=== This is the CP2K Performance-Test === Updating ce4b5920f..cd366a402 Fast-forward INSTALL.md | 7 +- arch/Linux-x86-64-mingw64-minimal.sopt | 2 +- src/CMakeLists.txt | 4 + src/cp2k_info.F | 8 +- src/hdf5_wrapper.F | 762 ++++++++ src/input_constants.F | 6 +- src/input_cp2k_dft.F | 125 +- src/ipi_driver.F | 314 +--- src/qcschema.F | 1036 +++++++++++ src/qs_active_space_methods.F | 2017 ++++++++++++---------- src/qs_active_space_types.F | 53 +- src/qs_active_space_utils.F | 158 ++ src/qs_energy.F | 4 + src/qs_scf_post_gpw.F | 4 - src/sockets.c | 119 +- src/sockets_interface.F | 304 ++++ tests/QS/regtest-as-1/TEST_FILES | 24 +- tests/QS/regtest-as-1/ch2_gapw_2-3.inp | 31 +- tests/QS/regtest-as-1/ch2_gapw_pp_2-3.inp | 31 +- tests/QS/regtest-as-1/ch2_gpw_pp_2-3.inp | 31 +- tests/QS/regtest-as-1/h2_gapw_2-2.inp | 30 +- tests/QS/regtest-as-1/h2_gapw_2-3.inp | 30 +- tests/QS/regtest-as-1/h2_gapw_2-4.inp | 30 +- tests/QS/regtest-as-1/h2_gapw_pp_2-2.inp | 30 +- tests/QS/regtest-as-1/h2_gapw_pp_2-3.inp | 30 +- tests/QS/regtest-as-1/h2_gapw_pp_2-4.inp | 30 +- tests/QS/regtest-as-1/h2_gpw_pp_2-2.inp | 30 +- tests/QS/regtest-as-1/h2_gpw_pp_2-3.inp | 30 +- tests/QS/regtest-as-1/h2_gpw_pp_2-4.inp | 30 +- tests/QS/regtest-as-1/h2o_gapw_2-2.inp | 30 +- tests/QS/regtest-as-dft/h2_2-2.inp | 38 +- tests/QS/regtest-as-dft/h2_iterative_density.inp | 42 +- tests/QS/regtest-as-dft/h2_manual_1+2.inp | 42 +- tests/QS/regtest-as-dft/h2_manual_1+2_1+3.inp | 43 +- tests/QS/regtest-as-dft/h2_manual_1+3.inp | 42 +- tests/QS/regtest-as-dft/h2_manual_1+3_1+2.inp | 43 +- tests/QS/regtest-as-dft/h2_manual_1+4.inp | 42 +- tests/QS/regtest-as-dft/h2_manual_UKS_1+2.inp | 43 +- tests/QS/regtest-as/be.inp | 36 +- tests/QS/regtest-as/h2o.inp | 44 +- tools/toolchain/scripts/stage7/install_hdf5.sh | 15 +- 41 files changed, 4085 insertions(+), 1685 deletions(-) create mode 100644 src/hdf5_wrapper.F create mode 100644 src/qcschema.F create mode 100644 src/qs_active_space_utils.F create mode 100644 src/sockets_interface.F Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: cd366a4022ad0ce9b8ea778ed4e993fe588c3c12 ################# 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 2.6.6, ELPA 2023.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, 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. # # Last update: 21.06.2023 # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/01 job id: 48544414 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/02 job id: 48544416 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/03 job id: 48544419 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/04 job id: 48544421 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/05 job id: 48544423 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/06 job id: 48544425 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/07 job id: 48544429 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/08 job id: 48544431 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/09 job id: 48544435 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/10 job id: 48544439 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/11 job id: 48544445 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/12 job id: 48544451 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/13 job id: 48544455 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/14 job id: 48544459 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/15 job id: 48544465 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/16 job id: 48544471 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/17 job id: 48544473 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/18 job id: 48544475 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/19 job id: 48544477 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/20 job id: 48544479 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/21 job id: 48544480 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/22 job id: 48544481 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/23 job id: 48544482 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/24 job id: 48544484 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/25 job id: 48544485 --- 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/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/26 job id: 48544486 --- 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 ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/27 job id: 48544488 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ 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]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 424 8. 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 0.026 0.036 138.170 138.171 farming_run 1 2.0 137.691 137.692 138.125 138.129 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488605E+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 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 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 0.009 0.030 116.644 116.644 qs_energies 1 2.0 0.000 0.000 116.434 116.436 mp2_main 1 3.0 0.000 0.000 113.834 113.837 mp2_gpw_main 1 4.0 0.028 0.034 112.514 112.516 mp2_ri_gpw_compute_in 1 5.0 0.173 0.178 93.023 93.383 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.707 55.066 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.163 41.089 46.151 get_2c_integrals 1 6.0 0.008 0.009 37.300 38.143 integrate_v_rspace 273 8.0 0.438 0.453 24.620 29.439 pw_transfer 6555 10.6 0.375 0.393 26.894 27.412 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.505 25.981 grid_integrate_task_list 273 9.0 20.520 25.799 20.520 25.799 fft_wrap_pw1pw2_100 2178 12.4 1.123 1.343 23.048 23.523 compute_2c_integrals 1 7.0 0.010 0.012 19.577 19.579 rpa_ri_compute_en 1 5.0 0.025 0.027 19.371 19.507 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.630 19.184 mp2_eri_2c_integrate_gpw 1 9.0 2.359 2.404 18.627 19.181 cp_fm_cholesky_decompose 12 8.2 17.926 18.796 17.926 18.796 cholesky_decomp 1 7.0 0.000 0.000 16.570 17.458 fft3d_s 5443 13.4 16.155 16.391 16.177 16.413 ao_to_mo_and_store_B_mult_1 272 7.0 10.783 15.321 10.783 15.321 calculate_wavefunction 272 8.0 5.421 5.575 12.363 13.053 rpa_num_int 1 6.0 0.001 0.011 10.974 10.975 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.941 10.960 calc_mat_Q 8 8.0 0.000 0.000 9.554 9.635 contract_S_to_Q 8 9.0 0.000 0.000 8.976 9.061 calc_potential_gpw 544 9.5 0.005 0.006 8.263 8.821 parallel_gemm_fm 14 9.1 0.000 0.000 8.570 8.677 parallel_gemm_fm_cosma 14 10.1 8.570 8.677 8.570 8.677 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.102 8.466 potential_pw2rs 545 10.0 0.106 0.108 7.505 8.171 create_integ_mat 1 6.0 0.022 0.028 7.936 7.936 collocate_single_gaussian 272 10.0 0.039 0.042 7.360 7.661 array2fm 1 7.0 0.000 0.000 6.658 7.162 pw_scatter_s 2720 13.7 4.342 4.458 4.342 4.458 pw_gather_s 2722 13.2 3.470 3.806 3.470 3.806 array2fm_buffer_send 1 8.0 2.899 3.111 2.899 3.111 pw_poisson_solve 545 10.5 1.116 1.172 2.212 2.503 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.516264, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2793.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 424 9. 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 0.038 420.265 420.266 farming_run 1 2.0 419.369 419.381 420.222 420.225 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.242403E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 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 0.012 0.034 217.153 217.153 qs_energies 1 2.0 0.000 0.000 216.934 216.941 scf_env_do_scf 1 3.0 0.000 0.000 114.940 114.940 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.043 114.051 rebuild_ks_matrix 4 6.0 0.000 0.000 114.042 114.050 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.063 114.041 114.050 hfx_ks_matrix 4 8.0 0.001 0.001 113.697 113.701 integrate_four_center 4 9.0 0.153 0.458 113.696 113.700 integrate_four_center_main 4 10.0 0.146 0.606 102.221 104.914 integrate_four_center_bin 265 11.0 102.074 104.912 102.074 104.912 mp2_main 1 3.0 0.000 0.000 101.703 101.710 mp2_gpw_main 1 4.0 0.047 0.071 100.862 100.872 init_scf_loop 1 4.0 0.000 0.000 97.053 97.053 mp2_ri_gpw_compute_in 1 5.0 0.064 0.067 73.998 75.115 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.861 54.978 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.160 41.400 46.479 integrate_v_rspace 95 8.0 0.396 0.567 27.869 32.767 pw_transfer 2240 10.6 0.147 0.162 29.448 29.825 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.464 28.850 mp2_ri_gpw_compute_en 1 5.0 0.055 0.064 26.707 28.423 grid_integrate_task_list 95 9.0 23.293 28.423 23.293 28.423 ao_to_mo_and_store_B_mult_1 91 7.0 10.781 28.109 10.781 28.109 fft_wrap_pw1pw2_100 730 12.4 1.259 1.471 26.227 26.614 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.905 24.931 24.941 get_2c_integrals 1 6.0 0.000 0.000 20.040 20.073 compute_2c_integrals 1 7.0 0.003 0.004 19.030 19.033 fft3d_s 1823 13.4 18.486 18.915 18.499 18.929 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.541 18.907 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.874 18.540 18.907 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.886 17.886 calculate_wavefunction 91 8.0 2.009 2.035 9.620 9.823 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.574 8.821 9.138 potential_pw2rs 186 10.0 0.033 0.034 8.389 9.005 local_gemm 172 8.0 8.266 8.580 8.266 8.580 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.104 8.518 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.512 7.892 8.245 mp_sync 37 10.5 3.440 8.102 3.440 8.102 calc_potential_gpw 182 9.5 0.002 0.002 7.880 8.068 collocate_single_gaussian 91 10.0 0.016 0.019 7.717 7.941 integrate_four_center_load 4 10.0 0.000 0.000 6.782 6.786 hfx_load_balance 1 11.0 0.000 0.000 6.782 6.786 mp2_ri_gpw_compute_en_ener 172 7.0 6.330 6.387 6.330 6.387 mp_sendrecv_dm3 2068 8.0 5.923 6.276 5.923 6.276 pw_gather_s 912 13.2 4.475 5.036 4.475 5.036 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.864213, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1502.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 451.944448E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. 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 0.051 0.140 55.506 55.510 qs_mol_dyn_low 1 2.0 0.016 0.040 54.785 54.946 qs_forces 11 3.9 0.003 0.004 54.567 54.569 qs_energies 11 4.9 0.003 0.012 53.090 53.104 scf_env_do_scf 11 5.9 0.000 0.001 47.116 47.117 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 45.113 45.113 qs_scf_new_mos 108 7.5 0.000 0.000 35.204 35.500 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.204 35.499 dbcsr_multiply_generic 2286 12.5 0.094 0.099 34.825 35.346 ot_scf_mini 108 9.5 0.002 0.002 33.461 33.641 multiply_cannon 2286 13.5 0.188 0.195 26.838 28.721 multiply_cannon_loop 2286 14.5 1.836 1.945 26.204 28.107 velocity_verlet 10 3.0 0.001 0.002 26.772 26.773 ot_mini 108 10.5 0.001 0.002 20.660 20.908 qs_ot_get_derivative 108 11.5 0.001 0.002 17.584 17.793 mp_waitall_1 245248 16.5 9.088 15.368 9.088 15.368 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.457 13.437 multiply_cannon_multrec 54864 15.5 3.621 5.740 7.557 10.915 qs_ot_get_p 119 10.4 0.001 0.001 8.055 8.339 mp_sum_l 7287 12.8 6.158 7.926 6.158 7.926 rebuild_ks_matrix 119 8.3 0.000 0.000 7.770 7.915 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 7.770 7.914 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.837 6.975 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.434 6.905 multiply_cannon_sync_h2d 54864 15.5 5.103 6.281 5.103 6.281 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.664 5.777 dbcsr_mm_accdrv_process 76910 16.1 1.829 2.870 3.850 5.478 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.210 5.264 init_scf_run 11 5.9 0.000 0.001 4.589 4.589 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.588 4.589 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.420 4.421 sum_up_and_integrate 119 10.3 0.012 0.017 4.388 4.395 integrate_v_rspace 119 11.3 0.002 0.003 4.376 4.384 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.232 4.232 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.056 4.214 calculate_rho_elec 119 8.7 0.011 0.016 4.055 4.214 cp_fm_redistribute_end 50 14.0 2.160 4.209 2.165 4.212 cp_fm_diag_elpa_base 50 14.0 2.041 4.104 2.045 4.113 apply_preconditioner_dbcsr 119 12.6 0.000 0.003 3.021 3.563 apply_single 119 13.6 0.000 0.000 3.021 3.563 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 2.020 3.258 calculate_dm_sparse 119 9.5 0.000 0.000 2.952 3.111 acc_transpose_blocks 54864 15.5 0.229 0.262 2.304 3.006 ot_diis_step 108 11.5 0.015 0.051 2.797 2.798 jit_kernel_multiply 13 15.8 1.958 2.608 1.958 2.608 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.533 2.592 calculate_first_density_matrix 1 7.0 0.000 0.002 2.425 2.429 density_rs2pw 119 9.7 0.004 0.004 2.085 2.211 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.005 2.195 2.196 grid_integrate_task_list 119 12.3 2.021 2.123 2.021 2.123 wfi_extrapolate 11 7.9 0.001 0.003 2.095 2.095 mp_sum_d 4135 12.0 1.402 2.048 1.402 2.048 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.919 1.990 init_scf_loop 11 6.9 0.000 0.001 1.988 1.988 potential_pw2rs 119 12.3 0.004 0.004 1.762 1.774 pw_transfer 1439 11.6 0.052 0.059 1.601 1.669 make_m2s 4572 13.5 0.054 0.056 1.576 1.621 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.526 1.597 make_images 4572 14.5 0.133 0.139 1.494 1.538 mp_alltoall_d11v 2130 13.8 1.282 1.519 1.282 1.519 transfer_rs2pw 487 10.6 0.005 0.006 1.391 1.506 mp_waitany 12084 13.8 1.312 1.473 1.312 1.473 acc_transpose_blocks_sync 164592 16.5 1.210 1.452 1.210 1.452 grid_collocate_task_list 119 9.7 1.345 1.405 1.345 1.405 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.378 1.402 acc_transpose_blocks_kernels 54864 16.5 0.247 0.378 0.819 1.331 fft3d_ps 1201 14.6 0.369 0.475 1.240 1.309 transfer_pw2rs 487 13.2 0.006 0.007 1.293 1.301 fft_wrap_pw1pw2_140 487 13.2 0.141 0.154 1.185 1.255 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.790 1.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.510000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 488.550400E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 922657. 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 0.028 0.051 40.140 40.147 qs_mol_dyn_low 1 2.0 0.003 0.003 39.771 39.809 qs_forces 11 3.9 0.003 0.004 39.446 39.447 qs_energies 11 4.9 0.002 0.002 37.774 37.780 scf_env_do_scf 11 5.9 0.000 0.001 32.648 32.649 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.160 30.160 dbcsr_multiply_generic 2286 12.5 0.100 0.102 22.927 23.281 qs_scf_new_mos 108 7.5 0.001 0.001 21.544 21.771 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.543 21.771 ot_scf_mini 108 9.5 0.002 0.003 20.628 20.793 multiply_cannon 2286 13.5 0.209 0.218 17.302 19.055 velocity_verlet 10 3.0 0.002 0.003 18.595 18.602 multiply_cannon_loop 2286 14.5 1.195 1.266 16.135 18.086 ot_mini 108 10.5 0.001 0.001 13.060 13.292 mp_waitall_1 200699 16.5 5.794 11.051 5.794 11.051 qs_ot_get_derivative 108 11.5 0.001 0.001 10.686 10.855 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.310 9.829 multiply_cannon_multrec 27432 15.5 1.803 3.993 6.187 9.056 rebuild_ks_matrix 119 8.3 0.000 0.000 6.856 6.995 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.856 6.995 dbcsr_mm_accdrv_process 47894 16.0 3.534 6.244 4.302 6.853 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.037 6.164 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.055 5.484 mp_sum_l 7287 12.8 2.614 4.958 2.614 4.958 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.589 4.953 apply_single 119 13.6 0.000 0.000 3.589 4.952 qs_ot_get_p 119 10.4 0.001 0.001 4.648 4.877 init_scf_run 11 5.9 0.000 0.001 3.932 3.932 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.931 3.932 sum_up_and_integrate 119 10.3 0.025 0.028 3.705 3.710 integrate_v_rspace 119 11.3 0.002 0.003 3.680 3.688 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.590 3.616 calculate_rho_elec 119 8.7 0.021 0.024 3.589 3.616 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.004 3.022 make_m2s 4572 13.5 0.052 0.054 2.577 2.898 make_images 4572 14.5 0.208 0.250 2.490 2.814 acc_transpose_blocks 27432 15.5 0.113 0.120 1.710 2.777 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.607 2.607 init_scf_loop 11 6.9 0.001 0.001 2.467 2.467 calculate_first_density_matrix 1 7.0 0.000 0.000 2.424 2.426 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.315 2.402 ot_diis_step 108 11.5 0.011 0.011 2.326 2.326 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.291 2.291 cp_fm_redistribute_end 50 14.0 1.162 2.266 1.165 2.269 multiply_cannon_sync_h2d 27432 15.5 1.708 2.246 1.708 2.246 cp_fm_diag_elpa_base 50 14.0 1.072 2.175 1.100 2.210 jit_kernel_multiply 9 16.4 0.710 2.201 0.710 2.201 calculate_dm_sparse 119 9.5 0.000 0.001 2.094 2.171 density_rs2pw 119 9.7 0.004 0.004 1.888 1.992 grid_integrate_task_list 119 12.3 1.833 1.938 1.833 1.938 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.932 1.934 pw_transfer 1439 11.6 0.065 0.068 1.832 1.861 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.741 1.772 acc_transpose_blocks_kernels 27432 16.5 0.188 0.274 0.755 1.722 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.673 1.714 make_images_data 4572 15.5 0.047 0.054 1.231 1.566 prepare_preconditioner 11 7.9 0.000 0.000 1.517 1.543 make_preconditioner 11 8.9 0.000 0.000 1.517 1.543 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.420 1.475 jit_kernel_transpose 5 15.5 0.567 1.454 0.567 1.454 potential_pw2rs 119 12.3 0.006 0.006 1.442 1.452 wfi_extrapolate 11 7.9 0.001 0.001 1.450 1.450 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.098 1.445 mp_alltoall_d11v 2130 13.8 1.214 1.410 1.214 1.410 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.372 1.405 fft3d_ps 1201 14.6 0.516 0.571 1.377 1.403 grid_collocate_task_list 119 9.7 1.289 1.351 1.289 1.351 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.288 1.297 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.203 1.252 transfer_rs2pw 487 10.6 0.005 0.005 1.068 1.140 mp_sum_d 4135 12.0 0.591 1.084 0.591 1.084 mp_allgather_i34 2286 14.5 0.591 0.987 0.591 0.987 acc_transpose_blocks_sync 82296 16.5 0.817 0.943 0.817 0.943 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.926 0.927 transfer_pw2rs 487 13.2 0.005 0.006 0.897 0.901 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.147000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 519.483392E+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 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.028 0.031 32.899 32.899 qs_mol_dyn_low 1 2.0 0.003 0.003 32.606 32.614 qs_forces 11 3.9 0.003 0.004 32.392 32.393 qs_energies 11 4.9 0.001 0.001 30.851 30.853 scf_env_do_scf 11 5.9 0.000 0.001 25.519 25.519 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.129 23.129 dbcsr_multiply_generic 2286 12.5 0.095 0.096 17.515 17.612 qs_scf_new_mos 108 7.5 0.001 0.001 15.439 15.458 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.438 15.458 velocity_verlet 10 3.0 0.002 0.003 15.272 15.278 multiply_cannon 2286 13.5 0.193 0.198 14.278 15.137 ot_scf_mini 108 9.5 0.002 0.002 14.681 14.698 multiply_cannon_loop 2286 14.5 0.862 0.914 13.504 14.393 ot_mini 108 10.5 0.001 0.001 9.049 9.068 multiply_cannon_multrec 18288 15.5 1.872 2.836 7.478 7.779 qs_ot_get_derivative 108 11.5 0.001 0.001 7.564 7.579 dbcsr_mm_accdrv_process 38222 16.0 5.455 6.400 5.513 6.465 rebuild_ks_matrix 119 8.3 0.000 0.000 5.958 5.973 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.957 5.973 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.247 5.262 init_scf_run 11 5.9 0.000 0.001 4.174 4.174 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.174 4.174 mp_waitall_1 158411 16.6 2.676 3.678 2.676 3.678 sum_up_and_integrate 119 10.3 0.030 0.031 3.559 3.564 integrate_v_rspace 119 11.3 0.002 0.003 3.529 3.537 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.844 3.448 qs_ot_get_p 119 10.4 0.001 0.001 3.400 3.426 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.286 3.291 calculate_rho_elec 119 8.7 0.031 0.031 3.285 3.290 calculate_first_density_matrix 1 7.0 0.000 0.000 2.951 2.953 calculate_dm_sparse 119 9.5 0.000 0.000 2.428 2.438 init_scf_loop 11 6.9 0.000 0.001 2.373 2.374 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.002 2.351 apply_single 119 13.6 0.000 0.000 2.002 2.350 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.274 2.282 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.476 2.249 density_rs2pw 119 9.7 0.004 0.004 1.854 1.980 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.974 1.975 make_m2s 4572 13.5 0.044 0.045 1.794 1.940 grid_integrate_task_list 119 12.3 1.804 1.905 1.804 1.905 make_images 4572 14.5 0.194 0.207 1.710 1.854 pw_transfer 1439 11.6 0.065 0.069 1.820 1.831 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.778 1.779 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.767 1.770 acc_transpose_blocks 18288 15.5 0.080 0.082 1.653 1.753 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.728 1.741 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.720 1.732 cp_fm_diag_elpa_base 50 14.0 1.696 1.709 1.718 1.731 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.718 1.727 prepare_preconditioner 11 7.9 0.000 0.000 1.625 1.627 make_preconditioner 11 8.9 0.000 0.000 1.625 1.627 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.591 1.596 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.486 1.566 ot_diis_step 108 11.5 0.011 0.011 1.466 1.467 mp_sum_l 7287 12.8 1.059 1.420 1.059 1.420 fft_wrap_pw1pw2_140 487 13.2 0.212 0.218 1.386 1.398 potential_pw2rs 119 12.3 0.007 0.008 1.342 1.346 grid_collocate_task_list 119 9.7 1.234 1.325 1.234 1.325 fft3d_ps 1201 14.6 0.526 0.542 1.296 1.308 multiply_cannon_sync_h2d 18288 15.5 1.048 1.200 1.048 1.200 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 transfer_rs2pw 487 10.6 0.005 0.005 1.008 1.141 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.943 0.961 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.936 0.937 make_images_data 4572 15.5 0.047 0.051 0.756 0.923 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.651 0.858 acc_transpose_blocks_sync 54864 16.5 0.740 0.836 0.740 0.836 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 0.815 0.824 transfer_pw2rs 487 13.2 0.004 0.004 0.818 0.821 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.463 0.797 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.781 mp_alltoall_d11v 2130 13.8 0.664 0.776 0.664 0.776 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.725 cp_fm_cholesky_invert 11 10.9 0.697 0.701 0.697 0.701 mp_alltoall_z22v 1201 16.6 0.618 0.681 0.618 0.681 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.899000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.000000, yerr=1.044466 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 558.137344E+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 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.015 0.033 36.678 36.679 qs_mol_dyn_low 1 2.0 0.003 0.003 36.500 36.509 qs_forces 11 3.9 0.002 0.003 36.442 36.443 qs_energies 11 4.9 0.002 0.002 34.735 34.740 scf_env_do_scf 11 5.9 0.000 0.001 29.204 29.206 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.867 25.867 dbcsr_multiply_generic 2286 12.5 0.098 0.101 20.290 20.415 velocity_verlet 10 3.0 0.002 0.002 18.283 18.285 qs_scf_new_mos 108 7.5 0.001 0.001 17.900 17.953 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.899 17.953 multiply_cannon 2286 13.5 0.217 0.226 16.601 16.962 ot_scf_mini 108 9.5 0.002 0.003 16.867 16.920 multiply_cannon_loop 2286 14.5 1.552 1.626 15.672 15.977 ot_mini 108 10.5 0.001 0.001 10.375 10.440 multiply_cannon_multrec 27432 15.5 2.464 3.156 9.151 9.581 qs_ot_get_derivative 108 11.5 0.001 0.001 8.516 8.570 dbcsr_mm_accdrv_process 47916 15.9 5.987 7.658 6.585 8.090 rebuild_ks_matrix 119 8.3 0.000 0.000 6.287 6.346 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.287 6.346 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.568 5.621 init_scf_run 11 5.9 0.000 0.001 4.197 4.198 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.197 4.197 qs_ot_get_p 119 10.4 0.001 0.001 3.644 3.721 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.116 3.521 sum_up_and_integrate 119 10.3 0.035 0.037 3.469 3.478 integrate_v_rspace 119 11.3 0.003 0.004 3.434 3.443 init_scf_loop 11 6.9 0.001 0.001 3.319 3.319 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.282 3.313 calculate_rho_elec 119 8.7 0.040 0.046 3.282 3.312 calculate_first_density_matrix 1 7.0 0.000 0.000 2.818 2.821 mp_waitall_1 137007 16.6 2.056 2.641 2.056 2.641 acc_transpose_blocks 27432 15.5 0.121 0.126 2.442 2.553 prepare_preconditioner 11 7.9 0.000 0.000 2.487 2.495 make_preconditioner 11 8.9 0.000 0.000 2.487 2.495 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.099 2.424 make_m2s 4572 13.5 0.054 0.056 2.280 2.385 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.346 2.347 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.318 2.346 make_images 4572 14.5 0.288 0.372 2.172 2.275 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.106 2.246 apply_single 119 13.6 0.000 0.000 2.106 2.246 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.234 2.245 calculate_dm_sparse 119 9.5 0.000 0.000 2.145 2.204 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.007 2.028 pw_transfer 1439 11.6 0.065 0.069 1.894 1.928 grid_integrate_task_list 119 12.3 1.836 1.915 1.836 1.915 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.802 1.839 density_rs2pw 119 9.7 0.004 0.004 1.748 1.822 ot_diis_step 108 11.5 0.012 0.012 1.819 1.819 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.792 1.792 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.971 1.722 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.554 1.563 cp_fm_diag_elpa_base 50 14.0 1.520 1.537 1.552 1.560 acc_transpose_blocks_sync 82296 16.5 1.453 1.555 1.453 1.555 fft_wrap_pw1pw2_140 487 13.2 0.246 0.259 1.491 1.530 fft3d_ps 1201 14.6 0.554 0.604 1.320 1.346 jit_kernel_multiply 6 16.4 0.532 1.346 0.532 1.346 grid_collocate_task_list 119 9.7 1.248 1.342 1.248 1.342 wfi_extrapolate 11 7.9 0.001 0.001 1.325 1.325 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.286 1.304 potential_pw2rs 119 12.3 0.008 0.009 1.252 1.256 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.238 1.247 mp_sum_l 7287 12.8 0.984 1.244 0.984 1.244 dbcsr_complete_redistribute 329 12.2 0.225 0.314 0.921 1.190 cp_fm_upper_to_full 72 14.2 0.806 1.151 0.806 1.151 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.074 1.075 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.720 0.984 make_images_data 4572 15.5 0.047 0.051 0.849 0.974 transfer_rs2pw 487 10.6 0.005 0.005 0.868 0.953 multiply_cannon_metrocomm1 27432 15.5 0.035 0.037 0.289 0.921 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.717 0.913 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.801 0.880 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.840 0.855 mp_alltoall_i22 627 13.8 0.431 0.827 0.431 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.812 mp_alltoall_d11v 2130 13.8 0.693 0.812 0.693 0.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.679000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.181818, yerr=4.130115 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 621.228032E+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 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.027 0.033 29.955 29.956 qs_mol_dyn_low 1 2.0 0.003 0.003 29.707 29.715 qs_forces 11 3.9 0.002 0.003 29.143 29.144 qs_energies 11 4.9 0.001 0.002 27.452 27.454 scf_env_do_scf 11 5.9 0.000 0.001 21.926 21.927 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.505 19.506 velocity_verlet 10 3.0 0.002 0.002 15.060 15.095 dbcsr_multiply_generic 2286 12.5 0.092 0.096 13.962 14.044 qs_scf_new_mos 108 7.5 0.001 0.001 11.947 11.972 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.947 11.972 multiply_cannon 2286 13.5 0.223 0.228 11.396 11.839 ot_scf_mini 108 9.5 0.002 0.002 11.238 11.263 multiply_cannon_loop 2286 14.5 0.645 0.668 10.479 10.680 multiply_cannon_multrec 9144 15.5 1.823 2.169 6.688 7.130 ot_mini 108 10.5 0.001 0.001 6.580 6.612 rebuild_ks_matrix 119 8.3 0.000 0.000 5.734 5.758 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.733 5.758 qs_ot_get_derivative 108 11.5 0.001 0.001 5.302 5.326 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.102 5.124 dbcsr_mm_accdrv_process 12550 15.8 3.930 4.809 4.755 4.955 init_scf_run 11 5.9 0.000 0.001 4.087 4.087 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.087 4.087 sum_up_and_integrate 119 10.3 0.037 0.041 3.423 3.428 integrate_v_rspace 119 11.3 0.003 0.003 3.386 3.390 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.296 3.308 calculate_rho_elec 119 8.7 0.060 0.061 3.296 3.308 calculate_first_density_matrix 1 7.0 0.000 0.000 2.924 2.926 qs_ot_get_p 119 10.4 0.001 0.001 2.782 2.819 init_scf_loop 11 6.9 0.001 0.001 2.402 2.403 calculate_dm_sparse 119 9.5 0.000 0.000 2.325 2.343 mp_waitall_1 115863 16.7 1.557 2.082 1.557 2.082 pw_transfer 1439 11.6 0.065 0.068 1.948 1.959 make_m2s 4572 13.5 0.034 0.035 1.789 1.957 grid_integrate_task_list 119 12.3 1.864 1.928 1.864 1.928 make_images 4572 14.5 0.273 0.305 1.701 1.867 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.855 1.867 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.804 1.807 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.772 1.790 density_rs2pw 119 9.7 0.003 0.004 1.690 1.778 jit_kernel_multiply 8 15.7 0.785 1.763 0.785 1.763 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.730 1.731 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.713 1.720 prepare_preconditioner 11 7.9 0.000 0.000 1.676 1.680 make_preconditioner 11 8.9 0.000 0.000 1.676 1.680 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.566 1.593 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.562 1.563 fft_wrap_pw1pw2_140 487 13.2 0.322 0.333 1.529 1.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.524 1.533 acc_transpose_blocks 9144 15.5 0.041 0.042 1.473 1.504 grid_collocate_task_list 119 9.7 1.297 1.373 1.297 1.373 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.342 1.356 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.293 1.301 cp_fm_diag_elpa_base 50 14.0 1.266 1.283 1.292 1.300 fft3d_ps 1201 14.6 0.558 0.569 1.269 1.276 ot_diis_step 108 11.5 0.013 0.013 1.266 1.266 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.205 1.242 apply_single 119 13.6 0.000 0.000 1.205 1.241 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.228 1.229 potential_pw2rs 119 12.3 0.010 0.011 1.208 1.211 wfi_extrapolate 11 7.9 0.001 0.001 1.112 1.113 hybrid_alltoall_any 4725 16.4 0.065 0.174 0.747 0.986 make_images_data 4572 15.5 0.042 0.045 0.774 0.961 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.437 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.920 cp_fm_cholesky_invert 11 10.9 0.846 0.849 0.846 0.849 transfer_rs2pw 487 10.6 0.004 0.004 0.773 0.848 mp_alltoall_d11v 2130 13.8 0.765 0.834 0.765 0.834 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.816 0.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.756 acc_transpose_blocks_sync 27432 16.5 0.722 0.752 0.722 0.752 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.675 0.723 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.694 0.696 mp_allgather_i34 2286 14.5 0.235 0.661 0.235 0.661 transfer_pw2rs 487 13.2 0.003 0.004 0.648 0.651 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.377 0.639 mp_alltoall_z22v 1201 16.6 0.584 0.613 0.584 0.613 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.956000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=589.818182, yerr=4.386305 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 777.605120E+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 63723. MP_Allreduce 10154 429. 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 0.019 0.039 104.659 104.660 qs_mol_dyn_low 1 2.0 0.003 0.003 104.429 104.456 velocity_verlet 10 3.0 0.002 0.002 78.114 85.895 mp_sum_dm 438 4.9 54.485 62.279 54.485 62.279 md_output 10 3.0 0.000 0.000 7.784 62.253 md_write_output 11 3.9 0.008 0.060 7.783 62.251 update_particle_set 20 4.0 0.000 0.000 54.466 62.247 write_trajectory 44 4.9 0.001 0.007 7.774 62.191 write_particle_coordinates 11 5.9 7.773 62.184 7.773 62.184 qs_forces 11 3.9 0.002 0.003 42.161 42.163 qs_energies 11 4.9 0.001 0.002 40.160 40.163 scf_env_do_scf 11 5.9 0.001 0.001 34.332 34.332 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.639 26.642 dbcsr_multiply_generic 2286 12.5 0.099 0.102 19.065 19.117 qs_scf_new_mos 108 7.5 0.001 0.001 17.325 17.427 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.324 17.427 ot_scf_mini 108 9.5 0.002 0.002 16.172 16.272 multiply_cannon 2286 13.5 0.296 0.299 15.251 15.880 multiply_cannon_loop 2286 14.5 0.871 0.894 14.014 14.674 ot_mini 108 10.5 0.001 0.001 9.975 10.091 multiply_cannon_multrec 9144 15.5 3.505 5.002 8.854 8.967 qs_ot_get_derivative 108 11.5 0.001 0.001 7.945 8.045 init_scf_loop 11 6.9 0.001 0.001 7.667 7.670 rebuild_ks_matrix 119 8.3 0.000 0.000 7.073 7.209 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.072 7.209 prepare_preconditioner 11 7.9 0.000 0.000 6.713 6.726 make_preconditioner 11 8.9 0.000 0.000 6.713 6.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.274 6.608 dbcsr_mm_accdrv_process 12550 15.8 4.660 6.523 5.213 6.556 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.370 6.495 cp_fm_upper_to_full 72 14.2 3.171 4.555 3.171 4.555 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.034 4.106 calculate_rho_elec 119 8.7 0.118 0.121 4.034 4.105 sum_up_and_integrate 119 10.3 0.064 0.066 3.793 3.799 init_scf_run 11 5.9 0.000 0.001 3.773 3.773 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.773 3.773 integrate_v_rspace 119 11.3 0.004 0.004 3.729 3.736 qs_ot_get_p 119 10.4 0.001 0.001 3.292 3.429 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.793 3.186 mp_waitall_1 94719 16.7 2.350 3.138 2.350 3.138 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.949 2.767 pw_transfer 1439 11.6 0.069 0.069 2.535 2.541 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.173 2.481 apply_single 119 13.6 0.000 0.000 2.173 2.481 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.658 2.465 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.437 2.443 make_m2s 4572 13.5 0.038 0.038 2.262 2.407 calculate_dm_sparse 119 9.5 0.000 0.000 2.308 2.378 calculate_first_density_matrix 1 7.0 0.000 0.000 2.272 2.342 mp_alltoall_i22 627 13.8 1.470 2.299 1.470 2.299 make_images 4572 14.5 0.361 0.396 2.143 2.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.434 2.233 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.499 2.192 grid_integrate_task_list 119 12.3 2.097 2.117 2.097 2.117 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=104.660000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.727273, yerr=17.705114 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 501.379072E+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 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.032 0.106 82.244 82.247 qs_mol_dyn_low 1 2.0 0.003 0.003 81.843 81.854 qs_forces 11 3.9 0.003 0.004 81.775 81.776 qs_energies 11 4.9 0.003 0.015 78.909 78.930 scf_env_do_scf 11 5.9 0.001 0.002 70.084 70.088 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.462 64.462 dbcsr_multiply_generic 2055 12.4 0.106 0.108 50.914 51.174 qs_scf_new_mos 99 7.5 0.000 0.001 47.344 47.461 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.343 47.460 ot_scf_mini 99 9.5 0.002 0.002 44.951 45.046 velocity_verlet 10 3.0 0.001 0.002 43.510 43.510 multiply_cannon 2055 13.4 0.184 0.187 42.528 43.198 multiply_cannon_loop 2055 14.4 1.803 1.843 41.562 42.222 ot_mini 99 10.5 0.001 0.001 26.594 26.678 qs_ot_get_derivative 99 11.5 0.001 0.001 19.744 19.823 multiply_cannon_multrec 49320 15.4 11.361 12.040 17.339 18.190 rebuild_ks_matrix 110 8.3 0.000 0.001 14.542 14.644 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 14.542 14.643 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.854 12.956 mp_waitall_1 220248 16.4 10.816 11.453 10.816 11.453 multiply_cannon_sync_h2d 49320 15.4 9.581 10.092 9.581 10.092 qs_ot_get_p 110 10.4 0.001 0.001 9.861 9.985 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.612 8.129 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.178 7.671 apply_single 110 13.6 0.000 0.000 7.178 7.670 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 6.562 7.396 sum_up_and_integrate 110 10.3 0.036 0.043 7.060 7.074 integrate_v_rspace 110 11.3 0.003 0.004 7.024 7.047 init_scf_run 11 5.9 0.000 0.001 6.759 6.760 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.759 6.759 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.608 6.623 ot_diis_step 99 11.5 0.006 0.006 6.557 6.557 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.375 6.524 calculate_rho_elec 110 8.6 0.021 0.024 6.375 6.524 dbcsr_mm_accdrv_process 87628 16.1 3.017 3.104 5.849 6.136 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.812 5.812 init_scf_loop 11 6.9 0.002 0.011 5.596 5.598 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.264 5.316 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.216 5.239 cp_fm_diag_elpa_base 48 14.0 5.203 5.226 5.215 5.237 mp_sum_l 6594 12.7 4.084 4.744 4.084 4.744 wfi_extrapolate 11 7.9 0.001 0.001 4.030 4.030 make_m2s 4110 13.4 0.061 0.066 3.863 3.971 make_images 4110 14.4 0.176 0.188 3.767 3.879 calculate_dm_sparse 110 9.5 0.001 0.001 3.771 3.859 density_rs2pw 110 9.6 0.004 0.005 3.387 3.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.539 3.544 grid_integrate_task_list 110 12.3 3.263 3.413 3.263 3.413 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.380 3.398 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.242 3.279 prepare_preconditioner 11 7.9 0.000 0.000 3.238 3.252 make_preconditioner 11 8.9 0.000 0.000 3.238 3.252 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.195 3.240 pw_transfer 1331 11.6 0.055 0.065 3.063 3.151 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.027 3.071 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.974 3.065 potential_pw2rs 110 12.3 0.006 0.007 2.638 2.670 fft_wrap_pw1pw2_140 451 13.1 0.388 0.430 2.538 2.635 calculate_first_density_matrix 1 7.0 0.000 0.003 2.630 2.633 acc_transpose_blocks 49320 15.4 0.213 0.219 2.572 2.629 jit_kernel_multiply 13 15.9 2.553 2.574 2.553 2.574 mp_alltoall_d11v 2046 13.8 2.063 2.526 2.063 2.526 fft3d_ps 1111 14.6 0.790 0.876 2.271 2.329 grid_collocate_task_list 110 9.6 2.162 2.276 2.162 2.276 mp_waitany 14300 13.8 1.845 2.159 1.845 2.159 transfer_rs2pw 451 10.6 0.005 0.006 1.939 2.118 mp_sum_d 3889 11.9 1.558 2.072 1.558 2.072 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.913 1.928 make_images_data 4110 15.4 0.043 0.047 1.739 1.861 transfer_pw2rs 451 13.1 0.006 0.007 1.812 1.831 cp_fm_cholesky_invert 11 10.9 1.783 1.786 1.783 1.786 hybrid_alltoall_any 4261 16.3 0.085 0.484 1.500 1.768 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.664 1.701 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.247000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.712079 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 593.002496E+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 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2769658. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.026 0.105 140.453 140.455 qs_mol_dyn_low 1 2.0 0.004 0.008 140.068 140.070 velocity_verlet 10 3.0 0.001 0.002 105.225 106.702 mp_sum_dm 438 4.9 69.461 70.957 69.461 70.957 md_write_output 11 3.9 1.477 70.884 1.479 70.903 update_particle_set 20 4.0 0.000 0.000 69.421 70.899 md_output 10 3.0 0.000 0.000 1.480 70.887 qs_forces 11 3.9 0.003 0.003 69.141 69.148 qs_energies 11 4.9 0.001 0.002 65.823 65.834 scf_env_do_scf 11 5.9 0.000 0.001 57.004 57.008 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 49.471 49.472 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.616 38.747 qs_scf_new_mos 99 7.5 0.001 0.001 33.704 33.824 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.704 33.823 multiply_cannon 2055 13.4 0.224 0.246 31.756 32.586 ot_scf_mini 99 9.5 0.003 0.003 32.037 32.156 multiply_cannon_loop 2055 14.4 1.169 1.202 30.618 31.683 ot_mini 99 10.5 0.001 0.001 18.953 19.063 multiply_cannon_multrec 24660 15.4 6.990 8.340 14.046 15.434 rebuild_ks_matrix 110 8.3 0.000 0.000 13.491 13.583 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.490 13.583 qs_ot_get_derivative 99 11.5 0.001 0.001 13.166 13.290 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.861 11.940 mp_waitall_1 176588 16.5 7.851 10.389 7.851 10.389 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.463 8.182 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.789 7.790 apply_single 110 13.6 0.000 0.001 6.789 7.789 init_scf_loop 11 6.9 0.000 0.001 7.499 7.499 multiply_cannon_sync_h2d 24660 15.4 6.443 7.272 6.443 7.272 dbcsr_mm_accdrv_process 52282 16.1 5.428 6.455 6.885 7.189 qs_ot_get_p 110 10.4 0.001 0.001 6.440 6.573 sum_up_and_integrate 110 10.3 0.052 0.059 6.351 6.366 init_scf_run 11 5.9 0.000 0.001 6.319 6.320 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.319 6.319 integrate_v_rspace 110 11.3 0.003 0.003 6.299 6.312 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.044 6.031 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.844 5.852 calculate_rho_elec 110 8.6 0.039 0.047 5.843 5.851 ot_diis_step 99 11.5 0.010 0.010 5.738 5.739 prepare_preconditioner 11 7.9 0.000 0.000 5.504 5.524 make_preconditioner 11 8.9 0.000 0.000 5.504 5.