Abstract: To clarify the comprehensive energy consumption and the size of flexible resources of data centers under the dual-carbon (carbon peaking and carbon neutrality) vision, this paper constructed a predictive model for the comprehensive energy consumption and flexibility of data centers nationwide. By comprehensively considering various factors such as computing power growth and technological advances, the model simulated and predicted the comprehensive energy consumption and flexibility requirements of data centers across China from 2023 to 2030. The research indicates that by 2030, China's data centers are projected to have an electric load of 110 million kW, electricity consumption of 525.76 billion kWh, water consumption of 262.88 billion liters, and carbon emissions of 310 million tons. Additionally, the computing power, electricity, and thermal flexibility resources within data centers can significantly reduce carbon emissions, energy consumption, and operational costs. In 2030, the total flexibility resources of China’s data centers are expected to range between 23 000 and 40 000 MW. Based on these findings, this paper proposes recommendations such as strengthening the separate accounting of data center energy consumption, promoting bidirectional coordination between computing power and electricity, and establishing related standards, norms, engineering demonstrations, and market mechanisms.