Last year, the total electricity consumption of national computing power centers reached 170 billion kilowatt-hours.

Accounting for 1.6% of total electricity consumption across the entire society.

  At the National “AI Plus” Energy On-site Promotion Conference held on the 26th, the National Energy Administration stated that by 2025, China will have built 42 exascale‑class intelligent computing clusters, with total electricity consumption at national computing centers reaching 170 billion kilowatt-hours—accounting for 1.6% of the nation’s overall electricity use. Among the eight major hub nodes of the national integrated computing power network, electricity demand from computing centers has become the primary driver of growth, with an average annual increase of approximately 39.5% over the past three years—far outpacing the average growth rate of total societal electricity consumption.

  At the conference, Wang Hongzhi, Director of the National Energy Administration, stated that with the rapid development of artificial intelligence, the continuous upgrading of ultra-large-scale intelligent computing clusters and the accelerated aggregation of new‑type productive forces such as chip manufacturing are giving rise to new characteristics and trends in electricity demand—namely, high growth rates, high density, centralized deployment, and increasingly complex operations—while the need for high‑quality power supply is steadily rising. It is projected that during the 15th Five-Year Plan period, national electricity consumption for computing will increase by more than 100 billion kilowatt-hours annually, reaching approximately 800 billion kilowatt-hours by 2030, accounting for around 6% of total societal electricity consumption.

  “Computing equipment has extremely stringent requirements for voltage stability, frequency deviation, and other performance metrics; even minor deviations in power quality can lead to data transmission errors and equipment malfunctions. Therefore, the electricity supply must not only meet demand in quantity but also ensure consistent quality,” said Wang Hongzhi. Judging from the current situation, China possesses a solid foundation for the mutual empowerment of artificial intelligence and the energy sector. In terms of quantitative reliability, China has built the world’s largest power system. On the qualitative front, the business environment for electricity consumption nationwide has improved markedly, with several “Getting Electricity” indicators reaching or approaching international best levels in the World Bank’s Doing Business enterprise surveys, cumulatively saving power users over RMB 300 billion in grid‑connection investment. Moreover, China’s green power supply continues to expand: last year, national green electricity transactions totaled 328.5 billion kWh, with cross‑provincial and inter‑regional green power trading steadily increasing. The country has also established the world’s largest green certificate market, with 930 million green certificates traded last year, and in newly built data centers in places such as Hohhot in Inner Mongolia and Qingyang in Gansu, the share of green electricity in total consumption has already exceeded 80%.

  Focusing on effectively advancing the two-way empowerment of artificial intelligence and the energy sector, Wang Hongzhi stated that efforts to open up energy‑related application scenarios will be continuously deepened. State‑owned energy enterprises should assume a leading role, proactively making high‑value application scenarios available while ensuring safety. Private enterprises, meanwhile, should serve as the main force, actively taking on challenges and engaging in scenario‑based openness and solution‑development initiatives. At the same time, it is essential to significantly enhance the quality of power supply for new‑type productive forces, strengthen coordination among all stakeholders, make breakthroughs in critical core technologies, foster a sound ecosystem for integrated development, and bolster the governance effectiveness of AI applications in the energy sector. (Reporter: Liao Ruiling)