Does this seem to help you?


PS: Didn't email the list at first, so sorry for the duplicate message.

On 3/5/07, Roy Stogner <> wrote:

Is there any way to tell PETSc to only build a preconditioner once,
rather than once per solve?  I've got an application for which I'm
doing transient linear solves at the moment and for which even the
nonlinear problem shouldn't be very nonlinear, and it would be great
to run "-pc_type lu -pc_build_once" or something similar to speed up
the solves.

Better still, is there any way to tell libMesh when to build a
preconditioner and when to just reuse an existing preconditioner?  It
would sometimes be useful to build new preconditioners often, but not
on every linear solve.

Take Surveys. Earn Cash. Influence the Future of IT
Join's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
Libmesh-users mailing list