Ireduce_scatter

The benchmark for MPI_Ireduce_scatter that measures communication and computation overlap. It reduces a vector of length L = X/sizeof(float) float items. The MPI data type is MPI_FLOAT. The MPI operation is MPI_SUM. In the scatter phase, the L items are split as evenly as possible. To be exact, for np number of processes:

L = r*np+s

where

In this case, the process with rank i gets:

Property

Description

Measured pattern

MPI_Ireduce_scatter/IMB_cpu_exploit/MPI_Wait

MPI data type

MPI_FLOAT

MPI operation

MPI_SUM

Reported timings

  • t_ovrl
  • t_pure
  • t_CPU
  • overlap=100.*max(0,min(1, (t_pure+t_CPU-t_ovrl) / min(t_pure, t_CPU))

For details, see Measuring Communication and Computation Overlap.

Reported throughput

None

Submit feedback on this help topic