HUT Q2 2024: 12 J/TH Fleet Upgrade to Cut Mining Costs, Boost Margins
- Fleet Upgrade for Efficiency Gains: Active discussions on upgrading the fleet to next-generation miners operating at 12 joules per terahash indicate a potential significant reduction in mining costs relative to current machines (down from approximately 17–18 joules per terahash), which could positively shift margins.
- Robust and Expanding Power Infrastructure: With a reported ~1 gigawatt of wholly owned power capacity (inclusive of joint ventures) across 4 operating sites and a strong development pipeline, the company is well positioned to scale operations and leverage growth opportunities.
- Diversified Revenue Streams via HPC Growth: The expansion into GPU-as-a-service, which is expected to generate around $20 million in annual revenue at current utilization rates, adds a valuable, non–Bitcoin mining revenue source, enhancing overall business diversification and resilience.
- Utilization Risk in HPC Business: The GPU-as-a-service revenue target of $20 million annual run rate is based on current on-demand rates and average utilization that are below 100%, meaning any further decline in actual utilization could lead to revenue shortfalls.
- Revenue Forecast Uncertainty: The reliance on partners' average utilization rates, rather than a guarantee of full utilization, introduces uncertainty whether the business model can consistently meet its projected financial outcomes.
- Execution Risk in Expansion Strategy: The dual approach of exploring both co-location deals and owned GPU solutions introduces strategic and operational risks, as market dynamics and customer adoption may significantly affect the ability to scale and generate anticipated revenue.
-
Fleet Upgrade
Q: When will a full fleet upgrade complete and what’s current capacity?
A: Management indicated they are actively discussing fleet upgrades with manufacturers to optimize efficiency. They noted that full upgrade timing depends on the purchase order execution, and highlighted having about 1 gigawatt of developed power capacity, including joint ventures, guiding their upgrade plans. -
GPU Agreement
Q: Has the GPU-as-a-service agreement changed?
A: Management confirmed the agreement remains as announced—operating under a 5-year fixed base plus revenue share model, with plans to explore additional project financing for future growth. -
HPC Connectivity
Q: Are pipeline assets fully grid connected and what’s the Texas site timeline?
A: Management clarified that all pipeline assets are grid connected, and the 205-megawatt Texas Panhandle site has an approved substation connection. They are in discussions with customers regarding its use for HPC, with site activity planned into calendar 2024. -
GPU Utilization
Q: What is current GPU utilization and its revenue assumptions?
A: Management explained that the $20 million annual run rate for GPU services is based on typical on-demand rates and average utilization, not assuming full 100% usage. -
HPC Go-to-Market
Q: What is the strategy for expanding HPC offerings?
A: They see distinct opportunities: one through the GPU-as-a-service model with dedicated teams, and another via co-location deals built-to-suit for large-scale customer demands. This dual approach leverages strong infrastructure and power availability to capture long-term market opportunities.
Research analysts covering Hut 8.