Implementation and performance analysis of non-blocking collective operations for MPI | IEEE Conference Publication | IEEE Xplore