- Changes in Energy Consumption Patterns
- AI Energy Consumption Growth
- Environmental Aspects
Bitcoin mining and artificial intelligence (AI) are becoming rivals in the race for energy resources, reshaping energy consumption patterns in the US.
Changes in Energy Consumption Patterns
The International Energy Agency estimates that combined energy consumption of Bitcoin mining and AI will surge to 1,050 TWh by 2026. This amount of energy is enough to power an entire country. Bitcoin mining, according to estimates, uses 120 TWh of energy annually. In 2023 alone, Bitcoin mining took 0.4% of global energy consumption. AI is expected to outperform Bitcoin miners’ energy consumption demand by 2027, shifting about 20% of their power capacity to AI.
AI Energy Consumption Growth
AI data centers are reportedly leading the energy consumption race and are projected to consume between 85 and 134 TWh of electricity by 2027. This amount is equal to the total annual energy consumption of Norway and Sweden, underscoring the significant energy needed to power complex AI models like OpenAI’s ChatGPT. These models run on large farms of servers. Half a million servers, with an estimated consumption of 29.2 TWh annually, are required to run ChatGPT.
Environmental Aspects
Bitcoin mining and AI are scrambling for the same energy sources. Tech firms like Amazon and Microsoft are reportedly after energy assets that were previously controlled by crypto miners. As competition heats up, some mining operators are making cash by leasing and selling power infrastructure. For others, the risk of losing access to electricity that keeps them in business becomes a reality. Data centers are projected to take up about 9% of all electricity consumed in the US, more than doubling their current consumption level. Interestingly, crypto mining operations rely mostly on renewable energy sources (about 70%), while AI data centers mostly rely on fossil fuels. With this in mind, tech firms are now forced to consider their carbon footprint and look into other energy sources such as nuclear power. For AI data centers, the challenge is not only in energy consumption but also in rising water consumption to cool generative AI servers.
The competition between Bitcoin mining and AI for energy resources affects energy consumption patterns, energy sources, and environmental aspects. These changes drive tech companies to reconsider their strategies and look for more sustainable solutions.
