Add RAJA view performance test to benchmark #1728
Merged
Conversation
This file contains 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
|
@artv3 move to benchmark folder. |
|
I ran the code on Pascal and got the following results: |
|
What gpus are on pascal? It seems strange that they are ~24x slower than the V100s on lassen? |
|
Tesla P100-PCIE-16GB I believe. |
|
Are we running the same code in the same way on both of these platforms? How did you build? |
|
We found that the issue was that my code built for debugging which was causing the slowdown. |
artv3
changed the title
johnbowen42
approved these changes
Sep 17, 2024
|
@johnbowen42 @rhornung67 can I get another review? I just pushed up the changes I thought I had pushed up. |
rhornung67
approved these changes
Sep 27, 2024
adayton1
approved these changes
Sep 27, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
After resolving issue #1718, this PR now adds the performance test into the bench mark folder.
//------------
This PR adds the code provided in issue #1718 in an effort to reproduce the slow down.
I don't have access to pascal but on lassen I see comparable performance:
Elapsed time with RAJA view : 0.0951086
Elapsed time with NO RAJA view : 0.0952884
To avoid measuring stream initialization I added an basic forall at the start of the program.
Compiler setup:
nvcc: nvcc11.2.0, cuda_arch=70, gcc8.3.1
I also tried: nvcc11.8.0, cuda_arch=70, gcc8.3.1
Elapsed time with RAJA view : 0.0949394
Elapsed time with NO RAJA view : 0.0949237