=== This is the CP2K Performance-Test === Updating 8b06a4d86..b41f84f0a Fast-forward src/hfx_ri.F | 88 ++++++++++++++++++++++++++++++------------------------------ 1 file changed, 44 insertions(+), 44 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: b41f84f0a640ce8588a35f300e725818d42616be ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ 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 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/01 job id: 41796833 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/02 job id: 41796835 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/03 job id: 41796837 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/04 job id: 41796839 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/05 job id: 41796840 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/06 job id: 41796841 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/07 job id: 41796842 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/08 job id: 41796843 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/09 job id: 41796844 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/10 job id: 41796845 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/11 job id: 41796846 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/12 job id: 41796847 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/13 job id: 41796848 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/14 job id: 41796849 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/15 job id: 41796850 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/16 job id: 41796851 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/17 job id: 41796852 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/18 job id: 41796853 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/19 job id: 41796854 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/20 job id: 41796855 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/21 job id: 41796856 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/22 job id: 41796857 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/23 job id: 41796858 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/24 job id: 41796859 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/25 job id: 41796860 --- 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/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/26 job id: 41796863 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.046 134.319 134.321 farming_run 1 2.0 132.853 132.855 134.283 134.287 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458287E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 115.134 115.136 qs_energies 1 2.0 0.000 0.000 114.823 114.825 mp2_main 1 3.0 0.000 0.000 112.961 112.963 mp2_gpw_main 1 4.0 0.098 0.106 112.158 112.160 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.169 93.669 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.442 55.943 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.170 41.731 47.184 get_2c_integrals 1 6.0 0.000 0.000 36.724 37.554 integrate_v_rspace 273 8.0 0.439 0.454 25.112 30.358 pw_transfer 6555 10.6 0.378 0.386 27.345 27.840 grid_integrate_task_list 273 9.0 20.920 26.637 20.920 26.637 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.021 26.488 fft_wrap_pw1pw2_100 2178 12.4 1.169 1.246 23.581 24.037 compute_2c_integrals 1 7.0 0.003 0.003 19.225 19.225 rpa_ri_compute_en 1 5.0 0.001 0.001 18.801 18.980 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.875 18.962 mp2_eri_2c_integrate_gpw 1 9.0 2.402 2.437 18.871 18.958 cp_fm_cholesky_decompose 12 8.2 17.436 18.268 17.436 18.268 cholesky_decomp 1 7.0 0.000 0.000 16.348 17.174 fft3d_s 5443 13.4 16.131 16.510 16.153 16.530 ao_to_mo_and_store_B_mult_1 272 7.0 10.872 15.604 10.872 15.604 calculate_wavefunction 272 8.0 5.409 5.567 12.530 13.156 rpa_num_int 1 6.0 0.000 0.000 10.473 10.474 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.457 10.470 calc_mat_Q 8 8.0 0.000 0.000 9.321 9.412 contract_S_to_Q 8 9.0 0.000 0.000 8.744 8.832 calc_potential_gpw 544 9.5 0.005 0.006 8.216 8.596 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.200 8.423 parallel_gemm_fm 14 9.1 0.000 0.000 8.330 8.398 parallel_gemm_fm_cosma 14 10.1 8.329 8.398 8.329 8.398 potential_pw2rs 545 10.0 0.108 0.110 7.720 8.374 create_integ_mat 1 6.0 0.022 0.028 7.732 7.733 collocate_single_gaussian 272 10.0 0.041 0.043 7.449 7.661 array2fm 1 7.0 0.000 0.000 6.684 7.162 pw_scatter_s 2720 13.7 4.426 4.651 4.426 4.651 pw_gather_s 2722 13.2 3.923 4.365 3.923 4.365 array2fm_buffer_send 1 8.0 2.982 3.195 2.982 3.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.157197, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.043 397.684 397.686 farming_run 1 2.0 395.822 395.827 397.638 397.642 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224221E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 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 0.013 0.047 209.619 209.620 qs_energies 1 2.0 0.000 0.000 209.365 209.377 scf_env_do_scf 1 3.0 0.000 0.000 106.591 106.592 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.705 105.719 rebuild_ks_matrix 4 6.0 0.000 0.000 105.368 105.376 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.368 105.376 hfx_ks_matrix 4 8.0 0.001 0.001 104.995 104.999 integrate_four_center 4 9.0 0.144 0.457 104.994 104.999 mp2_main 1 3.0 0.000 0.000 102.493 102.505 mp2_gpw_main 1 4.0 0.032 0.047 101.615 101.630 integrate_four_center_main 4 10.0 0.101 0.458 96.557 99.598 integrate_four_center_bin 261 11.0 96.455 99.555 96.455 99.555 init_scf_loop 1 4.0 0.000 0.000 91.926 91.926 mp2_ri_gpw_compute_in 1 5.0 0.067 0.073 74.985 76.038 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.492 55.545 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.974 47.129 integrate_v_rspace 95 8.0 0.398 0.572 28.342 33.261 pw_transfer 2240 10.6 0.143 0.164 29.931 30.296 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.924 29.313 grid_integrate_task_list 95 9.0 23.637 28.724 23.637 28.724 ao_to_mo_and_store_B_mult_1 91 7.0 10.834 28.396 10.834 28.396 mp2_ri_gpw_compute_en 1 5.0 0.069 0.082 26.486 28.193 fft_wrap_pw1pw2_100 730 12.4 1.293 1.448 26.670 27.089 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.919 24.773 24.785 get_2c_integrals 1 6.0 0.000 0.000 20.381 20.420 compute_2c_integrals 1 7.0 0.010 0.036 19.352 19.361 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.924 19.192 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.881 18.922 19.191 fft3d_s 1823 13.4 18.434 18.786 18.448 18.799 scf_env_do_scf_inner_loop 4 4.0 0.007 0.343 14.664 14.664 calculate_wavefunction 91 8.0 2.024 2.060 9.746 10.000 potential_pw2rs 186 10.0 0.034 0.036 8.653 9.223 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.588 8.757 9.175 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.277 8.618 local_gemm 172 8.0 8.201 8.597 8.201 8.597 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.522 