cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: FFTW 3.3.5: FFTW MPI Transposes cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127:
cannam@127:

cannam@127: Next: , Previous: , Up: Distributed-memory FFTW with MPI   [Contents][Index]

cannam@127:
cannam@127:
cannam@127: cannam@127:

6.7 FFTW MPI Transposes

cannam@127: cannam@127: cannam@127:

The FFTW’s MPI Fourier transforms rely on one or more global cannam@127: transposition step for their communications. For example, the cannam@127: multidimensional transforms work by transforming along some cannam@127: dimensions, then transposing to make the first dimension local and cannam@127: transforming that, then transposing back. Because global cannam@127: transposition of a block-distributed matrix has many other potential cannam@127: uses besides FFTs, FFTW’s transpose routines can be called directly, cannam@127: as documented in this section. cannam@127:

cannam@127: cannam@127: cannam@127: cannam@127: cannam@127:
• Basic distributed-transpose interface:   cannam@127:
• Advanced distributed-transpose interface:   cannam@127:
• An improved replacement for MPI_Alltoall:   cannam@127:
cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: cannam@127: