=== This is the CP2K Performance-Test === Updating e86548823..e64b33155 Fast-forward CMakeLists.txt | 4 +- README_cmake.md | 177 +++++++++++---------- cmake/FindBlis.cmake | 2 +- cmake/FindFftw.cmake | 40 +++-- cmake/FindGenericBLAS.cmake | 23 ++- cmake/FindLibSPG.cmake | 2 +- cmake/FindLibVORI.cmake | 2 +- cmake/FindLibXC.cmake | 20 ++- cmake/FindLibXSMM.cmake | 48 +++--- cmake/FindLibint2.cmake | 19 ++- cmake/FindMetis.cmake | 4 + cmake/FindOpenBLAS.cmake | 25 +-- cmake/FindQuip.cmake | 4 + cmake/FindSCI.cmake | 6 +- cmake/cp2k_utils.cmake | 16 +- src/CMakeLists.txt | 6 +- tools/docker/Dockerfile.test_cmake | 5 +- tools/docker/generate_dockerfiles.py | 5 +- tools/docker/scripts/install_dbcsr.sh | 24 +-- tools/toolchain/scripts/stage2/install_openblas.sh | 4 + tools/toolchain/scripts/stage3/install_fftw.sh | 4 +- tools/toolchain/scripts/stage3/install_libint.sh | 5 +- tools/toolchain/scripts/stage3/install_libxc.sh | 2 + tools/toolchain/scripts/stage4/install_cosma.sh | 2 + tools/toolchain/scripts/stage4/install_libxsmm.sh | 3 + .../toolchain/scripts/stage4/install_scalapack.sh | 3 + tools/toolchain/scripts/stage5/install_elpa.sh | 3 +- tools/toolchain/scripts/stage5/install_pexsi.sh | 2 + tools/toolchain/scripts/stage5/install_ptscotch.sh | 3 + tools/toolchain/scripts/stage5/install_superlu.sh | 3 + tools/toolchain/scripts/stage6/install_gsl.sh | 4 +- tools/toolchain/scripts/stage6/install_plumed.sh | 3 + tools/toolchain/scripts/stage6/install_quip.sh | 3 + tools/toolchain/scripts/stage7/install_hdf5.sh | 3 +- tools/toolchain/scripts/stage7/install_libtorch.sh | 2 + tools/toolchain/scripts/stage7/install_libvdwxc.sh | 4 +- tools/toolchain/scripts/stage7/install_libvori.sh | 2 + tools/toolchain/scripts/stage7/install_spglib.sh | 4 +- tools/toolchain/scripts/stage8/install_sirius.sh | 4 + tools/toolchain/scripts/stage8/install_spfft.sh | 4 +- tools/toolchain/scripts/stage8/install_spla.sh | 4 +- 41 files changed, 306 insertions(+), 197 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: e64b331552e1588eec11f575b9e771b24f7bf608 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/01 job id: 44305821 --- Point --- name: 10 plot: h2o_32_ri_rpa_mp2 regex: Total RI-RPA Time= label: RI-RPA (8n/2r/6t) --- Point --- name: 11 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-RPA (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-MP2.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-HF-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-MP2-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/02 job id: 44305824 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/03 job id: 44305825 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/04 job id: 44305826 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/05 job id: 44305828 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/06 job id: 44305829 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/07 job id: 44305830 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/08 job id: 44305831 --- Point --- name: 110 plot: h2o_64_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 111 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/09 job id: 44305832 --- Point --- name: 200 plot: h2o_128_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 201 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/10 job id: 44305833 --- Point --- name: 202 plot: h2o_128_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 203 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/11 job id: 44305834 --- Point --- name: 204 plot: h2o_128_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 205 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/12 job id: 44305835 --- Point --- name: 206 plot: h2o_128_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 207 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/13 job id: 44305837 --- Point --- name: 208 plot: h2o_128_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 209 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/14 job id: 44305840 --- Point --- name: 210 plot: h2o_128_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 211 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/15 job id: 44305842 --- Point --- name: 400 plot: h2o_256_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 401 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/16 job id: 44305843 --- Point --- name: 402 plot: h2o_256_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 403 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/17 job id: 44305844 --- Point --- name: 404 plot: h2o_256_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 405 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/18 job id: 44305845 --- Point --- name: 406 plot: h2o_256_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 407 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/19 job id: 44305846 --- Point --- name: 408 plot: h2o_256_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 409 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/20 job id: 44305847 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/21 job id: 44305851 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/22 job id: 44305852 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/23 job id: 44305854 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/24 job id: 44305855 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/25 job id: 44305857 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/26 job id: 44305859 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.036 134.988 134.989 farming_run 1 2.0 134.419 134.420 134.960 134.965 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 4194304 0.0% 0.0% 100.0% flops 14 x 32 x 32 154140672 0.0% 0.0% 100.0% flops 29 x 32 x 32 159645696 0.0% 0.0% 100.0% flops 14 x 14 x 32 208732160 0.0% 0.0% 100.0% flops 29 x 14 x 32 212860928 0.0% 0.0% 100.0% flops 14 x 29 x 32 212860928 0.0% 0.0% 100.0% flops 29 x 29 x 32 227352576 0.0% 0.0% 100.0% flops 14 x 32 x 14 896801644032 0.0% 0.0% 100.0% flops 29 x 32 x 14 928925089792 0.0% 0.0% 100.0% flops 14 x 32 x 29 928925089792 0.0% 0.0% 100.0% flops 29 x 32 x 29 962100985856 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693169221632 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753639550976 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.164741E+12 0.0% 0.0% 100.0% flops max/rank 447.801317E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249492158 0.0% 0.0% 100.0% number of processed stacks 164328 0.0% 0.0% 100.0% average stack size 0.0 0.0 1518.3 marketing flops 7.165779E+12 ------------------------------------------------------------------------------- # multiplications 1160 max memory usage/rank 1.458106E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.045 116.103 116.117 qs_energies 1 2.0 0.000 0.000 115.887 115.890 mp2_main 1 3.0 0.000 0.000 113.405 113.408 mp2_gpw_main 1 4.0 0.020 0.026 112.035 112.038 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.287 93.774 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.429 55.915 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.170 41.746 47.207 get_2c_integrals 1 6.0 0.000 0.001 37.100 37.685 integrate_v_rspace 273 8.0 0.436 0.448 25.105 30.273 pw_transfer 6555 10.6 0.373 0.379 27.361 27.865 grid_integrate_task_list 273 9.0 20.913 26.576 20.913 26.576 fft_wrap_pw1pw2 5465 11.4 0.044 0.048 26.054 26.557 fft_wrap_pw1pw2_100 2178 12.4 1.165 1.240 23.577 24.077 compute_2c_integrals 1 7.0 0.004 0.004 19.345 19.346 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.818 19.028 mp2_eri_2c_integrate_gpw 1 9.0 2.384 2.423 18.815 19.024 rpa_ri_compute_en 1 5.0 0.001 0.012 18.640 18.777 cp_fm_cholesky_decompose 12 8.2 17.726 18.296 17.726 18.296 cholesky_decomp 1 7.0 0.000 0.000 16.591 17.159 fft3d_s 5443 13.4 16.125 16.616 16.147 16.638 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.570 10.853 15.570 calculate_wavefunction 272 8.0 5.434 5.569 12.564 13.197 rpa_num_int 1 6.0 0.001 0.009 10.490 10.500 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.451 10.473 calc_mat_Q 8 8.0 0.000 0.000 9.283 9.382 contract_S_to_Q 8 9.0 0.000 0.000 8.708 8.805 calc_potential_gpw 544 9.5 0.005 0.006 8.248 8.589 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.188 8.403 parallel_gemm_fm 14 9.1 0.000 0.000 8.302 8.371 parallel_gemm_fm_cosma 14 10.1 8.302 8.371 8.302 8.371 potential_pw2rs 545 10.0 0.107 0.108 7.672 8.306 collocate_single_gaussian 272 10.0 0.040 0.042 7.427 7.659 create_integ_mat 1 6.0 0.013 0.028 7.560 7.569 array2fm 1 7.0 0.000 0.000 6.735 7.083 pw_scatter_s 2720 13.7 4.449 4.660 4.449 4.660 pw_gather_s 2722 13.2 3.907 4.227 3.907 4.227 array2fm_buffer_send 1 8.0 2.988 3.155 2.988 3.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.038117, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.038 397.824 397.825 farming_run 1 2.0 396.798 396.802 397.780 397.781 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.226256E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.050 209.862 209.863 qs_energies 1 2.0 0.000 0.000 209.571 209.579 scf_env_do_scf 1 3.0 0.000 0.000 106.774 106.774 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.856 105.864 rebuild_ks_matrix 4 6.0 0.000 0.000 105.855 105.862 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.855 105.862 hfx_ks_matrix 4 8.0 0.001 0.001 105.453 105.458 integrate_four_center 4 9.0 0.145 0.462 105.452 105.457 mp2_main 1 3.0 0.000 0.000 102.486 102.494 mp2_gpw_main 1 4.0 0.032 0.048 101.289 101.299 integrate_four_center_main 4 10.0 0.118 0.506 96.715 99.711 integrate_four_center_bin 264 11.0 96.596 99.695 96.596 99.695 init_scf_loop 1 4.0 0.000 0.000 92.469 92.469 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.824 75.865 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.397 55.438 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 42.180 47.137 integrate_v_rspace 95 8.0 0.399 0.572 28.552 33.341 pw_transfer 2240 10.6 0.143 0.163 29.947 30.350 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.951 29.355 grid_integrate_task_list 95 9.0 23.826 28.838 23.826 28.838 ao_to_mo_and_store_B_mult_1 91 7.0 10.534 28.732 10.534 28.732 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.310 28.001 fft_wrap_pw1pw2_100 730 12.4 1.286 1.438 26.656 27.113 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.824 1.886 24.623 24.633 get_2c_integrals 1 6.0 0.000 0.000 20.340 20.363 compute_2c_integrals 1 7.0 0.002 0.003 19.322 19.325 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.909 19.189 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.851 18.908 19.188 fft3d_s 1823 13.4 18.431 18.752 18.445 18.766 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.302 14.302 calculate_wavefunction 91 8.0 2.033 2.064 9.760 9.979 mp2_ri_gpw_compute_en_expansio 172 7.0 0.560 0.587 8.733 9.244 potential_pw2rs 186 10.0 0.034 0.035 8.634 9.238 local_gemm 172 8.0 8.173 8.672 8.173 8.672 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.266 8.522 calc_potential_gpw 182 9.5 0.002 0.002 7.952 8.170 collocate_single_gaussian 91 10.0 0.017 0.024 7.900 8.087 mp2_ri_gpw_compute_en_comm 22 7.0 0.542 0.560 7.678 8.070 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.449 6.344 6.449 mp_sendrecv_dm3 2068 8.0 5.603 5.974 5.603 5.974 pw_gather_s 912 13.2 4.902 5.470 4.902 5.470 mp_sync 38 10.4 3.401 5.335 3.401 5.335 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.283118, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.157440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 589622. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.044 53.705 53.711 qs_mol_dyn_low 1 2.0 0.003 0.003 53.469 53.477 qs_forces 11 3.9 0.002 0.002 53.406 53.407 qs_energies 11 4.9 0.002 0.004 51.815 51.839 scf_env_do_scf 11 5.9 0.000 0.001 45.360 45.360 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.188 43.189 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.877 34.316 qs_scf_new_mos 108 7.5 0.000 0.001 32.276 32.589 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.276 32.588 ot_scf_mini 108 9.5 0.002 0.002 30.624 30.828 multiply_cannon 2286 13.5 0.187 0.196 26.225 27.644 multiply_cannon_loop 2286 14.5 1.487 1.560 25.412 26.773 velocity_verlet 10 3.0 0.001 0.002 26.226 26.227 ot_mini 108 10.5 0.001 0.001 19.312 19.585 qs_ot_get_derivative 108 11.5 0.001 0.001 16.385 16.571 mp_waitall_1 245248 16.5 8.491 14.725 8.491 14.725 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 5.827 13.090 multiply_cannon_multrec 54864 15.5 4.245 6.486 7.906 11.515 rebuild_ks_matrix 119 8.3 0.000 0.000 8.300 8.436 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.299 8.436 multiply_cannon_sync_h2d 54864 15.5 5.889 7.478 5.889 7.478 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.338 7.468 mp_sum_l 7207 12.9 5.366 7.090 5.366 7.090 qs_ot_get_p 119 10.4 0.001 0.001 6.671 7.022 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.532 5.900 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.489 5.610 init_scf_run 11 5.9 0.000 0.001 5.166 5.166 scf_env_initial_rho_setup 11 6.9 0.009 0.012 5.166 5.166 dbcsr_mm_accdrv_process 76910 16.1 1.166 1.836 3.582 5.079 sum_up_and_integrate 119 10.3 0.012 0.015 4.836 4.842 integrate_v_rspace 119 11.3 0.002 0.002 4.823 4.830 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.636 4.734 calculate_rho_elec 119 8.7 0.011 0.017 4.636 4.733 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.732 3.844 calculate_dm_sparse 119 9.5 0.000 0.001 3.225 3.351 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.817 3.244 rs_pw_transfer 974 11.9 0.011 0.012 3.058 3.189 jit_kernel_multiply 13 15.8 2.354 3.159 2.354 3.159 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.871 3.117 apply_single 119 13.6 0.000 0.000 2.871 3.117 calculate_first_density_matrix 1 7.0 0.000 0.001 2.944 2.954 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.902 2.902 density_rs2pw 119 9.7 0.004 0.004 2.613 2.732 ot_diis_step 108 11.5 0.006 0.006 2.664 2.665 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.624 2.625 cp_fm_redistribute_end 50 14.0 2.381 2.594 2.387 2.595 cp_fm_diag_elpa_base 50 14.0 0.207 2.493 0.208 2.503 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.432 2.496 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.388 2.391 pw_transfer 1439 11.6 0.053 0.058 2.190 2.272 acc_transpose_blocks 54864 15.5 0.231 0.254 1.780 2.234 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.114 2.198 potential_pw2rs 119 12.3 0.004 0.004 2.168 2.192 wfi_extrapolate 11 7.9 0.001 0.001 2.153 2.153 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.111 2.151 init_scf_loop 11 6.9 0.000 0.000 2.150 2.150 grid_integrate_task_list 119 12.3 2.020 2.132 2.020 2.132 make_m2s 4572 13.5 0.054 0.057 1.978 2.033 mp_sum_d 4125 12.0 1.354 1.990 1.354 1.990 fft3d_ps 1201 14.6 0.369 0.475 1.885 1.970 make_images 4572 14.5 0.132 0.138 1.895 1.949 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.750 1.778 fft_wrap_pw1pw2_140 487 13.2 0.081 0.094 1.603 1.689 mp_alltoall_d11v 2130 13.8 1.419 1.595 1.419 1.595 grid_collocate_task_list 119 9.7 1.291 1.344 1.291 1.344 mp_waitany 12084 13.8 1.206 1.338 1.206 1.338 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.699 1.130 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.711000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.944192E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 100898. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 41.098 41.098 qs_mol_dyn_low 1 2.0 0.003 0.003 40.879 40.920 qs_forces 11 3.9 0.002 0.002 40.791 40.791 qs_energies 11 4.9 0.001 0.002 39.067 39.070 scf_env_do_scf 11 5.9 0.000 0.001 33.253 33.253 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.532 30.533 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.413 22.791 qs_scf_new_mos 108 7.5 0.001 0.001 20.786 21.036 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.785 21.036 ot_scf_mini 108 9.5 0.002 0.003 19.873 20.047 velocity_verlet 10 3.0 0.001 0.002 19.214 19.215 multiply_cannon 2286 13.5 0.209 0.219 17.080 18.742 multiply_cannon_loop 2286 14.5 0.900 0.972 15.880 17.447 ot_mini 108 10.5 0.001 0.001 12.230 12.467 mp_waitall_1 200699 16.5 5.974 11.031 5.974 11.031 qs_ot_get_derivative 108 11.5 0.001 0.001 9.726 9.905 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.243 9.514 multiply_cannon_multrec 27432 15.5 1.963 4.541 6.517 9.461 rebuild_ks_matrix 119 8.3 0.000 0.000 7.511 7.655 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.511 7.655 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.651 6.780 dbcsr_mm_accdrv_process 47894 16.0 3.365 5.629 4.484 6.321 qs_ot_get_p 119 10.4 0.001 0.001 4.688 4.912 init_scf_run 11 5.9 0.000 0.001 4.586 4.587 scf_env_initial_rho_setup 11 6.9 0.002 0.002 4.586 4.586 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.710 4.550 sum_up_and_integrate 119 10.3 0.024 0.026 4.473 4.479 integrate_v_rspace 119 11.3 0.002 0.002 4.449 4.455 mp_sum_l 7207 12.9 2.194 4.174 2.194 4.174 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.121 4.164 calculate_rho_elec 119 8.7 0.021 0.024 4.121 4.164 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.066 4.140 apply_single 119 13.6 0.000 0.000 3.065 4.140 rs_pw_transfer 974 11.9 0.010 0.011 3.005 3.409 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.124 3.146 calculate_first_density_matrix 1 7.0 0.000 0.000 2.956 2.958 make_m2s 4572 13.5 0.052 0.054 2.692 2.945 density_rs2pw 119 9.7 0.004 0.004 2.450 2.881 make_images 4572 14.5 0.198 0.235 2.604 2.853 multiply_cannon_sync_h2d 27432 15.5 2.147 2.819 2.147 2.819 calculate_dm_sparse 119 9.5 0.000 0.000 2.735 2.811 jit_kernel_multiply 11 16.1 1.067 2.800 1.067 2.800 init_scf_loop 11 6.9 0.000 0.000 2.692 2.692 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.663 2.663 ot_diis_step 108 11.5 0.011 0.011 2.441 2.442 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.300 2.395 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.256 2.257 cp_fm_redistribute_end 50 14.0 1.860 2.220 1.864 2.222 potential_pw2rs 119 12.3 0.006 0.006 2.199 2.207 pw_transfer 1439 11.6 0.065 0.073 2.127 2.158 cp_fm_diag_elpa_base 50 14.0 0.342 2.099 0.356 2.150 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.035 2.069 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.939 1.947 grid_integrate_task_list 119 12.3 1.838 1.933 1.838 1.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.895 1.897 prepare_preconditioner 11 7.9 0.000 0.000 1.748 1.775 make_preconditioner 11 8.9 0.000 0.000 1.748 1.775 fft3d_ps 1201 14.6 0.512 0.564 1.734 1.763 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.643 1.701 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.646 1.686 make_images_data 4572 15.5 0.045 0.051 1.179 1.634 wfi_extrapolate 11 7.9 0.001 0.001 1.570 1.570 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.023 1.532 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.468 1.503 acc_transpose_blocks 27432 15.5 0.110 0.115 1.189 1.470 mp_allgather_i34 2286 14.5 0.643 1.417 0.643 1.417 mp_alltoall_d11v 2130 13.8 1.264 1.409 1.264 1.409 grid_collocate_task_list 119 9.7 1.231 1.367 1.231 1.367 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.223 1.264 mp_sum_d 4125 12.0 0.620 1.064 0.620 1.064 make_images_sizes 4572 15.5 0.005 0.005 0.705 0.957 mp_alltoall_i44 4572 16.5 0.701 0.952 0.701 0.952 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.147 0.527 0.941 mp_waitany 5720 13.7 0.512 0.938 0.512 0.938 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.937 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.922 0.937 mp_alltoall_z22v 1201 16.6 0.761 0.870 0.761 0.870 acc_transpose_blocks_kernels 27432 16.5 0.183 0.270 0.649 0.855 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.606 0.825 0.844 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.098000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.838592E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1993547. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.106 35.223 35.225 qs_mol_dyn_low 1 2.0 0.003 0.004 34.813 34.820 qs_forces 11 3.9 0.005 0.015 34.699 34.701 qs_energies 11 4.9 0.003 0.016 33.039 33.044 scf_env_do_scf 11 5.9 0.004 0.011 27.260 27.261 scf_env_do_scf_inner_loop 108 6.5 0.005 0.018 24.443 24.444 dbcsr_multiply_generic 2286 12.5 0.108 0.122 17.745 17.859 velocity_verlet 10 3.0 0.005 0.013 16.245 16.248 qs_scf_new_mos 108 7.5 0.001 0.001 15.739 15.758 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.738 15.757 ot_scf_mini 108 9.5 0.003 0.003 14.980 15.000 multiply_cannon 2286 13.5 0.197 0.201 14.095 14.908 multiply_cannon_loop 2286 14.5 0.638 0.671 13.218 14.174 ot_mini 108 10.5 0.001 0.001 9.223 9.252 multiply_cannon_multrec 18288 15.5 1.946 2.904 7.619 7.880 qs_ot_get_derivative 108 11.5 0.001 0.001 7.680 7.700 rebuild_ks_matrix 119 8.3 0.000 0.000 6.697 6.719 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.697 6.718 dbcsr_mm_accdrv_process 38222 16.0 4.460 6.185 5.590 6.438 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.919 5.939 init_scf_run 11 5.9 0.000 0.001 4.438 4.438 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.437 4.438 mp_waitall_1 158411 16.6 2.953 4.323 2.953 4.323 sum_up_and_integrate 119 10.3 0.029 0.030 4.243 4.248 integrate_v_rspace 119 11.3 0.002 0.003 4.214 4.222 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.733 3.742 calculate_rho_elec 119 8.7 0.030 0.032 3.732 3.742 qs_ot_get_p 119 10.4 0.001 0.001 3.527 3.554 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.915 3.538 calculate_first_density_matrix 1 7.0 0.001 0.003 3.102 3.103 rs_pw_transfer 974 11.9 0.009 0.010 2.676 2.956 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.548 2.808 init_scf_loop 11 6.9 0.001 0.005 2.785 2.786 jit_kernel_multiply 10 15.8 1.080 2.673 1.080 2.673 calculate_dm_sparse 119 9.5 0.000 0.000 2.554 2.565 density_rs2pw 119 9.7 0.004 0.004 2.224 2.486 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.070 2.396 apply_single 119 13.6 0.000 0.000 2.070 2.396 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.352 2.358 make_m2s 4572 13.5 0.045 0.046 1.980 2.149 pw_transfer 1439 11.6 0.066 0.071 2.077 2.091 make_images 4572 14.5 0.190 0.202 1.895 2.064 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.061 2.062 potential_pw2rs 119 12.3 0.007 0.008 2.012 2.024 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.983 1.999 prepare_preconditioner 11 7.9 0.000 0.001 1.951 1.953 make_preconditioner 11 8.9 0.001 0.002 1.951 1.953 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.896 1.901 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.799 1.887 grid_integrate_task_list 119 12.3 1.798 1.879 1.798 1.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.875 1.878 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.788 1.790 cp_fm_redistribute_end 50 14.0 1.329 1.758 1.330 1.759 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.711 1.722 cp_fm_diag_elpa_base 50 14.0 0.411 1.670 0.426 1.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.691 1.699 fft3d_ps 1201 14.6 0.523 0.539 1.663 1.679 mp_sum_l 7207 12.9 1.279 1.666 1.279 1.666 ot_diis_step 108 11.5 0.011 0.011 1.513 1.513 multiply_cannon_sync_h2d 18288 15.5 1.344 1.512 1.344 1.512 fft_wrap_pw1pw2_140 487 13.2 0.089 0.093 1.484 1.499 grid_collocate_task_list 119 9.7 1.210 1.367 1.210 1.367 acc_transpose_blocks 18288 15.5 0.077 0.079 1.299 1.337 wfi_extrapolate 11 7.9 0.001 0.001 1.277 1.278 make_images_data 4572 15.5 0.045 0.049 0.856 1.050 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.369 0.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.985 0.988 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.934 0.957 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.746 0.944 acc_transpose_blocks_kernels 18288 16.5 0.211 0.220 0.851 0.893 mp_waitany 9880 13.7 0.569 0.865 0.569 0.865 mp_alltoall_d11v 2130 13.8 0.774 0.860 0.774 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.853 0.856 cp_fm_cholesky_invert 11 10.9 0.847 0.852 0.847 0.852 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.124 0.541 0.821 mp_alltoall_z22v 1201 16.6 0.715 0.812 0.715 0.812 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.662 0.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.225000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 550.621184E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.114 37.022 37.023 qs_mol_dyn_low 1 2.0 0.003 0.004 36.639 36.648 qs_forces 11 3.9 0.003 0.013 36.302 36.302 qs_energies 11 4.9 0.003 0.006 34.562 34.568 scf_env_do_scf 11 5.9 0.002 0.005 29.022 29.023 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 25.208 25.208 velocity_verlet 10 3.0 0.002 0.002 18.886 18.900 dbcsr_multiply_generic 2286 12.5 0.100 0.103 18.536 18.642 qs_scf_new_mos 108 7.5 0.001 0.001 16.462 16.517 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.461 16.516 ot_scf_mini 108 9.5 0.003 0.004 15.513 15.562 multiply_cannon 2286 13.5 0.227 0.273 14.733 15.212 multiply_cannon_loop 2286 14.5 0.938 0.970 13.720 14.160 ot_mini 108 10.5 0.001 0.001 9.463 9.526 multiply_cannon_multrec 27432 15.5 2.366 3.039 8.883 9.239 qs_ot_get_derivative 108 11.5 0.001 0.001 7.643 7.693 dbcsr_mm_accdrv_process 47916 15.9 5.482 7.200 6.424 7.688 rebuild_ks_matrix 119 8.3 0.000 0.000 6.755 6.807 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.755 6.806 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.015 6.059 init_scf_run 11 5.9 0.000 0.001 4.076 4.077 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.076 4.077 sum_up_and_integrate 119 10.3 0.035 0.038 3.972 3.981 integrate_v_rspace 119 11.3 0.002 0.002 3.937 3.946 init_scf_loop 11 6.9 0.001 0.003 3.789 3.790 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.736 3.772 calculate_rho_elec 119 8.7 0.040 0.046 3.736 3.772 qs_ot_get_p 119 10.4 0.001 0.001 3.412 3.485 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.825 3.309 prepare_preconditioner 11 7.9 0.000 0.001 2.882 2.890 make_preconditioner 11 8.9 0.003 0.010 2.881 2.890 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.483 2.817 calculate_first_density_matrix 1 7.0 0.000 0.002 2.618 2.619 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.134 2.597 apply_single 119 13.6 0.000 0.000 2.133 2.596 mp_waitall_1 137007 16.6 1.846 2.496 1.846 2.496 rs_pw_transfer 974 11.9 0.009 0.009 2.275 2.493 make_m2s 4572 13.5 0.054 0.056 2.286 2.439 density_rs2pw 119 9.7 0.004 0.004 2.141 2.332 make_images 4572 14.5 0.269 0.332 2.178 2.329 calculate_dm_sparse 119 9.5 0.000 0.000 2.201 2.260 pw_transfer 1439 11.6 0.066 0.071 2.126 2.165 jit_kernel_multiply 10 15.8 0.882 2.101 0.882 2.101 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.089 2.098 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.033 2.077 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.950 1.974 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.946 1.947 grid_integrate_task_list 119 12.3 1.827 1.914 1.827 1.914 ot_diis_step 108 11.5 0.012 0.012 1.779 1.779 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.765 1.