7.782 8.223 collocate_single_gaussian 91 10.0 0.017 0.020 7.910 8.163 calc_potential_gpw 182 9.5 0.002 0.002 7.925 8.154 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.431 6.346 6.431 mp_sendrecv_dm3 2068 8.0 5.819 6.295 5.819 6.295 mp_sync 38 10.4 3.101 5.860 3.101 5.860 pw_gather_s 912 13.2 4.908 5.400 4.908 5.400 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.617086, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 450.289664E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 1600075. MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.023 0.053 54.175 54.180 qs_mol_dyn_low 1 2.0 0.004 0.006 53.766 53.834 qs_forces 11 3.9 0.002 0.006 53.485 53.487 qs_energies 11 4.9 0.001 0.004 51.953 51.976 scf_env_do_scf 11 5.9 0.000 0.001 45.728 45.728 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.559 43.559 dbcsr_multiply_generic 2286 12.5 0.092 0.097 34.364 34.857 qs_scf_new_mos 108 7.5 0.000 0.001 33.034 33.310 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.034 33.309 ot_scf_mini 108 9.5 0.002 0.002 31.364 31.575 multiply_cannon 2286 13.5 0.187 0.194 26.319 27.904 velocity_verlet 10 3.0 0.028 0.213 27.258 27.259 multiply_cannon_loop 2286 14.5 1.501 1.572 25.537 27.201 ot_mini 108 10.5 0.001 0.002 20.209 20.449 qs_ot_get_derivative 108 11.5 0.001 0.001 17.243 17.428 mp_waitall_1 267858 16.1 9.447 15.607 9.447 15.607 multiply_cannon_metrocomm3 54864 15.5 0.066 0.070 6.269 13.077 multiply_cannon_multrec 54864 15.5 4.161 6.483 7.505 10.824 rebuild_ks_matrix 119 8.3 0.000 0.000 8.130 8.274 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 8.130 8.273 mp_sum_l 7207 12.9 5.846 7.572 5.846 7.572 multiply_cannon_sync_h2d 54864 15.5 5.786 7.537 5.786 7.537 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.163 7.295 qs_ot_get_p 119 10.4 0.001 0.001 6.390 6.664 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.919 6.383 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.780 5.903 init_scf_run 11 5.9 0.000 0.001 4.905 4.906 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.905 4.905 sum_up_and_integrate 119 10.3 0.013 0.018 4.712 4.720 integrate_v_rspace 119 11.3 0.002 0.003 4.699 4.708 dbcsr_mm_accdrv_process 76910 16.1 1.182 1.824 3.267 4.545 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.392 4.516 calculate_rho_elec 119 8.7 0.011 0.016 4.391 4.515 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.047 3.847 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.495 3.557 rs_pw_transfer 974 11.9 0.012 0.013 3.033 3.147 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.900 3.099 apply_single 119 13.6 0.000 0.000 2.900 3.099 calculate_dm_sparse 119 9.5 0.000 0.000 2.960 3.089 ot_diis_step 108 11.5 0.006 0.008 2.718 2.719 calculate_first_density_matrix 1 7.0 0.000 0.001 2.676 2.681 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.659 2.660 jit_kernel_multiply 13 15.8 2.024 2.636 2.024 2.636 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.543 2.615 density_rs2pw 119 9.7 0.004 0.004 2.423 2.499 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.430 2.430 cp_fm_redistribute_end 50 14.0 2.206 2.406 2.211 2.408 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.380 2.382 cp_fm_diag_elpa_base 50 14.0 0.195 2.360 0.196 2.369 wfi_extrapolate 11 7.9 0.001 0.004 2.166 2.167 init_scf_loop 11 6.9 0.000 0.000 2.142 2.142 grid_integrate_task_list 119 12.3 2.031 2.139 2.031 2.139 acc_transpose_blocks 54864 15.5 0.225 0.243 1.726 2.132 potential_pw2rs 119 12.3 0.004 0.005 2.044 2.062 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.998 2.057 pw_transfer 1439 11.6 0.052 0.056 1.905 1.981 mp_sum_d 4129 12.0 1.370 1.947 1.370 1.947 make_m2s 4572 13.5 0.053 0.055 1.898 1.946 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.830 1.907 make_images 4572 14.5 0.132 0.137 1.817 1.864 fft3d_ps 1201 14.6 0.355 0.460 1.609 1.679 mp_alltoall_d11v 2130 13.8 1.491 1.625 1.491 1.625 fft_wrap_pw1pw2_140 487 13.2 0.079 0.093 1.432 1.510 mp_waitany 12084 13.8 1.318 1.509 1.318 1.509 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.471 1.504 grid_collocate_task_list 119 9.7 1.286 1.342 1.286 1.342 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.753 1.171 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.180000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 485.146624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 863609. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 40.724 40.724 qs_mol_dyn_low 1 2.0 0.003 0.003 40.435 40.442 qs_forces 11 3.9 0.002 0.002 40.375 40.375 qs_energies 11 4.9 0.001 0.001 38.611 38.615 scf_env_do_scf 11 5.9 0.000 0.001 32.928 32.928 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.189 30.190 dbcsr_multiply_generic 2286 12.5 0.098 0.104 22.615 23.013 qs_scf_new_mos 108 7.5 0.001 0.001 20.837 21.081 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.836 21.081 ot_scf_mini 108 9.5 0.002 0.003 19.903 20.076 velocity_verlet 10 3.0 0.001 0.002 19.058 19.059 multiply_cannon 2286 13.5 0.209 0.220 16.954 18.530 multiply_cannon_loop 2286 14.5 0.905 0.987 15.741 17.335 ot_mini 108 10.5 0.001 0.001 12.456 12.693 mp_waitall_1 217478 16.2 6.727 11.831 6.727 11.831 qs_ot_get_derivative 108 11.5 0.001 0.001 9.949 10.125 multiply_cannon_metrocomm3 27432 15.5 0.068 0.072 4.280 9.820 multiply_cannon_multrec 27432 15.5 1.967 4.520 6.149 9.077 rebuild_ks_matrix 119 8.3 0.000 0.000 7.446 7.592 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.446 7.591 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.548 6.681 dbcsr_mm_accdrv_process 47894 16.0 3.343 5.500 4.111 6.169 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.743 4.853 qs_ot_get_p 119 10.4 0.001 0.001 4.482 4.721 init_scf_run 11 5.9 0.000 0.001 4.480 4.481 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.480 4.481 mp_sum_l 7207 12.9 2.445 4.470 2.445 4.470 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.334 4.450 apply_single 119 13.6 0.000 0.000 3.334 4.450 sum_up_and_integrate 119 10.3 0.025 0.028 4.424 4.430 integrate_v_rspace 119 11.3 0.002 0.002 4.399 4.406 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.830 3.869 calculate_rho_elec 119 8.7 0.021 0.024 3.830 3.869 make_m2s 4572 13.5 0.051 0.053 2.770 3.039 rs_pw_transfer 974 11.9 0.010 0.011 2.892 2.982 make_images 4572 14.5 0.200 0.237 2.684 2.950 calculate_first_density_matrix 1 7.0 0.000 0.000 2.871 2.873 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.813 2.832 multiply_cannon_sync_h2d 27432 15.5 2.189 2.814 2.189 2.814 init_scf_loop 11 6.9 0.000 0.000 2.718 2.719 calculate_dm_sparse 119 9.5 0.000 0.000 2.452 2.523 ot_diis_step 108 11.5 0.010 0.011 2.450 2.451 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.368 2.369 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.244 2.335 density_rs2pw 119 9.7 0.004 0.004 2.176 2.245 potential_pw2rs 119 12.3 0.006 0.007 2.142 2.161 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.063 2.065 