[CIG-CS] Variable viscosity Stokes solver
Walter Landry
walter at geodynamics.org
Mon Apr 11 02:58:51 PDT 2011
Ok. I have moved this to the cig-cs list.
As for which parts of Trilinos are used, I am not sure. I believe it
uses BoomerAMG. Wolfgang would know best. I am using Debian's
package which includes everything.
Cheers,
Walter Landry
walter at geodynamics.org
mspieg <mspieg at ldeo.columbia.edu> wrote:
> On Apr 8, 2011, at 12:18 PM, Walter Landry wrote:
>
>>
> mspieg <mspieg at ldeo.columbia.edu> wrote:
>
> Hi All,
> and thanks for keeping me in the loop.
>
> Maybe we should move this whole discussion to the cig-cs list?
>
> probably not a bad idea...
>
>>
>
>>
>
> I have a version of deal.ii built here (although I think I may need
> to upgrade again). Is there a repo for these benchmarks that I can
> pull?
>
> The deal II benchmarks I have been discussing are available via mercurial
>
> hg clone http://geodynamics.org/hg/cs/AMR/dealII_viscosity
>
> great...
>
>>
>
>>
> The SAMRAI code is available at
>
>>
> hg clone http://geodynamics.org/hg/cs/AMR/Gamr
> hg clone http://geodynamics.org/hg/cs/AMR/SAMRAI
>
>>
> SAMRAI is not too difficult to build, but Gamr is a bit tricky. It is
> still 2D at this point.
>
> I'll take a look...
>
>>
>
>>
>
> I have another developer working with me here (Cian Wilson) on some
> unstructured FEM codes in FEniCS and it would be fun to get the same
> benchmarks up and going there as well for comparison.
>
> >
> I already have a MADDs-1 version done in FEniCS/PETSc with block
> preconditioners I could clean up.
>
> >
> re: trilinos... I take it you need trilinos because you're building on Samrai?
>
> trilinos is only needed for the deal II code.
>
> Okay, I've only used the petsc back ends with deal.ii...which parts of
> trilinos are needed?
> m
>
>>
>
>>
> Cheers,
> Walter Landry
> walter at geodynamics.org
>
> ----------------------------------------------------
> Marc Spiegelman
> Lamont-Doherty Earth Observatory
> Dept. of Applied Physics/Applied Math
> Columbia University
> http://www.ldeo.columbia.edu/~mspieg
> tel: 845 704 2323 (SkypeIn)
> ----------------------------------------------------
More information about the CIG-CS
mailing list