首页 | 官方网站   微博 | 高级检索  
     

“天河一号”大规模并行应用程序测试
引用本文:朱小谦,孟祥飞,营晓东,冯景华. “天河一号”大规模并行应用程序测试[J]. 计算机科学, 2012, 39(3): 265-267
作者姓名:朱小谦  孟祥飞  营晓东  冯景华
作者单位:(国家超级计算天津中心 天津 300457);(国防科技大学计算机学院 长沙 410073)
基金项目:国家自然科学基金(61070041)资助
摘    要:目前安装在国家超级计算天津中心的"天河一号"是我国首台千万亿次超级计算机,在2010年11月世界超级计算机Top500排名中,位列世界第一。"天河一号"采用了CPU与GPU相结合的异构融合计算体系结构,自主设计实现了高速互连通信系统,在多个高性能计算应用领域具有应用适应性强、系统运行稳定可靠、性能可扩展性好等特点,为科学研究和应用提供了重要的高性能计算平台。采用石油地震数据处理、飞行器流场模拟、生物分子动力学模拟、磁约束聚变数值模拟、湍流数值模拟、晶体硅分子动力学模拟、全球大气浅水波全隐式数值模拟、地球外核热流动数值模拟等应用的典型算例对"天河一号"进行了大规模并行程序测试,结果表明,"天河一号"在上述应用领域具有良好的可扩展性和并行效率。

关 键 词:天河一号  高性能计算  应用  并行效率

Performance Test of Massively Parallel Program on TH-lA
ZHU Xiao-qian,MENG Xiang-fei,JIAN Xiao-dong,FENG Jing-hua. Performance Test of Massively Parallel Program on TH-lA[J]. Computer Science, 2012, 39(3): 265-267
Authors:ZHU Xiao-qian  MENG Xiang-fei  JIAN Xiao-dong  FENG Jing-hua
Affiliation:1(National Supercomputer Center in Tianjin,Tianjin 300457,China)1(College of Computer,National University of Defense Technology,Changsha 410073,China)2
Abstract:TH-lA is the china's first petaflops supercomputer and is now installed in national supercomputer center inI}ianjin,which is ranked No. 1 on Top500 list released in Nov. 2010. I}H-lA systcxn adopts a hybrid architecture of hetero-geneous integration of CPU}GPU and self-intellectual high-speed interconnection system,and shows distinguished sys-tem usability, performance stability and application scalability in quite a few high performance computing areas, providesan important platform for science research and technology innovation. In this paper,several largcscale application testson TH-lA were introduced, such as the oil seismic data processing, aircraft flow field simulation, biomolecular dynamicssimulation, magnetic confinement fusion numerical simulation, turbulent flow simulation, crystal silicon molecular dy-namics simulation, fully implicit simulation of global atmospheric shallow wave, heat flow simulation of earth outercore. These results show that TH-lA has good parallel efficiency and scalability in practical applications.
Keywords:TH-lA   High performance computing   Application   Parallel efficiency
本文献已被 CNKI 等数据库收录!
点击此处可从《计算机科学》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号