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.955 1.956 grid_integrate_task_list 119 12.3 1.836 1.952 1.836 1.952 cp_fm_redistribute_end 50 14.0 1.608 1.923 1.612 1.926 pw_transfer 1439 11.6 0.063 0.066 1.885 1.915 cp_fm_diag_elpa_base 50 14.0 0.303 1.853 0.311 1.884 jit_kernel_multiply 9 16.2 0.717 1.861 0.717 1.861 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.810 1.851 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.796 1.828 prepare_preconditioner 11 7.9 0.000 0.000 1.751 1.777 make_preconditioner 11 8.9 0.000 0.000 1.751 1.777 acc_transpose_blocks 27432 15.5 0.107 0.112 1.186 1.776 make_images_data 4572 15.5 0.045 0.051 1.232 1.713 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.634 1.691 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.627 1.634 hybrid_alltoall_any 4725 16.4 0.051 0.113 1.071 1.556 wfi_extrapolate 11 7.9 0.001 0.001 1.549 1.550 fft3d_ps 1201 14.6 0.495 0.551 1.502 1.532 mp_alltoall_d11v 2130 13.8 1.322 1.512 1.322 1.512 mp_allgather_i34 2286 14.5 0.656 1.433 0.656 1.433 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.331 1.363 grid_collocate_task_list 119 9.7 1.242 1.338 1.242 1.338 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.214 1.258 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.657 1.157 mp_sum_d 4129 12.0 0.616 1.056 0.616 1.056 make_images_sizes 4572 15.5 0.005 0.005 0.725 1.011 mp_alltoall_i44 4572 16.5 0.720 1.007 0.720 1.007 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.957 0.970 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 0.935 0.937 jit_kernel_transpose 5 15.5 0.476 0.891 0.476 0.891 multiply_cannon_metrocomm1 27432 15.5 0.031 0.036 0.455 0.887 rs_pw_transfer_PW2RS_50 119 14.3 0.591 0.611 0.824 0.875 mp_alltoall_z22v 1201 16.6 0.758 0.829 0.758 0.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.724000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.272727, yerr=1.354515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 520.585216E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.102 34.042 34.044 qs_mol_dyn_low 1 2.0 0.003 0.003 33.673 33.680 qs_forces 11 3.9 0.004 0.020 33.613 33.615 qs_energies 11 4.9 0.013 0.026 32.021 32.024 scf_env_do_scf 11 5.9 0.001 0.001 26.128 26.128 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.481 23.483 dbcsr_multiply_generic 2286 12.5 0.093 0.102 17.587 17.723 velocity_verlet 10 3.0 0.002 0.002 15.510 15.511 qs_scf_new_mos 108 7.5 0.001 0.001 15.393 15.414 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.392 15.413 multiply_cannon 2286 13.5 0.197 0.204 13.964 14.846 ot_scf_mini 108 9.5 0.002 0.003 14.648 14.666 multiply_cannon_loop 2286 14.5 0.639 0.667 13.144 14.063 ot_mini 108 10.5 0.001 0.001 9.167 9.187 multiply_cannon_multrec 18288 15.5 1.967 2.897 7.444 7.789 qs_ot_get_derivative 108 11.5 0.001 0.001 7.634 7.646 rebuild_ks_matrix 119 8.3 0.000 0.000 6.388 6.411 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.387 6.410 dbcsr_mm_accdrv_process 38222 16.0 4.220 5.961 5.393 6.355 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.654 5.675 init_scf_run 11 5.9 0.000 0.001 4.610 4.610 scf_env_initial_rho_setup 11 6.9 0.011 0.018 4.609 4.610 mp_waitall_1 169478 16.3 3.166 4.370 3.166 4.370 sum_up_and_integrate 119 10.3 0.031 0.032 4.012 4.016 integrate_v_rspace 119 11.3 0.002 0.003 3.981 3.990 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.949 3.654 calculate_first_density_matrix 1 7.0 0.000 0.000 3.341 3.350 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.308 3.328 calculate_rho_elec 119 8.7 0.030 0.031 3.308 3.328 qs_ot_get_p 119 10.4 0.001 0.001 3.242 3.271 jit_kernel_multiply 12 16.0 1.122 2.891 1.122 2.891 init_scf_loop 11 6.9 0.000 0.002 2.626 2.633 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.606 2.628 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.132 2.494 apply_single 119 13.6 0.000 0.000 2.132 2.493 rs_pw_transfer 974 11.9 0.009 0.010 2.370 2.465 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.422 2.424 calculate_dm_sparse 119 9.5 0.000 0.000 2.101 2.125 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.072 2.082 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.997 2.010 make_m2s 4572 13.5 0.044 0.045 1.889 2.003 density_rs2pw 119 9.7 0.004 0.004 1.883 1.977 make_images 4572 14.5 0.192 0.206 1.805 1.919 grid_integrate_task_list 119 12.3 1.799 1.886 1.799 1.886 mp_sum_l 7207 12.9 1.381 1.843 1.381 1.843 prepare_preconditioner 11 7.9 0.000 0.000 1.823 1.826 make_preconditioner 11 8.9 0.000 0.002 1.823 1.826 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.799 1.800 potential_pw2rs 119 12.3 0.007 0.008 1.788 1.795 pw_transfer 1439 11.6 0.063 0.066 1.757 1.770 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.674 1.754 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.666 1.682 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.617 1.622 multiply_cannon_sync_h2d 18288 15.5 1.356 1.533 1.356 1.533 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.531 1.532 ot_diis_step 108 11.5 0.011 0.011 1.507 1.507 cp_fm_redistribute_end 50 14.0 1.132 1.503 1.134 1.504 cp_fm_diag_elpa_base 50 14.0 0.353 1.434 0.368 1.473 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.452 1.456 acc_transpose_blocks 18288 15.5 0.074 0.076 1.361 1.379 fft3d_ps 1201 14.6 0.506 0.523 1.354 1.370 grid_collocate_task_list 119 9.7 1.203 1.329 1.203 1.329 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.292 1.305 wfi_extrapolate 11 7.9 0.001 0.001 1.212 1.212 qs_energies_init_hamiltonians 11 5.9 0.004 0.006 1.022 1.024 make_images_data 4572 15.5 0.045 0.049 0.848 0.989 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.335 0.983 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.946 0.980 acc_transpose_blocks_kernels 18288 16.5 0.209 0.216 0.924 0.937 dbcsr_complete_redistribute 329 12.2 0.220 0.334 0.784 0.904 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.731 0.904 mp_alltoall_d11v 2130 13.8 0.760 0.865 0.760 0.865 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.805 cp_fm_cholesky_invert 11 10.9 0.769 0.773 0.769 0.773 mp_alltoall_z22v 1201 16.6 0.686 0.764 0.686 0.764 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.640 0.730 jit_kernel_transpose 5 15.6 0.715 0.723 0.715 0.723 copy_fm_to_dbcsr 176 11.2 0.001 0.003 0.588 0.