Bitcoin mining is undergoing a significant narrative shift, championed by Nvidia CEO Jensen Huang. He recently described the process not as wasteful energy consumption, but as a method to convert excess or stranded energy into a portable store of value. "Energy is not wasted. It is stored as Bitcoin and can be moved anywhere," Huang explained, framing mining as a solution for power plants that produce more electricity than local grids can use.
This perspective aligns with the operational reality of major mining firms. For instance, Riot Platforms reported an average all-in power cost of about 3.9 cents per kilowatt-hour in December 2025, highlighting their reliance on cheap, abundant energy that would otherwise be curtailed. The timing is crucial as approximately 95% of all Bitcoin has been mined, making efficiency and access to low-cost power more critical than raw scale.
Concurrently, a separate geopolitical development could unlock massive new energy resources for mining. The White House has confirmed that a U.S. purchase of Greenland is an "active discussion." For the Bitcoin mining industry, the focus is on Greenland's untapped hydropower potential. The government plans a tender in the second half of 2026 for two major hydropower sites, Tasersiaq and Tarsartuup Tasersua, which together could produce over 9,500 gigawatt-hours annually.
This translates to a potential hashrate ceiling of roughly 56 exahashes per second (EH/s) at standard efficiency, which could account for 4–6% of the current global Bitcoin network hashrate. The ambition doesn't stop there; a scientific study suggests Greenland's onshore wind technical potential is about 333 GW, theoretically capable of supporting a hashrate far exceeding the entire current network.
The political connection is underscored by the involvement of Trump-linked capital. Hut 8 partnered with Eric Trump to launch American Bitcoin (ABTC), which reported an installed hashrate of about 24 EH/s as of September 2025. A fully developed 1.08 GW hydropower project in Greenland could power an ABTC-sized fleet more than once over. However, significant hurdles remain, including multi-year construction timelines, lack of a national grid, and competition from other industrial loads like AI data centers.
