Refactoring + Additional Optimizations #43
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@anandrdbz @wilfonba CI failed. Why have all the golden files changed? |
@sbryngelson should be fixed now, there was a deallocate statement in m_viscous that was causing the bug (which only shows on GCC) |
sbryngelson
approved these changes
Nov 30, 2022
| @@ -0,0 +1,4 @@ | |||
| #!bash | |||
|
|
|||
| qL_rsx_vf_flat, qL_rsy_vf_flat, qL_rsz_vf_flat, & | ||
| qR_rsx_vf_flat, qR_rsy_vf_flat, qR_rsz_vf_flat, & | ||
| qL_rsx_vf, qL_rsy_vf, qL_rsz_vf, & | ||
| qR_rsx_vf, qR_rsy_vf, qR_rsz_vf, & | ||
| id) | ||
|
|
||
| iv%beg = mom_idx%beg; iv%end = mom_idx%end | ||
| if (weno_Re_flux) then |
There was a problem hiding this comment.
it seems like this subroutine s_reconstruct_cell_boundary_values_visc_deriv should be moved to the viscous module?
There was a problem hiding this comment.
I must have missed that one. I'll move it.
There was a problem hiding this comment.
thanks @wilfonba -- i already pulled your request so please create a new one for it.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
Code is fully refactored (RHS only 2800 lines) and passes all test cases on GPUs. Dead code removed from m_cbc, m_weno, m_monopole and m_bubbles as well (FYPP, removing _flat, _alt etc)
Run times are identical (slightly quicker in fact) for the fully resolved cases.
Sub-grid bubbles, which was previously quite slow on GPUs, now runs 20X faster (had to modify the bubble source kernel). Now, they exhibit identical speedups to the fully resolved case.
QBMM cases also show slight (5%) improvement from previous state. However GPU speedup is limited (3-4X slower than fully resolved / bubbles) due to the vast number of power operations (This is quite slow on OpenACC + F90).
In summary, runtimes for all cases are either identical to previous state or significantly better, with a vast reduction in number of lines