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.109 5.265 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.564 4.583 make_m2s 4110 13.4 0.057 0.060 4.208 4.533 make_images 4110 14.4 0.414 0.475 4.100 4.422 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.097 4.097 pw_transfer 1331 11.6 0.066 0.071 3.497 3.638 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.532 3.548 cp_fm_diag_elpa_base 48 14.0 3.486 3.503 3.530 3.545 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.391 3.533 wfi_extrapolate 11 7.9 0.001 0.001 3.503 3.504 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.397 3.400 grid_integrate_task_list 110 12.3 3.160 3.337 3.160 3.337 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.262 3.320 density_rs2pw 110 9.6 0.004 0.004 3.102 3.298 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.067 3.098 fft_wrap_pw1pw2_140 451 13.1 0.459 0.472 2.918 3.070 calculate_dm_sparse 110 9.5 0.001 0.001 2.978 3.010 mp_sum_l 6594 12.7 2.040 2.835 2.040 2.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=140.455000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.545455, yerr=6.358439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 666.341376E+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 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.027 0.049 61.442 61.445 qs_mol_dyn_low 1 2.0 0.003 0.003 61.100 61.111 qs_forces 11 3.9 0.009 0.013 61.035 61.037 qs_energies 11 4.9 0.002 0.005 57.853 57.862 scf_env_do_scf 11 5.9 0.001 0.001 49.151 49.151 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.660 40.661 velocity_verlet 10 3.0 0.005 0.009 32.473 32.474 dbcsr_multiply_generic 2055 12.4 0.106 0.111 29.809 30.050 qs_scf_new_mos 99 7.5 0.001 0.001 26.023 26.125 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.023 26.124 ot_scf_mini 99 9.5 0.002 0.003 24.778 24.898 multiply_cannon 2055 13.4 0.213 0.223 22.992 24.638 multiply_cannon_loop 2055 14.4 0.814 0.864 21.840 23.405 ot_mini 99 10.5 0.001 0.001 14.383 14.498 rebuild_ks_matrix 110 8.3 0.000 0.000 12.172 12.301 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.171 12.301 multiply_cannon_multrec 16440 15.4 3.707 4.805 9.960 10.938 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.714 10.833 qs_ot_get_derivative 99 11.5 0.001 0.001 9.975 10.094 mp_waitall_1 139946 16.5 7.012 9.836 7.012 9.836 init_scf_loop 11 6.9 0.001 0.002 8.456 8.457 multiply_cannon_metrocomm3 16440 15.4 0.045 0.048 4.382 7.160 prepare_preconditioner 11 7.9 0.000 0.000 6.682 6.697 make_preconditioner 11 8.9 0.000 0.001 6.682 6.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.020 6.354 sum_up_and_integrate 110 10.3 0.060 0.061 6.304 6.320 dbcsr_mm_accdrv_process 34862 16.1 5.379 5.879 6.097 6.294 integrate_v_rspace 110 11.3 0.003 0.003 6.243 6.259 init_scf_run 11 5.9 0.000 0.001 6.098 6.098 scf_env_initial_rho_setup 11 6.9 0.003 0.014 6.097 6.098 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.414 5.998 apply_single 110 13.6 0.000 0.000 5.413 5.998 qs_rho_update_rho_low 110 7.6 0.000 0.001 5.727 5.739 calculate_rho_elec 110 8.6 0.058 0.059 5.727 5.739 qs_ot_get_p 110 10.4 0.001 0.001 5.556 5.704 make_m2s 4110 13.4 0.049 0.051 4.221 4.583 make_images 4110 14.4 0.401 0.522 4.108 4.470 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.322 4.459 ot_diis_step 99 11.5 0.011 0.011 4.378 4.378 multiply_cannon_sync_h2d 16440 15.4 3.269 4.030 3.269 4.030 qs_ot_p2m_diag 48 11.0 0.042 0.045 3.921 3.924 pw_transfer 1331 11.6 0.065 0.073 3.557 3.563 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.553 3.554 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.450 3.460 grid_integrate_task_list 110 12.3 3.207 3.393 3.207 3.393 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.315 3.319 density_rs2pw 110 9.6 0.004 0.004 2.975 3.177 calculate_first_density_matrix 1 7.0 0.009 0.071 3.013 3.017 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.003 3.012 cp_fm_diag_elpa_base 48 14.0 2.937 2.970 3.001 3.010 wfi_extrapolate 11 7.9 0.001 0.001 2.991 2.991 fft_wrap_pw1pw2_140 451 13.1 0.578 0.584 2.975 2.989 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.868 2.922 make_images_data 4110 15.4 0.045 0.049 2.391 2.876 hybrid_alltoall_any 4261 16.3 0.109 0.382 2.162 2.863 mp_sum_l 6594 12.7 1.979 2.755 1.979 2.755 cp_fm_cholesky_invert 11 10.9 2.608 2.614 2.608 2.614 calculate_dm_sparse 110 9.5 0.001 0.001 2.519 2.551 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.428 2.494 acc_transpose_blocks 16440 15.4 0.078 0.079 1.928 2.478 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.882 2.436 grid_collocate_task_list 110 9.6 2.212 2.405 2.212 2.405 fft3d_ps 1111 14.6 1.093 1.104 2.391 2.400 mp_irecv_dv 48980 15.7 0.807 2.302 0.807 2.302 qs_energies_init_hamiltonians 11 5.9 0.010 0.046 2.167 2.174 potential_pw2rs 110 12.3 0.011 0.011 2.125 2.132 mp_alltoall_d11v 2046 13.8 1.810 2.109 1.810 2.109 dbcsr_complete_redistribute 325 12.2 0.348 0.423 1.516 1.971 cp_fm_upper_to_full 70 14.2 1.358 1.723 1.358 1.723 acc_transpose_blocks_kernels 16440 16.4 0.216 0.220 1.149 1.698 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.651 1.665 transfer_rs2pw 451 10.6 0.005 0.006 1.381 1.616 mp_allgather_i34 2055 14.4 0.465 1.602 0.465 1.602 cp_fm_cholesky_decompose 22 10.9 1.574 1.592 1.574 1.592 copy_fm_to_dbcsr 174 11.2 0.002 0.008 1.077 1.511 jit_kernel_transpose 5 15.6 0.934 1.482 0.934 1.482 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.459 1.471 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.468 mp_waitany 17072 13.8 1.215 1.468 1.215 1.468 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.265 1.276 qs_env_update_s_mstruct 11 6.9 0.012 0.024 1.174 1.247 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.445000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.818182, yerr=8.767898 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 729.169920E+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 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.019 0.048 67.833 67.834 qs_mol_dyn_low 1 2.0 0.003 0.003 67.549 67.559 qs_forces 11 3.9 0.003 0.004 67.141 67.141 qs_energies 11 4.9 0.002 0.010 63.734 63.741 scf_env_do_scf 11 5.9 0.001 0.002 54.939 54.942 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 43.216 43.217 velocity_verlet 10 3.0 0.003 0.003 37.668 37.684 dbcsr_multiply_generic 2055 12.4 0.113 0.124 32.070 32.358 qs_scf_new_mos 99 7.5 0.001 0.001 28.643 28.757 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.643 28.756 ot_scf_mini 99 9.5 0.003 0.004 26.953 27.048 multiply_cannon 2055 13.4 0.239 0.258 24.872 26.092 multiply_cannon_loop 2055 14.4 1.421 1.501 23.502 24.224 ot_mini 99 10.5 0.001 0.001 15.861 15.984 multiply_cannon_multrec 24660 15.4 4.177 7.111 13.246 14.910 rebuild_ks_matrix 110 8.3 0.000 0.000 11.971 12.086 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 11.971 12.086 qs_ot_get_derivative 99 11.5 0.001 0.001 11.686 11.785 init_scf_loop 11 6.9 0.001 0.004 11.682 11.683 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.577 10.679 dbcsr_mm_accdrv_process 52304 16.0 7.855 9.400 8.911 10.424 prepare_preconditioner 11 7.9 0.000 0.001 9.934 9.951 make_preconditioner 11 8.9 0.000 0.001 9.934 9.951 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.157 9.648 mp_waitall_1 121746 16.5 4.987 6.849 4.987 6.849 sum_up_and_integrate 110 10.3 0.067 0.070 6.259 6.272 integrate_v_rspace 110 11.3 0.003 0.003 6.192 6.205 qs_ot_get_p 110 10.4 0.001 0.001 5.963 6.101 init_scf_run 11 5.9 0.000 0.001 5.978 5.978 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.977 5.978 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.803 5.817 calculate_rho_elec 110 8.6 0.078 0.081 5.802 5.816 make_m2s 4110 13.4 0.060 0.061 5.508 5.794 make_images 4110 14.4 0.586 0.707 5.367 5.647 cp_fm_upper_to_full 70 14.2 3.336 4.868 3.336 4.868 ot_diis_step 99 11.5 0.011 0.011 4.138 4.139 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.058 4.111 apply_single 110 13.6 0.000 0.000 4.058 4.111 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.073 4.094 dbcsr_complete_redistribute 325 12.2 0.425 0.493 2.701 3.872 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.850 3.860 pw_transfer 1331 11.6 0.065 0.075 3.635 3.670 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.528 3.567 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.499 3.548 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.504 3.505 grid_integrate_task_list 110 12.3 3.297 3.425 3.297 3.425 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.353 3.410 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.141 3.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.250 3.251 fft_wrap_pw1pw2_140 451 13.1 0.608 0.630 3.027 3.069 density_rs2pw 110 9.6 0.004 0.004 2.928 3.059 calculate_dm_sparse 110 9.5 0.001 0.001 3.012 3.053 make_images_data 4110 15.4 0.048 0.053 2.725 3.042 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.284 3.024 wfi_extrapolate 11 7.9 0.001 0.001 3.000 3.001 acc_transpose_blocks 24660 15.4 0.114 0.117 2.813 2.989 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.955 2.968 cp_fm_diag_elpa_base 48 14.0 2.808 2.872 2.953 2.966 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.872 2.921 calculate_first_density_matrix 1 7.0 0.000 0.003 2.879 2.882 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.767 2.881 mp_alltoall_i22 605 13.7 1.684 2.857 1.684 2.857 cp_fm_cholesky_invert 11 10.9 2.694 2.702 2.694 2.702 multiply_cannon_sync_h2d 24660 15.4 2.381 2.517 2.381 2.517 grid_collocate_task_list 110 9.6 2.266 2.426 2.266 2.426 fft3d_ps 1111 14.6 1.086 1.113 2.411 2.426 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.279 2.280 potential_pw2rs 110 12.3 0.012 0.013 2.034 2.042 mp_alltoall_d11v 2046 13.8 1.722 1.873 1.722 1.873 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.795 1.831 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.722 cp_fm_cholesky_decompose 22 10.9 1.677 1.714 1.677 1.714 mp_sum_l 6594 12.7 1.011 1.661 1.011 1.661 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.652 mp_allgather_i34 2055 14.4 0.452 1.587 0.452 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.565 1.585 acc_transpose_blocks_sync 73980 16.4 1.414 1.566 1.414 1.566 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.840 1.505 multiply_cannon_metrocomm1 24660 15.4 0.035 0.036 0.527 1.475 jit_kernel_multiply 8 15.7 0.709 1.424 0.709 1.424 mp_irecv_dv 62702 16.1 0.735 1.421 0.735 1.421 transfer_rs2pw 451 10.6 0.005 0.006 1.223 1.