Chris@10: Previous: Advanced distributed-transpose interface, Chris@10: Up: FFTW MPI Transposes Chris@10:
We close this section by noting that FFTW's MPI transpose routines can
Chris@10: be thought of as a generalization for the MPI_Alltoall function
Chris@10: (albeit only for floating-point types), and in some circumstances can
Chris@10: function as an improved replacement.
Chris@10:
Chris@10:
Chris@10:
MPI_Alltoall is defined by the MPI standard as:
Chris@10:
Chris@10:
int MPI_Alltoall(void *sendbuf, int sendcount, MPI_Datatype sendtype, Chris@10: void *recvbuf, int recvcnt, MPI_Datatype recvtype, Chris@10: MPI_Comm comm); Chris@10:Chris@10:
In particular, for double* arrays in and out,
Chris@10: consider the call:
Chris@10:
Chris@10:
MPI_Alltoall(in, howmany, MPI_DOUBLE, out, howmany MPI_DOUBLE, comm); Chris@10:Chris@10:
This is completely equivalent to: Chris@10: Chris@10:
MPI_Comm_size(comm, &P); Chris@10: plan = fftw_mpi_plan_many_transpose(P, P, howmany, 1, 1, in, out, comm, FFTW_ESTIMATE); Chris@10: fftw_execute(plan); Chris@10: fftw_destroy_plan(plan); Chris@10:Chris@10:
That is, computing a P × P transpose on P processes,
Chris@10: with a block size of 1, is just a standard all-to-all communication.
Chris@10:
Chris@10:
However, using the FFTW routine instead of MPI_Alltoall may
Chris@10: have certain advantages. First of all, FFTW's routine can operate
Chris@10: in-place (in == out) whereas MPI_Alltoall can only
Chris@10: operate out-of-place.
Chris@10:
Chris@10:
Chris@10:
Second, even for out-of-place plans, FFTW's routine may be faster,
Chris@10: especially if you need to perform the all-to-all communication many
Chris@10: times and can afford to use FFTW_MEASURE or
Chris@10: FFTW_PATIENT. It should certainly be no slower, not including
Chris@10: the time to create the plan, since one of the possible algorithms that
Chris@10: FFTW uses for an out-of-place transpose is simply to call
Chris@10: MPI_Alltoall. However, FFTW also considers several other
Chris@10: possible algorithms that, depending on your MPI implementation and
Chris@10: your hardware, may be faster.
Chris@10:
Chris@10:
Chris@10:
Chris@10:
Chris@10: