release/public-v2: bit-for-bit reproducibility for regional runs when changing MPI decomposition #409
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
These PRs solve the b4b reproducibility issues for regional runs when changing the MPI layout (decomposition) as described in #288.
With the changes in NOAA-GFDL/GFDL_atmos_cubed_sphere#68, results are b4b identical in DEBUG, REPRO and PROD mode when the dycore is compiled in 64bit, regardless of the physics suite.
When the dycore is compiled din 32bit, results are b4b identical in DEBUG, REPRO and PROD mode for the RRFSv1alpha suite (or, more generally, any suite that does not use GFDL-MP with saturation adjustment = fast physics turned on). Results are b4b identical in DEBUG and REPRO mode (but not in PROD mode) for the GFSv15p2 suite (which uses GFDL-MP with saturation adjustment = fast physics turned on).
Additional regression tests were added to
tests/rt.conf
to provide the necessary test coverage going forward. Note that for release/public-v2,32BIT=Y
is the default and thus note explicitly spelled out inrt.conf
.Also: added orion.intel to the tier-1 platforms for release/public-v2.
Issue(s) addressed
These PRs solve the "b4b reproducibility for regional runs on the ESG grid when changing the MPI decomposition" of issue #288.
Testing
Regression tests passed on all tier-1 platforms for release/public-v2, including the newly added orion.intel.
Dependencies
NOAA-GFDL/GFDL_atmos_cubed_sphere#68
NOAA-EMC/fv3atm#245
#409