MPI_Allreduce

Combines values from all processes and distributes the result back to all processes.

Syntax

• r = MPI_Allreduce(Value, Operation, Comm)

Input argument

• Value - value to send: numeric or logical array (sparse not supported).

• Operation - a string: MPI_SUM, MPI_MAX, MPI_MIN, MPI_SUM, MPI_PROD, MPI_LAND, MPI_LOR, MPI_BAND, MPI_BOR, MPI_LXOR or MPI_BXOR

• Comm - a MPI_Comm object.

Output argument

• r - received value

Description

Combines values from all processes and distributes the result back to all processes.

Nelson does not check to ensure that the reduction operation are all the same size across the various processes in the group.

Please be sure that each process passes the same sized array to the MPI_Allreduce operation.

See also

MPI_Reduce.

Example

mpiexec([modulepath('mpi'), '/examples/help_examples/MPI_Allreduce.m'], 4)

if ~MPI_Initialized()
  MPI_Init();
end
comm = MPI_Comm_object();
my_rank = MPI_Comm_rank ();
num_ranks = MPI_Comm_size();

A = [1 + my_rank:3 + my_rank]
B = MPI_Allreduce(A, 'MPI_PROD', comm);
if (my_rank == 0)
  disp('Result:')
  disp(B);
end
if MPI_Initialized()
  MPI_Finalize();
end

History

Version	Description
1.0.0	initial version

Author

Allan CORNET