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.734 1.747 fft3d_ps 1201 14.6 0.559 0.612 1.707 1.747 potential_pw2rs 119 12.3 0.008 0.009 1.733 1.739 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.629 1.676 mp_sum_l 7207 12.9 1.116 1.623 1.116 1.623 acc_transpose_blocks 27432 15.5 0.114 0.116 1.526 1.550 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.514 1.515 cp_fm_redistribute_end 50 14.0 1.003 1.490 1.004 1.491 cp_fm_diag_elpa_base 50 14.0 0.465 1.421 0.484 1.463 wfi_extrapolate 11 7.9 0.001 0.001 1.398 1.398 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.374 1.385 grid_collocate_task_list 119 9.7 1.221 1.325 1.221 1.325 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.720 1.239 qs_energies_init_hamiltonians 11 5.9 0.008 0.029 1.160 1.180 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.127 1.146 dbcsr_complete_redistribute 329 12.2 0.139 0.164 0.841 1.121 cp_fm_upper_to_full 72 13.5 0.804 1.121 0.804 1.121 multiply_cannon_sync_h2d 27432 15.5 0.983 1.055 0.983 1.055 make_images_data 4572 15.5 0.045 0.050 0.883 1.040 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.757 0.966 cp_fm_cholesky_invert 11 10.9 0.945 0.949 0.945 0.949 acc_transpose_blocks_kernels 27432 16.5 0.268 0.278 0.893 0.914 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.609 0.880 mp_alltoall_d11v 2130 13.8 0.770 0.878 0.770 0.878 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.832 0.837 mp_alltoall_z22v 1201 16.6 0.731 0.773 0.731 0.773 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.023000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=522.545455, yerr=2.965365 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 598.622208E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.077 29.956 29.957 qs_mol_dyn_low 1 2.0 0.003 0.003 29.567 29.576 qs_forces 11 3.9 0.002 0.002 29.508 29.509 qs_energies 11 4.9 0.001 0.002 27.761 27.764 scf_env_do_scf 11 5.9 0.000 0.001 22.641 22.641 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.929 19.930 velocity_verlet 10 3.0 0.002 0.002 14.994 14.997 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.842 12.981 qs_scf_new_mos 108 7.5 0.001 0.001 11.688 11.714 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.687 11.713 ot_scf_mini 108 9.5 0.002 0.002 10.992 11.018 multiply_cannon 2286 13.5 0.230 0.237 10.149 10.528 multiply_cannon_loop 2286 14.5 0.331 0.342 9.178 9.432 multiply_cannon_multrec 9144 15.5 1.681 1.923 6.017 6.213 rebuild_ks_matrix 119 8.3 0.000 0.000 6.143 6.167 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.143 6.166 ot_mini 108 10.5 0.001 0.001 6.105 6.139 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.474 5.496 qs_ot_get_derivative 108 11.5 0.001 0.001 4.789 4.815 dbcsr_mm_accdrv_process 12550 15.8 3.228 4.088 4.235 4.320 sum_up_and_integrate 119 10.3 0.038 0.040 3.790 3.793 integrate_v_rspace 119 11.3 0.002 0.003 3.752 3.756 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.678 3.707 calculate_rho_elec 119 8.7 0.060 0.061 3.677 3.707 init_scf_run 11 5.9 0.000 0.001 3.672 3.672 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.672 3.672 qs_ot_get_p 119 10.4 0.001 0.001 3.081 3.119 init_scf_loop 11 6.9 0.000 0.000 2.684 2.686 calculate_first_density_matrix 1 7.0 0.000 0.000 2.441 2.446 mp_waitall_1 115863 16.7 1.711 2.204 1.711 2.204 pw_transfer 1439 11.6 0.066 0.070 2.171 2.181 density_rs2pw 119 9.7 0.004 0.004 1.994 2.131 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.077 2.089 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.062 2.066 make_m2s 4572 13.5 0.034 0.036 1.851 2.046 rs_pw_transfer 974 11.9 0.008 0.008 1.847 1.982 jit_kernel_multiply 10 15.9 0.969 1.961 0.969 1.961 make_images 4572 14.5 0.266 0.299 1.761 1.954 grid_integrate_task_list 119 12.3 1.858 1.947 1.858 1.947 prepare_preconditioner 11 7.9 0.000 0.000 1.922 1.927 make_preconditioner 11 8.9 0.000 0.000 1.922 1.927 calculate_dm_sparse 119 9.5 0.000 0.000 1.882 1.904 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.854 1.854 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.810 1.837 fft3d_ps 1201 14.6 0.563 0.572 1.736 1.747 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.741 1.742 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.677 1.691 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.619 1.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.549 1.562 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.545 1.546 potential_pw2rs 119 12.3 0.010 0.010 1.519 1.525 cp_fm_redistribute_end 50 14.0 0.769 1.516 0.770 1.517 cp_fm_diag_elpa_base 50 14.0 0.699 1.439 0.744 1.496 grid_collocate_task_list 119 9.7 1.280 1.406 1.280 1.406 ot_diis_step 108 11.5 0.013 0.013 1.300 1.300 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.288 1.296 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.214 1.228 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.222 1.222 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.169 1.192 apply_single 119 13.6 0.000 0.000 1.169 1.192 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.825 1.107 make_images_data 4572 15.5 0.039 0.042 0.859 1.098 acc_transpose_blocks 9144 15.5 0.039 0.040 1.061 1.084 cp_fm_cholesky_invert 11 10.9 0.987 0.990 0.987 0.990 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.919 mp_alltoall_d11v 2130 13.8 0.808 0.895 0.808 0.895 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.818 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.792 0.795 multiply_cannon_sync_h2d 9144 15.5 0.709 0.795 0.709 0.795 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.748 0.756 mp_allgather_i34 2286 14.5 0.306 0.740 0.306 0.740 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.408 0.736 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.677 0.730 jit_kernel_transpose 5 15.6 0.701 0.718 0.701 0.718 mp_alltoall_z22v 1201 16.6 0.622 0.673 0.622 0.673 yz_to_x 606 15.1 0.264 0.274 0.590 0.606 dbcsr_complete_redistribute 329 12.2 0.162 0.168 0.566 0.602 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.957000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.454545, yerr=4.335134 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 776.146944E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 43.493 43.494 qs_mol_dyn_low 1 2.0 0.003 0.003 43.292 43.298 qs_forces 11 3.9 0.001 0.002 43.233 43.234 qs_energies 11 4.9 0.001 0.002 41.217 41.220 scf_env_do_scf 11 5.9 0.001 0.001 35.139 35.139 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.885 26.887 velocity_verlet 10 3.0 0.002 0.002 24.422 24.427 dbcsr_multiply_generic 2286 12.5 0.100 0.100 18.039 18.243 qs_scf_new_mos 108 7.5 0.001 0.001 16.749 16.842 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.748 16.842 ot_scf_mini 108 9.5 0.002 0.002 15.637 15.734 multiply_cannon 2286 13.5 0.302 0.313 13.812 14.787 multiply_cannon_loop 2286 14.5 0.343 0.347 12.498 13.512 ot_mini 108 10.5 0.001 0.001 9.123 9.245 multiply_cannon_multrec 9144 15.5 3.400 4.834 8.510 8.599 init_scf_loop 11 6.9 0.000 0.000 8.221 8.227 rebuild_ks_matrix 119 8.3 0.000 0.000 7.523 7.668 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.522 7.668 prepare_preconditioner 11 7.9 0.000 0.000 7.223 7.238 make_preconditioner 11 8.9 0.000 0.000 7.223 7.238 qs_ot_get_derivative 108 11.5 0.001 0.001 7.096 7.194 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.809 7.111 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.808 6.940 dbcsr_mm_accdrv_process 12550 15.8 4.070 5.610 4.986 6.332 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.495 4.555 calculate_rho_elec 119 8.7 0.118 0.121 4.494 4.555 cp_fm_upper_to_full 72 14.2 3.169 4.519 3.169 4.519 sum_up_and_integrate 119 10.3 0.065 0.066 4.192 4.201 integrate_v_rspace 119 11.3 0.003 0.003 4.126 4.135 init_scf_run 11 5.9 0.000 0.001 4.021 4.021 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.021 4.021 qs_ot_get_p 119 10.4 0.001 0.001 3.720 3.860 mp_waitall_1 94719 16.7 2.556 3.666 2.556 3.666 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.584 3.021 pw_transfer 1439 11.6 0.068 0.069 2.922 2.926 dbcsr_complete_redistribute 329 12.2 0.284 0.291 2.046 2.829 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.823 2.828 make_m2s 4572 13.5 0.038 0.038 2.498 2.695 make_images 4572 14.5 0.351 0.386 2.377 2.574 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.676 2.461 fft3d_ps 1201 14.6 0.595 0.601 2.447 2.452 calculate_first_density_matrix 1 7.0 0.000 0.000 2.390 2.452 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.178 2.426 apply_single 119 13.6 0.000 0.000 2.178 2.425 density_rs2pw 119 9.7 0.004 0.004 2.347 2.371 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.334 2.341 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.316 2.317 calculate_dm_sparse 119 9.5 0.000 0.000 2.242 2.299 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.307 2.286 mp_alltoall_i22 627 13.8 1.397 2.203 1.397 2.203 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.409 2.190 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.094 2.095 grid_integrate_task_list 119 12.3 2.071 2.084 2.071 2.084 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.026 2.027 ot_diis_step 108 11.5 0.014 0.014 1.994 1.995 mp_sum_l 7207 12.9 1.191 1.950 1.191 1.950 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.805 1.847 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.795 1.845 cp_fm_cholesky_invert 11 10.9 1.775 1.779 1.775 1.779 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.770 1.771 jit_kernel_multiply 10 15.5 0.889 1.726 0.889 1.726 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.679 1.679 cp_fm_diag_elpa_base 50 14.0 1.532 1.585 1.676 1.676 rs_pw_transfer 974 11.9 0.009 0.009 1.615 1.641 potential_pw2rs 119 12.3 0.014 0.014 1.572 1.576 grid_collocate_task_list 119 9.7 1.500 1.516 1.500 1.516 wfi_extrapolate 11 7.9 0.001 0.001 1.498 1.499 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.217 1.488 make_images_data 4572 15.5 0.043 0.046 1.195 1.442 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.286 1.342 mp_alltoall_d11v 2130 13.8 1.207 1.277 1.207 1.277 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.190 1.218 acc_transpose_blocks 9144 15.5 0.038 0.038 1.138 1.167 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.106 1.128 multiply_cannon_sync_h2d 9144 15.5 1.042 1.044 1.042 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.982 1.032 yz_to_x 606 15.1 0.461 0.471 0.996 1.008 qs_create_task_list 11 7.9 0.000 0.000 0.945 0.957 generate_qs_task_list 11 8.9 0.371 0.390 0.945 0.957 mp_alltoall_z22v 1201 16.6 0.898 0.945 0.898 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.927 0.940 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 0.887 0.916 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.494000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.818182, yerr=17.335293 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.157312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.033 85.255 85.256 qs_mol_dyn_low 1 2.0 0.003 0.004 84.908 84.918 qs_forces 11 3.9 0.002 0.003 84.833 84.834 qs_energies 11 4.9 0.002 0.004 81.941 81.957 scf_env_do_scf 11 5.9 0.001 0.001 72.686 72.689 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.778 66.779 dbcsr_multiply_generic 2055 12.4 0.106 0.111 52.714 53.057 qs_scf_new_mos 99 7.5 0.000 0.001 48.965 49.093 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.965 49.092 ot_scf_mini 99 9.5 0.002 0.003 46.509 46.620 multiply_cannon 2055 13.4 0.180 0.188 42.966 44.001 velocity_verlet 10 3.0 0.003 0.005 43.149 43.150 multiply_cannon_loop 2055 14.4 1.546 1.590 41.965 42.966 ot_mini 99 10.5 0.001 0.001 28.162 28.294 qs_ot_get_derivative 99 11.5 0.001 0.001 21.289 21.394 multiply_cannon_multrec 49320 15.4 12.130 12.952 17.457 18.074 rebuild_ks_matrix 110 8.3 0.000 0.000 14.799 14.901 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.798 14.900 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.000 13.089 mp_waitall_1 220248 16.4 11.622 12.727 11.622 12.727 multiply_cannon_sync_h2d 49320 15.4 9.945 10.722 9.945 10.722 qs_ot_get_p 110 10.4 0.001 0.001 9.848 9.944 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 6.971 8.278 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.281 7.861 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.205 7.705 apply_single 110 13.6 0.000 0.001 7.205 7.705 sum_up_and_integrate 110 10.3 0.036 0.044 7.290 7.303 integrate_v_rspace 110 11.3 0.003 0.003 7.254 7.274 init_scf_run 11 5.9 0.000 0.001 7.180 7.181 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.180 7.180 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.883 7.022 calculate_rho_elec 110 8.6 0.021 0.026 6.883 7.021 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.869 6.927 ot_diis_step 99 11.5 0.005 0.006 6.588 6.589 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.431 6.447 mp_sum_l 6514 12.8 5.262 6.231 5.262 6.231 init_scf_loop 11 6.9 0.000 0.003 5.877 5.877 dbcsr_mm_accdrv_process 87628 16.1 2.073 2.203 5.205 5.626 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.451 5.451 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.912 4.914 cp_fm_redistribute_end 48 14.0 4.284 4.881 4.287 4.883 cp_fm_diag_elpa_base 48 14.0 0.589 4.727 0.592 4.756 rs_pw_transfer 902 11.9 0.011 0.013 3.942 4.574 density_rs2pw 110 9.6 0.004 0.005 3.739 4.356 wfi_extrapolate 11 7.9 0.001 0.001 4.183 4.183 make_m2s 4110 13.4 0.062 0.065 4.041 4.137 make_images 4110 14.4 0.177 0.192 3.944 4.042 calculate_dm_sparse 110 9.5 0.000 0.001 3.916 4.026 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.748 3.752 multiply_cannon_metrocomm1 49320 15.4 0.061 0.065 2.703 3.706 prepare_preconditioner 11 7.9 0.000 0.001 3.652 3.677 make_preconditioner 11 8.9 0.000 0.001 3.652 3.677 pw_transfer 1331 11.6 0.055 0.066 3.534 3.620 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.446 3.535 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.429 3.491 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.399 3.440 grid_integrate_task_list 110 12.3 3.246 3.406 3.246 3.406 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.248 3.296 fft3d_ps 1111 14.6 0.778 0.883 2.976 3.049 fft_wrap_pw1pw2_140 451 13.1 0.169 0.189 2.868 2.958 jit_kernel_multiply 13 15.9 2.849 2.943 2.849 2.943 calculate_first_density_matrix 1 7.0 0.000 0.002 2.905 2.908 potential_pw2rs 110 12.3 0.006 0.007 2.867 2.897 mp_waitany 14300 13.8 1.766 2.400 1.766 2.400 acc_transpose_blocks 49320 15.4 0.225 0.239 2.270 2.373 mp_alltoall_d11v 2046 13.8 1.985 2.357 1.985 2.357 grid_collocate_task_list 110 9.6 2.084 2.261 2.084 2.261 mp_sum_d 3879 11.9 1.565 2.116 1.565 2.116 cp_fm_cholesky_invert 11 10.9 1.962 1.966 1.962 1.966 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.941 1.960 make_images_data 4110 15.4 0.043 0.045 1.752 1.865 mp_alltoall_z22v 1111 16.6 1.513 1.795 1.513 1.795 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.735 1.759 hybrid_alltoall_any 4261 16.3 0.081 0.477 1.519 1.