Workflow
Applied Digital (APLD) - 2024 Q1 - Earnings Call Transcript
APLDApplied Digital (APLD)2023-10-09 15:47
Filings

Financial Data and Key Metrics - Revenue for Q1 2024 was 36.3million,upfrom36.3 million, up from 6.9 million in Q1 2023, driven by increased hosting capacity and AI cloud service contracts [12] - Cost of revenue increased to 24.4millionfrom24.4 million from 6.1 million, primarily due to higher energy costs, depreciation, and personnel expenses [12] - Operating expenses rose to 17.1millionfrom17.1 million from 5 million, mainly due to increased headcount and depreciation [13] - Net loss for Q1 2024 was 9.6million,or9.6 million, or 0.10 per share, compared to a net loss of 4.7million,or4.7 million, or 0.05 per share, in Q1 2023 [13] - Adjusted EBITDA for Q1 2024 was 10million,comparedtoanadjustedEBITDAlossof10 million, compared to an adjusted EBITDA loss of 1.9 million in Q1 2023 [14] - The company ended the quarter with 31.2millionincashand31.2 million in cash and 44 million in debt [15] Business Line Data and Key Metrics - Blockchain hosting operations: The 100-megawatt Jamestown facility operated at full capacity for the fourth consecutive quarter, and the 180-megawatt Ellendale facility became fully operational, bringing total hosting capacity to 280 megawatts [6][7] - AI cloud services: The company activated the first cluster of GPUs for Character.AI in July and received a second cluster in September, with additional GPUs expected in October [8] - The total annual contract value of AI cloud services contracts at full capacity is approximately 378million[9]−Thecompanyhas300megawattsofcapacityindevelopmentforpurpose−builtHPCdatacenters,withinitialgroundworkunderwayfortheEllendalefacility[11]MarketDataandKeyMetrics−Thecompanyexpectsitsthreesitestoproducearound378 million [9] - The company has 300 megawatts of capacity in development for purpose-built HPC data centers, with initial groundwork underway for the Ellendale facility [11] Market Data and Key Metrics - The company expects its three sites to produce around 300 million in revenue and 100millioninEBITDAannually[7]−TheGardenCity,TexasfacilityisexpectedtobefullyenergizedbyOctober23,bringingtotalhostingcapacitytoapproximately500megawatts[7]CompanyStrategyandIndustryCompetition−Thecompanyisfocusedonprovidingdigitalinfrastructuresolutionsforthehigh−performancecomputingindustry,withastrongemphasisonAIcloudservicesandblockchainhosting[6][8]−PartnershipswithleadingOEMslikeSuperMicro,HewlettPackardEnterprise,andDell,alongwithNVIDIA′sElitePartnerstatus,providevisibilityintoGPUdeliverytimelines[10]−ThecompanyisactivelypursuinganchortenantsforitsHPCdatacenters,aimingtosecurecredit−ratedtenantstosupportconstructionandfinancing[11]ManagementCommentaryonOperatingEnvironmentandFutureOutlook−Managementremainsconfidentinthecompany′scompetitiveadvantagesanddifferentiatedcapabilitiestomeetthedemandsofAIworkloadsandotherHPCapplications[17]−ThecompanyexpectsasteeprevenuerampintheAIcloudbusiness,particularlyinthelasttwoquartersofthefiscalyear,asGPUdeliveriesincrease[36]−Managementisoptimisticaboutthefuture,withastrongpipelineofopportunitiesintheAIcloudservicebusinessandongoingdevelopmentofHPCdatacenters[17]OtherImportantInformation−Thecompanyreceived100 million in EBITDA annually [7] - The Garden City, Texas facility is expected to be fully energized by October 23, bringing total hosting capacity to approximately 500 megawatts [7] Company Strategy and Industry Competition - The company is focused on providing digital infrastructure solutions for the high-performance computing industry, with a strong emphasis on AI cloud services and blockchain hosting [6][8] - Partnerships with leading OEMs like Super Micro, Hewlett Packard Enterprise, and Dell, along with NVIDIA's Elite Partner status, provide visibility into GPU delivery timelines [10] - The company is actively pursuing anchor tenants for its HPC data centers, aiming to secure credit-rated tenants to support construction and financing [11] Management Commentary on Operating Environment and Future Outlook - Management remains confident in the company's competitive advantages and differentiated capabilities to meet the demands of AI workloads and other HPC applications [17] - The company expects a steep revenue ramp in the AI cloud business, particularly in the last two quarters of the fiscal year, as GPU deliveries increase [36] - Management is optimistic about the future, with a strong pipeline of opportunities in the AI cloud service business and ongoing development of HPC data centers [17] Other Important Information - The company received 39.5 million in customer prepayments during Q1 2024, with an additional 15millionreceivedpost−quarterand15 million received post-quarter and 23 million expected this week [15][16] - The company reaffirmed its full-year fiscal 2024 guidance, expecting revenue in the range of 385millionto385 million to 405 million and adjusted EBITDA of 195millionto195 million to 205 million [16] Q&A Session Summary Question: Anchor Tenant Requirements for AI Business - The company needs to contract roughly 70% of the capacity on 7 to 10-year contracts to kick off construction, with plans to start with 100 megawatts in North Dakota, followed by 100 megawatts in Utah [20] Question: GPU Delivery Timeline - The company expects to receive large volumes of GPUs in November, December, and January, with deliveries ramping up significantly during these months [22] Question: Garden City Facility Energization - The Garden City facility is expected to start energizing on October 23, with a faster ramp-up compared to previous facilities due to completed construction and ready-to-use miners [25] Question: AI Cloud Service Capacity and Contracts - The company has some additional capacity for AI cloud services but will need to bring its own facilities online to support further growth [30] - The company increased its GPU order from 26,000 to 34,000 due to strong demand, with the additional 8,000 GPUs expected to be deployed post-April 2024 [33] Question: Capital Intensity and Financing for HPC Data Centers - The company redesigned its data centers to increase density, with construction costs now estimated at $6 million per megawatt [39] - The company is exploring various financing options, including construction financing and equity financing at the site level [39] Question: GPU Delivery Cadence - The company received its second cluster of 1,024 GPUs in September and expects deliveries to double in October, with significant increases in November, December, and January [44] Question: Anchor Tenant Demand and Data Center Financing - The company is in advanced discussions with several potential anchor tenants and expects to finalize agreements in the coming weeks [47] Question: AI Cloud vs. AI Hosting Strategy - The company is pursuing both AI cloud services and AI hosting, with anchor tenants expected to bring their own equipment for hosting, while the company retains 30% of capacity for its own cloud services [57] Question: Data Center Redesign and Heat Management - The new data center design focuses on density and liquid cooling, with the ability to support up to 150 kW per rack [64]