Skip to content

[test] Add simple FFTW benchmark #776

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.

Sign up for GitHub

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

Jump to bottom
Merged
merged 6 commits into from
May 27, 2019
Merged

[test] Add simple FFTW benchmark #776

merged 6 commits into from
May 27, 2019

Conversation

@ajocksch
Copy link
Contributor

@ajocksch ajocksch commented May 9, 2019

No description provided.

@ajocksch ajocksch self-assigned this May 9, 2019
@vkarak
Copy link
Contributor

vkarak commented May 9, 2019

@ajocksch Please read the PR guidelines.

@ajocksch ajocksch changed the title WIP: first version of fftw benchmark WIP: [test] first version of fftw benchmark May 9, 2019
@vkarak vkarak added this to the ReFrame sprint 2019w20 milestone May 13, 2019
@vkarak vkarak requested review from victorusu and vkarak May 14, 2019 11:30
@ajocksch ajocksch changed the title WIP: [test] first version of fftw benchmark [test] first version of fftw benchmark May 14, 2019
@ajocksch ajocksch changed the title [test] first version of fftw benchmark [test] This PR contains a simple fftw benchmark May 14, 2019
@vkarak vkarak changed the title [test] This PR contains a simple fftw benchmark [test] Add simple FFTW benchmark May 14, 2019
vkarak
vkarak requested changes May 14, 2019
View reviewed changes
vkarak
vkarak reviewed May 14, 2019
View reviewed changes
teojgo
teojgo requested changes May 14, 2019
View reviewed changes
@ajocksch
Copy link
Contributor Author

ajocksch commented May 15, 2019 •
edited by vkarak
Loading

On kesch fftw is much slower for PrgEnv-gnu than for other programming environments (more than factor 10). This is without the MPI option on one node.

==============================================================================
SUMMARY OF FAILURES
------------------------------------------------------------------------------
FAILURE INFO for FFTWTest_withoutmpi 
  * System partition: kesch:cn
  * Environment: PrgEnv-gnu
  * Stage directory: /scratch/ajocksch/stage/kesch/cn/PrgEnv-gnu/FFTWTest_withoutmpi
  * Job type: batch job (id=1598028)
  * Maintainers: ['AJ']
  * Failing phase: performance
  * Reason: performance error: 7.374048 is beyond reference value 0.61 (l=-inf, u=0.6405)
------------------------------------------------------------------------------
==============================================================================
PERFORMANCE REPORT
------------------------------------------------------------------------------
FFTWTest_withoutmpi
- kesch:cn
   - PrgEnv-cray
      * fftw_exec_time: 0.614295 s
   - PrgEnv-pgi
      * fftw_exec_time: 0.6145959 s
   - PrgEnv-gnu
      * fftw_exec_time: 7.374048 s
------------------------------------------------------------------------------
FFTWTest_withmpi
   - PrgEnv-cray
      * fftw_exec_time: 1.610656 s
   - PrgEnv-pgi
      * fftw_exec_time: 1.577487 s
   - PrgEnv-gnu
      * fftw_exec_time: 1.594661 s
------------------------------------------------------------------------------

vkarak
vkarak requested changes May 17, 2019
View reviewed changes
Copy link
Contributor

@vkarak vkarak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You should not let the check fail for Kesch with PrgEnv-gnu. Either adapt the reference or remove PrgEnv-gnu from the supported environments or investigate why there is such a difference.

vkarak
vkarak requested changes May 17, 2019
View reviewed changes
@vkarak
Copy link
Contributor

vkarak commented May 21, 2019

@teojgo Can you approve this PR?

vkarak
vkarak requested changes May 24, 2019
View reviewed changes
Copy link
Contributor

@vkarak vkarak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ajocksch You should make this test generic, so that it can be launched on unknown systems.

vkarak
vkarak approved these changes May 25, 2019
View reviewed changes
Copy link
Contributor

@vkarak vkarak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lgtm

@vkarak
Copy link
Contributor

vkarak commented May 27, 2019

@jenkins-cscs retry none

@codecov-io
Copy link

codecov-io commented May 27, 2019 •
edited
Loading

Codecov Report

Merging #776 into master will decrease coverage by 0.02%.
The diff coverage is n/a.

Impacted file tree graph

@@            Coverage Diff             @@
##           master     #776      +/-   ##
==========================================
- Coverage   91.61%   91.59%   -0.03%     
==========================================
  Files          78       78              
  Lines        9884     9884              
==========================================
- Hits         9055     9053       -2     
- Misses        829      831       +2
Impacted Files Coverage Δ
reframe/core/config.py 84.07% <0%> (-1.77%) ⬇️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 12b7110...84fee46. Read the comment docs.

@vkarak vkarak merged commit d8517e5 into reframe-hpc:master May 27, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@vkarak vkarak vkarak approved these changes

@teojgo teojgo teojgo approved these changes

@victorusu victorusu Awaiting requested review from victorusu

Assignees

@ajocksch ajocksch

Labels

prio: normal

Projects

None yet

Milestone

ReFrame sprint 2019w20

Development

Successfully merging this pull request may close these issues.

4 participants

@ajocksch @vkarak @codecov-io @teojgo