For this multiplication it is a dense matrix of size Nxp (Where N >> p) with a matrix of size pxp. If dense-sparse works my workaround could be to store the pxp matrix as sparse since it never gets very large anyways (and its triangular). And I don't need direct access to its data array.
So to summarize what I think I have learned: For MatMatMult in parallel sparse-dense, dense-sparse, and sparse-sparse works? And dense-dense does not because you have switched to Elemental and that has a different API that must be used? ----- Original Message ----- From: "Jed Brown" <[email protected]> To: "James A Charles" <[email protected]>, "Matthew Knepley" <[email protected]> Cc: [email protected] Sent: Friday, October 18, 2013 1:42:04 PM Subject: Re: [petsc-users] Dense Matrix setting local rows / local columns Question James A Charles <[email protected]> writes: > Thank you all for the responses. I was able to figure out the issue > with the dense matrix. I had another question related to dense > matrices. Does there exist support for multiplication of a sparse A > and dense B matrix matrix multiplication in parallel? I'm using Petsc > Release Version 3.4.2 and I am getting an error "No support for this > operation for this object type! MatMatMult not supported for B of type > mpidense" It should be supported. Please send the entire error message. (Always send the entire error message.)
