Geant4 Profiling and Benchmarking

Geant4 CPU Performance by Version (Geant4.10.x-series)

1) The Current profiling activity is a part of Geant4 Computing Performance Task

2) Profiling Results

Profiled on the Wilson CPU cluster using AMD 6128HE Opteron 2GHz
(green: gcc 7.1.0 -O3), (pink: gcc 6.3.0 -O2), (yellow: gcc 4.9.2 -O2)
Geant4 Version Application Performance Summary
Shielding.10.6.r00.g4ndl4.5 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
Shielding.10.6.r00 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.6 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.6.c01 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.6.c00 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r10c SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r10 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r09rr SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r09 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r08rr SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r08 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r07rr SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r07 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.master.july.09.19 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r06 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r04rrr SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.6.b01 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.6.b00 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r05+f1e50db SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r05noverb SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r05 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r04rr SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r04 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r03rr(static) SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r02static SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r03 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r02rerun SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r02 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r01rr SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.r01 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.p01static SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5.p01 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.5 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.4.p03 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.3.p03 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.2.p03 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.4 cmsExp-2018 Open|Speedshop IgProf(Memory) CPU MEM
10.3.p03 cmsExp-2018 Open|Speedshop IgProf(Memory) CPU MEM
10.2.p03 cmsExp-2018 Open|Speedshop IgProf(Memory) CPU MEM
Geant4 Version Application Performance Summary
10.4 LArTest Open|Speedshop IgProf(Memory) CPU MEM

Old Profiling Results: 9.4 9.5 9.6 10.0 10.1 10.2 10.3 10.4

3) CPU per Event: Summary Plots by Versions

SimplifiedCalo PYTHIA H->ZZ electrons pions protons anti-protons gamma

4) Total Memory Count: Summary Plots by Versions

SimplifiedCalo PYTHIA H->ZZ electrons pions protons anti-protons gamma

6) Geant4 MT Performance

Geant4 Version Application Performance
10.6 cmsExpMT AMD Intel Open|SpeedShop
10.6.c01 cmsExpMT AMD Intel Open|SpeedShop
10.6.c00 cmsExpMT AMD Intel Open|SpeedShop
10.5.r10 cmsExpMT AMD Intel Open|SpeedShop
10.5.r09 cmsExpMT AMD Intel Open|SpeedShop
10.5.r08 cmsExpMT AMD Intel Open|SpeedShop
10.5.r07 cmsExpMT AMD Intel Open|SpeedShop
10.5.r06 cmsExpMT AMD Intel Open|SpeedShop
10.5.r05+f1e50db cmsExpMT AMD Intel Open|SpeedShop
10.5.r04 cmsExpMT AMD Intel Open|SpeedShop
10.5.r03 cmsExpMT AMD Intel Open|SpeedShop
10.5.r02 cmsExpMT AMD Intel Open|SpeedShop
10.5.r01 cmsExpMT AMD Intel Open|SpeedShop
10.5.p01static cmsExpMT AMD Intel Open|SpeedShop
10.5.p01 cmsExpMT AMD Intel Open|SpeedShop
10.5 cmsExpMT AMD Intel Open|SpeedShop
10.4.p03 cmsExpMT AMD Intel Open|SpeedShop
10.3.p03 cmsExpMT AMD Intel Open|SpeedShop
ParFullCMS XeonPhi XeonPhi
10.2.p03 cmsExpMT AMD Intel Open|SpeedShop
ParFullCMS XeonPhi XeonPhi
Geant4 Version Application Performance
10.3.r06dev lArTestMT AMD Intel Open|SpeedShop

7) Other Test Results and Activities

Geant4 Version Application Performance Summary
10.5.cf274b0 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.4.r02+event-V10-04-04 SimplifiedCalo Open|Speedshop IgProf(Memory) CPU MEM
10.4.Beta (VecGeom+cms2018) cmsExp (Vector) Open|Speedshop IgProf(Memory) CPU MEM
10.4.Beta (VecGeom+cms2018) cmsExp (Scalar) Open|Speedshop IgProf(Memory) CPU MEM
10.4.Beta (cms2018**) cmsExp Open|Speedshop IgProf(Memory) CPU MEM
10.4.Beta (VecGeom*) cmsExp Open|Speedshop IgProf(Memory) CPU MEM
**: cms2018 (the upgraded CMS pixel tracker and muon system) *: VecGeom v00.04.00
  • Code reviews (summary pdf files): CHIPS Physics Field Propagation EM Physics

  • KISTI IgProf Results - Beam.PhysList.Energy=Iodine.QGSP_BIC_EMY.35
  • 8) Useful Links for Performance Tools and Optimization

    1. Open|Speedshop: Home page
    2. IgProf: Ignominous Profiler is a simple tool for measuring and analysing application memory and
      performance characteristics. For more information, see IgProf home page

    3. Other HPC Performance Tools: HPCToolkits TAU
    4. Software Optimization Resources: Agner Fog