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.256000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=1.975051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.243520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3517269. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.037 71.855 71.857 qs_mol_dyn_low 1 2.0 0.003 0.003 71.500 71.510 qs_forces 11 3.9 0.003 0.005 71.173 71.174 qs_energies 11 4.9 0.044 0.128 67.821 67.827 scf_env_do_scf 11 5.9 0.000 0.001 58.232 58.235 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.181 50.182 dbcsr_multiply_generic 2055 12.4 0.114 0.119 39.189 39.400 velocity_verlet 10 3.0 0.001 0.002 37.275 37.282 qs_scf_new_mos 99 7.5 0.001 0.001 33.593 33.761 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.593 33.761 multiply_cannon 2055 13.4 0.224 0.247 32.179 33.489 ot_scf_mini 99 9.5 0.003 0.003 31.923 32.084 multiply_cannon_loop 2055 14.4 0.929 0.951 30.760 31.811 ot_mini 99 10.5 0.001 0.001 18.894 19.057 multiply_cannon_multrec 24660 15.4 7.636 9.448 14.646 16.352 rebuild_ks_matrix 110 8.3 0.000 0.001 13.927 14.041 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.927 14.041 qs_ot_get_derivative 99 11.5 0.001 0.001 13.043 13.201 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.276 12.372 mp_waitall_1 176588 16.5 7.841 10.598 7.841 10.598 multiply_cannon_sync_h2d 24660 15.4 7.028 8.158 7.028 8.158 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.325 8.089 init_scf_loop 11 6.9 0.000 0.000 8.011 8.012 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.713 7.626 apply_single 110 13.6 0.000 0.001 6.713 7.626 dbcsr_mm_accdrv_process 52282 16.1 5.242 6.317 6.851 7.290 init_scf_run 11 5.9 0.000 0.001 7.067 7.068 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.067 7.068 sum_up_and_integrate 110 10.3 0.051 0.057 6.743 6.754 integrate_v_rspace 110 11.3 0.002 0.003 6.692 6.706 qs_ot_get_p 110 10.4 0.001 0.001 6.344 6.521 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.319 6.331 calculate_rho_elec 110 8.6 0.039 0.047 6.319 6.330 prepare_preconditioner 11 7.9 0.000 0.000 5.967 5.993 make_preconditioner 11 8.9 0.000 0.000 5.967 5.993 ot_diis_step 99 11.5 0.010 0.010 5.800 5.801 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.872 5.739 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.535 5.700 make_m2s 4110 13.4 0.057 0.060 4.280 4.808 make_images 4110 14.4 0.395 0.441 4.171 4.695 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.338 4.359 density_rs2pw 110 9.6 0.004 0.004 3.483 3.968 pw_transfer 1331 11.6 0.066 0.074 3.799 3.944 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.881 3.881 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.692 3.840 rs_pw_transfer 902 11.9 0.012 0.013 3.177 3.702 calculate_dm_sparse 110 9.5 0.001 0.001 3.583 3.617 wfi_extrapolate 11 7.9 0.001 0.001 3.604 3.604 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.532 3.534 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.354 3.444 calculate_first_density_matrix 1 7.0 0.000 0.001 3.373 3.374 grid_integrate_task_list 110 12.3 3.142 3.323 3.142 3.323 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.271 3.273 cp_fm_redistribute_end 48 14.0 2.442 3.237 2.445 3.239 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.149 3.212 cp_fm_diag_elpa_base 48 14.0 0.759 3.093 0.790 3.174 fft3d_ps 1111 14.6 1.101 1.318 3.042 3.173 fft_wrap_pw1pw2_140 451 13.1 0.202 0.221 3.020 3.173 make_images_data 4110 15.4 0.047 0.051 2.325 2.861 cp_fm_cholesky_invert 11 10.9 2.848 2.856 2.848 2.856 mp_sum_l 6514 12.8 2.111 2.848 2.111 2.848 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.027 2.785 jit_kernel_multiply 10 16.4 1.261 2.695 1.261 2.695 potential_pw2rs 110 12.3 0.008 0.008 2.585 2.598 grid_collocate_task_list 110 9.6 2.054 2.483 2.054 2.483 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.147 2.157 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.002 2.025 mp_alltoall_d11v 2046 13.8 1.715 1.961 1.715 1.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.877 1.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.818 1.832 mp_allgather_i34 2055 14.4 0.751 1.749 0.751 1.749 mp_waitany 10164 13.8 1.208 1.706 1.208 1.706 acc_transpose_blocks 24660 15.4 0.111 0.114 1.620 1.637 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.779 1.614 mp_irecv_dv 57340 16.2 0.655 1.501 0.655 1.501 dbcsr_complete_redistribute 325 12.2 0.244 0.309 1.214 1.487 cp_fm_cholesky_decompose 22 10.9 1.478 1.484 1.478 1.484 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.474 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.220 0.950 1.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.857000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=7.183544 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 661.110784E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.039 63.377 63.379 qs_mol_dyn_low 1 2.0 0.003 0.006 62.972 62.987 qs_forces 11 3.9 0.007 0.043 62.898 62.903 qs_energies 11 4.9 0.012 0.027 59.641 59.651 scf_env_do_scf 11 5.9 0.001 0.001 50.444 50.445 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.288 41.289 velocity_verlet 10 3.0 0.006 0.008 33.678 33.680 dbcsr_multiply_generic 2055 12.4 0.107 0.110 29.719 29.962 qs_scf_new_mos 99 7.5 0.001 0.001 26.007 26.110 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.006 26.110 ot_scf_mini 99 9.5 0.002 0.003 24.775 24.884 multiply_cannon 2055 13.4 0.212 0.225 23.199 24.496 multiply_cannon_loop 2055 14.4 0.616 0.638 21.929 22.917 ot_mini 99 10.5 0.001 0.001 14.072 14.190 rebuild_ks_matrix 110 8.3 0.000 0.000 12.584 12.693 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 12.584 12.692 multiply_cannon_multrec 16440 15.4 4.005 5.306 10.712 11.951 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.086 11.179 mp_waitall_1 139946 16.5 7.100 10.480 7.100 10.480 qs_ot_get_derivative 99 11.5 0.001 0.001 9.592 9.704 init_scf_loop 11 6.9 0.001 0.010 9.110 9.111 multiply_cannon_metrocomm3 16440 15.4 0.044 0.044 4.240 7.390 prepare_preconditioner 11 7.9 0.000 0.001 7.283 7.298 make_preconditioner 11 8.9 0.000 0.002 7.283 7.298 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.630 6.979 sum_up_and_integrate 110 10.3 0.059 0.060 6.695 6.711 dbcsr_mm_accdrv_process 34862 16.1 5.427 6.181 6.561 6.707 integrate_v_rspace 110 11.3 0.002 0.003 6.635 6.651 init_scf_run 11 5.9 0.000 0.001 6.417 6.417 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.416 6.417 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.077 6.085 calculate_rho_elec 110 8.6 0.058 0.059 6.076 6.085 qs_ot_get_p 110 10.4 0.001 0.001 5.833 5.956 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.012 5.440 apply_single 110 13.6 0.000 0.000 5.012 5.440 make_m2s 4110 13.4 0.050 0.051 4.320 4.742 make_images 4110 14.4 0.389 0.508 4.205 4.629 ot_diis_step 99 11.5 0.011 0.011 4.444 4.445 density_rs2pw 110 9.6 0.004 0.005 3.221 4.379 multiply_cannon_sync_h2d 16440 15.4 3.720 4.378 3.720 4.378 qs_ot_p2m_diag 48 11.0 0.041 0.044 4.120 4.125 rs_pw_transfer 902 11.9 0.010 0.011 2.993 4.078 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.187 3.826 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.765 3.765 pw_transfer 1331 11.6 0.065 0.073 3.678 3.689 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.572 3.584 grid_integrate_task_list 110 12.3 3.160 3.361 3.160 3.361 calculate_first_density_matrix 1 7.0 0.000 0.004 3.264 3.265 calculate_dm_sparse 110 9.5 0.001 0.001 3.165 3.191 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.176 3.178 cp_fm_redistribute_end 48 14.0 1.978 3.140 1.980 3.140 cp_fm_diag_elpa_base 48 14.0 1.095 2.971 1.154 3.090 wfi_extrapolate 11 7.9 0.001 0.001 3.058 3.058 cp_fm_cholesky_invert 11 10.9 3.024 3.030 3.024 3.030 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 3.015 3.029 make_images_data 4110 15.4 0.043 0.048 2.508 2.998 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.944 2.945 fft3d_ps 1111 14.6 1.086 1.097 2.872 2.882 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.201 2.877 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.599 2.635 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.420 2.481 potential_pw2rs 110 12.3 0.010 0.011 2.462 2.472 grid_collocate_task_list 110 9.6 2.079 2.458 2.079 2.458 mp_waitany 17072 13.8 1.290 2.426 1.290 2.426 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.852 2.281 mp_sum_l 6514 12.8 1.595 2.224 1.595 2.224 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.183 2.191 mp_irecv_dv 48980 15.7 0.783 2.158 0.783 2.158 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.118 2.131 mp_alltoall_d11v 2046 13.8 1.747 2.091 1.747 2.091 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.957 2.030 dbcsr_complete_redistribute 325 12.2 0.319 0.356 1.520 1.984 jit_kernel_multiply 8 16.4 0.748 1.933 0.748 1.933 cp_fm_upper_to_full 70 13.6 1.390 1.875 1.390 1.875 cp_fm_cholesky_decompose 22 10.9 1.771 1.792 1.771 1.792 mp_allgather_i34 2055 14.4 0.598 1.760 0.598 1.760 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.703 1.721 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.363 1.471 copy_fm_to_dbcsr 174 11.2 0.001 0.002 0.988 1.442 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.221 1.316 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.305 1.313 acc_transpose_blocks 16440 15.4 0.071 0.074 1.270 1.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.379000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.090909, yerr=9.567507 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 729.980928E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.087 66.510 66.517 qs_mol_dyn_low 1 2.0 0.003 0.003 66.138 66.168 qs_forces 11 3.9 0.002 0.002 66.066 66.068 qs_energies 11 4.9 0.003 0.012 62.640 62.644 scf_env_do_scf 11 5.9 0.001 0.001 53.882 53.885 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.869 41.869 velocity_verlet 10 3.0 0.002 0.006 37.390 37.392 dbcsr_multiply_generic 2055 12.4 0.114 0.116 29.893 30.090 qs_scf_new_mos 99 7.5 0.001 0.001 26.696 26.806 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.695 26.805 ot_scf_mini 99 9.5 0.003 0.004 25.068 25.169 multiply_cannon 2055 13.4 0.242 0.265 22.882 23.947 multiply_cannon_loop 2055 14.4 0.888 0.903 21.454 22.032 multiply_cannon_multrec 24660 15.4 4.233 6.974 13.093 14.388 ot_mini 99 10.5 0.001 0.001 14.247 14.362 rebuild_ks_matrix 110 8.3 0.000 0.000 12.269 12.385 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 12.269 12.385 init_scf_loop 11 6.9 0.002 0.011 11.970 11.971 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.851 10.948 qs_ot_get_derivative 99 11.5 0.001 0.001 10.105 10.210 prepare_preconditioner 11 7.9 0.000 0.001 10.192 10.209 make_preconditioner 11 8.9 0.000 0.002 10.192 10.209 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.450 9.881 dbcsr_mm_accdrv_process 52304 16.0 7.093 8.567 8.711 9.725 sum_up_and_integrate 110 10.3 0.066 0.069 6.599 6.611 integrate_v_rspace 110 11.3 0.003 0.003 6.532 6.547 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.199 6.207 calculate_rho_elec 110 8.6 0.077 0.081 6.198 6.206 mp_waitall_1 121746 16.5 4.186 6.114 4.186 6.114 qs_ot_get_p 110 10.4 0.001 0.001 5.843 5.985 init_scf_run 11 5.9 0.000 0.001 5.930 5.931 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.930 5.931 make_m2s 4110 13.4 0.060 0.061 5.360 5.674 make_images 4110 14.4 0.573 0.696 5.219 5.529 cp_fm_upper_to_full 70 13.8 3.248 4.605 3.248 4.605 ot_diis_step 99 11.5 0.011 0.012 4.103 4.104 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.017 4.103 apply_single 110 13.6 0.000 0.000 4.017 4.102 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.015 4.030 pw_transfer 1331 11.6 0.065 0.074 3.840 3.868 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.734 3.767 dbcsr_complete_redistribute 325 12.2 0.416 0.459 2.643 3.757 density_rs2pw 110 9.6 0.004 0.004 3.162 3.667 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.525 3.525 grid_integrate_task_list 110 12.3 3.261 3.474 3.261 3.474 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.378 3.429 multiply_cannon_sync_h2d 24660 15.4 3.183 3.395 3.183 3.395 calculate_dm_sparse 110 9.5 0.001 0.001 3.196 3.227 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.102 3.212 fft_wrap_pw1pw2_140 451 13.1 0.202 0.211 3.156 3.190 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.072 3.128 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.361 3.092 rs_pw_transfer 902 11.9 0.010 0.011 2.551 3.055 fft3d_ps 1111 14.6 1.091 1.130 3.024 3.054 wfi_extrapolate 11 7.9 0.001 0.001 2.995 2.995 make_images_data 4110 15.4 0.047 0.050 2.651 2.983 cp_fm_cholesky_invert 11 10.9 2.958 2.969 2.958 2.969 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.961 2.963 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.924 2.925 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.255 2.925 cp_fm_redistribute_end 48 14.0 1.477 2.923 1.479 2.924 cp_fm_diag_elpa_base 48 14.0 1.363 2.783 1.442 2.894 calculate_first_density_matrix 1 7.0 0.000 0.003 2.834 2.837 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.732 2.819 mp_alltoall_i22 605 13.7 1.613 2.779 1.613 2.779 jit_kernel_multiply 12 15.7 1.291 2.600 1.291 2.600 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.562 2.591 grid_collocate_task_list 110 9.6 2.181 2.444 2.181 2.444 potential_pw2rs 110 12.3 0.012 0.013 2.285 2.295 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.289 2.291 mp_alltoall_d11v 2046 13.8 1.702 2.242 1.702 2.242 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.807 1.826 cp_fm_cholesky_decompose 22 10.9 1.778 1.812 1.778 1.812 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.717 1.751 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.596 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.670 1.682 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.833 1.636 acc_transpose_blocks 24660 15.4 0.107 0.111 1.604 1.634 mp_sum_l 6514 12.8 1.030 1.631 1.030 1.631 mp_waitany 13376 13.8 1.098 1.577 1.098 1.577 mp_irecv_dv 62702 16.1 0.733 1.557 0.733 1.557 mp_allgather_i34 2055 14.4 0.523 1.500 0.523 1.500 rs_gather_matrices 110 12.3 0.266 0.296 0.909 1.418 rs_pw_transfer_RS2PW_140 121 11.5 0.170 0.179 0.807 1.