Jan Blechta <[email protected]> writes:
> Jed, don't you build BLAS in PETSc configure by something like
> --with-openblas=1. If so, I suppose you build it without threads.

We only build reference BLAS because optimized BLAS is more of a hassle
to build and nearly irrelevant for what is used in typical sparse
iterative methods.  On the other hand, sparse direct solvers benefit
significantly from tuned Level 3.

Attachment: signature.asc
Description: PGP signature

_______________________________________________
fenics mailing list
[email protected]
http://fenicsproject.org/mailman/listinfo/fenics

Reply via email to