DER: A Novel Metric Based on Dynamic Efficiency Ratio for Evaluating Computing Resource Efficiency

The digitization of the power sector has become a significant global trend. As the core infrastructure supporting cloud computing and big data processing, data centers face intense scrutiny regarding their energy efficiency. Consequently, accurately assessing the energy efficiency of data centers based on cloud platforms is urgent. Traditional assessment metrics, such as PUE and DCiE, account for computing equipment's power consumption into full consideration and achieve considerable results. However, these metrics are static and do not reflect dynamic changes in computing loads, leading to inaccurate evaluations. To address this issue, we propose a novel metric named the Dynamic Efficiency Ratio (DER). DER combines computing load and energy consumption while considering temporal and spatial dimensions. By real-time monitoring and analysis, DER provides precise and dynamic energy efficiency data for resource scheduling. Therefore, this metric more accurately reflects the energy efficiency of data centers in actual operation. Furthermore, we validated this metric with a detailed case study of a data center. The results demonstrate that DER assists in improving system performance and reducing operational costs, which has practical value.