-
Notifications
You must be signed in to change notification settings - Fork 578
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ShyLu: use MueLu as coarse solver for ShyLy/FROSch #5851
Comments
Be careful of circular dependencies. ShyLu depending on MueLu directly (even optionally) means that MueLu can't depend on ShyLu directly. (Using ShyLu as a smoother or coarse-grid solver for MueLu would also make sense.) It's probably healthier to put this kind of code fully downstream, in TrilinosCouplings. |
In this first draft, a 3D elasticity example is done. Certain problem dimensions, e.g. spatial dimension and number of nullspace vectors, is read from the xml file. That's not a great solution in the long run, but is ok to get our feet on the ground. Part of trilinos#5851.
@searhein I've made a first attempt at using MueLu as a subdomain solver for an elasticity problem. I've pushed the implementation to the branch |
@mayrmt This is going to be interesting ! I have never thought this would even be a use case. (MueLu as a subdomain solver). Are your subdomains that large that you can't use the standard approach (direct methods or incomplete methods) ? As @mhoemmen said above, we need to make sure we can enable all packages correctly. This may not be a problem if just ShyLU_DD will depend on MueLu, and not ShyLU itself. |
@srajama1 Our primary goal is actually to use MueLu as the coarse solver. However, FROSch uses the same class ( |
This issue has had no activity for 365 days and is marked for closure. It will be closed after an additional 30 days of inactivity. |
@searhein and I are still actively pursuing this. |
…6af130 From repository at [email protected]:kokkos/kokkos.git At commit: commit 36f65d03f48e504e63ad124f7f5cf74dd46af130 Merge: f4cc47a31 0f820d0d2 Author: Damien L-G <[email protected]> Date: Fri Feb 3 12:47:26 2023 -0500 Merge pull request #5851 from crtrott/no-deprecated-3-in-makefile-400 [4.0.0] Don't enable deprecated code 3 in Makefile builds anymore
This issue has had no activity for 365 days and is marked for closure. It will be closed after an additional 30 days of inactivity. |
@searhein and I are still working on this. |
…6af130 From repository at [email protected]:kokkos/kokkos.git At commit: commit 36f65d03f48e504e63ad124f7f5cf74dd46af130 Merge: f4cc47a31 0f820d0d2 Author: Damien L-G <[email protected]> Date: Fri Feb 3 12:47:26 2023 -0500 Merge pull request trilinos#5851 from crtrott/no-deprecated-3-in-makefile-400 [4.0.0] Don't enable deprecated code 3 in Makefile builds anymore
This issue has had no activity for 365 days and is marked for closure. It will be closed after an additional 30 days of inactivity. |
@searhein and I are still working on this. |
Enhancement
@trilinos/shylu
@searhein wants to experiment with MueLu as a coarse solver for his ShyLu/FROSch preconditioner. This issue is just meant to track and discuss some progress.
The text was updated successfully, but these errors were encountered: