Template Function combigrid::CombiCom::distributedGlobalSparseGridReduce

Function Documentation

template<typename SparseGridType>
void combigrid::CombiCom::distributedGlobalSparseGridReduce(SparseGridType &dsg, uint32_t maxMiBToSendPerThread, RankType globalReduceRankThatCollects = MPI_PROC_NULL, MPI_Comm globalComm = theMPISystem()->getGlobalReduceComm())

perform distributed global sparse grid reduce

sparse grid reduce: entire sparse grid is used for reduction global = between process groups; this assumes that each process group has already performed its local reduction INTO the sparse grid

Template Parameters:

SparseGridType – (derived from) DistributedSparseGridUniform

Parameters:

  • dsg – the DistributedSparseGrid

  • maxMiBToSendPerThread – the maximum number of MiB to send per OpenMP thread (~= per core) at once

  • globalReduceRankThatCollects – the rank that collects the data, or MPI_PROC_NULL if allreduce

  • globalComm – the communicator for the global reduction