I am trying to use MKL to solve sparse matrix multiplication (i.e. C=A*B) in parallel. I found there is sparse blas for sequential code but I am wondering if there is a parallel version of it.
one clarification -- are you asking about mkl_scsrmultcsr(...) routine?
Thanks for your reply. Yes, this is the subroutine I refer to. However, I notice that mkl_scrsrmultcsr() is for sequential codes only?
Hi Likun, Looking at the sources code I see that the general case is threaded but the 'trans' case has serial implementation. What is your case?
Hi Gennady, thanks. I am wondering if mkl_csrmultcsr can be used in mpi (distributed memory) setting.
Hi Likun. MKL supports Sparse BLAS routines for shared memory system only. thanks, Gennady
For more complete information about compiler optimizations, see our Optimization Notice.