Question about higher resolution topography at global scale and MPI interfaces. #1698
Closed
lsawade
started this conversation in
SPECFEM3D_GLOBE [General]
Replies: 1 comment 1 reply
-
hi Lucas, is this for the globe version? maybe adding this as an issue there would be a better place. in general, if something with the MPI interface is messed up, it might be a numerical precision thing. these interface points get sorted on the different processes and then a match on both sides is found. thus, if the sorting is different due to slightly different positioning, it might fail. to reproduce this error, it will depend on your topo file, OS and compilers - you would need to add all those infos... |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I have been playing around with higher resolution topography and associated smoothing function. When I use anything above ETOPO1 (e.g. ETOPO2, ETOPO1), something weird happens at the MPI interfaces. Since the error is caught, I have to assume that is expected. Are elements deleted/added depending on stretching?
ERROR:
PS: Happy to move this to issues.
Beta Was this translation helpful? Give feedback.
All reactions