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.044000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.090909, yerr=1.975051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 544.649216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.080 36.151 36.152 qs_mol_dyn_low 1 2.0 0.004 0.009 35.743 35.750 qs_forces 11 3.9 0.002 0.004 35.657 35.658 qs_energies 11 4.9 0.005 0.022 33.882 33.894 scf_env_do_scf 11 5.9 0.001 0.001 27.836 27.838 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.316 24.318 dbcsr_multiply_generic 2286 12.5 0.102 0.120 18.738 18.917 velocity_verlet 10 3.0 0.005 0.016 17.820 17.825 qs_scf_new_mos 108 7.5 0.001 0.001 16.138 16.186 qs_scf_loop_do_ot 108 8.5 0.001 0.002 16.138 16.185 multiply_cannon 2286 13.5 0.335 0.660 14.565 15.318 ot_scf_mini 108 9.5 0.003 0.003 15.200 15.253 multiply_cannon_loop 2286 14.5 0.935 0.971 13.432 13.735 ot_mini 108 10.5 0.001 0.001 9.326 9.389 multiply_cannon_multrec 27432 15.5 2.322 2.991 8.563 8.918 qs_ot_get_derivative 108 11.5 0.067 0.327 7.503 7.557 dbcsr_mm_accdrv_process 47916 15.9 5.271 6.920 6.147 7.391 rebuild_ks_matrix 119 8.3 0.000 0.000 6.554 6.610 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.554 6.609 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.768 5.816 init_scf_run 11 5.9 0.000 0.001 4.492 4.492 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.492 4.492 sum_up_and_integrate 119 10.3 0.035 0.040 3.826 3.842 integrate_v_rspace 119 11.3 0.002 0.003 3.791 3.808 init_scf_loop 11 6.9 0.001 0.002 3.500 3.505 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.307 3.337 calculate_rho_elec 119 8.7 0.040 0.046 3.307 3.337 qs_ot_get_p 119 10.4 0.001 0.001 3.244 3.314 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.799 3.176 calculate_first_density_matrix 1 7.0 0.000 0.002 3.108 3.110 mp_waitall_1 145218 16.4 2.511 3.098 2.511 3.098 make_m2s 4572 13.5 0.054 0.056 2.732 2.920 make_images 4572 14.5 0.277 0.335 2.625 2.812 prepare_preconditioner 11 7.9 0.000 0.000 2.641 2.648 make_preconditioner 11 8.9 0.000 0.001 2.640 2.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.229 2.569 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.553 2.554 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.300 2.323 rs_pw_transfer 974 11.9 0.009 0.009 2.123 2.280 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.969 2.170 apply_single 119 13.6 0.000 0.000 1.969 2.170 calculate_dm_sparse 119 9.5 0.000 0.000 2.033 2.092 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.002 2.027 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.938 1.948 density_rs2pw 119 9.7 0.003 0.004 1.793 1.897 grid_integrate_task_list 119 12.3 1.821 1.875 1.821 1.875 pw_transfer 1439 11.6 0.063 0.066 1.829 1.866 ot_diis_step 108 11.5 0.013 0.015 1.781 1.781 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.740 1.780 jit_kernel_multiply 9 16.1 0.816 1.646 0.816 1.646 potential_pw2rs 119 12.3 0.008 0.010 1.612 1.624 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.617 1.618 make_images_data 4572 15.5 0.045 0.051 1.219 1.468 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.418 1.462 acc_transpose_blocks 27432 15.5 0.111 0.113 1.440 1.456 fft3d_ps 1201 14.6 0.533 0.588 1.421 1.449 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.377 1.377 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.983 1.374 mp_sum_l 7207 12.9 1.017 1.374 1.017 1.374 cp_fm_redistribute_end 50 14.0 0.907 1.351 0.908 1.352 wfi_extrapolate 11 7.9 0.001 0.004 1.330 1.331 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.774 1.329 cp_fm_diag_elpa_base 50 14.0 0.421 1.288 0.441 1.328 grid_collocate_task_list 119 9.7 1.222 1.289 1.222 1.289 qs_energies_init_hamiltonians 11 5.9 0.017 0.025 1.258 1.266 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.225 dbcsr_complete_redistribute 329 12.2 0.147 0.220 0.883 1.154 cp_fm_upper_to_full 72 13.5 0.821 1.144 0.821 1.144 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.117 1.136 multiply_cannon_sync_h2d 27432 15.5 0.989 1.045 0.989 1.045 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.670 0.937 mp_alltoall_d11v 2130 13.8 0.783 0.926 0.783 0.926 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.880 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.873 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.819 0.826 mp_alltoall_z22v 1201 16.6 0.770 0.805 0.770 0.805 cp_fm_cholesky_invert 11 10.9 0.799 0.802 0.799 0.802 mp_alltoall_i22 627 13.8 0.467 0.774 0.467 0.774 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.691 0.766 rs_pw_transfer_RS2PW_140 130 11.5 0.109 0.115 0.580 0.743 mp_waitany 7280 13.7 0.572 0.736 0.572 0.736 mp_allgather_i34 2286 14.5 0.322 0.727 0.322 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.152000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.818182, yerr=2.854603 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 607.842304E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.413 29.935 29.936 qs_mol_dyn_low 1 2.0 0.003 0.003 29.225 29.233 qs_forces 11 3.9 0.002 0.002 29.161 29.162 qs_energies 11 4.9 0.001 0.001 27.468 27.470 scf_env_do_scf 11 5.9 0.000 0.001 21.431 21.431 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.918 18.919 velocity_verlet 10 3.0 0.005 0.006 14.163 14.166 dbcsr_multiply_generic 2286 12.5 0.091 0.093 13.242 13.354 qs_scf_new_mos 108 7.5 0.001 0.001 11.307 11.330 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.307 11.330 multiply_cannon 2286 13.5 0.229 0.236 10.724 11.121 ot_scf_mini 108 9.5 0.002 0.002 10.637 10.661 multiply_cannon_loop 2286 14.5 0.330 0.342 9.817 10.034 multiply_cannon_multrec 9144 15.5 1.697 1.933 6.358 6.550 ot_mini 108 10.5 0.001 0.001 6.061 6.091 rebuild_ks_matrix 119 8.3 0.000 0.000 5.833 5.857 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.833 5.856 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.203 5.224 qs_ot_get_derivative 108 11.5 0.001 0.001 4.770 4.795 dbcsr_mm_accdrv_process 12550 15.8 3.161 3.848 4.560 4.642 init_scf_run 11 5.9 0.000 0.001 4.622 4.622 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.622 4.622 sum_up_and_integrate 119 10.3 0.038 0.042 3.574 3.578 integrate_v_rspace 119 11.3 0.003 0.003 3.536 3.541 calculate_first_density_matrix 1 7.0 0.000 0.000 3.501 3.503 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.247 3.253 calculate_rho_elec 119 8.7 0.059 0.061 3.246 3.252 qs_ot_get_p 119 10.4 0.001 0.001 2.812 2.855 init_scf_loop 11 6.9 0.000 0.000 2.492 2.493 calculate_dm_sparse 119 9.5 0.000 0.000 2.327 2.346 jit_kernel_multiply 10 15.7 1.361 2.278 1.361 2.278 mp_waitall_1 121218 16.5 1.703 2.089 1.703 2.089 grid_integrate_task_list 119 12.3 1.849 1.924 1.849 1.924 make_m2s 4572 13.5 0.034 0.035 1.747 1.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.881 1.883 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.852 1.854 make_images 4572 14.5 0.268 0.302 1.659 1.809 prepare_preconditioner 11 7.9 0.000 0.000 1.754 1.758 make_preconditioner 11 8.9 0.000 0.000 1.754 1.758 rs_pw_transfer 974 11.9 0.008 0.009 1.674 1.754 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.749 1.754 pw_transfer 1439 11.6 0.063 0.065 1.746 1.754 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.731 1.743 density_rs2pw 119 9.7 0.004 0.004 1.648 1.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.644 1.670 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.656 1.665 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.658 1.