Differents number of iterations with the same problem

Barry Smith bsmith at mcs.anl.gov
Wed Feb 14 16:47:03 CST 2007

  With jacobi it probably requires a lot of iterations? Then it would
not supprise me to have a slight difference in iteration count. If
it is taking 1000's of iterations then I would expect large differences
in iteration count.


On Wed, 14 Feb 2007, LECAS Dimitri wrote:

> Hello
> I'm surprised to not have the same numbers of iterations when i run
> several instance of my program with the same number of processors and
> the same matrix/right hand side. Is there some random or it's
> asynchronous message passing fault ?
> PS:  My code used KSPSolve with bicg and jacobi for pc)

More information about the petsc-users mailing list