Parallelizing Quicksort Algorithm using JOMP and its performance analysis with OpenMP

摘要

This paper makes an effort on parallelizing the quicksort algorithm using openMP on java platform(JOMP) to sort data by sharing the partitions generated from regular sampling. The basic idea is to perform the initial partitioning of data sequentially and calling the recursive function in parallel manner, and finally to conduct a performance evaluation for the implementation. The evaluation is based on comparing sorting times with an algorithm which uses initial partitioned set between threads and to the simple sequential sorting algorithm. Comparision study is also made with the OpenMP parallel quicksort implemented using C and the results shows that JOMP gives better result with respect to time complexity than OpenMP in C.