Abstract: We consider the distributed memory parallel multiplication of a sparse matrix by a dense matrix (SpMM). The dense matrix is often a collection of dense vectors. Standard implementations will ...