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.517000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.454545, yerr=6.919848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 827.195392E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 57.033 57.035 qs_mol_dyn_low 1 2.0 0.003 0.003 56.745 56.762 qs_forces 11 3.9 0.002 0.002 56.679 56.680 qs_energies 11 4.9 0.001 0.001 52.955 52.960 scf_env_do_scf 11 5.9 0.000 0.001 44.673 44.673 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 36.654 36.655 velocity_verlet 10 3.0 0.005 0.006 31.843 31.846 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.033 23.157 qs_scf_new_mos 99 7.5 0.001 0.001 21.200 21.259 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.199 21.259 ot_scf_mini 99 9.5 0.002 0.002 19.957 20.005 multiply_cannon 2055 13.4 0.246 0.258 17.475 18.685 multiply_cannon_loop 2055 14.4 0.321 0.333 16.127 16.343 rebuild_ks_matrix 110 8.3 0.000 0.000 12.203 12.266 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.203 12.266 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.843 10.899 ot_mini 99 10.5 0.001 0.001 10.590 10.632 multiply_cannon_multrec 8220 15.4 3.234 4.544 7.538 8.743 mp_waitall_1 103326 16.6 6.192 7.978 6.192 7.978 init_scf_loop 11 6.9 0.000 0.000 7.968 7.969 qs_ot_get_derivative 99 11.5 0.001 0.001 6.731 6.779 sum_up_and_integrate 110 10.3 0.079 0.080 6.692 6.705 integrate_v_rspace 110 11.3 0.003 0.003 6.613 6.626 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.482 6.496 calculate_rho_elec 110 8.6 0.115 0.116 6.481 6.495 prepare_preconditioner 11 7.9 0.000 0.000 6.268 6.276 make_preconditioner 11 8.9 0.000 0.000 6.268 6.276 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.842 5.913 qs_ot_get_p 110 10.4 0.001 0.001 5.411 5.471 dbcsr_mm_accdrv_process 17442 15.9 2.794 3.920 4.172 5.157 init_scf_run 11 5.9 0.000 0.001 5.118 5.119 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.118 5.118 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.150 4.584 make_m2s 4110 13.4 0.039 0.040 4.233 4.499 make_images 4110 14.4 0.635 0.683 4.103 4.372 pw_transfer 1331 11.6 0.066 0.072 4.244 4.263 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.137 4.158 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.974 3.978 ot_diis_step 99 11.5 0.012 0.012 3.834 3.835 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.738 3.798 apply_single 110 13.6 0.000 0.000 3.738 3.797 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.660 3.661 density_rs2pw 110 9.6 0.004 0.004 3.210 3.585 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 3.560 3.585 grid_integrate_task_list 110 12.3 3.355 3.531 3.355 3.531 fft3d_ps 1111 14.6 1.144 1.180 3.380 3.402 cp_fm_cholesky_invert 11 10.9 3.118 3.123 3.118 3.123 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.098 3.099 cp_fm_redistribute_end 48 14.0 0.794 3.060 0.801 3.061 multiply_cannon_sync_h2d 8220 15.4 2.896 3.053 2.896 3.053 cp_fm_diag_elpa_base 48 14.0 2.072 2.849 2.250 3.009 make_images_data 4110 15.4 0.039 0.044 2.413 2.840 wfi_extrapolate 11 7.9 0.001 0.001 2.803 2.804 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.334 2.729 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.704 2.704 rs_pw_transfer 902 11.9 0.010 0.010 2.251 2.642 grid_collocate_task_list 110 9.6 2.277 2.540 2.277 2.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.507 2.508 calculate_dm_sparse 110 9.5 0.001 0.001 2.473 2.507 calculate_first_density_matrix 1 7.0 0.000 0.000 2.217 2.218 potential_pw2rs 110 12.3 0.015 0.015 2.198 2.203 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.142 2.164 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.990 2.010 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.982 mp_alltoall_d11v 2046 13.8 1.726 1.967 1.726 1.967 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.850 1.876 cp_fm_cholesky_decompose 22 10.9 1.706 1.717 1.706 1.717 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.696 1.703 jit_kernel_multiply 9 15.9 1.071 1.700 1.071 1.700 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.545 1.669 mp_allgather_i34 2055 14.4 0.538 1.642 0.538 1.642 dbcsr_complete_redistribute 325 12.2 0.558 0.595 1.478 1.570 mp_waitany 9240 13.8 1.066 1.486 1.066 1.486 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.433 1.444 qs_create_task_list 11 7.9 0.000 0.000 1.230 1.334 generate_qs_task_list 11 8.9 0.379 0.447 1.230 1.334 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.474 1.300 mp_irecv_dv 24056 15.7 0.450 1.258 0.450 1.258 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.184 0.803 1.206 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.165 1.192 rs_gather_matrices 110 12.3 0.324 0.364 0.970 1.181 mp_alltoall_z22v 1111 16.6 1.125 1.151 1.125 1.151 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.035000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.636364, yerr=11.080472 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/14/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 1.612391E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1464624 0.0% 0.0% 100.0% average stack size 0.0 0.0 672.0 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 1.364816E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.040 89.721 89.723 qs_mol_dyn_low 1 2.0 0.003 0.003 88.992 89.309 qs_forces 11 3.9 0.002 0.002 88.880 88.881 qs_energies 11 4.9 0.001 0.002 84.665 84.667 scf_env_do_scf 11 5.9 0.001 0.001 74.171 74.172 velocity_verlet 10 3.0 0.002 0.002 56.653 56.660 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 45.410 45.411 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.880 28.971 init_scf_loop 11 6.9 0.000 0.000 28.685 28.687 qs_scf_new_mos 99 7.5 0.001 0.001 26.891 26.943 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.890 26.943 prepare_preconditioner 11 7.9 0.000 0.000 26.644 26.653 make_preconditioner 11 8.9 0.000 0.000 26.644 26.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.922 26.106 ot_scf_mini 99 9.5 0.002 0.002 25.137 25.176 multiply_cannon 2055 13.4 0.338 0.359 21.672 22.486 multiply_cannon_loop 2055 14.4 0.344 0.377 19.839 20.112 cp_fm_upper_to_full 70 14.2 12.550 17.981 12.550 17.981 rebuild_ks_matrix 110 8.3 0.001 0.001 14.248 14.298 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.247 14.297 ot_mini 99 10.5 0.001 0.001 13.681 13.721 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.912 12.957 dbcsr_complete_redistribute 325 12.2 1.022 1.046 7.348 10.529 multiply_cannon_multrec 8220 15.4 4.369 4.555 9.858 9.988 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.294 9.486 qs_ot_get_derivative 99 11.5 0.001 0.001 9.204 9.243 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.706 8.876 mp_alltoall_i22 605 13.7 5.326 8.499 5.326 8.499 mp_waitall_1 84994 16.7 7.308 8.122 7.308 8.122 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.890 7.926 calculate_rho_elec 110 8.6 0.227 0.228 7.890 7.925 sum_up_and_integrate 110 10.3 0.150 0.151 7.485 7.504 integrate_v_rspace 110 11.3 0.003 0.003 7.335 7.353 qs_ot_get_p 110 10.4 0.001 0.001 6.268 6.368 init_scf_run 11 5.9 0.000 0.001 6.186 6.186 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.185 6.186 make_m2s 4110 13.4 0.043 0.044 5.309 5.815 cp_fm_cholesky_invert 11 10.9 5.713 5.718 5.713 5.718 dbcsr_mm_accdrv_process 11614 15.7 3.261 3.688 5.346 5.628 make_images 4110 14.4 0.878 0.928 5.120 5.625 pw_transfer 1331 11.6 0.075 0.075 5.362 5.370 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.245 5.253 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.541 5.025 apply_single 110 13.6 0.000 0.000 4.540 5.025 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.469 4.783 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.593 4.599 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.526 4.532 fft3d_ps 1111 14.6 1.298 1.306 4.438 4.445 ot_diis_step 99 11.5 0.015 0.016 4.445 4.445 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.140 4.141 density_rs2pw 110 9.6 0.004 0.004 3.929 3.963 multiply_cannon_sync_h2d 8220 15.4 3.947 3.951 3.947 3.951 grid_integrate_task_list 110 12.3 3.658 3.721 3.658 3.721 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.694 3.695 hybrid_alltoall_any 4261 16.3 0.257 0.555 2.859 3.577 make_images_data 4110 15.4 0.042 0.045 2.872 3.531 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.495 3.495 cp_fm_diag_elpa_base 48 14.0 2.963 3.156 3.493 3.493 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.918 3.364 wfi_extrapolate 11 7.9 0.001 0.001 3.355 3.355 calculate_dm_sparse 110 9.5 0.001 0.001 3.306 3.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.115 3.117 calculate_first_density_matrix 1 7.0 0.000 0.000 2.708 2.709 grid_collocate_task_list 110 9.6 2.632 2.654 2.632 2.654 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.606 2.623 potential_pw2rs 110 12.3 0.021 0.021 2.606 2.608 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.486 2.506 rs_pw_transfer 902 11.9 0.011 0.011 2.422 2.438 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.222 2.290 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.183 cp_fm_cholesky_decompose 22 10.9 2.135 2.151 2.135 2.151 mp_alltoall_d11v 2046 13.8 1.979 2.077 1.979 2.077 jit_kernel_multiply 10 15.3 1.886 2.048 1.886 2.048 qs_create_task_list 11 7.9 0.000 0.000 1.911 1.959 generate_qs_task_list 11 8.9 0.736 0.790 1.911 1.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.945 1.951 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.861 1.877 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.796 1.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.723000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1206.181818, yerr=60.380337 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.904512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.046 206.638 206.640 qs_mol_dyn_low 1 2.0 0.004 0.007 206.151 206.165 qs_forces 11 3.9 0.003 0.004 206.058 206.061 qs_energies 11 4.9 0.001 0.002 200.396 200.409 scf_env_do_scf 11 5.9 0.001 0.001 183.730 183.734 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 162.442 162.444 dbcsr_multiply_generic 2507 12.6 0.177 0.183 123.869 124.651 velocity_verlet 10 3.0 0.001 0.002 124.346 124.347 qs_scf_new_mos 117 7.6 0.001 0.001 121.836 122.049 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.835 122.049 ot_scf_mini 117 9.6 0.003 0.003 115.278 115.477 multiply_cannon 2507 13.6 0.236 0.249 100.922 102.661 multiply_cannon_loop 2507 14.6 2.101 2.193 98.686 100.403 ot_mini 117 10.6 0.001 0.001 65.610 65.831 multiply_cannon_multrec 60168 15.6 33.471 35.270 41.847 43.599 qs_ot_get_derivative 117 11.6 0.001 0.001 40.638 40.836 rebuild_ks_matrix 128 8.3 0.001 0.001 34.037 34.291 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.036 34.290 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.605 30.829 mp_waitall_1 267128 16.5 27.849 30.407 27.849 30.407 multiply_cannon_sync_h2d 60168 15.6 27.548 29.429 27.548 29.429 qs_ot_get_p 128 10.4 0.001 0.001 27.866 28.083 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.498 25.352 apply_single 128 13.6 0.001 0.001 24.498 25.352 ot_diis_step 117 11.6 0.007 0.008 24.676 24.677 init_scf_loop 11 6.9 0.000 0.001 21.210 21.211 qs_ot_p2m_diag 83 11.4 0.077 0.091 21.084 21.137 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.729 18.933 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.612 18.614 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 15.666 17.384 prepare_preconditioner 11 7.9 0.000 0.000 16.584 16.638 make_preconditioner 11 8.9 0.000 0.000 16.584 16.