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.624 1.635 acc_transpose_blocks 9144 15.5 0.039 0.040 1.481 1.490 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.362 1.363 grid_collocate_task_list 119 9.7 1.271 1.363 1.271 1.363 potential_pw2rs 119 12.3 0.010 0.011 1.352 1.357 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.337 1.346 cp_fm_redistribute_end 50 14.0 0.676 1.338 0.678 1.338 fft3d_ps 1201 14.6 0.539 0.550 1.322 1.330 cp_fm_diag_elpa_base 50 14.0 0.614 1.261 0.659 1.319 ot_diis_step 108 11.5 0.012 0.012 1.277 1.277 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 1.242 1.249 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.209 1.210 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.146 1.170 apply_single 119 13.6 0.000 0.000 1.146 1.169 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.146 1.160 jit_kernel_transpose 5 15.6 1.126 1.134 1.126 1.134 wfi_extrapolate 11 7.9 0.001 0.001 1.075 1.075 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.790 1.015 make_images_data 4572 15.5 0.038 0.042 0.814 0.989 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.919 mp_alltoall_d11v 2130 13.8 0.823 0.906 0.823 0.906 cp_fm_cholesky_invert 11 10.9 0.866 0.869 0.866 0.869 mp_sum_l 7207 12.9 0.700 0.816 0.700 0.816 multiply_cannon_sync_h2d 9144 15.5 0.708 0.784 0.708 0.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.755 0.757 parallel_gemm_fm 81 9.0 0.000 0.000 0.742 0.756 parallel_gemm_fm_cosma 81 10.0 0.742 0.755 0.742 0.755 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.714 0.724 make_basis_sm 11 9.8 0.000 0.000 0.719 0.719 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.663 0.713 mp_alltoall_z22v 1201 16.6 0.658 0.687 0.658 0.687 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.356 0.685 dbcsr_complete_redistribute 329 12.2 0.201 0.268 0.639 0.669 mp_allgather_i34 2286 14.5 0.240 0.647 0.240 0.647 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.936000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=576.272727, yerr=1.813631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 747.327488E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 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 0.057 0.359 41.678 41.678 qs_mol_dyn_low 1 2.0 0.003 0.003 40.783 41.157 qs_forces 11 3.9 0.002 0.002 40.671 40.672 qs_energies 11 4.9 0.001 0.001 38.712 38.716 scf_env_do_scf 11 5.9 0.001 0.001 32.823 32.823 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.022 25.023 velocity_verlet 10 3.0 0.002 0.002 22.802 22.807 dbcsr_multiply_generic 2286 12.5 0.099 0.100 17.799 17.872 qs_scf_new_mos 108 7.5 0.001 0.001 16.175 16.283 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.174 16.283 ot_scf_mini 108 9.5 0.002 0.002 15.088 15.197 multiply_cannon 2286 13.5 0.309 0.323 13.954 14.736 multiply_cannon_loop 2286 14.5 0.342 0.347 12.689 13.517 ot_mini 108 10.5 0.001 0.001 9.066 9.189 multiply_cannon_multrec 9144 15.5 3.448 5.109 8.585 8.789 init_scf_loop 11 6.9 0.000 0.000 7.774 7.776 qs_ot_get_derivative 108 11.5 0.001 0.001 7.034 7.144 rebuild_ks_matrix 119 8.3 0.000 0.000 6.815 6.962 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.815 6.961 prepare_preconditioner 11 7.9 0.000 0.000 6.855 6.869 make_preconditioner 11 8.9 0.000 0.000 6.855 6.869 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.415 6.751 dbcsr_mm_accdrv_process 12550 15.8 4.018 5.591 5.013 6.313 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.160 6.292 cp_fm_upper_to_full 72 14.2 3.182 4.567 3.182 4.567 mp_waitall_1 97218 16.6 2.741 3.888 2.741 3.888 init_scf_run 11 5.9 0.000 0.001 3.870 3.870 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.869 3.869 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.703 3.758 calculate_rho_elec 119 8.7 0.118 0.121 3.702 3.757 sum_up_and_integrate 119 10.3 0.064 0.066 3.666 3.672 integrate_v_rspace 119 11.3 0.003 0.004 3.601 3.609 qs_ot_get_p 119 10.4 0.001 0.001 3.283 3.434 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.760 3.168 dbcsr_complete_redistribute 329 12.2 0.288 0.296 1.983 2.783 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.540 2.593 calculate_first_density_matrix 1 7.0 0.000 0.000 2.408 2.465 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.665 2.465 make_m2s 4572 13.5 0.037 0.038 2.275 2.445 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.171 2.432 apply_single 119 13.6 0.000 0.000 2.170 2.432 mp_alltoall_i22 627 13.8 1.542 2.411 1.542 2.411 make_images 4572 14.5 0.355 0.388 2.156 2.326 calculate_dm_sparse 119 9.5 0.000 0.000 2.202 2.255 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.435 2.228 mp_sum_l 7207 12.9 1.312 2.103 1.312 2.103 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.043 2.043 pw_transfer 1439 11.6 0.066 0.067 2.034 2.037 grid_integrate_task_list 119 12.3 2.010 2.021 2.010 2.021 ot_diis_step 108 11.5 0.014 0.014 2.005 2.006 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.998 1.999 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.939 1.941 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.711 1.766 density_rs2pw 119 9.7 0.003 0.003 1.755 1.764 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.758 1.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.743 jit_kernel_multiply 10 15.5 0.968 1.723 0.968 1.723 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.705 1.705 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.626 1.628 fft3d_ps 1201 14.6 0.569 0.580 1.573 1.575 grid_collocate_task_list 119 9.7 1.449 1.469 1.449 1.469 cp_fm_cholesky_invert 11 10.9 1.445 1.448 1.445 1.448 rs_pw_transfer 974 11.9 0.009 0.009 1.370 1.398 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.394 1.395 cp_fm_diag_elpa_base 50 14.0 1.248 1.299 1.391 1.392 wfi_extrapolate 11 7.9 0.001 0.001 1.349 1.350 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.117 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.261 1.314 make_images_data 4572 15.5 0.043 0.046 1.065 1.281 mp_alltoall_d11v 2130 13.8 1.187 1.257 1.187 1.257 potential_pw2rs 119 12.3 0.014 0.015 1.213 1.215 acc_transpose_blocks 9144 15.5 0.038 0.039 1.030 1.176 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.154 1.175 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.097 1.117 multiply_cannon_sync_h2d 9144 15.5 1.043 1.045 1.043 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.987 1.035 qs_create_task_list 11 7.9 0.001 0.001 0.951 0.962 generate_qs_task_list 11 8.9 0.374 0.394 0.950 0.961 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.833 0.937 acc_transpose_blocks_kernels 9144 16.5 0.117 0.118 0.782 0.926 mp_alltoall_z22v 1201 16.6 0.868 0.888 0.868 0.888 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.867 0.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.678000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=696.181818, yerr=14.646242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 504.025088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 941135. