next up previous contents
Next: MPI_Send-MPI_Iprobe_Recv Up: Point-to-point-pattern Previous: MPI_Isend-MPI_Recv   Contents

MPI_Issend-MPI_Recv

Now we use MPI_Issend for sending and MPI_Recv for receiving. After the non-blocking synchronous send we use an MPI_Wait. So we can determine the costs of the wait's synchronization compared to MPI_Isend.
\epsfig{file=p2p_MPI_Issend-MPI_Recv.eps}
  message length
X axis scale logarithmic
Param. refinement automatic x wide adaption
Argument range 1 - 8388608 units (s. below)
Argument stepwidth 1.414214

next up previous contents
Next: MPI_Send-MPI_Iprobe_Recv Up: Point-to-point-pattern Previous: MPI_Isend-MPI_Recv   Contents
Per Ekman 2002-01-29