Q4 2024 Earnings Summary
- The company's cloud services business is generating strong revenue, with the 6 clusters online driving an annualized revenue run rate of $100 million to $110 million , indicating significant growth potential.
- Applied Digital has secured over 30 megawatts of capacity ready for deployment in its AI cloud business, which is a big asset in the market due to the scarcity of open capacity ready for immediate deployment. This positions the company well to meet the strong demand coming from the enterprise side.
- The company is close to finalizing a lease agreement with a U.S.-based hyperscaler for a 400-megawatt capacity, being north of 90% of the way there. This contract would solidify Applied Digital's position as a leader in the HPC data center market and unlock significant growth opportunities.
- Applied Digital's growth is heavily dependent on securing new financing structures, which are not yet in place. The company is limiting expansion of its cloud services business to 6 clusters until better financing is secured, as the current capital lease structure forces accelerated depreciation over 2 years instead of the industry standard of 5 or 6 years, negatively impacting financials.
- Intense competition from established private data center companies that have been building for hyperscalers for years or even decades may challenge Applied Digital's ability to win contracts. These competitors are already building hundreds of megawatts or over a gigawatt of power for hyperscalers, posing a significant threat to Applied Digital's market share.
- There is uncertainty regarding the timing of finalizing lease agreements with hyperscale customers and when they might go live, potentially delaying revenue generation. The company mentioned it's "not clear when the customer might want to go live with the lease," even though they are "north of 90% of the way there" in completing the contract.
-
AI Cloud Revenue and Growth
Q: What's the revenue run rate from the 6 clusters?
A: The 6 clusters are generating an annual revenue run rate of $100 million to $110 million. Demand is strong, particularly from the enterprise market, and with the right financing, the company expects to significantly grow this business. -
Financing Structure for AI Cloud Business
Q: Will you expand clusters after improving financing?
A: Yes, the company plans to expand beyond 6 clusters once a better financing structure is in place. They're working on financing that aligns depreciation with the equipment's useful life, improving financials by reducing the impact of accelerated depreciation. Interest from financiers has grown, and costs have come down, depending on the quality of customers. -
Completion and Funding of Ellendale HPC Campus
Q: What's the completion percentage and remaining capital needs for Ellendale?
A: With over $200 million invested to date , the Ellendale campus is progressing well, with the building fully enclosed and mechanical, electrical, and plumbing work underway. The facility will cost over $1 billion for the first 100 megawatts. The company expects to fund the remaining capital through project-level financing, having selected a bank and seen 80% to 90% loan-to-cost quotations. They hope to close financing shortly after finalizing the lease. -
Lease Agreement Timing and Details
Q: How close are you to finalizing the lease agreement?
A: The company is over 90% complete with the lease agreement. They're finalizing the last details, which include both large and small items like fiber connectivity and power redundancy. Completion timelines also depend on when the customer wants the facility ready, which isn't completely locked in yet. -
Future Pipeline and Expansion Plans
Q: What are your plans for other campuses and customer diversification?
A: The company recently began marketing additional sites, focusing on hyperscalers for campuses with at least 200-plus megawatts of critical IT capacity. They are open to being a one-stop shop for existing and new hyperscaler customers, given significant demand for near-term power. -
Competitive Advantages and Environment
Q: How do you compete against established data center companies?
A: Applied locates power availability first, ensuring it's suitable for development with the necessary fiber connectivity and latency requirements. This approach, combined with experience managing supply chain bottlenecks, positions the company favorably in a market with strong demand for near-term power. -
GPU Financing Structure and Interest
Q: What's the status of GPU financing and market interest?
A: Interest from financiers has grown, and costs have decreased. Financing structures depend on customer quality, with lower costs for enterprise and hyperscaler clients. The company expects to implement a financing structure involving a bankruptcy-remote SPV for deploying GPUs and customer contracts. -
Not Pursuing Texas Sites for HPC
Q: Why is Garden City not suitable for HPC?
A: The company believes that having a large flexible load in the ERCOT market isn't suitable for the customers they're pursuing. Proposed legislative changes in Texas may make it even harder to serve these customers with a flexible load structure. They are focusing on sites with firm power and redundancy elsewhere. -
Secured Megawatt Capacity for Cloud AI Business
Q: How much megawatt capacity do you have ready now?
A: The company has secured just over 30 megawatts of capacity ready for deployment in locations like Minnesota, Jamestown, Denver, and Salt Lake City. This capacity is significant given the difficulty in finding ready-to-deploy sites.
Research analysts covering Applied Digital.