基于梯度提升回归树的处理器性能数据挖掘研究-集成技术-中国.PDF

第 7 卷 第5期 集 成 技 术 Vol. 7 No. 5 2018年9月 JOURNAL OF INTEGRATION TECHNOLOGY Sep. 2018 基于梯度提升回归树的处理器性能数据挖掘研究 1,2 3 1 吕依蓉 孙 斌 喻之斌 1 (中国科学院深圳先进技术研究院 深圳 518055) 2 (中国科学院大学 北京 100049) 3 (首都师范大学信息工程学院 北京 100048) 摘  要  现代处理器一般只内置了4～8 个性能计数器，但可以监测多达上千个时钟周期级别的性能 事件。这些事件可以轻易地产生大量数据，称为处理器性能大数据。然而，如何从这些性能大数据中 提取有价值的信息面临着许多挑战。该文提出一种处理器性能数据分析方法，通过迭代使用梯度提升 回归树算法构建性能模型，为云计算负载的性能事件进行重要性排序，从而指导云计算平台的性能调 优。 关键词  性能计数器；梯度提升回归树；云计算 中图分类号  TP 302.7 文献标志码  A Mining CPU Performance Data Based on Gradient Boosting Regression Tree Lü Yirong1,2 SUN Bin3 YU Zhibin1 1( Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China ) 2( University of Chinese Academy of Sciences, Beijing 100049, China ) 3( Capital Normal University Information Engineering College, Beijing 100048, China ) Abstract Modern processors typically have only 4-8 performance counters which can be programmed to measure up to thousands of cycle-level performance events. These events can easily generate large amount of data, which is called central processing unit (CPU) big performance data. However, how to extract value from the big performance data faces many challenges. This paper presents a performance data analysis approach, which builds a performance model by iteratively using the gradient boosting regression tree algorithm and quantifies the importance of the performance events of workloads in cloud to guide their performance optimization