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.851 16.041 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.549 15.556 cp_fm_redistribute_end 83 14.4 12.276 15.452 12.290 15.456 cp_fm_diag_elpa_base 83 14.4 3.116 15.055 3.151 15.174 sum_up_and_integrate 128 10.3 0.089 0.107 14.551 14.564 integrate_v_rspace 128 11.3 0.003 0.004 14.461 14.478 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.043 14.167 calculate_rho_elec 128 8.7 0.045 0.064 14.042 14.166 make_m2s 5014 13.6 0.106 0.112 13.708 14.088 make_images 5014 14.6 0.403 0.421 13.525 13.912 init_scf_run 11 5.9 0.000 0.001 12.455 12.456 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.455 12.456 density_rs2pw 128 9.7 0.006 0.007 7.521 10.780 cp_fm_cholesky_invert 11 10.9 9.666 9.674 9.666 9.674 rs_pw_transfer 1046 11.9 0.016 0.018 6.249 9.516 wfi_extrapolate 11 7.9 0.001 0.001 9.199 9.199 mp_sum_l 7870 13.0 7.986 9.128 7.986 9.128 dbcsr_mm_accdrv_process 124484 16.2 3.253 3.482 7.929 8.483 calculate_dm_sparse 128 9.5 0.001 0.001 8.361 8.464 pw_transfer 1547 11.6 0.076 0.101 8.052 8.305 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.083 8.190 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.846 8.104 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.765 7.894 grid_integrate_task_list 128 12.3 7.001 7.583 7.001 7.583 multiply_cannon_metrocomm1 60168 15.6 0.090 0.095 5.656 7.470 make_images_data 5014 15.6 0.068 0.073 6.585 7.438 fft_wrap_pw1pw2_140 523 13.2 0.442 0.513 6.587 6.883 fft3d_ps 1291 14.7 2.104 2.861 6.638 6.835 hybrid_alltoall_any 5200 16.5 0.291 2.259 5.755 6.830 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.690 6.698 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.872 5.993 mp_waitany 16020 13.9 2.671 5.934 2.671 5.934 grid_collocate_task_list 128 9.7 4.557 5.817 4.557 5.817 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.296 2.116 5.411 mp_alltoall_d11v 2415 14.1 4.023 5.274 4.023 5.274 potential_pw2rs 128 12.3 0.009 0.011 4.983 5.021 cp_fm_cholesky_decompose 22 10.9 4.772 4.784 4.772 4.784 mp_sum_d 4455 12.2 3.594 4.407 3.594 4.407 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.640000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.727273, yerr=6.620779 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 824.709120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473827979536 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56635. MP_Allreduce 11357 942. MP_Sync 172 MP_Alltoall 1983 5621114. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_comm_split 84 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 198.678 198.679 qs_mol_dyn_low 1 2.0 0.003 0.004 198.230 198.245 qs_forces 11 3.9 0.003 0.006 197.011 197.012 qs_energies 11 4.9 0.001 0.002 190.191 190.204 scf_env_do_scf 11 5.9 0.001 0.001 170.263 170.273 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.845 136.848 velocity_verlet 10 3.0 0.002 0.005 123.462 123.487 dbcsr_multiply_generic 2529 12.6 0.187 0.195 101.491 102.598 qs_scf_new_mos 118 7.6 0.001 0.001 96.273 96.718 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.272 96.717 ot_scf_mini 118 9.6 0.004 0.004 91.441 91.973 multiply_cannon 2529 13.6 0.486 0.533 81.816 85.436 multiply_cannon_loop 2529 14.6 1.268 1.308 78.409 80.866 ot_mini 118 10.6 0.001 0.001 50.244 50.721 multiply_cannon_multrec 30348 15.6 22.423 26.712 34.475 40.510 mp_waitall_1 216598 16.6 25.122 38.211 25.122 38.211 rebuild_ks_matrix 129 8.3 0.001 0.001 33.679 34.082 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.678 34.081 init_scf_loop 11 6.9 0.000 0.000 33.324 33.325 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.341 30.704 prepare_preconditioner 11 7.9 0.000 0.000 28.918 28.979 make_preconditioner 11 8.9 0.000 0.000 28.918 28.979 qs_ot_get_derivative 118 11.6 0.001 0.002 28.176 28.705 multiply_cannon_metrocomm3 30348 15.6 0.095 0.101 15.875 28.391 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.653 28.166 qs_ot_get_p 129 10.4 0.001 0.001 22.836 23.485 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.067 23.149 apply_single 129 13.6 0.001 0.001 22.067 23.149 ot_diis_step 118 11.6 0.014 0.015 21.893 21.895 multiply_cannon_sync_h2d 30348 15.6 19.716 21.870 19.716 21.870 qs_ot_p2m_diag 84 11.4 0.189 0.218 17.968 18.008 cp_fm_cholesky_invert 11 10.9 17.066 17.079 17.066 17.079 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.783 16.784 make_m2s 5058 13.6 0.091 0.095 13.919 15.554 make_images 5058 14.6 1.152 1.338 13.708 15.343 init_scf_run 11 5.9 0.000 0.001 15.143 15.144 scf_env_initial_rho_setup 11 6.9 0.001 0.001 15.142 15.144 sum_up_and_integrate 129 10.3 0.114 0.132 14.959 14.987 integrate_v_rspace 129 11.3 0.003 0.004 14.844 14.876 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.232 14.273 calculate_rho_elec 129 8.7 0.089 0.106 14.231 14.273 cp_fm_diag_elpa 84 13.4 0.000 0.001 13.509 13.519 cp_fm_redistribute_end 84 14.4 7.885 13.432 7.898 13.433 dbcsr_mm_accdrv_process 62758 16.2 6.520 8.917 11.505 13.340 cp_fm_diag_elpa_base 84 14.4 5.289 12.903 5.515 13.270 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.143 11.546 density_rs2pw 129 9.7 0.006 0.007 7.641 10.408 multiply_cannon_metrocomm4 27819 15.6 0.095 0.106 3.745 10.228 make_images_data 5058 15.6 0.065 0.072 8.256 10.145 calculate_dm_sparse 129 9.5 0.001 0.001 9.946 10.092 mp_irecv_dv 70084 16.3 3.551 9.848 3.551 9.848 hybrid_alltoall_any 5245 16.5 0.346 1.505 7.043 9.243 pw_transfer 1559 11.6 0.088 0.102 8.950 9.024 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.722 8.799 rs_pw_transfer 1054 12.0 0.014 0.016 5.965 8.770 wfi_extrapolate 11 7.9 0.001 0.001 8.476 8.476 fft_wrap_pw1pw2_140 527 13.2 0.475 0.539 7.595 7.692 grid_integrate_task_list 129 12.3 7.207 7.628 7.207 7.628 cp_fm_cholesky_decompose 22 10.9 7.049 7.125 7.049 7.125 fft3d_ps 1301 14.7 2.768 2.936 7.076 7.122 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.234 6.973 calculate_first_density_matrix 1 7.0 0.000 0.001 6.492 6.494 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.260 6.269 grid_collocate_task_list 129 9.7 4.737 5.899 4.737 5.899 jit_kernel_multiply 12 15.9 1.970 5.774 1.970 5.774 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 5.688 5.719 mp_sum_l 7936 13.1 3.963 5.613 3.963 5.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.414 5.554 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.387 5.455 mp_waitany 11836 13.9 2.433 5.262 2.433 5.262 potential_pw2rs 129 12.3 0.015 0.018 5.211 5.233 mp_alltoall_d11v 2429 14.1 3.956 4.940 3.956 4.940 mp_allgather_i34 2529 14.6 1.979 4.839 1.979 4.839 rs_pw_transfer_RS2PW_140 140 11.5 0.356 0.380 2.072 4.832 mp_sum_d 4499 12.2 2.687 4.141 2.687 4.141 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=198.679000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=785.818182, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 931.856384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57277. MP_Allreduce 11236 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.042 179.386 179.398 qs_mol_dyn_low 1 2.0 0.003 0.005 178.823 178.837 qs_forces 11 3.9 0.007 0.022 178.702 178.706 qs_energies 11 4.9 0.006 0.031 172.007 172.019 scf_env_do_scf 11 5.9 0.004 0.010 156.053 156.054 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 120.256 120.256 velocity_verlet 10 3.0 0.010 0.015 114.049 114.051 dbcsr_multiply_generic 2485 12.5 0.181 0.185 81.668 82.900 qs_scf_new_mos 116 7.6 0.001 0.001 81.693 82.063 qs_scf_loop_do_ot 116 8.6 0.001 0.002 81.692 82.062 ot_scf_mini 116 9.6 0.004 0.004 77.558 77.972 multiply_cannon 2485 13.5 0.497 0.515 61.725 65.873 multiply_cannon_loop 2485 14.5 0.852 0.879 58.442 60.885 ot_mini 116 10.6 0.001 0.001 42.538 42.963 init_scf_loop 11 6.9 0.001 0.004 35.682 35.683 mp_waitall_1 169034 16.6 25.130 34.766 25.130 34.766 prepare_preconditioner 11 7.9 0.000 0.001 31.593 31.647 make_preconditioner 11 8.9 0.000 0.002 31.593 31.646 rebuild_ks_matrix 127 8.3 0.001 0.001 31.026 31.542 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 31.026 31.542 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.222 30.653 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.920 28.392 multiply_cannon_multrec 19880 15.5 13.431 16.553 22.292 25.370 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 15.209 24.895 qs_ot_get_derivative 116 11.6 0.001 0.002 22.782 23.195 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.938 21.146 apply_single 127 13.6 0.001 0.001 19.937 21.145 qs_ot_get_p 127 10.4 0.001 0.001 20.398 20.969 ot_diis_step 116 11.6 0.018 0.018 19.655 19.656 qs_ot_p2m_diag 82 11.4 0.261 0.268 16.156 16.164 make_m2s 4970 13.5 0.080 0.085 14.939 15.968 multiply_cannon_sync_h2d 19880 15.5 14.087 15.855 14.087 15.855 make_images 4970 14.5 1.149 1.249 14.707 15.734 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.149 15.149 sum_up_and_integrate 127 10.3 0.130 0.140 14.848 14.873 cp_fm_cholesky_invert 11 10.9 14.844 14.853 14.844 14.853 integrate_v_rspace 127 11.3 0.003 0.004 14.718 14.743 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.513 14.549 calculate_rho_elec 127 8.7 0.131 0.146 14.512 14.548 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.974 11.976 cp_fm_redistribute_end 82 14.4 4.524 11.901 4.541 11.903 cp_fm_diag_elpa_base 82 14.4 6.929 11.288 7.339 11.764 init_scf_run 11 5.9 0.000 0.001 10.810 10.810 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.809 10.810 make_images_data 4970 15.5 0.061 0.068 9.309 10.713 hybrid_alltoall_any 5155 16.4 0.433 1.984 8.187 9.893 density_rs2pw 127 9.7 0.006 0.006 7.422 9.661 pw_transfer 1535 11.6 0.086 0.105 9.291 9.407 multiply_cannon_metrocomm4 17395 15.5 0.061 0.070 3.402 9.330 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.068 9.189 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.797 9.099 mp_irecv_dv 49801 16.2 3.281 9.083 3.281 9.083 dbcsr_mm_accdrv_process 41158 16.2 4.450 5.484 8.322 8.442 fft_wrap_pw1pw2_140 519 13.2 0.474 0.519 7.988 8.114 grid_integrate_task_list 127 12.3 7.221 7.800 7.221 7.800 wfi_extrapolate 11 7.9 0.001 0.001 7.537 7.537 cp_fm_upper_to_full 104 14.5 5.843 7.537 5.843 7.537 cp_fm_cholesky_decompose 22 10.9 7.486 7.509 7.486 7.509 rs_pw_transfer 1038 11.9 0.013 0.014 5.285 7.502 fft3d_ps 1281 14.7 2.646 2.870 7.346 7.427 dbcsr_complete_redistribute 393 12.7 1.168 1.197 4.650 6.412 calculate_dm_sparse 127 9.5 0.001 0.001 5.869 5.957 grid_collocate_task_list 127 9.7 4.861 5.611 4.861 5.611 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.604 5.608 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.626 5.370 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.428 5.185 mp_alltoall_d11v 2401 14.1 4.451 5.170 4.451 5.170 mp_allgather_i34 2485 14.5 1.830 5.027 1.830 5.027 potential_pw2rs 127 12.3 0.020 0.022 4.980 5.002 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.686 4.839 mp_waitany 11660 13.9 2.339 4.653 2.339 4.653 mp_sum_l 7804 13.0 3.221 4.635 3.221 4.635 rs_pw_transfer_RS2PW_140 138 11.5 0.330 0.353 1.907 4.129 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.351 4.092 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.018 4.045 mp_alltoall_i22 712 14.1 1.924 3.824 1.924 3.824 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.789 3.789 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.615 3.660 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.398000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.727273, yerr=10.855170 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.138360E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.064 192.286 192.287 qs_mol_dyn_low 1 2.0 0.003 0.004 191.819 191.832 qs_forces 11 3.9 0.003 0.006 191.704 191.708 qs_energies 11 4.9 0.002 0.005 184.491 184.501 scf_env_do_scf 11 5.9 0.001 0.002 165.382 165.390 velocity_verlet 10 3.0 0.002 0.002 125.542 125.544 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.385 118.386 dbcsr_multiply_generic 2507 12.6 0.189 0.196 81.260 81.922 qs_scf_new_mos 117 7.6 0.001 0.001 80.928 81.257 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.928 81.256 ot_scf_mini 117 9.6 0.003 0.005 76.470 76.766 multiply_cannon 2507 13.6 0.558 0.590 56.999 59.882 multiply_cannon_loop 2507 14.6 1.184 1.215 53.163 54.857 init_scf_loop 11 6.9 0.001 0.004 46.872 46.874 prepare_preconditioner 11 7.9 0.000 0.001 42.758 42.776 make_preconditioner 11 8.9 0.000 0.000 42.758 42.776 ot_mini 117 10.6 0.001 0.001 42.365 42.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.487 41.394 multiply_cannon_multrec 30084 15.6 14.224 19.358 28.648 33.254 rebuild_ks_matrix 128 8.3 0.001 0.001 29.877 30.158 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.877 30.157 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.911 27.155 mp_waitall_1 147882 16.7 16.549 26.397 16.549 26.397 qs_ot_get_derivative 117 11.6 0.001 0.002 22.738 23.042 make_m2s 5014 13.6 0.096 0.099 19.888 20.984 qs_ot_get_p 128 10.4 0.001 0.001 20.417 20.750 make_images 5014 14.6 1.943 2.204 19.581 20.677 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.051 19.527 apply_single 128 13.6 0.001 0.001 19.051 19.527 ot_diis_step 117 11.6 0.017 0.018 19.494 19.496 cp_fm_cholesky_invert 11 10.9 16.686 16.697 16.686 16.697 qs_ot_p2m_diag 83 11.4 0.342 0.390 16.285 16.337 cp_fm_upper_to_full 105 14.7 10.916 16.162 10.916 16.162 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.984 14.986 sum_up_and_integrate 128 10.3 0.140 0.153 14.819 14.842 integrate_v_rspace 128 11.3 0.003 0.004 14.678 14.706 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.581 14.629 calculate_rho_elec 128 8.7 0.175 0.190 14.581 14.629 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.140 14.491 dbcsr_mm_accdrv_process 62264 16.2 8.775 10.508 13.998 14.479 init_scf_run 11 5.9 0.000 0.001 13.157 13.159 scf_env_initial_rho_setup 11 6.9 0.001 0.004 13.157 13.159 dbcsr_complete_redistribute 395 12.7 1.512 1.622 8.865 12.619 multiply_cannon_sync_h2d 30084 15.6 11.738 12.397 11.738 12.397 make_images_data 5014 15.6 0.065 0.069 10.539 12.332 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.727 11.731 cp_fm_redistribute_end 83 14.4 2.037 11.616 2.056 11.622 cp_fm_diag_elpa_base 83 14.4 8.940 10.960 9.520 11.435 copy_fm_to_dbcsr 209 11.7 0.002 0.003 7.481 11.234 hybrid_alltoall_any 5200 16.5 0.526 2.219 9.509 11.188 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.248 9.893 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.334 9.551 pw_transfer 1547 11.6 0.087 0.106 9.220 9.326 mp_alltoall_i22 716 14.1 5.479 9.213 5.479 9.213 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.993 9.101 density_rs2pw 128 9.7 0.006 0.006 7.090 8.538 calculate_dm_sparse 128 9.5 0.001 0.001 8.421 8.528 fft_wrap_pw1pw2_140 523 13.2 0.482 0.493 7.920 8.033 cp_fm_cholesky_decompose 22 10.9 7.810 7.916 7.810 7.916 grid_integrate_task_list 128 12.3 7.533 7.907 7.533 7.907 wfi_extrapolate 11 7.9 0.001 0.001 7.647 7.647 fft3d_ps 1291 14.7 2.776 2.881 7.217 7.306 multiply_cannon_metrocomm4 25070 15.6 0.075 0.083 2.767 6.858 mp_irecv_dv 76098 16.2 2.622 6.599 2.622 6.599 rs_pw_transfer 1046 11.9 0.013 0.014 4.900 6.409 grid_collocate_task_list 128 9.7 5.080 5.798 5.080 5.