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.834000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.727273, yerr=6.689081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 851.025920E+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 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.017 0.036 56.009 56.010 qs_mol_dyn_low 1 2.0 0.003 0.003 55.732 55.742 qs_forces 11 3.9 0.003 0.003 55.188 55.189 qs_energies 11 4.9 0.001 0.001 51.520 51.524 scf_env_do_scf 11 5.9 0.000 0.001 43.274 43.274 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.570 35.571 velocity_verlet 10 3.0 0.003 0.008 31.565 31.598 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.533 23.670 qs_scf_new_mos 99 7.5 0.001 0.001 21.016 21.059 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.016 21.059 ot_scf_mini 99 9.5 0.002 0.002 19.760 19.778 multiply_cannon 2055 13.4 0.237 0.245 17.827 19.098 multiply_cannon_loop 2055 14.4 0.606 0.628 16.524 16.771 rebuild_ks_matrix 110 8.3 0.000 0.000 11.676 11.688 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.675 11.687 ot_mini 99 10.5 0.001 0.001 10.948 10.959 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.374 10.382 multiply_cannon_multrec 8220 15.4 3.173 4.487 7.585 8.515 init_scf_loop 11 6.9 0.001 0.001 7.652 7.653 mp_waitall_1 103326 16.6 5.806 7.390 5.806 7.390 qs_ot_get_derivative 99 11.5 0.001 0.001 7.268 7.284 sum_up_and_integrate 110 10.3 0.080 0.082 6.274 6.289 integrate_v_rspace 110 11.3 0.003 0.004 6.194 6.208 prepare_preconditioner 11 7.9 0.000 0.000 5.999 6.002 make_preconditioner 11 8.9 0.000 0.000 5.999 6.002 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.952 5.968 calculate_rho_elec 110 8.6 0.114 0.114 5.951 5.967 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.573 5.651 dbcsr_mm_accdrv_process 17442 15.9 3.117 4.179 4.272 5.225 init_scf_run 11 5.9 0.000 0.001 5.098 5.098 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.098 5.098 qs_ot_get_p 110 10.4 0.001 0.001 4.831 4.859 make_m2s 4110 13.4 0.038 0.040 4.320 4.607 make_images 4110 14.4 0.654 0.716 4.191 4.476 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.941 4.331 pw_transfer 1331 11.6 0.066 0.072 3.861 3.886 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.784 3.878 apply_single 110 13.6 0.000 0.000 3.783 3.878 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.753 3.783 ot_diis_step 99 11.5 0.012 0.012 3.658 3.658 grid_integrate_task_list 110 12.3 3.391 3.481 3.391 3.481 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.420 3.424 fft_wrap_pw1pw2_140 451 13.1 0.773 0.787 3.239 3.280 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.076 3.077 density_rs2pw 110 9.6 0.004 0.004 2.940 3.024 cp_fm_cholesky_invert 11 10.9 2.870 2.874 2.870 2.874 hybrid_alltoall_any 4261 16.3 0.201 0.852 2.303 2.848 make_images_data 4110 15.4 0.040 0.046 2.311 2.727 wfi_extrapolate 11 7.9 0.001 0.001 2.695 2.695 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.683 2.684 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.551 2.552 calculate_dm_sparse 110 9.5 0.001 0.001 2.485 2.524 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.503 2.510 cp_fm_diag_elpa_base 48 14.0 2.446 2.469 2.501 2.508 grid_collocate_task_list 110 9.6 2.369 2.494 2.369 2.494 multiply_cannon_sync_h2d 8220 15.4 2.379 2.474 2.379 2.474 fft3d_ps 1111 14.6 1.136 1.182 2.422 2.437 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.297 2.379 calculate_first_density_matrix 1 7.0 0.000 0.000 2.302 2.304 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.191 2.204 potential_pw2rs 110 12.3 0.015 0.015 1.998 2.003 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.997 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.947 1.956 mp_alltoall_d11v 2046 13.8 1.710 1.798 1.710 1.798 acc_transpose_blocks 8220 15.4 0.039 0.040 1.621 1.756 cp_fm_cholesky_decompose 22 10.9 1.685 1.701 1.685 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.631 1.635 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.512 1.631 mp_allgather_i34 2055 14.4 0.452 1.575 0.452 1.575 dbcsr_complete_redistribute 325 12.2 0.562 0.580 1.443 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.445 jit_kernel_multiply 8 15.6 0.837 1.428 0.837 1.428 qs_create_task_list 11 7.9 0.000 0.001 1.231 1.333 generate_qs_task_list 11 8.9 0.389 0.452 1.231 1.333 transfer_rs2pw 451 10.6 0.005 0.005 1.141 1.239 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.193 1.214 mp_waitany 9240 13.8 1.053 1.144 1.053 1.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.010000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.181818, yerr=11.519297 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 1.426985E+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 67098. MP_Allreduce 9752 812. 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 0.029 0.043 86.661 86.662 qs_mol_dyn_low 1 2.0 0.003 0.003 86.330 86.342 qs_forces 11 3.9 0.003 0.003 86.246 86.247 qs_energies 11 4.9 0.001 0.001 82.106 82.107 scf_env_do_scf 11 5.9 0.001 0.001 72.128 72.128 velocity_verlet 10 3.0 0.002 0.002 55.546 55.554 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.810 43.812 dbcsr_multiply_generic 2055 12.4 0.120 0.125 29.284 29.381 init_scf_loop 11 6.9 0.001 0.001 28.247 28.249 qs_scf_new_mos 99 7.5 0.001 0.001 26.586 26.620 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.585 26.619 prepare_preconditioner 11 7.9 0.000 0.000 26.243 26.250 make_preconditioner 11 8.9 0.000 0.000 26.243 26.250 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.276 25.713 ot_scf_mini 99 9.5 0.002 0.002 24.787 24.811 multiply_cannon 2055 13.4 0.335 0.363 22.311 23.064 multiply_cannon_loop 2055 14.4 0.808 0.817 20.529 20.802 cp_fm_upper_to_full 70 14.2 12.746 18.424 12.746 18.424 ot_mini 99 10.5 0.001 0.001 13.976 14.003 rebuild_ks_matrix 110 8.3 0.000 0.001 13.628 13.661 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.627 13.660 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.363 12.393 dbcsr_complete_redistribute 325 12.2 1.022 1.048 7.516 10.848 multiply_cannon_multrec 8220 15.4 4.070 4.226 9.748 9.879 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.511 9.840 qs_ot_get_derivative 99 11.5 0.001 0.001 9.417 9.444 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.951 9.247 mp_alltoall_i22 605 13.7 5.569 8.923 5.569 8.923 mp_waitall_1 84994 16.7 7.347 8.148 7.347 8.148 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.025 7.060 calculate_rho_elec 110 8.6 0.225 0.225 7.024 7.059 sum_up_and_integrate 110 10.3 0.150 0.151 6.804 6.820 integrate_v_rspace 110 11.3 0.004 0.004 6.654 6.671 make_m2s 4110 13.4 0.044 0.044 5.296 5.793 dbcsr_mm_accdrv_process 11614 15.7 3.839 4.100 5.531 5.737 init_scf_run 11 5.9 0.000 0.001 5.681 5.681 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.681 5.681 make_images 4110 14.4 0.895 0.949 5.108 5.604 qs_ot_get_p 110 10.4 0.001 0.001 5.542 5.577 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.717 5.182 apply_single 110 13.6 0.000 0.000 4.717 5.181 cp_fm_cholesky_invert 11 10.9 5.077 5.081 5.077 5.081 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.762 5.073 pw_transfer 1331 11.6 0.075 0.075 4.880 4.888 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.763 4.772 ot_diis_step 99 11.5 0.015 0.015 4.543 4.543 fft_wrap_pw1pw2_140 451 13.1 1.280 1.286 4.165 4.177 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.011 4.019 grid_integrate_task_list 110 12.3 3.675 3.751 3.675 3.751 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.690 3.690 density_rs2pw 110 9.6 0.004 0.004 3.488 3.506 hybrid_alltoall_any 4261 16.3 0.262 0.562 2.774 3.503 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.498 3.498 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.978 3.428 make_images_data 4110 15.4 0.045 0.048 2.760 3.371 wfi_extrapolate 11 7.9 0.001 0.001 3.300 3.300 calculate_dm_sparse 110 9.5 0.001 0.001 3.176 3.200 multiply_cannon_sync_h2d 8220 15.4 3.154 3.172 3.154 3.172 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.917 2.917 cp_fm_diag_elpa_base 48 14.0 2.382 2.580 2.916 2.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.881 2.883 fft3d_ps 1111 14.6 1.299 1.310 2.814 2.827 grid_collocate_task_list 110 9.6 2.666 2.679 2.666 2.679 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.623 2.640 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.374 2.389 calculate_first_density_matrix 1 7.0 0.000 0.000 2.268 2.270 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.206 2.263 potential_pw2rs 110 12.3 0.021 0.021 2.211 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.104 2.200 mp_alltoall_d11v 2046 13.8 2.019 2.095 2.019 2.095 cp_fm_cholesky_decompose 22 10.9 1.953 1.977 1.953 1.977 qs_create_task_list 11 7.9 0.010 0.011 1.904 1.947 generate_qs_task_list 11 8.9 0.749 0.798 1.894 1.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.862 1.865 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.721 1.758 jit_kernel_multiply 10 15.2 1.494 1.740 1.494 1.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.662000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1279.909091, yerr=66.267302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 630.571008E+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 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1595625. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 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 0.046 0.075 208.924 208.926 qs_mol_dyn_low 1 2.0 0.003 0.003 208.389 208.402 qs_forces 11 3.9 0.005 0.005 208.304 208.305 qs_energies 11 4.9 0.001 0.002 202.690 202.707 scf_env_do_scf 11 5.9 0.001 0.001 185.623 185.627 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.910 164.912 dbcsr_multiply_generic 2507 12.6 0.180 0.184 126.322 127.263 qs_scf_new_mos 117 7.6 0.001 0.001 125.384 125.665 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.383 125.664 velocity_verlet 10 3.0 0.001 0.002 124.717 124.718 ot_scf_mini 117 9.6 0.003 0.003 118.747 118.989 multiply_cannon 2507 13.6 0.239 0.247 102.201 104.163 multiply_cannon_loop 2507 14.6 2.414 2.464 100.022 101.917 ot_mini 117 10.6 0.001 0.001 66.301 66.549 multiply_cannon_multrec 60168 15.6 31.951 34.106 42.068 43.929 qs_ot_get_derivative 117 11.6 0.001 0.001 41.349 41.595 rebuild_ks_matrix 128 8.3 0.001 0.001 33.683 33.993 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.683 33.992 mp_waitall_1 267128 16.5 28.917 31.737 28.917 31.737 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.256 30.537 qs_ot_get_p 128 10.4 0.001 0.001 30.161 30.366 multiply_cannon_sync_h2d 60168 15.6 26.438 28.341 26.438 28.341 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.422 25.534 apply_single 128 13.6 0.001 0.001 24.422 25.534 ot_diis_step 117 11.6 0.008 0.008 24.723 24.724 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.548 23.606 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.682 20.683 init_scf_loop 11 6.9 0.000 0.001 20.637 20.639 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.825 20.023 multiply_cannon_metrocomm3 60168 15.6 0.120 0.126 16.235 18.261 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.670 17.702 cp_fm_diag_elpa_base 83 14.4 17.602 17.642 17.666 17.700 prepare_preconditioner 11 7.9 0.000 0.000 16.018 16.061 make_preconditioner 11 8.9 0.000 0.000 16.018 16.061 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.260 15.428 make_m2s 5014 13.6 0.105 0.115 13.864 14.249 make_images 5014 14.6 0.397 0.417 13.681 14.078 sum_up_and_integrate 128 10.3 0.089 0.107 14.057 14.076 integrate_v_rspace 128 11.3 0.004 0.005 13.967 13.991 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.220 13.343 calculate_rho_elec 128 8.7 0.045 0.064 13.220 13.342 init_scf_run 11 5.9 0.000 0.001 12.918 12.919 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.918 12.918 dbcsr_mm_accdrv_process 124484 16.2 4.683 4.787 9.684 10.272 mp_sum_l 7950 12.9 9.022 10.035 9.022 10.035 cp_fm_cholesky_invert 11 10.9 9.184 9.192 9.184 9.192 wfi_extrapolate 11 7.9 0.001 0.001 9.172 9.172 calculate_dm_sparse 128 9.5 0.001 0.001 8.958 9.058 multiply_cannon_metrocomm1 60168 15.6 0.096 0.101 6.153 8.702 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.095 8.219 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.126 8.214 pw_transfer 1547 11.6 0.075 0.087 7.681 7.872 make_images_data 5014 15.6 0.068 0.073 6.807 7.698 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.478 7.666 grid_integrate_task_list 128 12.3 7.047 7.509 7.047 7.509 density_rs2pw 128 9.7 0.006 0.007 6.873 7.392 hybrid_alltoall_any 5200 16.5 0.297 2.279 5.952 7.141 fft_wrap_pw1pw2_140 523 13.2 1.133 1.180 6.526 6.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.687 6.698 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.869 fft3d_ps 1291 14.7 2.192 2.836 5.477 5.815 mp_alltoall_d11v 2415 14.1 4.292 5.560 4.292 5.560 grid_collocate_task_list 128 9.7 4.847 5.272 4.847 5.272 cp_fm_cholesky_decompose 22 10.9 4.663 4.677 4.663 4.677 potential_pw2rs 128 12.3 0.009 0.010 4.541 4.565 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.926000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.909091, yerr=6.316710 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 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 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 845.172736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 5056073. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 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 0.017 0.033 194.234 194.236 qs_mol_dyn_low 1 2.0 0.003 0.003 193.839 193.854 qs_forces 11 3.9 0.004 0.005 193.424 193.425 qs_energies 11 4.9 0.002 0.007 186.636 186.649 scf_env_do_scf 11 5.9 0.001 0.001 169.552 169.562 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.067 136.069 velocity_verlet 10 3.0 0.001 0.002 122.236 122.245 dbcsr_multiply_generic 2527 12.6 0.190 0.196 98.786 100.000 qs_scf_new_mos 118 7.6 0.001 0.001 96.841 97.384 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.840 97.383 ot_scf_mini 118 9.6 0.004 0.005 91.970 92.594 multiply_cannon 2527 13.6 0.505 0.559 78.159 82.927 multiply_cannon_loop 2527 14.6 1.594 1.673 75.079 77.896 ot_mini 118 10.6 0.001 0.001 50.819 51.364 mp_waitall_1 216438 16.6 24.963 38.931 24.963 38.931 multiply_cannon_multrec 30324 15.6 21.181 26.099 31.976 37.644 init_scf_loop 11 6.9 0.001 0.006 33.395 33.398 rebuild_ks_matrix 129 8.3 0.001 0.001 32.751 33.349 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.751 33.348 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.487 30.037 multiply_cannon_metrocomm3 30324 15.6 0.097 0.103 16.084 29.315 qs_ot_get_derivative 118 11.6 0.001 0.002 28.545 29.157 prepare_preconditioner 11 7.9 0.000 0.001 29.014 29.075 make_preconditioner 11 8.9 0.000 0.002 29.013 29.075 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.673 28.223 qs_ot_get_p 129 10.4 0.001 0.002 23.053 23.593 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.338 23.496 apply_single 129 13.6 0.001 0.001 22.338 23.496 ot_diis_step 118 11.6 0.015 0.015 22.094 22.096 multiply_cannon_sync_h2d 30324 15.6 18.167 20.323 18.167 20.323 qs_ot_p2m_diag 83 11.4 0.189 0.217 18.033 18.075 cp_fm_cholesky_invert 11 10.9 16.894 16.906 16.894 16.906 cp_dbcsr_syevd 83 12.4 0.006 0.006 16.802 16.803 make_m2s 5054 13.6 0.091 0.097 14.549 15.875 make_images 5054 14.6 1.198 1.394 14.338 15.663 sum_up_and_integrate 129 10.3 0.117 0.133 14.345 14.374 integrate_v_rspace 129 11.3 0.004 0.004 14.228 14.264 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.664 13.695 calculate_rho_elec 129 8.7 