北理工报告：建议全国统一算力定价，绿电算力享折扣、向小微企业发算力券

Core Insights - The unprecedented demand for computing power driven by AI large models has become a critical topic in energy research, highlighting the interconnection between computing power and energy supply [1] - The report from Beijing Institute of Technology emphasizes the need for a low-carbon computing power service system to support China's transition from a "big computing power country" to a "strong computing power country" [1] Group 1: Current Challenges in Computing Power Services - The report identifies five major challenges in China's computing power services, including structural imbalances in demand and supply, insufficient autonomy in key hardware, and technical bottlenecks in transmission [2] - The demand side shows a coexistence of high-end computing shortages and low-end computing idleness, while the supply side faces dual constraints of hardware autonomy and uneven distribution [2] Group 2: Regional Disparities in Computing Power - Eastern core urban areas have a high concentration of computing power demand, with some hotspots experiencing load rates exceeding 85%, while western regions have low utilization rates due to insufficient demand [3] - The report notes that the cost of cross-regional transmission can reach 2-5 times that of local resources, undermining the economic attractiveness of western computing power [3] Group 3: Recommendations for Optimizing Computing Power Services - The report suggests strengthening the "East Data West Computing" initiative by implementing differentiated access requirements and establishing a compensation mechanism for green electricity [4] - It recommends creating a unified national pricing mechanism for computing power and providing subsidies for green electricity usage [4] Group 4: Future Projections - The report forecasts that by 2030, the average power load of data centers nationwide will reach 105 million kW, with total electricity consumption averaging 525.76 billion kWh [4] - By 2040, data centers are expected to account for 10.84% of national electricity demand, driven by medium-speed growth in intelligent computing power [4] Group 5: Integration of Computing Power and Energy Systems - Experts emphasize the need for coordinated planning between computing centers and the overall power grid to optimize both computing and energy transmission networks [5] - There is a call for improved price transmission channels between the electricity market and the computing power market to enhance synergy [6]