Anhui Province recently completed a valley-filling dispatch test for computing-power coordination: China Telecom Anhui transferred real-time computing tasks from the intelligent computing center in Hefei High-tech Zone to the Huaibei Telecom Cloud Computing Center. The Huaibei center took over the tasks completely after three minutes, and the power load of a single machine rose by 50%, effectively improving local consumption of green electricity.
Policy Deployment Continues to Advance
This year’s Government Work Report proposed implementing new infrastructure projects such as ultra-large-scale intelligent computing clusters and computing-power coordination, strengthening national integrated computing monitoring and dispatch, and supporting the development of public cloud. The Action Plan on Promoting the Two-Way Empowerment of Artificial Intelligence and Energy, issued by the National Development and Reform Commission, the National Energy Administration, the Ministry of Industry and Information Technology and the National Data Administration, proposed promoting efficient and economical coordination of computing power and electricity, giving full play to the scale effect of computing-power coordination, tapping the flexible adjustment potential of computing facilities, and using market-based electricity trading to improve the overall operating efficiency of computing facilities and the allocation of energy across society. The outline of the 15th Five-Year Plan proposes “promoting the coordinated layout of green electricity and computing power”; the 15th Five-Year Plan for Building a New Energy System proposes “improving the energy efficiency of new high energy-consuming industries such as computing” and “accelerating the coverage of incremental electricity consumption with incremental clean energy electricity”.
Computing-power coordination refers to a new infrastructure project that deeply integrates computing infrastructure with the power system through digital technology, intelligent algorithms and information networks, promoting dynamic matching and optimal allocation of resources, with the goal of achieving a virtuous cycle of “strengthening computing with electricity and promoting electricity with computing”. Under the Action Plan, at the operational level, a computing-electricity interaction mechanism will be established, using electricity market price signals to guide computing facilities in optimizing energy management and multi-form computing dispatch across networks and regions, and encouraging computing facilities to participate in grid operation as flexible and dispatchable load-side resources. At the market level, new computing facilities are encouraged to sign multi-year green power trading contracts with renewable energy generators, and computing facilities are supported in participating in electricity energy, ancillary services and demand response market trading in various forms.
Rapidly Growing Electricity Demand from Computing
According to the National Energy Administration, total electricity consumption of computing centers in China reached 170 billion kWh in 2025, accounting for 1.6% of total social electricity consumption. Computing electricity consumption at the eight hub nodes of the national integrated computing network has become the main driver of incremental demand, with an average growth rate of about 39.5% over the past three years, far higher than the average growth rate of total social electricity consumption. According to the forecast of the International Energy Agency, global electricity consumption of computing centers in 2030 will be nearly double that of 2025. For AI large models, the national daily average number of token calls was 100 billion in early 2024, jumped to 100 trillion by the end of 2025, and exceeded 140 trillion in March 2026, an increase of more than 1,000 times in just over two years.
Computing-Power Coordination Practices in Multiple Regions
In Suqian, Jiangsu, the computing center of the Suqian Digital Economy Center consumes about 17,000 kWh per day on average. State Grid Suqian Power Supply Company upgraded distribution lines and optimized the local grid structure, channeling surplus PV power from the surrounding area to the computing center, which both supports nearby consumption of distributed PV and provides a continuous green power supply for the computing facility. In Ningbo, Zhejiang, the Fangjia 110 kV transmission and substation project built by State Grid Ningbo Yinzhou Power Supply Company was commissioned with two 50 MVA main transformers, diverting load from the original substations and improving the power supply guarantee for the Ningbo AI supercomputing center. In Shizuishan, Ningxia, the Fuxiang No. 1 energy storage station (200 MW/800 MWh) in Huinong District was commissioned in mid-September; according to the plan, the city’s installed energy storage capacity will reach 1.5 GW by the end of 2026 and is expected to rise to 3.6 GW by the end of the 15th Five-Year Plan period.
As for benchmark projects, the large-scale computing-power coordination green power direct supply project in Zhongwei, Ningxia, was commissioned on May 2, 2026, with 500 MW of supporting PV capacity, adopting a “physical direct supply plus bilateral trading” dual-track model. Zhangjiakou, Hebei, is exploring an “aggregated green power supply” model that aggregates wind and solar plants in three counties within about 150 km for nearby, local consumption at the municipal level. The data center green power direct connection source-grid-load-storage integration project in Ulanqab, Inner Mongolia, was commissioned in July 2025, and after the multiple source-grid-load-storage integration projects subsequently approved there are fully built, their annual power generation could reach 4.5 billion kWh.