Geant4 Profiling and Benchmarking

Other Test Results

Geant4 Version Application Performance Summary
10.0.r03gcc447 SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r03 33 SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r03 32 SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r03 31 SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r03 30 SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r03 GT SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r03 T1 SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r01b SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r01a SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r01em SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
10.0.r00a SimplifiedCalo Simple Profiler Memory Profiler CPU MEM
(10.0.r03 33) rerun 32 again starting from 9:30am
(10.0.r03 32) after reconfiguing crob jobs
(10.0.r03 31) fix the sar counter and collect sar profiling results
(10.0.r03 30) using the new pbs configuration with the maximum 30 queues per node
(10.0.r03 GT) write outputs to local (worker) and copy them to the sever once each pbs job is done
(10.0.r03 T1) test a couple of samples with N_queue=30 and process them one at a time
(10.0.r01b) r01 + inline G4Track::G4Track + inline G4StepPoint::operator= + inline G4NavigationLevel::operator=
(10.0.r01a) r01 + uninline FindBinLocation (to measure the effect of the inlining)
(10.0.r01em) r01 + emstand-V10-00-09 (contributed by V. Ivantchenko/U. Laszlo)
(10.0.r00a) EM review r00a (contributed by K. Genser)