Research And Optimization In Astrophysical Clustering Studies On Multi-core CPU

Halo-based Galaxy Group Finder(HGGF)tries to find galaxies in the same dark matter halo which is not directly visible.It plays a very important role in the research of large-scale structure of the universe.However,because of the growth of the data scale,it's extremely necessary to increase the running speed by optimizing the group finder coding algorithm.After a thorough investigation on the original HGGF code,we found that the kernel part of the algorithm was seriously affected by the irregular memory access,we thus proposed a specific “Data Pre-sorting” approach and analyzed how it affect the process of memory access according to the structure of the algorithm and the irregular memory access pattern.Moreover,we optimized the memory access with “Array Fission”,“Data Alignment” and “Loop Fission” and enhanced the efficiency of OpenMP with “Load Balance” and “Mutex Privatization”.Eventually the HGGF application got 11.6 times speedup on 12 threads,and got better weak scalability compared with the original code.The following are our original contributions:(1)Summarized and analyzed the structure and the memory access pattern of the HGGF application.(2)Proposed and applied the “Data Pre-sorting” to improve the locality;(3)Enhanced the running speed of HGGF application with another four approaches including data alignment,loop fission,load balance and mutex privatization based on multi-core CPU.