高通量计算机的图算法优化技术研究  

Research on Graph algorithm optimization for high throughput computer

作　　者：贾子昂 Jia Zi’ang(School of Electrical and Control Engineering,North China University of Technology,Beijing 110144,China)

机构地区：[1]北方工业大学电气与控制工程学院,北京100144

出　　处：《无线互联科技》2022年第4期68-69,共2页Wireless Internet Technology

摘　　要：高通量计算机具有高并发、强实时、低功耗等适于大数据计算特点,在单节点上,BFS算法的优化已取得一系列进展。文章对BFS算法的性能进行了系统的评估,优化后的BFS算法在高通量计算机上评价性能为24.26 GTEPS和两路X86构建服务器相比,单节点更具性能优势。High throughput computers have the characteristics of high concurrency,strong real-time and low power consumption,which are suitable for big data computing.On a single node,BFS algorithm optimization has made a series of progress.In this paper,the performance of BFS algorithm is evaluated systematically.The performance of the optimized BFS algorithm is 24.26 GTEPS on high-throughput computers.Compared with two-way X86 construction servers,single node has more performance advantages.

关 键 词：宽度优先搜索 高通量 Graph500 图算法 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]