Sparse matrix vector product performance

Sparse matrix vector product performance

Hello,

mkl sparse matrix dense vector product is about 30% slower than my own routine.
Tested with a symmetric matrix from fem unstructured mesh, on amd64 and xeon platforms and mkl version 8.1, 9.1 and 10.

I am puzzled by these results, has anybody benchmarked those mkl routines ?

Regards,
Boolegue.

1 contribution / 0 nouveau(x)
Reportez-vous à notre Notice d'optimisation pour plus d'informations sur les choix et l'optimisation des performances dans les produits logiciels Intel.