标题:Million-Core-Scalable Simulation of the Elastic Migration Algorithm on Sunway TaihuLight Supercomputer
作者:Gan, Lin; Xu, Jingheng; Wang, Xin; Wu, Sihai; Duan, Xiaohui; Li, Yuxuan; Fu, Haohuan; Yang, Guangwen
通讯作者:Xu, JH;Xu, JH;Xu, JH;Xu, JH
作者机构:[Gan, Lin; Xu, Jingheng; Wang, Xin; Li, Yuxuan; Yang, Guangwen] Tsinghua Univ, Dept Comp Sci & Technol, Beijing, Peoples R China.; [Gan, Lin; Xu, Ji 更多
会议名称:19th Annual IEEE/ACM International Symposium on Cluster, Cloud, and Grid Computing (CCGRID)
会议日期:MAY 14-17, 2019
来源:2019 19TH IEEE/ACM INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND GRID COMPUTING (CCGRID)
出版年:2019
页码:62-69
DOI:10.1109/CCGRID.2019.00016
关键词:HPC; geophysics exploration; migration algorithm; Elastic RTM; Sunway; TaihuLight supercomputer
摘要:Migration algorithm is one of the most essential methods in seismic application to image the underground geology, and to help scientists and researchers in geophysics exploration better understand the earth system. However, due to the desire in migration algorithm 14 covering lager region and acquiring better resolution, many tough challenges have to be tackled for current state-of-the-art computing systems. This work optimized and scaled the elastic migration algorithm onto the Sunway TaihuLight supercomputer, one of the most powerful systems of the world. Targeting at the major process, the reverse time migration (RTM) algorithm, a set of algorithmic, process-level, and thread-level optimizations is proposed, to significantly improve the performance (up to 163 x speedup in time-to-solution) on Sunway CPU. Our design is successfully scaled to over two million cores (2,662,400 cores in total) on the Sunway TaihuLight supercomputer, with nearly ideal weak-scaling efficiency. The largest run is able to achieve a sustainable performance of processing over 859 billion cells per second.
收录类别:CPCI-S
资源类型:会议论文
TOP