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.123 0.225 84.171 84.173 qs_mol_dyn_low 1 2.0 0.004 0.007 83.355 83.366 qs_forces 11 3.9 0.005 0.024 83.279 83.280 qs_energies 11 4.9 0.004 0.018 80.423 80.433 scf_env_do_scf 11 5.9 0.001 0.001 70.368 70.370 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.724 64.725 dbcsr_multiply_generic 2055 12.4 0.102 0.107 52.369 52.586 qs_scf_new_mos 99 7.5 0.000 0.001 47.876 47.993 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.876 47.992 ot_scf_mini 99 9.5 0.002 0.003 45.457 45.543 multiply_cannon 2055 13.4 0.192 0.206 42.860 43.899 multiply_cannon_loop 2055 14.4 1.562 1.609 41.888 42.854 velocity_verlet 10 3.0 0.001 0.002 41.714 41.715 ot_mini 99 10.5 0.001 0.001 27.742 27.837 qs_ot_get_derivative 99 11.5 0.001 0.002 20.917 20.993 multiply_cannon_multrec 49320 15.4 12.082 12.831 17.413 18.134 rebuild_ks_matrix 110 8.3 0.000 0.001 14.452 14.545 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.024 14.451 14.544 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.666 12.754 mp_waitall_1 241148 16.1 11.706 12.405 11.706 12.405 multiply_cannon_sync_h2d 49320 15.4 9.934 10.549 9.934 10.549 qs_ot_get_p 110 10.4 0.001 0.001 9.260 9.419 multiply_cannon_metrocomm3 49320 15.4 0.077 0.082 6.967 7.944 init_scf_run 11 5.9 0.000 0.001 7.811 7.812 scf_env_initial_rho_setup 11 6.9 0.014 0.029 7.810 7.812 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.067 7.616 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.161 7.611 apply_single 110 13.6 0.000 0.001 7.161 7.610 sum_up_and_integrate 110 10.3 0.038 0.047 7.036 7.056 integrate_v_rspace 110 11.3 0.003 0.004 6.998 7.026 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.742 6.797 ot_diis_step 99 11.5 0.006 0.008 6.526 6.527 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.134 6.260 calculate_rho_elec 110 8.6 0.020 0.024 6.134 6.259 mp_sum_l 6514 12.8 5.230 6.084 5.230 6.084 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.051 6.083 init_scf_loop 11 6.9 0.013 0.102 5.616 5.619 dbcsr_mm_accdrv_process 87628 16.1 2.072 2.183 5.209 5.538 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.068 5.069 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.600 4.601 cp_fm_redistribute_end 48 14.0 4.009 4.578 4.014 4.578 cp_fm_diag_elpa_base 48 14.0 0.559 4.482 0.562 4.505 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.031 4.035 rs_pw_transfer 902 11.9 0.012 0.014 3.832 4.017 wfi_extrapolate 11 7.9 0.001 0.003 3.994 3.995 calculate_dm_sparse 110 9.5 0.000 0.001 3.881 3.989 make_m2s 4110 13.4 0.060 0.066 3.870 3.979 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.676 3.890 make_images 4110 14.4 0.177 0.191 3.775 3.888 calculate_first_density_matrix 1 7.0 0.000 0.001 3.717 3.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.584 3.620 grid_integrate_task_list 110 12.3 3.253 3.394 3.253 3.394 density_rs2pw 110 9.6 0.004 0.004 3.216 3.386 prepare_preconditioner 11 7.9 0.000 0.000 3.339 3.355 make_preconditioner 11 8.9 0.000 0.001 3.339 3.355 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.224 3.304 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.118 3.160 pw_transfer 1331 11.6 0.054 0.063 2.861 2.932 jit_kernel_multiply 13 15.9 2.854 2.909 2.854 2.909 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.774 2.847 potential_pw2rs 110 12.3 0.006 0.007 2.620 2.652 mp_alltoall_d11v 2046 13.8 2.113 2.526 2.113 2.526 fft_wrap_pw1pw2_140 451 13.1 0.170 0.188 2.348 2.420 fft3d_ps 1111 14.6 0.743 0.827 2.310 2.365 mp_sum_d 3883 11.9 1.567 2.364 1.567 2.364 acc_transpose_blocks 49320 15.4 0.213 0.222 2.248 2.331 mp_waitany 14300 13.8 1.947 2.268 1.947 2.268 grid_collocate_task_list 110 9.6 2.084 2.223 2.084 2.223 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.925 1.941 make_images_data 4110 15.4 0.044 0.047 1.774 1.896 cp_fm_cholesky_invert 11 10.9 1.811 1.814 1.811 1.814 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.546 1.807 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.733 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.173000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=3.249921 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 586.133504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1389819. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 68.878 68.879 qs_mol_dyn_low 1 2.0 0.003 0.003 68.602 68.611 qs_forces 11 3.9 0.002 0.002 68.511 68.512 qs_energies 11 4.9 0.001 0.001 65.211 65.216 scf_env_do_scf 11 5.9 0.000 0.001 55.979 55.983 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.362 48.363 dbcsr_multiply_generic 2055 12.4 0.111 0.114 38.453 38.670 velocity_verlet 10 3.0 0.001 0.002 35.900 35.902 multiply_cannon 2055 13.4 0.221 0.248 31.851 33.090 qs_scf_new_mos 99 7.5 0.001 0.001 32.598 32.733 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.597 32.733 multiply_cannon_loop 2055 14.4 0.929 0.951 30.545 31.321 ot_scf_mini 99 9.5 0.003 0.003 30.914 31.026 ot_mini 99 10.5 0.001 0.001 18.291 18.421 multiply_cannon_multrec 24660 15.4 7.653 9.545 13.995 15.811 rebuild_ks_matrix 110 8.3 0.000 0.001 13.572 13.715 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.572 13.715 qs_ot_get_derivative 99 11.5 0.001 0.001 12.456 12.572 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.958 12.083 mp_waitall_1 186928 16.3 8.120 10.321 8.120 10.321 multiply_cannon_sync_h2d 24660 15.4 7.015 8.000 7.015 8.000 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.332 7.843 init_scf_loop 11 6.9 0.000 0.000 7.574 7.574 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.548 7.198 apply_single 110 13.6 0.000 0.001 6.548 7.198 init_scf_run 11 5.9 0.000 0.001 6.895 6.896 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.895 6.896 dbcsr_mm_accdrv_process 52282 16.1 4.747 5.636 6.182 6.531 sum_up_and_integrate 110 10.3 0.053 0.062 6.469 6.482 integrate_v_rspace 110 11.3 0.002 0.003 6.416 6.433 qs_ot_get_p 110 10.4 0.001 0.001 5.995 6.138 ot_diis_step 99 11.5 0.010 0.010 5.784 5.785 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.710 5.726 calculate_rho_elec 110 8.6 0.039 0.047 5.709 5.726 prepare_preconditioner 11 7.9 0.000 0.000 5.577 5.596 make_preconditioner 11 8.9 0.000 0.000 5.576 5.596 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.676 5.391 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.186 5.339 make_m2s 4110 13.4 0.056 0.058 4.198 4.664 make_images 4110 14.4 0.402 0.443 4.095 4.560 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.135 4.156 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.711 3.711 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.673 3.675 wfi_extrapolate 11 7.9 0.001 0.001 3.493 3.493 calculate_dm_sparse 110 9.5 0.001 0.001 3.374 3.412 pw_transfer 1331 11.6 0.065 0.069 3.236 3.376 calculate_first_density_matrix 1 7.0 0.000 0.000 3.317 3.320 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.253 3.310 grid_integrate_task_list 110 12.3 3.148 3.285 3.148 3.285 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.130 3.269 