PCAIRSetProcessorAgglom#

Sets whether PCAIR uses a graph partitioner to repartition coarse grids and reduce the number of active MPI ranks

Synopsis#

PETSC_EXTERN PetscErrorCode PCAIRSetProcessorAgglom(PC pc, PetscBool input_bool)

Logically Collective

Input Parameters#

  • pc - the PCAIR preconditioner context

  • input_bool - PETSC_TRUE to repartition the coarse grids onto fewer MPI ranks as the hierarchy coarsens

Options Database Key#

  • -pc_air_processor_agglom (true|false) - use a graph partitioner to repartition coarse grids and reduce the number of active MPI ranks; defaults to true

Note#

Processor agglomeration reduces the number of active MPI ranks by PCAIRSetProcessorAgglomFactor() whenever the local to non-local nonzero ratio drops below PCAIRSetProcessorAgglomRatio(), or the average number of equations per rank drops below PCAIRSetProcessEqLimit(); it is only performed where necessary, not on every level, and the entire hierarchy remains on the original communicator.

See Also#

KSP: Linear System Solvers, PCAIR, PCAIRGetProcessorAgglom(), PCAIRSetProcessorAgglomRatio(), PCAIRSetProcessorAgglomFactor(), PCAIRSetProcessEqLimit()

Level#

advanced

Index of all PC routines
Table of Contents for all manual pages
Index of all manual pages