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.479 5.524 mp_alltoall_d11v 2415 14.1 4.616 5.429 4.616 5.429 calculate_first_density_matrix 1 7.0 0.000 0.002 5.308 5.311 potential_pw2rs 128 12.3 0.022 0.023 4.722 4.737 jit_kernel_multiply 12 15.8 2.513 4.712 2.513 4.712 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.606 4.627 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.442 4.496 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.475 4.476 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.384 4.469 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.201 4.260 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.287000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1072.636364, yerr=20.825128 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1919850381312 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.792282E+12 0.0% 0.0% 100.0% flops max/rank 5.819790E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705435008 0.0% 0.0% 100.0% number of processed stacks 1943760 0.0% 0.0% 100.0% average stack size 0.0 0.0 3449.7 marketing flops 143.503447E+12 ------------------------------------------------------------------------------- # multiplications 2483 max memory usage/rank 1.507324E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238368 MPI messages size (bytes): total size 1.321004E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.541866E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112608 59039023104 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8846 52. MP_Alltoall 9592 804121. MP_ISend 39684 2106261. MP_IRecv 39684 2105362. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3956 58872. MP_Allreduce 10887 1178. MP_Sync 87 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 176.773 176.775 qs_mol_dyn_low 1 2.0 0.003 0.003 176.346 176.358 qs_forces 11 3.9 0.003 0.003 176.252 176.255 qs_energies 11 4.9 0.002 0.002 168.641 168.650 scf_env_do_scf 11 5.9 0.001 0.001 151.091 151.096 velocity_verlet 10 3.0 0.002 0.003 115.083 115.087 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.915 114.917 qs_scf_new_mos 116 7.6 0.001 0.001 75.946 76.041 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.945 76.041 dbcsr_multiply_generic 2483 12.5 0.181 0.188 74.757 75.289 ot_scf_mini 116 9.6 0.004 0.004 71.549 71.682 multiply_cannon 2483 13.5 0.575 0.601 54.738 59.408 multiply_cannon_loop 2483 14.5 0.441 0.456 50.138 51.248 ot_mini 116 10.6 0.001 0.001 39.891 40.035 init_scf_loop 11 6.9 0.000 0.000 36.024 36.026 mp_waitall_1 124584 16.7 26.504 33.984 26.504 33.984 prepare_preconditioner 11 7.9 0.000 0.000 32.032 32.059 make_preconditioner 11 8.9 0.000 0.000 32.032 32.059 rebuild_ks_matrix 127 8.3 0.001 0.001 30.175 30.323 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.174 30.322 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.859 30.128 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.383 27.520 multiply_cannon_multrec 9932 15.5 10.306 14.741 17.894 22.668 multiply_cannon_metrocomm3 9932 15.5 0.023 0.024 12.764 20.272 qs_ot_get_derivative 116 11.6 0.001 0.002 20.018 20.149 ot_diis_step 116 11.6 0.019 0.020 19.800 19.800 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.395 19.712 apply_single 127 13.6 0.001 0.001 19.395 19.712 qs_ot_get_p 127 10.4 0.001 0.001 18.729 18.855 make_m2s 4966 13.5 0.066 0.071 16.344 18.731 make_images 4966 14.5 2.243 2.587 16.040 18.424 cp_fm_cholesky_invert 11 10.9 18.410 18.416 18.410 18.416 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.525 15.573 calculate_rho_elec 127 8.7 0.257 0.266 15.524 15.572 sum_up_and_integrate 127 10.3 0.178 0.188 15.177 15.229 integrate_v_rspace 127 11.3 0.004 0.004 14.998 15.054 qs_ot_p2m_diag 83 11.4 0.495 0.500 14.904 14.920 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.748 13.749 make_images_data 4966 15.5 0.052 0.059 10.056 12.656 hybrid_alltoall_any 5152 16.4 0.832 3.774 9.844 12.295 multiply_cannon_sync_h2d 9932 15.5 11.419 12.213 11.419 12.213 init_scf_run 11 5.9 0.000 0.001 10.704 10.704 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.704 10.704 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.627 10.639 cp_fm_diag_elpa_base 83 14.4 10.384 10.463 10.622 10.634 pw_transfer 1535 11.6 0.085 0.094 10.037 10.080 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.815 9.864 fft_wrap_pw1pw2_140 519 13.2 0.493 0.514 8.639 8.695 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.114 8.218 grid_integrate_task_list 127 12.3 7.708 8.055 7.708 8.055 fft3d_ps 1281 14.7 2.691 2.770 7.988 8.024 density_rs2pw 127 9.7 0.005 0.006 7.161 8.023 cp_fm_cholesky_decompose 22 10.9 7.913 8.014 7.913 8.014 dbcsr_mm_accdrv_process 20582 16.1 2.602 3.570 7.225 7.901 multiply_cannon_metrocomm1 9932 15.5 0.028 0.029 4.392 7.810 mp_allgather_i34 2483 14.5 2.839 7.600 2.839 7.600 wfi_extrapolate 11 7.9 0.001 0.001 7.511 7.511 calculate_dm_sparse 127 9.5 0.001 0.001 6.148 6.251 mp_alltoall_d11v 2407 14.1 5.143 6.111 5.143 6.111 grid_collocate_task_list 127 9.7 5.351 6.001 5.351 6.001 dbcsr_complete_redistribute 395 12.7 2.114 2.187 5.281 5.678 multiply_cannon_metrocomm4 7449 15.5 0.024 0.027 1.859 5.616 mp_irecv_dv 28602 15.9 1.823 5.529 1.823 5.529 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.371 5.384 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.315 5.317 rs_pw_transfer 1038 11.9 0.012 0.013 4.219 5.130 potential_pw2rs 127 12.3 0.026 0.026 4.679 4.691 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.278 4.309 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.866 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.454 3.784 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.733 3.773 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.689 3.700 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.601 3.685 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.775000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1399.545455, yerr=50.526975 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.006079E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.088 299.309 299.314 qs_mol_dyn_low 1 2.0 0.003 0.003 298.631 298.643 qs_forces 11 3.9 0.003 0.003 298.535 298.538 qs_energies 11 4.9 0.002 0.002 289.237 289.245 scf_env_do_scf 11 5.9 0.001 0.001 266.813 266.824 velocity_verlet 10 3.0 0.002 0.002 215.340 215.348 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 140.296 140.298 init_scf_loop 11 6.9 0.000 0.000 126.260 126.263 prepare_preconditioner 11 7.9 0.000 0.000 121.216 121.246 make_preconditioner 11 8.9 0.000 0.000 121.216 121.246 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.273 118.382 qs_scf_new_mos 116 7.6 0.001 0.001 89.830 89.952 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.829 89.951 ot_scf_mini 116 9.6 0.004 0.004 85.198 85.226 dbcsr_multiply_generic 2485 12.5 0.212 0.219 80.738 81.261 cp_fm_upper_to_full 104 14.8 52.831 75.768 52.831 75.768 multiply_cannon 2485 13.5 0.702 0.744 56.664 57.256 multiply_cannon_loop 2485 14.5 0.467 0.480 53.180 54.512 ot_mini 116 10.6 0.001 0.001 43.140 43.168 dbcsr_complete_redistribute 393 12.7 3.968 4.018 29.600 42.591 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.208 39.132 rebuild_ks_matrix 127 8.3 0.001 0.001 37.983 38.023 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 37.982 38.022 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.896 36.628 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.017 35.060 cp_fm_cholesky_invert 11 10.9 34.905 34.912 34.905 34.912 mp_alltoall_i22 712 14.1 21.764 34.485 21.764 34.485 mp_waitall_1 102768 16.8 26.093 29.479 26.093 29.479 qs_ot_get_p 127 10.4 0.001 0.001 27.134 27.155 qs_ot_get_derivative 116 11.6 0.002 0.002 23.649 23.682 qs_ot_p2m_diag 82 11.4 0.867 0.872 23.148 23.179 cp_dbcsr_syevd 82 12.4 0.005 0.006 21.414 21.416 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.562 20.578 calculate_rho_elec 127 8.7 0.479 0.481 20.561 20.577 make_m2s 4970 13.5 0.077 0.079 19.420 20.187 sum_up_and_integrate 127 10.3 0.319 0.322 19.724 19.810 make_images 4970 14.5 3.701 3.816 18.944 19.715 integrate_v_rspace 127 11.3 0.004 0.004 19.404 19.491 ot_diis_step 116 11.6 0.022 0.022 19.460 19.461 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.653 18.854 apply_single 127 13.6 0.001 0.001 18.652 18.853 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 17.071 18.192 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.133 18.135 cp_fm_diag_elpa_base 82 14.4 13.844 15.409 18.129 18.130 multiply_cannon_multrec 9940 15.5 10.382 12.178 17.986 18.065 multiply_cannon_sync_h2d 9940 15.5 15.529 15.537 15.529 15.537 init_scf_run 11 5.9 0.000 0.001 12.370 12.371 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.370 12.371 pw_transfer 1535 11.6 0.092 0.093 12.181 12.188 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.948 11.955 hybrid_alltoall_any 5155 16.4 1.295 3.012 10.293 11.781 make_images_data 4970 15.5 0.061 0.066 10.235 11.679 fft_wrap_pw1pw2_140 519 13.2 0.536 0.538 10.558 10.571 fft3d_ps 1281 14.7 2.719 2.727 9.998 10.006 cp_fm_cholesky_decompose 22 10.9 9.223 9.252 9.223 9.252 dbcsr_mm_accdrv_process 20590 16.0 3.796 5.741 7.367 9.190 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.124 9.150 mp_alltoall_d11v 2401 14.1 8.298 9.020 8.298 9.020 wfi_extrapolate 11 7.9 0.001 0.001 8.976 8.976 grid_integrate_task_list 127 12.3 8.502 8.676 8.502 8.676 density_rs2pw 127 9.7 0.005 0.005 8.315 8.456 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.055 8.056 calculate_dm_sparse 127 9.5 0.001 0.001 6.611 6.732 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.280 6.363 grid_collocate_task_list 127 9.7 6.306 6.339 6.306 6.339 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.074 6.158 rs_scatter_matrices 138 9.7 3.548 4.486 5.900 6.135 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.314000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2653.909091, yerr=176.639054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/21/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 2.766000E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 419739 0.0% 0.0% 100.0% average stack size 0.0 0.0 22952.9 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 1.262141E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.037 84.482 84.484 qs_energies 1 2.0 0.000 0.000 84.034 84.041 ls_scf 1 3.0 0.000 0.000 83.117 83.123 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.153 72.312 multiply_cannon 111 7.7 0.017 0.021 55.657 56.837 multiply_cannon_loop 111 8.7 0.212 0.226 52.226 53.523 ls_scf_main 1 4.0 0.000 0.000 52.060 52.060 density_matrix_trs4 2 5.0 0.002 0.003 46.744 46.821 ls_scf_init_scf 1 4.0 0.000 0.000 28.105 28.107 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.991 27.055 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.855 24.871 mp_waitall_1 11031 10.9 21.280 24.051 21.280 24.051 multiply_cannon_multrec 2664 9.7 8.253 8.936 15.583 17.226 multiply_cannon_sync_h2d 2664 9.7 14.334 16.136 14.334 16.136 make_m2s 222 7.7 0.008 0.010 12.959 13.521 make_images 222 8.7 0.099 0.108 12.937 13.502 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 8.841 11.156 make_images_data 222 9.7 0.004 0.005 7.541 8.085 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.226 7.974 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.609 6.944 7.903 hybrid_alltoall_any 227 10.6 0.215 1.835 6.475 7.601 dbcsr_mm_accdrv_process_sort 4760 11.4 6.230 7.115 6.230 7.115 calculate_norms 4752 9.8 5.606 6.192 5.606 6.192 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.796 4.906 mp_sum_l 807 5.4 2.996 4.671 2.996 4.671 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.025 3.881 mp_irecv_dv 6231 10.9 2.008 3.854 2.008 3.854 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.252 3.612 arnoldi_extremal 4 6.8 0.000 0.000 3.282 3.300 arnoldi_normal_ev 4 7.8 0.001 0.002 3.282 3.300 build_subspace 16 8.4 0.009 0.012 3.189 3.192 make_images_sizes 222 9.7 0.000 0.000 0.705 3.175 mp_alltoall_i44 222 10.7 0.705 3.175 0.705 3.175 ls_scf_post 1 4.0 0.000 0.000 2.952 2.958 ls_scf_store_result 1 5.0 0.000 0.000 2.767 2.811 dbcsr_special_finalize 555 9.7 0.005 0.006 2.362 2.762 dbcsr_merge_single_wm 555 10.7 0.466 0.592 2.354 2.754 make_images_pack 222 9.7 2.206 2.624 2.208 2.625 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.326 2.566 dbcsr_sort_data 658 11.4 2.148 2.501 2.148 2.501 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.462 2.072 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.250 2.297 buffer_matrices_ensure_size 222 8.7 1.767 2.076 1.767 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.785 1.787 rebuild_ks_matrix 3 7.3 0.000 0.000 1.776 1.777 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.776 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.484000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/22/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 5.588524E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 368848 0.0% 0.0% 100.0% average stack size 0.0 0.0 26119.8 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 2.103837E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.048 90.729 90.730 qs_energies 1 2.0 0.000 0.000 90.279 90.284 ls_scf 1 3.0 0.000 0.000 88.965 88.970 