition for power and grid access is intensifying, and any regulatory or permitting setbacks could slow expansion. Finally, while fixed-price hosting contracts insulate near-term revenue, longer-term shifts in AI pricing or crypto economics could impact renewals.
Forward Outlook
For Q3 and Q4, Applied Digital guided to:
- Revenue and EBITDA below the low end of prior full-year guidance due to networking component delays.
- Exiting fiscal 2024 at an annualized revenue run rate of approximately $500 million and annualized adjusted EBITDA of $250 million, assuming 10 GPU clusters deployed.
For full-year 2024, management lowered guidance:
- Cluster deployment targets reduced to 10,000 GPUs by fiscal year-end, with upside possible if supply chain accelerates.
Management emphasized:
- Cash flow from hosting will inflect positively in March as prepayment credits burn off.
- Demand pipeline for both hosting and cloud remains robust, especially from new enterprise customers.
Takeaways
APLD’s quarter underscores the company’s unique positioning at the intersection of AI infrastructure scarcity and digital asset hosting, but also highlights the operational bottlenecks that can slow revenue realization even in a high-demand environment.
- Supply Chain and Grid Remain Gating Factors: Revenue inflection is now more a function of component and power delivery than customer demand, with management proactively pacing deployments to avoid idle asset drag.
- Strategic Asset Base Growing in Value: The buildout of power-rich, scalable data center sites in North Dakota and Utah is emerging as a durable competitive advantage as industry-wide capacity tightens.
- Watch for Enterprise Contracting and Cluster Ramp: Investors should monitor the pace of enterprise customer conversions and the resolution of networking bottlenecks as the key swing factors for 2024 results.
Conclusion
Applied Digital’s Q2 results reflect a business with surging demand but constrained by supply chain and grid realities. The company’s strategic focus on scalable power, disciplined capital allocation, and evolving customer mix positions it for significant upside if operational execution keeps pace with market opportunity.
Industry Read-Through
The quarter sends a clear signal across the digital infrastructure sector: AI and HPC compute demand is outstripping available power and data center capacity, with both enterprise and startup customers now competing for limited resources. Operators with access to scalable, low-cost power and the ability to rapidly deploy GPU clusters are best positioned to capture this wave, but must navigate acute supply chain and financing complexity. Cloud service providers, hardware vendors, and data center REITs should expect continued upward pressure on pricing and a premium on power-rich locations, while investors should scrutinize the pace at which contracted demand converts to revenue amid ongoing component shortages.