Skip to content
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.

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

Update GPU workflows #35331

Merged
merged 4 commits into from
Sep 24, 2021
Merged

Conversation

fwyzard
Copy link
Contributor

@fwyzard fwyzard commented Sep 18, 2021

PR description:

Add new GPU workflows, that run the Patatrack pixel local and pixel-only track reconstruction in addition to the full reconstruction, with the possibility of offloading to GPUs also the ECAL and HCAL local reconstruction.

PR validation:

Validated with

$ runTheMatrix.py -w upgrade -j 4 -t 8 -l 11634.591,11634.592,11634.595,11634.596
...
Running up to 4 concurrent jobs, each with 8 threads per process
...
11634.591_TTbar_14TeV+2021_Patatrack_CPU+TTbar_14TeV_TuneCP5_GenSim+Digi+Reco+HARVEST Step0-PASSED Step1-PASSED Step2-PASSED Step3-PASSED  - time date Sat Sep 18 12:26:05 2021-date Sat Sep 18 12:21:24 2021; exit: 0 0 0 0
11634.592_TTbar_14TeV+2021_Patatrack_GPU+TTbar_14TeV_TuneCP5_GenSim+Digi+Reco+HARVEST Step0-PASSED Step1-PASSED Step2-PASSED Step3-PASSED  - time date Sat Sep 18 12:26:12 2021-date Sat Sep 18 12:21:25 2021; exit: 0 0 0 0
11634.595_TTbar_14TeV+2021_Patatrack_TripletsCPU+TTbar_14TeV_TuneCP5_GenSim+Digi+Reco+HARVEST Step0-PASSED Step1-PASSED Step2-PASSED Step3-PASSED  - time date Sat Sep 18 12:26:15 2021-date Sat Sep 18 12:21:25 2021; exit: 0 0 0 0
11634.596_TTbar_14TeV+2021_Patatrack_TripletsGPU+TTbar_14TeV_TuneCP5_GenSim+Digi+Reco+HARVEST Step0-PASSED Step1-PASSED Step2-PASSED Step3-PASSED  - time date Sat Sep 18 12:26:10 2021-date Sat Sep 18 12:21:26 2021; exit: 0 0 0 0
4 4 4 4 tests passed, 0 0 0 0 failed

Note: this PR includes the changes from #34978; it can be rebased once that is merged.

When the gpu modifier is enabled, make the Pixel, ECAL and HCAL
reconstruction offloadable to a GPU.
Simplify the Pixel local reconstruction running on GPU, taking advatage of cms-sw#34259.
@cmsbuild
Copy link
Contributor

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-35331/25377

@cmsbuild
Copy link
Contributor

A new Pull Request was created by @fwyzard (Andrea Bocci) for master.

It involves the following packages:

  • Configuration/PyReleaseValidation (pdmv, upgrade)
  • HLTrigger/Configuration (hlt)

@cmsbuild, @jordan-martins, @bbilin, @wajidalikhan, @Martin-Grunewald, @AdrianoDee, @srimanob, @kskovpen can you please review it and eventually sign? Thanks.
@makortel, @Martin-Grunewald, @fabiocos, @slomeo, @kpedro88 this is something you requested to watch as well.
@perrotta, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

@fwyzard
Copy link
Contributor Author

fwyzard commented Sep 18, 2021

please test

@fwyzard fwyzard changed the title Update gpu workflows 121x Update GPU workflows Sep 18, 2021
# there is no customisation for enabling the Patatrack pixel quadruplets running only on the CPU
},
reco = {
# skip the @pixelTrackingOnlyValidation which cannot run together with the full reconstruction
Copy link
Contributor Author

Choose a reason for hiding this comment

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

Copy link
Contributor

Choose a reason for hiding this comment

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

@sroychow @arossi83 @tsusa @connorpa please take note of this issue triggered in the phase1 pixel validation code when using the pixelonly reconstruction (by the way if the issue is only in SiPixelPhase1HitsV I suppose one could cook temporarily a lighter ad hoc validation sequence, just removing that module).

@cmsbuild
Copy link
Contributor

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-335860/18730/summary.html
COMMIT: f776cab
CMSSW: CMSSW_12_1_X_2021-09-18-1100/slc7_amd64_gcc900
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/35331/18730/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

  • No significant changes to the logs found
  • Reco comparison results: 0 differences found in the comparisons
  • DQMHistoTests: Total files compared: 40
  • DQMHistoTests: Total histograms compared: 3211080
  • DQMHistoTests: Total failures: 0
  • DQMHistoTests: Total nulls: 0
  • DQMHistoTests: Total successes: 3211058
  • DQMHistoTests: Total skipped: 22
  • DQMHistoTests: Total Missing objects: 0
  • DQMHistoSizes: Histogram memory added: 0.0 KiB( 39 files compared)
  • Checked 169 log files, 37 edm output root files, 40 DQM output files
  • TriggerResults: no differences found

@Martin-Grunewald
Copy link
Contributor

+1

@srimanob
Copy link
Contributor

+Upgrade

This PR introduces new GPU workflows as the title said.

@qliphy
Copy link
Contributor

qliphy commented Sep 23, 2021

@cms-sw/pdmv-l2 Any comment?

@kskovpen
Copy link
Contributor

+1

@cmsbuild
Copy link
Contributor

This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @perrotta, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2)

@qliphy
Copy link
Contributor

qliphy commented Sep 24, 2021

+1

@cmsbuild cmsbuild merged commit 701c067 into cms-sw:master Sep 24, 2021
@fwyzard fwyzard deleted the update_GPU_workflows_121x branch September 30, 2021 08:41
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

7 participants