Applying Collective Operations to Intercommunicators (original) (raw)

Up: Communicator Argument Next: Specifics for Intercommunicator Collective Operations Previous: Specifics for Intracommunicator Collective Operations

To understand how collective operations apply to intercommunicators, we can view most MPI intracommunicator collective operations as fitting one of the following categories (see, for instance, [56]):

All-To-All

All processes contribute to the result. All processes receive the result.

• MPI_ALLGATHER, MPI_IALLGATHER, MPI_ALLGATHERV, MPI_IALLGATHERV
• MPI_ALLTOALL, MPI_IALLTOALL, MPI_ALLTOALLV, MPI_IALLTOALLV, MPI_ALLTOALLW, MPI_IALLTOALLW
• MPI_ALLREDUCE, MPI_IALLREDUCE, MPI_REDUCE_SCATTER_BLOCK, MPI_IREDUCE_SCATTER_BLOCK, MPI_REDUCE_SCATTER, MPI_IREDUCE_SCATTER
• MPI_BARRIER, MPI_IBARRIER

All-To-One

All processes contribute to the result. One process receives the result.

• MPI_GATHER, MPI_IGATHER, MPI_GATHERV, MPI_IGATHERV
• MPI_REDUCE, MPI_IREDUCE

One-To-All

One process contributes to the result. All processes receive the result.

• MPI_BCAST, MPI_IBCAST
• MPI_SCATTER, MPI_ISCATTER, MPI_SCATTERV, MPI_ISCATTERV

Other

Collective operations that do not fit into one of the above categories.

• MPI_SCAN, MPI_ISCAN, MPI_EXSCAN, MPI_IEXSCAN

The data movement patterns of MPI_SCAN, MPI_ISCAN, MPI_EXSCAN, and MPI_IEXSCANdo not fit this taxonomy.

The application of collective communication to intercommunicators is best described in terms of two groups. For example, an all-to-all MPI_ALLGATHER operation can be described as collecting data from all members of one group with the result appearing in all members of the other group (see Figure 2 ). As another example, a one-to-all MPI_BCAST operation sends data from one member of one group to all members of the other group. Collective computation operations such as MPI_REDUCE_SCATTER have a similar interpretation (see Figure 3 ). For intracommunicators, these two groups are the same. For intercommunicators, these two groups are distinct. For the all-to-all operations, each such operation is described in two phases, so that it has a symmetric, full-duplex behavior.

The following collective operations also apply to intercommunicators:

• MPI_BARRIER, MPI_IBARRIER
• MPI_BCAST, MPI_IBCAST
• MPI_GATHER, MPI_IGATHER, MPI_GATHERV, MPI_IGATHERV,
• MPI_SCATTER, MPI_ISCATTER, MPI_SCATTERV, MPI_ISCATTERV,
• MPI_ALLGATHER, MPI_IALLGATHER, MPI_ALLGATHERV, MPI_IALLGATHERV,
• MPI_ALLTOALL, MPI_IALLTOALL, MPI_ALLTOALLV, MPI_IALLTOALLV, MPI_ALLTOALLW, MPI_IALLTOALLW,
• MPI_ALLREDUCE, MPI_IALLREDUCE, MPI_REDUCE, MPI_IREDUCE,
• MPI_REDUCE_SCATTER_BLOCK, MPI_IREDUCE_SCATTER_BLOCK, MPI_REDUCE_SCATTER, MPI_IREDUCE_SCATTER.

Figure 2: Intercommunicator allgather. The focus of data to one process is represented, not mandated by the semantics. The two phases do allgathers in both directions.

Figure 3: Intercommunicator reduce-scatter. The focus of data to one process is represented, not mandated by the semantics. The two phases do reduce-scatters in both directions.

Up: Communicator Argument Next: Specifics for Intercommunicator Collective Operations Previous: Specifics for Intracommunicator Collective Operations

Return to MPI-3.1 Standard Index
Return to MPI Forum Home Page

(Unofficial) MPI-3.1 of June 4, 2015
HTML Generated on June 4, 2015