A Federal Energy Regulatory Commission (FERC) ruling on November 1 has blocked an Amazon-backed deal that would have allowed Amazon’s AI data center in Pennsylvania to tap power from the nearby Susquehanna nuclear plant. While the decision is seen as a temporary setback for Amazon, it signals a potential shift in the battle for energy resources—one that could have far-reaching consequences for Bitcoin miners.
The Power Struggle: Tech Giants vs. Bitcoin Miners
The deal, which had raised eyebrows in the energy and tech industries, was seen as a way for Amazon to quickly access large amounts of power, bypassing the long wait times typically involved in building new power plants. As AI data centers ramp up their energy consumption in line with the rapid growth of generative AI technologies, the need for reliable, abundant power is at a premium. The rejected agreement, which would have allowed Amazon’s data center to pull power from the Susquehanna nuclear plant, highlighted the growing competition for energy between AI facilities and other power-hungry industries, including Bitcoin mining.
Bitcoin mining, which requires substantial amounts of energy to run proof-of-work algorithms, has long been criticized for its environmental impact. However, the increased appetite for power from AI companies could push Bitcoin miners to the fringes of the energy market.
Jaran Mellerud, a Bitcoin mining expert, pointed out that AI facilities are aggressively expanding across the U.S., particularly targeting prime locations with abundant energy and fiber-optic infrastructure. As AI operations can generate significantly higher revenues per kilowatt-hour than Bitcoin mining, these tech giants have the financial muscle to outbid miners for the same electricity, further intensifying competition in the energy market.
The Future of Bitcoin Mining in the U.S.
Mellerud predicted that over the next five years, AI operations will increasingly push Bitcoin miners to less accessible and less developed areas. By 2030, he forecasts the U.S. share of global Bitcoin mining could drop from the current 40% to under 20%, as AI’s growing energy consumption forces mining operations into remote regions, primarily in developing countries where energy resources may be cheaper and less contested.
This shift in energy priorities is already reflected in the growing energy consumption of AI systems. According to the Bitcoin Policy Institute, AI systems may already be consuming more power than Bitcoin mining operations, a trend that’s expected to accelerate. In 2024, generative AI is expected to consume 169 TWh (terawatt-hours) of electricity, surpassing Bitcoin’s estimated 160 TWh in the same year. By 2027, AI is expected to consume 240 TWh, while Bitcoin mining will only require 160 TWh, further tilting the balance of power in favor of AI.
The Financial Advantage of AI over Bitcoin Mining
The financial dynamics at play are also influencing the energy competition. AI data centers generate up to 25 times more revenue per kilowatt-hour than Bitcoin mining operations, making it far more lucrative for energy suppliers to prioritize AI contracts over Bitcoin mining deals. As a result, tech giants like Amazon and Microsoft are rapidly securing power procurement deals, locking in long-term access to the energy resources needed to support their AI infrastructure.
Bitcoin miners, on the other hand, are finding themselves squeezed as they compete for increasingly scarce energy supplies. Some miners are attempting to pivot by integrating AI processing into their data centers or switching entirely from Bitcoin mining to more profitable AI operations. Margot Paez, a Bitcoin Policy Institute researcher, noted that this trend would likely continue as long as AI provides higher revenue per megawatt-hour than Bitcoin.
Challenges for Bitcoin Miners Pivoting to AI
While some Bitcoin miners are exploring ways to shift to AI, this is not a straightforward transition. Anibal Garrido, a crypto assets adviser, pointed out that Bitcoin mining hardware is highly specialized and tailored for the proof-of-work protocol. Most miners use application-specific integrated circuit (ASIC) machines, which are designed specifically for mining and cannot be repurposed for AI or other types of data processing.
This specialization means that Bitcoin miners would have to invest in new hardware if they wanted to switch to AI mining—an expensive and logistically challenging process.
Conclusion: A Tightening Energy Market
The ongoing battle for energy between AI tech giants and Bitcoin miners could lead to a significant shift in the global mining landscape. As AI’s power demands continue to surge, tech companies will likely continue to outbid miners for access to the best energy sources, potentially pushing Bitcoin mining operations into less developed regions.
With energy procurement now at the forefront of both AI development and Bitcoin mining, the next decade may see a dramatic shift in the U.S. mining sector, as AI-driven industries increasingly take precedence. For Bitcoin miners, this could mean higher costs, more competition, and a continued struggle to secure sustainable and affordable energy sources.
