According to a report by the International Energy Agency (IEA), by 2030, energy demands of data centers in the US could reach 945 TWh, matching Japan's current consumption. The primary factor driving this growth is artificial intelligence (AI).
Energy Requirements of AI
The IEA report indicates that in 2024, data centers consumed 415 TWh, about 1.5% of global electricity consumption. AI servers accounted for 24% of server electricity use and 15% of total data center energy demand last year.
State of Energy Infrastructure
Analysis reveals that the current energy infrastructure in the US is already heavily taxed and may struggle to meet further demands. Many transmission lines and substations in key states are over 50 years old, posing additional risks to providing the necessary electricity for AI development.
Position of Authorities and Developers
Government authorities have begun to act by introducing new support programs at both federal and local levels. For example, the Trump administration signed orders aimed at fast-tracking coal energy production to meet AI needs. However, this has faced criticism from analysts and environmentalists who doubt the effectiveness of such measures.
The question of whether the US can meet the rising energy demands of artificial intelligence remains open. The government needs to enhance the energy infrastructure and consider alternative energy sources.