0.088 0.104 13.663 13.695 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.555 13.592 cp_fm_diag_elpa_base 83 14.4 13.298 13.396 13.549 13.580 init_scf_run 11 5.9 0.000 0.001 12.246 12.247 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.245 12.247 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.292 11.736 dbcsr_mm_accdrv_process 62734 16.2 5.409 6.141 10.247 11.108 multiply_cannon_metrocomm4 27797 15.6 0.106 0.121 3.869 10.857 mp_irecv_dv 70031 16.3 3.664 10.453 3.664 10.453 make_images_data 5054 15.6 0.068 0.076 8.611 10.215 hybrid_alltoall_any 5240 16.5 0.354 1.551 7.270 9.636 pw_transfer 1559 11.6 0.085 0.097 8.792 8.848 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.567 8.625 wfi_extrapolate 11 7.9 0.001 0.001 8.418 8.418 density_rs2pw 129 9.7 0.006 0.007 7.293 7.750 grid_integrate_task_list 129 12.3 7.242 7.630 7.242 7.630 fft_wrap_pw1pw2_140 527 13.2 1.234 1.270 7.525 7.601 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.411 7.183 cp_fm_cholesky_decompose 22 10.9 7.017 7.100 7.017 7.100 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.768 6.777 calculate_dm_sparse 129 9.5 0.001 0.001 6.543 6.694 fft3d_ps 1301 14.7 2.857 3.018 6.094 6.131 mp_sum_l 8010 12.9 4.266 6.100 4.266 6.100 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.851 5.995 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.461 5.551 grid_collocate_task_list 129 9.7 5.043 5.389 5.043 5.389 mp_alltoall_d11v 2423 14.1 4.294 5.060 4.294 5.060 mp_allgather_i34 2527 14.6 1.594 4.760 1.594 4.760 potential_pw2rs 129 12.3 0.015 0.017 4.696 4.716 dbcsr_complete_redistribute 395 12.7 0.781 0.901 3.424 4.277 mp_sum_d 4499 12.1 2.701 4.025 2.701 4.025 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.236000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.545455, yerr=6.035706 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/17/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 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 948.224000E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+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 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.074 177.084 177.087 qs_mol_dyn_low 1 2.0 0.003 0.003 176.295 176.569 qs_forces 11 3.9 0.006 0.013 176.117 176.121 qs_energies 11 4.9 0.004 0.013 169.574 169.584 scf_env_do_scf 11 5.9 0.001 0.002 153.970 153.972 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 118.919 118.920 velocity_verlet 10 3.0 0.001 0.002 112.914 112.919 dbcsr_multiply_generic 2507 12.6 0.179 0.184 81.952 83.122 qs_scf_new_mos 117 7.6 0.001 0.001 81.757 82.227 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.756 82.226 ot_scf_mini 117 9.6 0.004 0.005 77.528 78.039 multiply_cannon 2507 13.6 0.504 0.531 61.970 66.303 multiply_cannon_loop 2507 14.6 1.132 1.203 59.209 62.080 ot_mini 117 10.6 0.001 0.001 42.740 43.249 mp_waitall_1 170520 16.6 25.087 35.177 25.087 35.177 init_scf_loop 11 6.9 0.006 0.023 34.952 34.954 prepare_preconditioner 11 7.9 0.000 0.000 30.927 30.981 make_preconditioner 11 8.9 0.000 0.001 30.927 30.981 rebuild_ks_matrix 128 8.3 0.001 0.001 30.265 30.786 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 30.265 30.786 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.620 29.965 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.287 27.772 multiply_cannon_multrec 20056 15.6 12.989 16.804 22.311 26.133 multiply_cannon_metrocomm3 20056 15.6 0.062 0.066 15.596 25.138 qs_ot_get_derivative 117 11.6 0.001 0.002 22.995 23.504 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.739 20.766 apply_single 128 13.6 0.001 0.001 19.738 20.765 qs_ot_get_p 128 10.4 0.001 0.001 20.115 20.608 ot_diis_step 117 11.6 0.018 0.018 19.633 19.634 qs_ot_p2m_diag 83 11.4 0.266 0.274 15.817 15.826 make_m2s 5014 13.6 0.080 0.085 14.926 15.723 multiply_cannon_sync_h2d 20056 15.6 13.459 15.566 13.459 15.566 make_images 5014 14.6 1.158 1.252 14.694 15.489 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.736 14.737 cp_fm_cholesky_invert 11 10.9 14.595 14.604 14.595 14.604 sum_up_and_integrate 128 10.3 0.132 0.146 14.097 14.120 integrate_v_rspace 128 11.3 0.004 0.004 13.965 13.989 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.605 13.633 calculate_rho_elec 128 8.7 0.130 0.145 13.605 13.632 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.443 11.464 cp_fm_diag_elpa_base 83 14.4 11.034 11.185 11.440 11.460 init_scf_run 11 5.9 0.000 0.001 10.543 10.544 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.543 10.544 make_images_data 5014 15.6 0.063 0.072 8.944 10.180 hybrid_alltoall_any 5200 16.5 0.451 2.043 7.623 9.679 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.105 9.473 multiply_cannon_metrocomm4 17549 15.6 0.067 0.078 3.533 9.454 mp_irecv_dv 50230 16.2 3.405 9.204 3.405 9.204 dbcsr_mm_accdrv_process 41502 16.2 5.593 5.910 8.786 8.926 pw_transfer 1547 11.6 0.085 0.104 8.743 8.843 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.519 8.623 grid_integrate_task_list 128 12.3 7.314 7.791 7.314 7.791 fft_wrap_pw1pw2_140 523 13.2 1.313 1.351 7.539 7.653 wfi_extrapolate 11 7.9 0.001 0.001 7.477 7.478 cp_fm_cholesky_decompose 22 10.9 7.340 7.371 7.340 7.371 density_rs2pw 128 9.7 0.006 0.006 6.998 7.369 cp_fm_upper_to_full 105 14.8 5.612 7.044 5.612 7.044 dbcsr_complete_redistribute 395 12.7 1.169 1.198 4.591 6.295 fft3d_ps 1291 14.7 2.754 2.979 5.884 5.954 calculate_dm_sparse 128 9.5 0.001 0.001 5.827 5.894 grid_collocate_task_list 128 9.7 5.155 5.542 5.155 5.542 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.462 5.466 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.706 5.355 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.395 5.094 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.551 4.704 mp_alltoall_d11v 2415 14.1 4.334 4.695 4.334 4.695 mp_sum_l 7950 12.9 3.241 4.563 3.241 4.563 mp_allgather_i34 2507 14.6 1.260 4.562 1.260 4.562 potential_pw2rs 128 12.3 0.020 0.022 4.456 4.473 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.013 4.040 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.287 3.939 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.765 3.767 mp_alltoall_i22 716 14.1 1.883 3.695 1.883 3.695 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.087000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.363636, yerr=8.336914 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 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 11528903135232 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 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 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 1.151242E+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 1.770320E+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 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57949. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 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 0.052 190.256 190.256 qs_mol_dyn_low 1 2.0 0.003 0.004 189.647 189.660 qs_forces 11 3.9 0.004 0.005 188.467 188.475 qs_energies 11 4.9 0.003 0.007 181.334 181.345 scf_env_do_scf 11 5.9 0.001 0.002 163.207 163.219 velocity_verlet 10 3.0 0.014 0.028 125.986 126.033 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 116.999 117.000 dbcsr_multiply_generic 2507 12.6 0.187 0.198 81.926 82.740 qs_scf_new_mos 117 7.6 0.001 0.001 80.430 80.763 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.429 80.762 ot_scf_mini 117 9.6 0.003 0.004 75.869 76.227 multiply_cannon 2507 13.6 0.552 0.591 57.134 60.537 multiply_cannon_loop 2507 14.6 1.847 1.941 53.305 55.114 init_scf_loop 11 6.9 0.001 0.002 46.086 46.087 ot_mini 117 10.6 0.001 0.001 43.099 43.455 prepare_preconditioner 11 7.9 0.000 0.000 41.992 42.038 make_preconditioner 11 8.9 0.000 0.002 41.992 42.038 make_full_inverse_cholesky 11 9.9 0.011 0.021 35.607 40.660 multiply_cannon_multrec 30084 15.6 13.537 19.242 27.409 32.699 rebuild_ks_matrix 128 8.3 0.001 0.001 29.374 29.694 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.373 29.694 mp_waitall_1 147882 16.7 17.403 26.817 17.403 26.817 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.482 26.761 qs_ot_get_derivative 117 11.6 0.002 0.002 23.250 23.612 make_m2s 5014 13.6 0.097 0.102 20.430 21.752 make_images 5014 14.6 1.964 2.323 20.121 21.441 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.283 19.935 apply_single 128 13.6 0.001 0.001 19.283 19.934 ot_diis_step 117 11.6 0.018 0.018 19.735 19.737 qs_ot_get_p 128 10.4 0.001 0.001 18.852 19.255 cp_fm_upper_to_full 105 14.8 11.392 16.825 11.392 16.825 cp_fm_cholesky_invert 11 10.9 16.074 16.083 16.074 16.083 multiply_cannon_metrocomm3 30084 15.6 0.049 0.053 6.555 15.534 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.711 14.762 sum_up_and_integrate 128 10.3 0.140 0.153 14.087 14.116 dbcsr_mm_accdrv_process 62264 16.2 8.445 9.330 13.439 13.992 integrate_v_rspace 128 11.3 0.004 0.004 13.947 13.980 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.692 13.733 calculate_rho_elec 128 8.7 0.173 0.188 13.691 13.733 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.301 13.302 make_images_data 5014 15.6 0.076 0.153 10.956 13.059 dbcsr_complete_redistribute 395 12.7 1.499 1.633 8.982 12.844 hybrid_alltoall_any 5200 16.5 0.520 2.080 9.829 12.285 init_scf_run 11 5.9 0.000 0.001 11.959 11.960 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.958 11.960 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.608 11.396 multiply_cannon_sync_h2d 30084 15.6 10.487 11.280 10.487 11.280 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.237 10.250 cp_fm_diag_elpa_base 83 14.4 9.278 9.589 10.231 10.242 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.771 10.034 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.363 10.024 mp_alltoall_i22 716 14.1 5.659 9.393 5.659 9.393 pw_transfer 1547 11.6 0.086 0.102 8.920 9.002 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.694 8.785 fft_wrap_pw1pw2_140 523 13.2 1.453 1.490 7.749 7.841 grid_integrate_task_list 128 12.3 7.559 7.835 7.559 7.835 wfi_extrapolate 11 7.9 0.001 0.001 7.652 7.652 cp_fm_cholesky_decompose 22 10.9 7.392 7.504 7.392 7.504 multiply_cannon_metrocomm4 25070 15.6 0.085 0.098 2.817 7.197 density_rs2pw 128 9.7 0.006 0.006 6.717 7.108 mp_irecv_dv 76098 16.2 2.661 6.914 2.661 6.914 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.531 6.611 calculate_dm_sparse 128 9.5 0.001 0.001 6.387 6.480 fft3d_ps 1291 14.7 2.847 2.943 5.860 5.947 grid_collocate_task_list 128 9.7 5.277 5.703 5.277 5.703 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.453 5.566 mp_alltoall_d11v 2415 14.1 4.782 5.315 4.782 5.315 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.628 4.628 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.465 4.571 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.252 4.294 potential_pw2rs 128 12.3 0.023 0.024 4.269 4.281 mp_allgather_i34 2507 14.6 1.910 4.199 1.910 4.199 calculate_first_density_matrix 1 7.0 0.000 0.002 4.087 4.090 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.256000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1088.727273, yerr=16.954972 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 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 11528903135232 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 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.525473E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.038 170.358 170.360 qs_mol_dyn_low 1 2.0 0.003 0.003 169.925 169.939 qs_forces 11 3.9 0.004 0.005 169.812 169.815 qs_energies 11 4.9 0.002 0.002 162.404 162.411 scf_env_do_scf 11 5.9 0.001 0.001 145.039 145.050 velocity_verlet 10 3.0 0.005 0.005 111.843 111.848 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 109.394 109.395 qs_scf_new_mos 117 7.6 0.001 0.001 72.556 72.714 qs_scf_loop_do_ot 117 8.6 0.001 0.001 72.555 72.713 dbcsr_multiply_generic 2507 12.6 0.181 0.192 72.139 72.436 ot_scf_mini 117 9.6 0.003 0.004 68.088 68.196 multiply_cannon 2507 13.6 0.561 0.593 53.367 56.985 multiply_cannon_loop 2507 14.6 0.811 0.846 50.361 51.037 ot_mini 117 10.6 0.001 0.001 37.452 37.555 init_scf_loop 11 6.9 0.001 0.001 35.493 35.494 prepare_preconditioner 11 7.9 0.000 0.000 31.599 31.628 make_preconditioner 11 8.9 0.000 0.000 31.599 31.628 mp_waitall_1 125778 16.7 24.661 30.883 24.661 30.883 make_full_inverse_cholesky 11 9.9 0.013 0.027 29.473 29.751 rebuild_ks_matrix 128 8.3 0.001 0.001 29.171 29.298 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.170 29.298 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.601 26.726 multiply_cannon_multrec 10028 15.6 10.405 14.517 18.015 20.932 qs_ot_get_derivative 117 11.6 0.001 0.002 20.534 20.641 multiply_cannon_metrocomm3 10028 15.6 0.026 0.027 12.418 19.613 qs_ot_get_p 128 10.4 0.001 0.001 17.887 17.979 cp_fm_cholesky_invert 11 10.9 17.939 17.946 17.939 17.946 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.933 17.155 apply_single 128 13.6 0.001 0.001 16.933 17.155 ot_diis_step 117 11.6 0.020 0.020 16.848 16.848 make_m2s 5014 13.6 0.066 0.072 14.870 15.738 make_images 5014 14.6 2.184 2.618 14.565 15.437 sum_up_and_integrate 128 10.3 0.182 0.194 14.220 14.273 qs_ot_p2m_diag 83 11.4 0.496 0.502 14.130 14.149 integrate_v_rspace 128 11.3 0.004 0.004 14.038 14.101 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.011 14.049 calculate_rho_elec 128 8.7 0.256 0.267 14.011 14.048 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.934 12.935 multiply_cannon_sync_h2d 10028 15.6 10.893 11.122 10.893 11.122 init_scf_run 11 5.9 0.000 0.001 10.486 10.487 