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.019 75.367 multiply_cannon 111 7.7 0.028 0.036 52.896 56.724 ls_scf_main 1 4.0 0.000 0.000 54.753 54.757 multiply_cannon_loop 111 8.7 0.116 0.122 49.667 53.328 density_matrix_trs4 2 5.0 0.002 0.003 49.022 49.220 ls_scf_init_scf 1 4.0 0.000 0.000 30.685 30.686 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.501 29.579 mp_waitall_1 9105 10.9 20.850 29.219 20.850 29.219 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.099 27.112 multiply_cannon_multrec 1332 9.7 13.216 16.419 22.442 26.809 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.279 20.217 make_m2s 222 7.7 0.006 0.007 15.473 16.148 make_images 222 8.7 1.566 1.924 15.443 16.120 dbcsr_mm_accdrv_process 4041 10.4 0.269 0.453 8.826 10.422 dbcsr_mm_accdrv_process_sort 4041 11.4 8.411 9.969 8.411 9.969 make_images_data 222 9.7 0.004 0.004 8.925 9.816 hybrid_alltoall_any 227 10.6 0.518 2.425 8.346 9.282 mp_sum_l 807 5.4 5.357 8.392 5.357 8.392 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.188 7.460 mp_irecv_dv 3311 11.0 3.169 7.411 3.169 7.411 calculate_norms 2376 9.8 6.020 6.828 6.020 6.828 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.149 6.585 multiply_cannon_sync_h2d 1332 9.7 4.886 5.933 4.886 5.933 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.210 arnoldi_extremal 4 6.8 0.000 0.000 4.630 4.650 arnoldi_normal_ev 4 7.8 0.001 0.004 4.630 4.650 build_subspace 16 8.4 0.014 0.021 4.376 4.379 ls_scf_post 1 4.0 0.000 0.000 3.528 3.533 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.140 3.373 ls_scf_store_result 1 5.0 0.000 0.000 3.240 3.352 dbcsr_matrix_vector_mult_local 304 10.0 2.731 3.211 2.733 3.213 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.275 2.817 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.614 2.695 mp_allgather_i34 111 8.7 0.929 2.421 0.929 2.421 make_images_pack 222 9.7 2.018 2.396 2.021 2.398 dbcsr_sort_data 436 11.2 1.846 2.086 1.846 2.086 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.876 1.878 dbcsr_data_new 4174 10.1 1.610 1.849 1.610 1.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.730000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1709.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/23/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 8.404608E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 353133 0.0% 0.0% 100.0% average stack size 0.0 0.0 27282.1 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 2.702148E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.118 94.071 94.073 qs_energies 1 2.0 0.000 0.000 93.314 93.318 ls_scf 1 3.0 0.000 0.001 91.916 91.921 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.659 76.934 ls_scf_main 1 4.0 0.000 0.001 57.737 57.740 multiply_cannon 111 7.7 0.038 0.096 52.780 56.664 multiply_cannon_loop 111 8.7 0.100 0.111 49.077 53.862 density_matrix_trs4 2 5.0 0.002 0.003 51.770 51.958 mp_waitall_1 7281 11.0 23.178 33.790 23.178 33.790 ls_scf_init_scf 1 4.0 0.000 0.001 30.656 30.659 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.460 29.527 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.072 27.084 multiply_cannon_multrec 888 9.7 12.653 15.269 21.199 24.394 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.766 23.363 make_m2s 222 7.7 0.006 0.007 16.708 18.123 make_images 222 8.7 1.968 2.269 16.669 18.084 make_images_data 222 9.7 0.003 0.004 9.338 10.503 hybrid_alltoall_any 227 10.6 0.623 2.872 8.912 10.468 mp_sum_l 807 5.4 5.631 9.639 5.631 9.639 dbcsr_mm_accdrv_process 3754 10.4 0.263 0.434 8.046 9.268 dbcsr_mm_accdrv_process_sort 3754 11.4 7.672 8.834 7.672 8.834 multiply_cannon_sync_h2d 888 9.7 6.078 7.702 6.078 7.702 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.331 7.501 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.452 7.021 mp_irecv_dv 2335 11.1 2.436 6.964 2.436 6.964 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.700 6.606 arnoldi_extremal 4 6.8 0.000 0.000 5.111 5.126 arnoldi_normal_ev 4 7.8 0.001 0.005 5.111 5.126 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.916 5.117 build_subspace 16 8.4 0.014 0.020 4.807 4.812 calculate_norms 1584 9.8 4.262 4.620 4.262 4.620 mp_allgather_i34 111 8.7 1.544 3.905 1.544 3.905 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.472 3.788 dbcsr_matrix_vector_mult_local 304 10.0 3.003 3.576 3.005 3.578 ls_scf_post 1 4.0 0.000 0.001 3.523 3.529 ls_scf_store_result 1 5.0 0.000 0.000 3.269 3.346 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.837 2.948 dbcsr_sort_data 325 11.1 1.897 2.160 1.897 2.160 make_images_pack 222 9.7 1.813 2.131 1.816 2.133 dbcsr_data_release 9322 10.9 1.339 2.088 1.339 2.088 make_images_sizes 222 9.7 0.000 0.000 1.052 1.982 mp_alltoall_i44 222 10.7 1.052 1.981 1.052 1.981 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.923 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.905 1.907 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.905 1.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.073000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2180.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/24/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 10.747127E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 369794 0.0% 0.0% 100.0% average stack size 0.0 0.0 26053.0 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 3.349246E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.084 0.142 96.039 96.048 qs_energies 1 2.0 0.000 0.000 95.224 95.230 ls_scf 1 3.0 0.000 0.001 93.580 93.585 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.343 77.614 ls_scf_main 1 4.0 0.000 0.003 57.994 57.995 multiply_cannon 111 7.7 0.061 0.179 50.876 55.778 density_matrix_trs4 2 5.0 0.002 0.003 51.900 52.027 multiply_cannon_loop 111 8.7 0.114 0.126 45.743 49.091 ls_scf_init_scf 1 4.0 0.001 0.005 32.328 32.330 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.105 31.185 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.660 28.674 mp_waitall_1 6369 11.0 21.428 27.451 21.428 27.451 multiply_cannon_multrec 1332 9.7 14.307 17.916 22.162 25.438 make_m2s 222 7.7 0.007 0.008 20.685 22.207 make_images 222 8.7 3.134 3.594 20.635 22.159 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.504 16.552 make_images_data 222 9.7 0.004 0.004 11.414 13.019 hybrid_alltoall_any 227 10.6 0.801 3.821 10.607 12.669 dbcsr_mm_accdrv_process 3641 10.4 0.228 0.412 7.487 9.004 dbcsr_mm_accdrv_process_sort 3641 11.4 7.114 8.584 7.114 8.584 mp_sum_l 807 5.4 4.255 7.968 4.255 7.968 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.252 6.358 multiply_cannon_sync_h2d 1332 9.7 5.537 6.068 5.537 6.068 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.067 5.947 mp_irecv_dv 3229 10.9 2.045 5.884 2.045 5.884 arnoldi_extremal 4 6.8 0.000 0.000 5.234 5.245 arnoldi_normal_ev 4 7.8 0.001 0.005 5.234 5.245 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.343 5.127 build_subspace 16 8.4 0.014 0.021 4.898 4.904 mp_allgather_i34 111 8.7 2.268 4.822 2.268 4.822 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.529 4.718 calculate_norms 2376 9.8 4.190 4.551 4.190 4.551 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.610 3.894 dbcsr_matrix_vector_mult_local 304 10.0 3.164 3.669 3.166 3.671 dbcsr_sort_data 658 11.4 3.048 3.486 3.048 3.486 dbcsr_special_finalize 555 9.7 0.006 0.007 2.788 3.280 dbcsr_merge_single_wm 555 10.7 0.532 0.669 2.780 3.272 ls_scf_post 1 4.0 0.000 0.001 3.258 3.265 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.010 3.064 ls_scf_store_result 1 5.0 0.000 0.000 2.953 3.032 dbcsr_data_release 10477 10.7 1.582 2.412 1.582 2.412 dbcsr_finalize 304 7.8 0.049 0.061 1.802 1.988 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.956 1.958 rebuild_ks_matrix 3 7.3 0.000 0.000 1.933 1.935 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.933 1.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.048000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2719.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/25/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 15.383312E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 336818 0.0% 0.0% 100.0% average stack size 0.0 0.0 28603.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 4.627677E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.058 99.128 99.129 qs_energies 1 2.0 0.000 0.000 98.334 98.340 ls_scf 1 3.0 0.000 0.000 96.395 96.405 dbcsr_multiply_generic 111 6.7 0.017 0.019 77.482 77.710 ls_scf_main 1 4.0 0.000 0.000 62.218 62.219 multiply_cannon 111 7.7 0.085 0.179 55.066 60.616 density_matrix_trs4 2 5.0 0.002 0.003 54.932 55.058 multiply_cannon_loop 111 8.7 0.070 0.079 50.484 52.473 ls_scf_init_scf 1 4.0 0.000 0.000 30.631 30.636 mp_waitall_1 5436 11.0 25.873 30.417 25.873 30.417 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.273 29.310 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.009 27.019 multiply_cannon_multrec 444 9.7 14.049 16.076 21.073 22.988 make_m2s 222 7.7 0.004 0.005 17.321 19.897 make_images 222 8.7 3.712 4.410 17.260 19.838 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.514 16.516 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.561 14.132 make_images_data 222 9.7 0.003 0.004 9.602 12.067 hybrid_alltoall_any 227 10.6 0.791 3.784 9.382 11.936 multiply_cannon_sync_h2d 444 9.7 6.502 8.133 6.502 8.133 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.341 6.732 7.882 dbcsr_mm_accdrv_process_sort 3003 11.4 6.417 7.533 6.417 7.533 mp_allgather_i34 111 8.7 2.818 6.981 2.818 6.981 arnoldi_extremal 4 6.8 0.000 0.000 5.946 5.952 arnoldi_normal_ev 4 7.8 0.001 0.005 5.946 5.952 mp_sum_l 807 5.4 3.146 5.571 3.146 5.571 build_subspace 16 8.4 0.015 0.019 5.551 5.562 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.578 4.756 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.213 4.541 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.224 4.428 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.664 4.201 mp_irecv_dv 1241 11.2 1.644 4.171 1.644 4.171 dbcsr_matrix_vector_mult_local 304 10.0 3.656 4.122 3.658 4.124 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.678 3.759 calculate_norms 792 9.8 3.530 3.689 3.530 3.689 ls_scf_post 1 4.0 0.000 0.000 3.545 3.551 ls_scf_store_result 1 5.0 0.000 0.000 3.332 3.371 make_images_sizes 222 9.7 0.000 0.000 1.087 3.225 mp_alltoall_i44 222 10.7 1.087 3.225 1.087 3.225 dbcsr_finalize 304 7.8 0.062 0.077 2.198 2.263 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.198 2.199 rebuild_ks_matrix 3 7.3 0.000 0.000 2.166 2.167 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.166 2.167 dbcsr_merge_all 275 8.9 0.472 0.515 2.046 2.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.129000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3636.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e64b331552e1588eec11f575b9e771b24f7bf608_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.731542E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.098 106.307 106.308 qs_energies 1 2.0 0.000 0.000 104.897 104.911 ls_scf 1 3.0 0.000 0.000 101.957 101.969 dbcsr_multiply_generic 111 6.7 0.023 0.028 75.268 75.369 ls_scf_main 1 4.0 0.000 0.000 64.531 64.532 density_matrix_trs4 2 5.0 0.002 0.003 55.327 55.380 multiply_cannon 111 7.7 0.178 0.222 48.176 49.512 multiply_cannon_loop 111 8.7 0.068 0.070 44.921 45.884 ls_scf_init_scf 1 4.0 0.000 0.000 33.813 33.814 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.162 32.177 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.356 29.362 make_m2s 222 7.7 0.005 0.005 23.431 24.180 make_images 222 8.7 4.584 4.953 23.324 24.071 mp_waitall_1 4527 11.1 20.292 23.818 20.292 23.818 multiply_cannon_multrec 444 9.7 17.837 18.614 22.431 23.007 hybrid_alltoall_any 227 10.6 1.665 3.633 12.472 15.010 make_images_data 222 9.7 0.003 0.003 12.676 14.818 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.071 9.583 multiply_cannon_sync_h2d 444 9.7 8.850 8.894 8.850 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.468 7.477 arnoldi_normal_ev 4 7.8 0.002 0.009 7.468 7.477 build_subspace 16 8.4 0.026 0.037 6.909 6.919 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.474 5.622 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.235 5.326 dbcsr_matrix_vector_mult_local 304 10.0 4.949 5.246 4.952 5.249 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.936 5.184 dbcsr_mm_accdrv_process 1814 10.4 0.161 0.320 4.424 4.554 dbcsr_mm_accdrv_process_sort 1814 11.4 4.126 4.258 4.126 4.258 ls_scf_post 1 4.0 0.000 0.000 3.612 3.626 make_images_sizes 222 9.7 0.000 0.000 1.427 3.487 mp_alltoall_i44 222 10.7 1.426 3.487 1.426 3.487 ls_scf_store_result 1 5.0 0.000 0.000 3.367 3.380 calculate_norms 792 9.8 3.240 3.274 3.240 3.274 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.163 dbcsr_merge_all 275 8.9 0.889 0.913 2.873 2.939 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.910 2.910 dbcsr_complete_redistribute 5 7.6 1.437 1.466 2.782 2.892 dbcsr_data_release 12724 10.6 2.327 2.833 2.327 2.833 mp_allgather_i34 111 8.7 0.889 2.799 0.889 2.799 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.571 2.574 matrix_ls_to_qs 2 6.0 0.000 0.000 2.424 2.543 rebuild_ks_matrix 3 7.3 0.000 0.000 2.506 2.509 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.506 2.509 dbcsr_sort_data 325 11.1 2.446 2.502 2.446 2.502 dbcsr_new_transposed 4 7.5 0.242 0.250 2.302 2.320 dbcsr_frobenius_norm 74 6.6 2.056 2.128 2.201 2.243 dbcsr_add_d 103 6.2 0.000 0.000 2.156 2.227 dbcsr_add_anytype 103 7.2 0.860 0.894 2.156 2.227 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.308000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6818.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e64b331552e1588eec11f575b9e771b24f7bf608 Summary: empty Status: OK