Chris@42: Next: FFTW MPI Wisdom, Previous: Other Multi-dimensional Real-data MPI Transforms, Up: Distributed-memory FFTW with MPI [Contents][Index]
Chris@42:The FFTW’s MPI Fourier transforms rely on one or more global Chris@42: transposition step for their communications. For example, the Chris@42: multidimensional transforms work by transforming along some Chris@42: dimensions, then transposing to make the first dimension local and Chris@42: transforming that, then transposing back. Because global Chris@42: transposition of a block-distributed matrix has many other potential Chris@42: uses besides FFTs, FFTW’s transpose routines can be called directly, Chris@42: as documented in this section. Chris@42:
Chris@42:| • Basic distributed-transpose interface: | Chris@42: | |
| • Advanced distributed-transpose interface: | Chris@42: | |
| • An improved replacement for MPI_Alltoall: | Chris@42: |