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.486 10.487 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.013 10.023 cp_fm_diag_elpa_base 83 14.4 9.775 9.851 10.009 10.019 make_images_data 5014 15.6 0.056 0.064 8.562 9.903 hybrid_alltoall_any 5200 16.5 0.844 3.785 8.391 9.845 pw_transfer 1547 11.6 0.085 0.093 9.317 9.342 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.096 9.128 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.071 8.154 fft_wrap_pw1pw2_140 523 13.2 1.817 1.854 8.041 8.079 cp_fm_cholesky_decompose 22 10.9 7.967 8.075 7.967 8.075 grid_integrate_task_list 128 12.3 7.764 7.987 7.764 7.987 dbcsr_mm_accdrv_process 20762 16.1 3.148 4.420 7.258 7.932 multiply_cannon_metrocomm1 10028 15.6 0.030 0.030 4.429 7.524 wfi_extrapolate 11 7.9 0.001 0.001 7.362 7.362 density_rs2pw 128 9.7 0.005 0.006 6.742 7.023 calculate_dm_sparse 128 9.5 0.001 0.001 6.160 6.220 fft3d_ps 1291 14.7 2.762 2.830 5.807 5.857 grid_collocate_task_list 128 9.7 5.569 5.784 5.569 5.784 dbcsr_complete_redistribute 395 12.7 2.115 2.160 5.291 5.680 qs_energies_init_hamiltonians 11 5.9 0.010 0.015 5.319 5.320 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.252 5.261 mp_alltoall_d11v 2415 14.1 4.625 4.886 4.625 4.886 mp_allgather_i34 2507 14.6 1.188 4.615 1.188 4.615 potential_pw2rs 128 12.3 0.026 0.026 4.251 4.264 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.150 4.202 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.586 3.866 multiply_cannon_metrocomm4 7521 15.6 0.027 0.029 1.732 3.749 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.408 3.739 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.644 3.727 mp_irecv_dv 28860 15.9 1.692 3.672 1.692 3.672 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.595 3.627 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.482 3.556 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.475 3.481 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.360000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1438.818182, yerr=21.565200 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 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 11444702699520 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 92.796573E+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 6705499744 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 3.080552E+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 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59121. MP_Allreduce 11085 1504. 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 0.027 0.045 289.849 289.851 qs_mol_dyn_low 1 2.0 0.003 0.003 289.283 289.329 qs_forces 11 3.9 0.005 0.005 289.184 289.187 qs_energies 11 4.9 0.002 0.003 280.386 280.398 scf_env_do_scf 11 5.9 0.001 0.001 257.532 257.542 velocity_verlet 10 3.0 0.005 0.006 209.377 209.385 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 133.061 133.063 init_scf_loop 11 6.9 0.023 0.053 124.197 124.200 prepare_preconditioner 11 7.9 0.000 0.000 119.336 119.376 make_preconditioner 11 8.9 0.000 0.000 119.336 119.376 make_full_inverse_cholesky 11 9.9 0.035 0.039 95.204 116.530 qs_scf_new_mos 116 7.6 0.001 0.001 89.823 89.900 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.822 89.900 ot_scf_mini 116 9.6 0.004 0.004 84.958 84.994 dbcsr_multiply_generic 2485 12.5 0.210 0.222 81.524 81.919 cp_fm_upper_to_full 104 14.8 52.146 75.240 52.146 75.240 multiply_cannon 2485 13.5 0.680 0.729 58.922 59.987 multiply_cannon_loop 2485 14.5 1.042 1.053 54.950 56.067 ot_mini 116 10.6 0.001 0.001 44.032 44.067 dbcsr_complete_redistribute 393 12.7 4.014 4.062 30.157 43.499 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.603 39.856 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.089 37.127 mp_alltoall_i22 712 14.1 21.887 35.309 21.887 35.309 rebuild_ks_matrix 127 8.3 0.001 0.001 33.863 33.906 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.862 33.905 cp_fm_cholesky_invert 11 10.9 33.675 33.681 33.675 33.681 mp_waitall_1 102768 16.8 27.854 32.196 27.854 32.196 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.393 31.434 qs_ot_get_p 127 10.4 0.001 0.001 26.010 26.063 qs_ot_get_derivative 116 11.6 0.002 0.002 24.547 24.583 qs_ot_p2m_diag 82 11.4 0.868 0.873 21.922 21.951 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 19.128 20.293 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.133 20.135 make_m2s 4970 13.5 0.076 0.078 18.046 19.458 ot_diis_step 116 11.6 0.021 0.022 19.452 19.452 make_images 4970 14.5 3.060 3.245 17.570 18.987 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.795 18.926 apply_single 127 13.6 0.001 0.001 18.795 18.926 multiply_cannon_multrec 9940 15.5 10.233 12.012 17.969 18.056 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.954 16.955 cp_fm_diag_elpa_base 82 14.4 12.634 14.218 16.948 16.949 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.329 16.347 calculate_rho_elec 127 8.7 0.477 0.478 16.329 16.346 sum_up_and_integrate 127 10.3 0.318 0.320 16.199 16.285 integrate_v_rspace 127 11.3 0.004 0.005 15.880 15.966 multiply_cannon_sync_h2d 9940 15.5 14.196 14.234 14.196 14.234 init_scf_run 11 5.9 0.000 0.001 12.261 12.261 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.261 12.261 hybrid_alltoall_any 5155 16.4 1.301 3.036 10.190 12.241 make_images_data 4970 15.5 0.064 0.070 9.982 12.076 pw_transfer 1535 11.6 0.092 0.092 11.282 11.285 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.049 11.051 fft_wrap_pw1pw2_140 519 13.2 3.027 3.060 9.845 9.849 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.662 9.690 dbcsr_mm_accdrv_process 20590 16.0 4.143 5.947 7.489 9.362 cp_fm_cholesky_decompose 22 10.9 9.136 9.231 9.136 9.231 wfi_extrapolate 11 7.9 0.001 0.001 9.027 9.027 grid_integrate_task_list 127 12.3 8.499 8.701 8.499 8.701 qs_energies_init_hamiltonians 11 5.9 0.006 0.014 8.389 8.390 density_rs2pw 127 9.7 0.005 0.006 7.432 7.483 mp_alltoall_d11v 2401 14.1 6.922 7.058 6.922 7.058 calculate_dm_sparse 127 9.5 0.001 0.001 6.734 6.810 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.538 6.596 grid_collocate_task_list 127 9.7 6.364 6.448 6.364 6.448 fft3d_ps 1281 14.7 2.828 2.834 6.328 6.361 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.223 6.301 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.851000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2821.818182, yerr=139.561533 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 1.259442E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 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 0.022 0.044 85.641 85.642 qs_energies 1 2.0 0.000 0.000 85.172 85.178 ls_scf 1 3.0 0.000 0.000 84.281 84.287 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.397 73.586 multiply_cannon 111 7.7 0.017 0.020 56.631 57.671 multiply_cannon_loop 111 8.7 0.227 0.244 53.200 54.480 ls_scf_main 1 4.0 0.000 0.000 52.365 52.366 density_matrix_trs4 2 5.0 0.002 0.003 46.863 46.972 ls_scf_init_scf 1 4.0 0.000 0.000 28.877 28.879 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.744 27.803 mp_waitall_1 11031 10.9 22.571 25.834 22.571 25.834 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.669 25.691 multiply_cannon_multrec 2664 9.7 8.123 8.949 15.562 17.388 multiply_cannon_sync_h2d 2664 9.7 13.554 15.605 13.554 15.605 make_m2s 222 7.7 0.009 0.012 13.200 13.753 make_images 222 8.7 0.098 0.108 13.178 13.732 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.674 13.260 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.457 8.492 make_images_data 222 9.7 0.004 0.005 7.775 8.284 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.701 7.056 8.002 hybrid_alltoall_any 227 10.6 0.215 1.839 6.644 7.988 dbcsr_mm_accdrv_process_sort 4760 11.4 6.272 7.145 6.272 7.145 calculate_norms 4752 9.8 5.509 6.228 5.509 6.228 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.007 5.127 mp_sum_l 887 5.1 3.029 4.260 3.029 4.260 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.044 3.753 mp_irecv_dv 6231 10.9 2.027 3.735 2.027 3.735 make_images_sizes 222 9.7 0.000 0.000 0.680 3.412 mp_alltoall_i44 222 10.7 0.680 3.412 0.680 3.412 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.273 3.260 arnoldi_extremal 4 6.8 0.000 0.000 3.167 3.201 arnoldi_normal_ev 4 7.8 0.001 0.003 3.167 3.201 build_subspace 16 8.4 0.009 0.012 3.075 3.077 ls_scf_post 1 4.0 0.000 0.000 3.038 3.045 ls_scf_store_result 1 5.0 0.000 0.000 2.865 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.391 2.839 dbcsr_merge_single_wm 555 10.7 0.458 0.589 2.382 2.830 make_images_pack 222 9.7 2.208 2.633 2.210 2.635 dbcsr_sort_data 658 11.4 2.181 2.575 2.181 2.575 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.306 2.552 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.462 2.068 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.285 2.352 buffer_matrices_ensure_size 222 8.7 1.757 2.069 1.757 2.069 compute_matrix_preconditioner 1 6.0 0.000 0.001 1.796 1.800 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.767 1.768 rebuild_ks_matrix 3 7.3 0.000 0.000 1.757 1.758 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.007 1.757 1.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.642000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 2.143425E+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 266673. MP_Allreduce 3138 10075. 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 0.034 0.054 89.939 89.942 qs_energies 1 2.0 0.000 0.001 89.405 89.409 ls_scf 1 3.0 0.000 0.001 88.044 88.047 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.274 74.507 multiply_cannon 111 7.7 0.027 0.042 52.846 57.082 ls_scf_main 1 4.0 0.004 0.056 54.487 54.489 multiply_cannon_loop 111 8.7 0.135 0.147 50.030 53.138 density_matrix_trs4 2 5.0 0.002 0.008 48.773 49.039 ls_scf_init_scf 1 4.0 0.000 0.002 29.848 29.850 mp_waitall_1 9105 10.9 20.777 29.677 20.777 29.677 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.644 28.748 multiply_cannon_multrec 1332 9.7 13.302 17.113 22.636 27.591 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.286 26.295 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.522 20.033 make_m2s 222 7.7 0.006 0.008 14.975 15.563 make_images 222 8.7 1.372 1.703 14.944 15.534 dbcsr_mm_accdrv_process 4041 10.4 0.348 0.540 8.932 10.476 dbcsr_mm_accdrv_process_sort 4041 11.4 8.452 9.936 8.452 9.936 make_images_data 222 9.7 0.004 0.005 8.703 9.633 hybrid_alltoall_any 227 10.6 0.541 2.556 8.076 9.447 mp_sum_l 887 5.1 4.979 7.887 4.979 7.887 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.194 7.752 mp_irecv_dv 3311 11.0 3.174 7.694 3.174 7.694 multiply_cannon_sync_h2d 1332 9.7 4.905 6.861 4.905 6.861 calculate_norms 2376 9.8 6.072 6.808 6.072 6.808 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.739 6.062 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.970 5.203 arnoldi_extremal 4 6.8 0.000 0.000 4.583 4.600 arnoldi_normal_ev 4 7.8 0.001 0.005 4.583 4.600 build_subspace 16 8.4 0.014 0.021 4.331 4.334 ls_scf_post 1 4.0 0.000 0.002 3.709 3.712 ls_scf_store_result 1 5.0 0.000 0.000 3.410 3.534 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.118 3.351 dbcsr_matrix_vector_mult_local 304 10.0 2.738 3.221 2.740 3.222 mp_allgather_i34 111 8.7 0.799 2.976 0.799 2.976 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.200 2.893 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.617 2.700 dbcsr_data_new 4174 10.1 2.115 2.398 2.115 2.398 make_images_pack 222 9.7 1.818 2.123 1.820 2.125 dbcsr_sort_data 436 11.2 1.792 2.001 1.792 2.001 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.859 1.861 rebuild_ks_matrix 3 7.3 0.000 0.000 1.846 1.848 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.004 1.846 1.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.942000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 2.881233E+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 265448. MP_Allreduce 3138 10896. 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 0.038 0.063 93.613 93.614 qs_energies 1 2.0 0.000 0.001 92.953 92.976 ls_scf 1 3.0 0.000 0.001 91.525 91.546 dbcsr_multiply_generic 111 6.7 0.017 0.021 76.201 76.583 ls_scf_main 1 4.0 0.000 0.005 57.006 57.009 multiply_cannon 111 7.7 0.034 0.077 52.584 56.800 multiply_cannon_loop 111 8.7 0.117 0.130 49.805 54.039 density_matrix_trs4 2 5.0 0.002 0.004 51.096 51.299 mp_waitall_1 7281 11.0 24.037 33.944 24.037 33.944 ls_scf_init_scf 1 4.0 0.000 0.001 30.846 30.849 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.564 29.686 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.211 27.228 multiply_cannon_multrec 888 9.7 12.712 15.365 21.488 24.856 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.201 23.478 make_m2s 222 7.7 0.006 0.008 16.580 17.373 make_images 222 8.7 1.590 1.857 16.542 17.331 make_images_data 222 9.7 0.004 0.005 9.740 10.779 hybrid_alltoall_any 227 10.6 0.642 2.950 9.181 10.436 dbcsr_mm_accdrv_process 3754 10.4 0.438 1.069 8.299 9.848 mp_sum_l 887 5.1 5.336 8.979 5.336 8.979 dbcsr_mm_accdrv_process_sort 3754 11.4 7.724 8.904 7.724 8.904 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.741 7.309 multiply_cannon_sync_h2d 888 9.7 6.131 7.121 6.131 7.121 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.430 7.088 mp_irecv_dv 2335 11.1 2.415 7.046 2.415 7.046 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.140 7.042 arnoldi_extremal 4 6.8 0.000 0.000 5.048 5.070 arnoldi_normal_ev 4 7.8 0.001 0.006 5.048 5.070 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.766 5.017 calculate_norms 1584 9.8 4.339 4.755 4.339 4.755 build_subspace 16 8.4 0.014 0.021 4.742 4.748 mp_allgather_i34 111 8.7 0.905 3.833 0.905 3.833 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.423 3.754 ls_scf_post 1 4.0 0.000 0.000 3.673 3.695 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.589 3.023 3.591 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.495 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.841 2.931 dbcsr_data_new 4116 9.9 2.105 2.449 2.105 2.449 dbcsr_sort_data 325 11.1 1.901 2.167 1.901 2.167 make_images_sizes 222 9.7 0.000 0.000 0.990 2.133 mp_alltoall_i44 222 10.7 0.989 2.132 0.989 2.132 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.930 1.932 dbcsr_finalize 304 7.8 0.026 0.032 1.618 1.917 rebuild_ks_matrix 3 7.3 0.000 0.000 1.911 1.913 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.911 1.913 make_images_pack 222 9.7 1.624 1.882 1.627 1.