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.171 3.238 density_rs2pw 110 9.6 0.004 0.004 2.986 3.174 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.169 3.171 cp_fm_redistribute_end 48 14.0 2.363 3.143 2.365 3.144 cp_fm_diag_elpa_base 48 14.0 0.744 3.022 0.775 3.107 rs_pw_transfer 902 11.9 0.012 0.013 2.877 3.009 hybrid_alltoall_any 4261 16.3 0.103 0.440 2.103 2.878 make_images_data 4110 15.4 0.048 0.052 2.383 2.857 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.666 2.806 fft3d_ps 1111 14.6 1.077 1.271 2.485 2.613 cp_fm_cholesky_invert 11 10.9 2.599 2.605 2.599 2.605 mp_sum_l 6514 12.8 1.782 2.406 1.782 2.406 potential_pw2rs 110 12.3 0.009 0.009 2.328 2.346 jit_kernel_multiply 11 16.4 1.088 2.307 1.088 2.307 grid_collocate_task_list 110 9.6 2.120 2.252 2.120 2.252 acc_transpose_blocks 24660 15.4 0.106 0.109 1.894 2.215 mp_alltoall_d11v 2046 13.8 1.803 2.103 1.803 2.103 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.949 1.965 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.923 1.935 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.838 1.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.778 1.793 mp_allgather_i34 2055 14.4 0.635 1.702 0.635 1.702 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.798 1.660 acc_transpose_blocks_kernels 24660 16.4 0.309 0.316 1.281 1.601 dbcsr_complete_redistribute 325 12.2 0.285 0.379 1.212 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.372 1.480 mp_irecv_dv 57340 16.2 0.675 1.451 0.675 1.451 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.879000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.545455, yerr=6.867097 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 658.657280E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.126 60.739 60.740 qs_mol_dyn_low 1 2.0 0.004 0.006 60.154 60.170 qs_forces 11 3.9 0.003 0.007 60.045 60.047 qs_energies 11 4.9 0.023 0.168 56.876 56.880 scf_env_do_scf 11 5.9 0.001 0.001 48.930 48.951 scf_env_do_scf_inner_loop 99 6.5 0.095 0.373 39.862 39.865 velocity_verlet 10 3.0 0.001 0.002 32.354 32.357 dbcsr_multiply_generic 2055 12.4 0.105 0.107 28.277 28.516 qs_scf_new_mos 99 7.5 0.001 0.001 24.983 25.077 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.983 25.076 ot_scf_mini 99 9.5 0.003 0.003 23.767 23.873 multiply_cannon 2055 13.4 0.216 0.239 21.986 23.255 multiply_cannon_loop 2055 14.4 0.619 0.634 20.796 21.799 ot_mini 99 10.5 0.001 0.001 13.442 13.554 rebuild_ks_matrix 110 8.3 0.000 0.000 12.214 12.331 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.019 12.214 12.331 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.947 11.108 multiply_cannon_multrec 16440 15.4 4.013 5.075 9.770 10.742 mp_waitall_1 146766 16.3 7.010 9.902 7.010 9.902 qs_ot_get_derivative 99 11.5 0.001 0.002 9.026 9.134 init_scf_loop 11 6.9 0.054 0.214 9.034 9.058 prepare_preconditioner 11 7.9 0.000 0.000 6.902 6.920 make_preconditioner 11 8.9 0.000 0.002 6.901 6.920 multiply_cannon_metrocomm3 16440 15.4 0.042 0.043 4.105 6.860 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.205 6.557 sum_up_and_integrate 110 10.3 0.061 0.065 6.440 6.454 integrate_v_rspace 110 11.3 0.003 0.004 6.379 6.392 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.799 5.900 calculate_rho_elec 110 8.6 0.058 0.058 5.798 5.900 dbcsr_mm_accdrv_process 34862 16.1 4.383 4.946 5.608 5.710 qs_ot_get_p 110 10.4 0.001 0.002 5.532 5.661 init_scf_run 11 5.9 0.000 0.001 5.485 5.486 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.485 5.485 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.929 5.367 apply_single 110 13.6 0.000 0.000 4.928 5.367 make_m2s 4110 13.4 0.049 0.051 4.232 4.590 make_images 4110 14.4 0.394 0.511 4.121 4.480 ot_diis_step 99 11.5 0.011 0.012 4.387 4.388 multiply_cannon_sync_h2d 16440 15.4 3.744 4.269 3.744 4.269 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.875 3.879 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.103 3.740 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.533 3.533 grid_integrate_task_list 110 12.3 3.201 3.443 3.201 3.443 pw_transfer 1331 11.6 0.065 0.070 3.193 3.199 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.086 3.095 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.001 3.002 wfi_extrapolate 11 7.9 0.001 0.002 2.973 2.973 cp_fm_redistribute_end 48 14.0 1.864 2.970 1.866 2.972 make_images_data 4110 15.4 0.044 0.048 2.485 2.942 cp_fm_diag_elpa_base 48 14.0 1.040 2.831 1.100 2.938 density_rs2pw 110 9.6 0.004 0.004 2.803 2.931 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.252 2.903 rs_pw_transfer 902 11.9 0.011 0.013 2.722 2.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.773 2.775 cp_fm_cholesky_invert 11 10.9 2.705 2.710 2.705 2.710 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.624 2.634 mp_alltoall_d11v 2046 13.8 2.093 2.565 2.093 2.565 calculate_dm_sparse 110 9.5 0.001 0.001 2.524 2.557 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.426 2.469 calculate_first_density_matrix 1 7.0 0.001 0.010 2.427 2.429 grid_collocate_task_list 110 9.6 2.176 2.422 2.176 2.422 fft3d_ps 1111 14.6 1.064 1.073 2.401 2.413 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.318 2.377 multiply_cannon_metrocomm4 14385 15.4 0.046 0.048 0.852 2.277 potential_pw2rs 110 12.3 0.010 0.011 2.258 2.266 mp_irecv_dv 48980 15.7 0.781 2.149 0.781 2.149 mp_sum_l 6514 12.8 1.512 2.135 1.512 2.135 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.001 2.004 dbcsr_complete_redistribute 325 12.2 0.329 0.381 1.486 1.949 cp_fm_upper_to_full 70 13.6 1.402 1.879 1.402 1.879 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.633 1.647 cp_fm_cholesky_decompose 22 10.9 1.614 1.640 1.614 1.640 mp_allgather_i34 2055 14.4 0.508 1.608 0.508 1.608 jit_kernel_multiply 8 16.4 0.840 1.529 0.840 1.529 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.496 1.508 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.345 1.483 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.005 1.457 mp_waitany 17072 13.8 1.191 1.387 1.191 1.387 rs_gather_matrices 110 12.3 0.137 0.151 0.871 1.265 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.239 1.245 mp_alltoall_i22 605 13.7 0.698 1.230 0.698 1.230 acc_transpose_blocks 16440 15.4 0.073 0.074 1.199 1.215 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.740000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=620.181818, yerr=9.173263 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 736.768000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.058 0.132 66.764 66.765 qs_mol_dyn_low 1 2.0 0.003 0.005 66.303 66.312 qs_forces 11 3.9 0.002 0.006 66.135 66.135 qs_energies 11 4.9 0.002 0.005 62.771 62.775 scf_env_do_scf 11 5.9 0.001 0.001 53.821 53.824 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.877 41.879 velocity_verlet 10 3.0 0.001 0.002 37.091 37.097 dbcsr_multiply_generic 2055 12.4 0.116 0.133 30.783 31.089 qs_scf_new_mos 99 7.5 0.001 0.001 27.416 27.523 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.416 