China Targets 80% Green Power for AI Data Centers by 2030

The Government of China has established a strategic priority in its 2026 government work report to increase renewable energy use in the artificial intelligence data center sector. The plan targets a shift to 80% green power consumption by 2030, a significant increase from the 11% recorded in 2023. This initiative seeks to integrate computing infrastructure with power networks to reduce emissions, including the prioritization of data centers in western China and the addition of nearly 90 TWh of solar and wind electricity.

Industry officials warn that the plan faces significant technical and economic hurdles. AI data centers maintain rigid, high power consumption loads to maximize the utility of expensive GPUs, making peak demand difficult to forecast. Additionally, grid operators resist direct green-power links due to fears of declining electricity sales and unrecovered investments in transmission infrastructure. Power demand for these centers is projected to grow by 300 billion to 500 billion kilowatt-hours between 2026 and 2030, representing 18% of total electricity demand growth.

To address these challenges, research suggests that adjusting just 15% of power consumption loads could significantly ease grid capacity pressure. In a move toward innovation, China recently launched the Shanghai Lingang demonstration, the world's first offshore wind-powered underwater data center with a 24 MW capacity. This comes as the International Energy Agency reports that coal still dominated the sector's power supply at nearly 70% in 2025.