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.614000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2294.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 3.293651E+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 266673. MP_Allreduce 3138 13030. 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 0.042 0.054 98.093 98.096 qs_energies 1 2.0 0.000 0.001 97.395 97.402 ls_scf 1 3.0 0.000 0.001 95.631 95.635 dbcsr_multiply_generic 111 6.7 0.018 0.021 79.432 79.672 ls_scf_main 1 4.0 0.001 0.007 59.417 59.418 multiply_cannon 111 7.7 0.056 0.131 52.250 56.503 density_matrix_trs4 2 5.0 0.002 0.005 53.323 53.445 multiply_cannon_loop 111 8.7 0.153 0.165 47.250 50.763 ls_scf_init_scf 1 4.0 0.000 0.001 32.946 32.948 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.731 31.802 mp_waitall_1 6369 11.0 23.286 29.867 23.286 29.867 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.906 28.927 multiply_cannon_multrec 1332 9.7 14.104 17.370 22.177 24.793 make_m2s 222 7.7 0.007 0.008 21.322 22.679 make_images 222 8.7 3.152 3.630 21.271 22.630 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.606 17.400 make_images_data 222 9.7 0.004 0.004 11.872 13.395 hybrid_alltoall_any 227 10.6 0.803 3.761 11.262 12.869 dbcsr_mm_accdrv_process 3641 10.4 0.285 0.480 7.698 9.142 dbcsr_mm_accdrv_process_sort 3641 11.4 7.157 8.642 7.157 8.642 mp_sum_l 887 5.1 4.271 8.209 4.271 8.209 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.341 6.548 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.082 6.173 mp_irecv_dv 3229 10.9 2.056 6.100 2.056 6.100 multiply_cannon_sync_h2d 1332 9.7 5.458 5.960 5.458 5.960 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.640 5.335 arnoldi_extremal 4 6.8 0.000 0.000 5.204 5.223 arnoldi_normal_ev 4 7.8 0.001 0.005 5.204 5.223 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.872 5.046 build_subspace 16 8.4 0.015 0.021 4.864 4.872 calculate_norms 2376 9.8 4.184 4.540 4.184 4.540 mp_allgather_i34 111 8.7 2.119 4.303 2.119 4.303 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.577 3.867 dbcsr_matrix_vector_mult_local 304 10.0 3.193 3.680 3.195 3.682 dbcsr_sort_data 658 11.4 3.101 3.500 3.101 3.500 ls_scf_post 1 4.0 0.000 0.001 3.268 3.275 dbcsr_special_finalize 555 9.7 0.006 0.007 2.845 3.252 dbcsr_merge_single_wm 555 10.7 0.537 0.666 2.836 3.243 ls_scf_store_result 1 5.0 0.000 0.000 3.005 3.070 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.001 3.056 dbcsr_data_release 10477 10.7 1.595 2.401 1.595 2.401 dbcsr_finalize 304 7.8 0.050 0.061 1.814 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.096000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2711.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 4.759831E+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 265536. MP_Allreduce 3129 15263. 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 0.041 0.056 91.631 91.632 qs_energies 1 2.0 0.000 0.000 90.891 90.893 ls_scf 1 3.0 0.000 0.000 88.878 88.879 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.488 70.682 ls_scf_main 1 4.0 0.000 0.000 56.314 56.315 multiply_cannon 111 7.7 0.075 0.167 52.192 55.817 multiply_cannon_loop 111 8.7 0.088 0.095 49.634 51.164 density_matrix_trs4 2 5.0 0.002 0.003 49.324 49.421 ls_scf_init_scf 1 4.0 0.000 0.000 29.138 29.138 mp_waitall_1 5436 11.0 24.079 28.395 24.079 28.395 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.908 27.943 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.806 25.824 multiply_cannon_multrec 444 9.7 13.686 16.100 20.818 22.958 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.249 15.149 make_m2s 222 7.7 0.005 0.005 13.580 14.496 make_images 222 8.7 2.044 2.483 13.512 14.428 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.326 14.427 hybrid_alltoall_any 227 10.6 0.803 3.847 8.101 9.726 make_images_data 222 9.7 0.003 0.004 8.324 9.605 multiply_cannon_sync_h2d 444 9.7 6.723 7.984 6.723 7.984 dbcsr_mm_accdrv_process 3003 10.4 0.362 0.399 6.829 7.966 dbcsr_mm_accdrv_process_sort 3003 11.4 6.468 7.569 6.468 7.569 arnoldi_extremal 4 6.8 0.000 0.000 5.843 5.861 arnoldi_normal_ev 4 7.8 0.002 0.005 5.843 5.861 build_subspace 16 8.4 0.015 0.019 5.445 5.453 mp_sum_l 887 5.1 2.715 4.886 2.715 4.886 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.481 4.679 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.179 4.397 dbcsr_matrix_vector_mult_local 304 10.0 3.703 4.178 3.705 4.180 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.901 3.869 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.555 3.798 mp_irecv_dv 1241 11.2 1.541 3.772 1.541 3.772 calculate_norms 792 9.8 3.621 3.723 3.621 3.723 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.531 3.623 mp_allgather_i34 111 8.7 1.135 3.603 1.135 3.603 ls_scf_post 1 4.0 0.000 0.000 3.426 3.428 make_images_sizes 222 9.7 0.000 0.000 0.875 3.335 mp_alltoall_i44 222 10.7 0.875 3.335 0.875 3.335 ls_scf_store_result 1 5.0 0.000 0.000 3.209 3.244 dbcsr_finalize 304 7.8 0.062 0.077 2.194 2.288 dbcsr_data_new 4608 9.7 1.790 2.217 1.790 2.217 dbcsr_merge_all 275 8.9 0.479 0.525 2.052 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.022 2.023 qs_energies_init_hamiltonians 1 3.0 0.005 0.026 1.997 1.997 rebuild_ks_matrix 3 7.3 0.000 0.000 1.989 1.990 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.989 1.990 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.632000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3691.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_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 8.840839E+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 284089. MP_Allreduce 3123 21388. 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 0.098 0.109 105.534 105.538 qs_energies 1 2.0 0.000 0.000 104.082 104.095 ls_scf 1 3.0 0.000 0.000 101.149 101.161 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.881 75.023 ls_scf_main 1 4.0 0.000 0.000 63.820 63.821 density_matrix_trs4 2 5.0 0.002 0.003 54.869 54.936 multiply_cannon 111 7.7 0.115 0.157 48.686 50.985 multiply_cannon_loop 111 8.7 0.098 0.100 45.700 46.078 ls_scf_init_scf 1 4.0 0.000 0.000 33.533 33.534 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.030 32.042 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.324 29.335 mp_waitall_1 4527 11.1 22.013 25.803 22.013 25.803 make_m2s 222 7.7 0.005 0.005 22.593 23.690 make_images 222 8.7 3.579 3.889 22.486 23.580 multiply_cannon_multrec 444 9.7 17.849 18.427 22.540 23.170 hybrid_alltoall_any 227 10.6 1.656 3.629 12.759 15.669 make_images_data 222 9.7 0.003 0.004 12.996 15.128 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.434 10.844 multiply_cannon_sync_h2d 444 9.7 8.793 8.831 8.793 8.831 arnoldi_extremal 4 6.8 0.000 0.000 7.412 7.424 arnoldi_normal_ev 4 7.8 0.003 0.009 7.412 7.424 build_subspace 16 8.4 0.026 0.035 6.856 6.868 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.497 5.644 dbcsr_matrix_vector_mult_local 304 10.0 5.062 5.355 5.065 5.358 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.094 5.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.854 5.099 dbcsr_mm_accdrv_process 1814 10.4 0.294 0.366 4.502 4.624 dbcsr_mm_accdrv_process_sort 1814 11.4 4.162 4.291 4.162 4.291 ls_scf_post 1 4.0 0.000 0.000 3.796 3.809 make_images_sizes 222 9.7 0.000 0.000 1.460 3.619 mp_alltoall_i44 222 10.7 1.460 3.619 1.460 3.619 mp_allgather_i34 111 8.7 1.106 3.604 1.106 3.604 ls_scf_store_result 1 5.0 0.000 0.000 3.516 3.551 calculate_norms 792 9.8 3.230 3.275 3.230 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.081 3.184 dbcsr_merge_all 275 8.9 0.887 0.917 2.866 2.959 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.903 2.903 dbcsr_complete_redistribute 5 7.6 1.437 1.478 2.754 2.871 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.523 dbcsr_sort_data 325 11.1 2.440 2.513 2.440 2.513 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.420 2.422 dbcsr_data_new 6591 9.6 1.880 2.360 1.880 2.360 rebuild_ks_matrix 3 7.3 0.000 0.000 2.354 2.356 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.354 2.356 dbcsr_new_transposed 4 7.5 0.240 0.252 2.305 2.328 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.182 2.203 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.203 dbcsr_add_anytype 103 7.2 0.859 0.890 2.126 2.202 dbcsr_data_release 12724 10.6 1.983 2.195 1.983 2.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.538000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6957.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/cd366a4022ad0ce9b8ea778ed4e993fe588c3c12_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 592.891904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4037197. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.229 230.120 230.153 qs_mol_dyn_low 1 2.0 0.005 0.030 228.840 228.874 qs_forces 5 3.8 0.005 0.026 228.732 228.750 qs_energies 5 4.8 0.003 0.033 225.663 225.684 scf_env_do_scf 5 5.8 0.001 0.003 212.472 212.475 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 183.947 183.948 qs_scf_new_mos 105 7.6 0.000 0.001 143.135 143.288 qs_scf_loop_do_ot 105 8.6 0.001 0.001 143.135 143.288 ot_scf_mini 105 9.6 0.003 0.004 133.279 133.421 dbcsr_multiply_generic 1445 12.2 0.126 0.137 132.862 133.303 multiply_cannon 1445 13.2 0.274 0.287 113.507 115.470 multiply_cannon_loop 1445 14.2 2.855 3.014 111.838 113.022 velocity_verlet 4 3.0 0.001 0.013 107.788 107.789 ot_mini 105 10.6 0.001 0.003 60.705 60.826 multiply_cannon_multrec 69360 15.2 29.782 34.868 39.737 45.079 qs_ot_get_p 112 10.4 0.001 0.001 42.868 43.206 mp_waitall_1 488190 16.1 35.560 42.358 35.560 42.358 qs_ot_get_derivative 55 11.6 0.001 0.001 38.997 39.139 multiply_cannon_metrocomm3 69360 15.2 0.201 0.210 26.228 33.358 multiply_cannon_sync_h2d 69360 15.2 29.035 33.198 29.035 33.198 qs_ot_p2m_diag 40 11.0 0.020 0.031 31.754 31.844 rebuild_ks_matrix 110 8.4 0.000 0.000 29.409 29.607 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.015 29.409 29.606 cp_dbcsr_syevd 40 12.0 0.002 0.005 28.532 28.533 init_scf_loop 7 6.6 0.001 0.006 28.488 28.490 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.953 27.135 cp_fm_syevd 40 13.0 0.000 0.002 23.377 23.529 prepare_preconditioner 7 7.6 0.000 0.000 23.479 23.510 make_preconditioner 7 8.6 0.000 0.004 23.479 23.510 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.004 23.251 apply_single 62 13.6 0.000 0.000 23.004 23.250 ot_new_cg_direction 55 11.6 0.001 0.003 21.035 21.035 cp_fm_redistribute_end 40 14.0 9.332 18.615 9.337 18.616 cp_fm_syevd_base 40 14.0 9.272 18.557 9.272 18.557 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.415 17.758 calculate_rho_elec 110 8.6 0.029 0.032 17.414 17.758 make_full_inverse_cholesky 7 9.6 0.000 0.001 15.233 15.297 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.522 14.651 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.982 14.075 mp_sum_l 4764 12.2 12.342 13.085 12.342 13.085 pw_transfer 1645 12.4 0.080 0.094 12.249 12.448 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.110 12.313 density_rs2pw 110 9.6 0.005 0.007 11.670 11.991 calculate_dm_sparse 110 9.5 0.000 0.001 11.428 11.622 dbcsr_mm_accdrv_process 154766 15.8 6.177 6.365 9.822 11.257 fft_wrap_pw1pw2_240 915 15.0 1.149 1.224 10.630 10.816 qs_vxc_create 110 10.4 0.003 0.009 10.726 10.761 cp_fm_cholesky_invert 7 10.6 10.696 10.704 10.696 10.704 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.590 10.667 init_scf_run 5 5.8 0.000 0.001 10.501 10.502 scf_env_initial_rho_setup 5 6.8 0.001 0.003 10.501 10.502 check_diag 80 13.5 8.594 8.892 9.538 9.689 fft3d_pb 915 16.0 2.384 2.577 8.832 9.026 sum_up_and_integrate 60 10.3 0.028 0.034 8.660 8.672 integrate_v_rspace 60 11.3 0.002 0.002 8.632 8.645 transfer_rs2pw 445 10.6 0.007 0.008 8.209 8.608 make_full_single_inverse 7 9.6 0.001 0.007 7.889 7.919 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 7.495 7.736 multiply_cannon_metrocomm1 69360 15.2 0.099 0.105 4.830 7.540 xc_vxc_pw_create 60 11.3 0.039 0.050 7.156 7.192 make_m2s 2890 13.2 0.078 0.088 6.600 7.174 make_images 2890 14.2 0.240 0.259 6.494 7.070 acc_transpose_blocks 69360 15.2 0.356 0.377 6.053 6.573 xc_pw_derive 510 13.4 0.005 0.006 6.436 6.508 mp_alltoall_z22v 2340 17.7 5.855 6.156 5.855 6.156 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.019 6.036 calculate_first_density_matrix 1 7.0 0.000 0.003 5.562 5.580 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 5.502 5.551 multiply_cannon_metrocomm4 67915 15.2 0.185 0.202 2.016 4.994 mp_waitany 7680 13.5 4.461 4.948 4.461 4.948 potential_pw2rs 60 12.3 0.003 0.004 4.845 4.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.153000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: cd366a4022ad0ce9b8ea778ed4e993fe588c3c12 Summary: empty Status: OK