Minimizing communication in bitonic sorting software

Two parallel sorting algorithms, GENERAL-BS and MINIMIZING-BS, which are implemented on shared-memory parallel computers, are presented in this paper. A parity strategy which gives an idea for the efficient usage of the local memory associated with each processor is introduced. The number of network accesses(or communications) of the algorithm MINIMIZING-BS is reduced by approximately one half compared with the algorithm GENERAL-BS. On the basis of decreasing the communication, the algorithm MINIMIZING-BS results in a significant improvement of performance.