A Collaborative Join Scheme on a MIC-Based Heterogeneous Platform

摘要：

　　Join is one of the most important operations in data analytics systems.Prior works focus mainly on join optimization using GPUs,but little is known about performance impact on the MICs.In order to investigate potential benefits of the use of MIC accelerators in improving performance of join operation,in this paper we design a join scheme with a CPU and MICs working collaboratively.This scheme includes task partitioning,a data transfer mode,join algorithm design.Experimental results show that our collective join scheme is effective for a heterogeneous platform with two Xeon Phi cards,and can improve performance by up to 30% over the CPU-only platform.

关键词： Join Heterogeneous platform CPU-MIC Xeon Phi Coprocessor

作者: Kailai Zhou Hui Sun Hong Chen Tianzhen Wu Cuiping Li

作者单位: Key Lab of Data Engineering and Knowledge Engineering of MOE,and School of Information,Renmin Univer Key Lab of Data Engineering and Knowledge Engineering of MOE,and School of Information,Renmin Univer

会议类型: 国际会议

会议名称: International Asia-Pacific Web Conference(第18届国际亚太互联网大会)

会议地点: 苏州

会议语种:英文

页码: 454-458

在线出版日期: 2016-09-23（万方平台首次上网日期，不代表论文的发表时间）