用机器学习精准预测数据中心服务器功耗与性能，误差低于10%。
Generalizable Machine Learning Models for Predicting Data Center Server Power, Efficiency, and Throughput
- 基于SPECPower_ssj2008数据集，构建可通用的服务器建模方法
- 模型在测试集上误差控制在10%以内，具备实用性和泛化能力
- 揭示硬件发布时间、负载水平等关键影响因素，适合运维优化
在快速发展的数字时代，理解影响服务器功耗、效率和性能的复杂动态对可持续的数据中心运营至关重要。然而，现有模型难以提供对这些复杂关系的详尽且可靠的解读。本研究采用基于机器学习的方法，利用SPECPower_ssj2008数据库，实现用户友好且可泛化的服务器建模。所得模型在测试数据集上误差约为10%，展现出良好的实用性与泛化能力。通过细致分析，识别出与硬件上市时间、服务器负载水平及配置相关的预测特征，为优化服务器部署与运行中的节能、效率与性能提供洞见。通过系统性测量偏差与不确定性，研究强调在使用历史数据进行未来服务器建模时需谨慎，须考虑技术环境的动态变化。总体而言，该工作为数据中心服务器的可持续部署与运行提供了重要参考，推动资源利用效率提升与更环保的实践。
原文摘要 · Abstract (English)
In the rapidly evolving digital era, comprehending the intricate dynamics influencing server power consumption, efficiency, and performance is crucial for sustainable data center operations. However, existing models lack the ability to provide a detailed and reliable understanding of these intricate relationships. This study employs a machine learning-based approach, using the SPECPower_ssj2008 database, to facilitate user-friendly and generalizable server modeling. The resulting models demonstrate high accuracy, with errors falling within approximately 10% on the testing dataset, showcasing their practical utility and generalizability. Through meticulous analysis, predictive features related to hardware availability date, server workload level, and specifications are identified, providing insights into optimizing energy conservation, efficiency, and performance in server deployment and operation. By systematically measuring biases and uncertainties, the study underscores the need for caution when employing historical data for prospective server modeling, considering the dynamic nature of technology landscapes. Collectively, this work offers valuable insights into the sustainable deployment and operation of servers in data centers, paving the way for enhanced resource use efficiency and more environmentally conscious practices.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。