options

gmx_mpi - 2024-08-05 19:41:40 - MAQAO 2.20.7

Help is available by moving the cursor above any symbol or by checking MAQAO website.

Global Metrics

Total Time (s)18.23
Profiled Time (s)16.60
Time in analyzed loops (%)48.9
Time in analyzed innermost loops (%)41.1
Time in user code (%)52.0
Compilation Options Score (%)100
Array Access Efficiency (%)52.1
Potential Speedups
Perfect Flow Complexity1.02
Perfect OpenMP + MPI + Pthread1.23
Perfect OpenMP + MPI + Pthread + Perfect Load Distribution2.14
No Scalar IntegerPotential Speedup1.03
Nb Loops to get 80%14
FP VectorisedPotential Speedup1.02
Nb Loops to get 80%8
Fully VectorisedPotential Speedup1.10
Nb Loops to get 80%33
FP Arithmetic OnlyPotential Speedup1.10
Nb Loops to get 80%32

CQA Potential Speedups Summary

Loop Based Profile⏎

Innermost Loop Based Profile⏎

Application Categorization⏎

Compilation Options⏎

Source ObjectIssue
▼libgromacs_mpi.so.9.0.0–
○fft5d.cpp
○threaded_force_buffer.cpp
○pme_pp.cpp
○pme_gather.cpp
○listed_forces.cpp
○simd_prune_kernel.cpp
○partition.cpp
○settle.cpp
○pairlist.cpp
○update.cpp
○md_support.cpp
○kernel_common.cpp
○mdatoms.cpp
○lincs.cpp
○pbc.cpp
○calc_verletbuf.h
○pme_redistribute.cpp
○domdec_specatomcomm.cpp
○pme_grid.cpp
○localtopology.cpp
○pme_solve.cpp
○pme_spread.cpp
○calc_verletbuf.cpp
○simd_kernel.h
○fft_mkl.cpp
○bonded.cpp
○sim_util.cpp
○grid.cpp
○vec.h
○domdec_constraints.cpp
○pairs.cpp
○domdec.cpp
○atomdata.cpp

Loop Path Count Profile⏎

Cumulated Speedup If No Scalar Integer⏎

Cumulated Speedup If FP Vectorized⏎

Cumulated Speedup If Fully Vectorized⏎

Cumulated Speedup If FP Arithmetic Only⏎

Experiment Summary

Application../../install_gcc/bin/gmx_mpi
Timestamp2024-08-05 19:41:40 Universal Timestamp1722879700
Number of processes observed48 Number of threads observed192
Experiment TypeMPI; OpenMP;
Machineins01.benchmarkcenter.megware.com
Model NameAMD EPYC 9654 96-Core Processor
Architecturex86_64 Micro ArchitectureZEN_V4
Cache Size1024 KB Number of Cores96
OS VersionLinux 5.14.0-427.18.1.el9_4.x86_64 #1 SMP PREEMPT_DYNAMIC Tue May 28 06:27:02 EDT 2024
Architecture used during static analysisx86_64 Micro Architecture used during static analysisZEN_V4
Frequency Driveracpi-cpufreq Frequency Governorperformance
Huge Pagesalways Hyperthreadingon
Number of sockets2 Number of cores per socket96
Compilation Optionslibgromacs_mpi.so.9.0.0: GNU C++17 13.2.0 -march=skylake-avx512 -g -O3 -std=c++17 -fno-omit-frame-pointer -fPIC -fexcess-precision=fast -funroll-all-loops -fopenmp
CommentsGROMACS 2024.2 compiled with gcc 13.2 running on two 96 cores AMD Zen 4 processors, using 48 MPI ranks and 4 OMP threads per MPI rank. Pinning is controlled by GROMACS.

Configuration Summary

Dataset
Run Command<executable> mdrun -s ion_channel.tpr -nsteps 10000 -pin on -deffnm gcc
MPI Commandmpirun -genv I_MPI_FABRICS=shm -n <number_processes>
Number Processes48
Number Nodes1
Number Processes per Nodes48
FilterNot Used
Profile StartNot Used
×