d@0: d@0: d@0: FFTW MPI Transposes - FFTW 3.2alpha3 d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0: d@0:
d@0:

d@0: d@0: Next: , d@0: Previous: Other Multi-dimensional Real-data MPI Transforms, d@0: Up: Distributed-memory FFTW with MPI d@0:

d@0:
d@0: d@0:

6.7 FFTW MPI Transposes

d@0: d@0:

d@0: The FFTW's MPI Fourier transforms rely on one or more global d@0: transposition step for their communications. For example, the d@0: multidimensional transforms work by transforming along some d@0: dimensions, then transposing to make the first dimension local and d@0: transforming that, then transposing back. Because global d@0: transposition of a block-distributed matrix has many other potential d@0: uses besides FFTs, FFTW's transpose routines can be called directly, d@0: as documented in this section. d@0: d@0:

d@0: d@0: d@0: