Parallel cluster sparse direct solver on very large matrix

Jason_d_ — Wed, 19 Aug 2015 22:19:49 GMT

Hi there,

We have some C code working for solving a big linear system that is quite sparse using the cluster sparse solver routines in MKL. We generally followed the `cl_solver_unsym_distr_c.c` example code.

We want to now use this code on an even larger sparse matrix that has over 3 billion non-zero entries (but it is, in fact, sparse). It appears that the cluster sparse solver insists on the number of non-zeros being specified by an MKL_INT, which we are obviously overflowing. Passing in a long instead of an int generates an error.

Is there a solution to this problem? Are we limited to considering matrices with nnz that is bounded by the maximum size of an integer?

Thanks, Jason

Did you link with ILP64

Gennady_F_Intel — Thu, 20 Aug 2015 04:18:44 GMT

Did you link with ILP64 libraries?

No, but that solved the

Jason_d_ — Thu, 20 Aug 2015 16:05:31 GMT

No, but that solved the problem. Thanks, life is better now!

topic Did you link with ILP64 in Intel® oneAPI Math Kernel Library

Parallel cluster sparse direct solver on very large matrix

Did you link with ILP64

No, but that solved the