Bitcoin Miners Explore AI and HPC Opportunities

Bitcoin (BTC) miners are increasingly shifting toward artificial intelligence (AI) and high-performance computing (HPC) due to inconsistent transaction fee revenues, as noted in a recent VanEck report.

The report highlights that while network congestion can cause temporary fee spikes, the long-term growth of on-chain revenue remains unpredictable. This uncertainty stems from the emergence of off-chain solutions like exchange-traded funds (ETFs), futures markets, layer-2 (L2) solutions, and centralized exchanges.

As block rewards halve every four years, Bitcoin miners face a necessity to adapt, as BTC’s price needs to double to sustain revenue levels. This scenario underscores the importance of pivoting to AI and HPC as diversification strategies.

Pivoting to New Strategies

Successful Bitcoin miners are adopting alternative strategies to optimize their revenue streams. Some miners support grid expansion in remote energy markets, while others use existing power infrastructure for AI and HPC workloads that yield higher margins and new financing opportunities.

Innovations in liquid cooling systems, chip design, and co-location technologies improve operational efficiencies for miners.

The shift toward AI and HPC has gained traction, especially following CoreWeave’s 700MW agreement for AI/HPC in 2024, prompting many Bitcoin miners to investigate similar revenue streams. For instance, Bitfarms (BITF) is consulting AI/HPC experts for feasibility studies at North American sites, and Bitdeer (BTDR) is exploring partnerships in the AI/HPC sector after its recent data center consulting.

Cipher Mining (CIFR) received a $50 million investment from SoftBank in January for HPC data center expansion, while Riot Platforms (RIOT) expanded its board to include professionals in AI/HPC investment banking and data centers. HIVE Digital Technologies (HIVE) appointed Craig Tavares to lead its growth in HPC and GPU cloud services.

Iris Energy (IREN) is advancing its AI and HPC initiatives with a 75 MW liquid-cooled AI/HPC data center planned for launch in Childress, Texas, by the second half of 2025. The company also plans to expand its Sweetwater site to a total capacity of 2 GW, making it one of North America’s largest eligible sites for AI/HPC.

The report emphasized the rarity of such large-scale sites, highlighting their strategic advantage for AI model training and iterative workloads.

Scaling Electrical Capacity

The move towards AI has increased the need for Bitcoin miners to scale their electrical capacity. A review of 13 public Bitcoin miners shows their current operational capacity is 7.1 gigawatts (GW), with projections to expand it to 11.7 GW by 2025, 15.9 GW by 2026, and 20.4 GW by 2027, reflecting a 42% compound annual growth rate over three years.

An additional 7.3 GW is planned beyond 2028, although this is a conservative estimate considering competitive power procurement in the industry. Significant investments will be required for scaling operations. If expansion to 2027 utilizes a modernized Bitmain Antminer S21 Pros fleet at $5,000 each, along with $450,000 per MW of supporting infrastructure, total capital expenses may reach $24.8 billion.

However, miners are not expected to use all of this capacity solely for mining. Securing electrical capacity has become critical amidst rising AI power demands. Goldman Sachs indicates that AI currently consumes about 7.7 GW of global data center power, accounting for 14% of total usage, and this is forecasted to rise to 22.7 GW (27%) by 2027.

As a result, Bitcoin miners are anticipated to allocate 20-30% of their electrical capacity to AI and HPC workloads, marking a significant strategic shift toward more sustainable and diversified revenue streams.