27.522 ot_scf_mini 99 9.5 0.003 0.003 25.783 25.880 multiply_cannon 2055 13.4 0.244 0.267 23.721 24.663 multiply_cannon_loop 2055 14.4 0.879 0.900 22.167 22.892 ot_mini 99 10.5 0.001 0.001 15.007 15.120 multiply_cannon_multrec 24660 15.4 4.278 6.746 13.294 14.425 rebuild_ks_matrix 110 8.3 0.000 0.000 11.983 12.092 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 11.982 12.091 init_scf_loop 11 6.9 0.002 0.006 11.900 11.903 qs_ot_get_derivative 99 11.5 0.001 0.002 10.826 10.921 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.616 10.709 prepare_preconditioner 11 7.9 0.000 0.000 10.148 10.162 make_preconditioner 11 8.9 0.000 0.001 10.148 10.162 dbcsr_mm_accdrv_process 52304 16.0 6.797 8.195 8.867 9.995 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.350 9.828 mp_waitall_1 126806 16.4 4.918 7.089 4.918 7.089 sum_up_and_integrate 110 10.3 0.069 0.073 6.382 6.396 integrate_v_rspace 110 11.3 0.003 0.003 6.313 6.329 init_scf_run 11 5.9 0.000 0.001 6.044 6.045 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.044 6.045 qs_ot_get_p 110 10.4 0.001 0.001 5.743 5.871 make_m2s 4110 13.4 0.059 0.060 5.418 5.831 make_images 4110 14.4 0.580 0.699 5.278 5.688 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.620 5.629 calculate_rho_elec 110 8.6 0.077 0.081 5.620 5.628 cp_fm_upper_to_full 70 13.8 3.339 4.752 3.339 4.752 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.042 4.146 apply_single 110 13.6 0.000 0.000 4.042 4.146 ot_diis_step 99 11.5 0.011 0.013 4.140 4.140 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.786 4.091 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.918 3.973 qs_ot_p2m_diag 48 11.0 0.055 0.065 3.917 3.933 dbcsr_complete_redistribute 325 12.2 0.422 0.488 2.773 3.929 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.430 3.430 grid_integrate_task_list 110 12.3 3.285 3.428 3.285 3.428 multiply_cannon_sync_h2d 24660 15.4 3.193 3.377 3.193 3.377 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.210 3.354 pw_transfer 1331 11.6 0.064 0.071 3.273 3.298 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.168 3.195 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.149 3.150 make_images_data 4110 15.4 0.046 0.050 2.706 3.149 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.074 3.129 calculate_dm_sparse 110 9.5 0.001 0.001 3.075 3.103 hybrid_alltoall_any 4261 16.3 0.121 0.458 2.346 3.103 wfi_extrapolate 11 7.9 0.001 0.006 3.006 3.006 calculate_first_density_matrix 1 7.0 0.000 0.001 2.945 2.948 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.789 2.910 density_rs2pw 110 9.6 0.004 0.004 2.778 2.905 mp_alltoall_i22 605 13.7 1.681 2.895 1.681 2.895 cp_fm_cholesky_invert 11 10.9 2.851 2.863 2.851 2.863 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.855 2.858 cp_fm_redistribute_end 48 14.0 1.424 2.824 1.426 2.825 cp_fm_diag_elpa_base 48 14.0 1.317 2.683 1.396 2.797 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.740 2.782 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.665 2.696 rs_pw_transfer 902 11.9 0.010 0.011 2.512 2.633 jit_kernel_multiply 12 15.8 1.740 2.605 1.740 2.605 fft3d_ps 1111 14.6 1.065 1.094 2.470 2.488 qs_energies_init_hamiltonians 11 5.9 0.034 0.040 2.382 2.388 grid_collocate_task_list 110 9.6 2.225 2.383 2.225 2.383 potential_pw2rs 110 12.3 0.013 0.013 2.150 2.163 mp_alltoall_d11v 2046 13.8 1.808 1.956 1.808 1.956 cp_fm_cholesky_decompose 22 10.9 1.714 1.771 1.714 1.771 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.727 1.763 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.706 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.684 1.700 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.614 1.629 acc_transpose_blocks 24660 15.4 0.104 0.105 1.582 1.621 mp_sum_l 6514 12.8 0.962 1.577 0.962 1.577 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.832 1.435 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.305 1.404 mp_irecv_dv 62702 16.1 0.734 1.360 0.734 1.360 mp_allgather_i34 2055 14.4 0.649 1.354 0.649 1.354 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.765000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.636364, yerr=14.290955 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 829.607936E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.034 56.111 56.112 qs_mol_dyn_low 1 2.0 0.003 0.003 55.849 55.858 qs_forces 11 3.9 0.002 0.002 55.564 55.565 qs_energies 11 4.9 0.001 0.001 51.959 51.964 scf_env_do_scf 11 5.9 0.001 0.001 43.568 43.568 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.723 35.724 velocity_verlet 10 3.0 0.199 0.229 31.107 31.109 dbcsr_multiply_generic 2055 12.4 0.103 0.105 23.940 24.081 qs_scf_new_mos 99 7.5 0.001 0.001 21.525 21.586 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.524 21.586 ot_scf_mini 99 9.5 0.002 0.002 20.278 20.334 multiply_cannon 2055 13.4 0.247 0.260 18.331 19.623 multiply_cannon_loop 2055 14.4 0.322 0.335 16.964 17.197 rebuild_ks_matrix 110 8.3 0.000 0.000 11.496 11.538 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.496 11.538 ot_mini 99 10.5 0.001 0.001 11.283 11.332 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.253 10.290 multiply_cannon_multrec 8220 15.4 3.234 4.553 8.294 9.467 mp_waitall_1 106626 16.5 6.426 8.066 6.426 8.066 init_scf_loop 11 6.9 0.000 0.000 7.797 7.799 qs_ot_get_derivative 99 11.5 0.001 0.001 7.385 7.442 sum_up_and_integrate 110 10.3 0.080 0.082 6.200 6.211 prepare_preconditioner 11 7.9 0.000 0.000 6.143 6.152 make_preconditioner 11 8.9 0.000 0.000 6.143 6.152 integrate_v_rspace 110 11.3 0.003 0.003 6.119 6.130 dbcsr_mm_accdrv_process 17442 15.9 2.864 4.574 4.928 5.857 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.718 5.791 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.659 5.673 calculate_rho_elec 110 8.6 0.115 0.115 5.659 5.672 init_scf_run 11 5.9 0.000 0.001 5.263 5.263 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.263 5.263 qs_ot_get_p 110 10.4 0.001 0.001 5.034 5.101 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.154 4.657 make_m2s 4110 13.4 0.038 0.040 4.289 4.556 make_images 4110 14.4 0.640 0.690 4.162 4.430 ot_diis_step 99 11.5 0.012 0.012 3.873 3.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.788 3.830 apply_single 110 13.6 0.000 0.000 3.788 3.830 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.663 3.667 grid_integrate_task_list 110 12.3 3.373 3.507 3.373 3.507 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.346 3.347 pw_transfer 1331 11.6 0.065 0.068 3.273 3.283 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.166 3.181 multiply_cannon_sync_h2d 8220 15.4 2.900 3.009 2.900 3.009 cp_fm_cholesky_invert 11 10.9 2.980 2.984 2.980 2.984 make_images_data 4110 15.4 0.038 0.043 2.455 2.899 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.792 2.793 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.394 2.790 density_rs2pw 110 9.6 0.004 0.004 2.695 2.790 cp_fm_redistribute_end 48 14.0 0.709 2.761 0.713 2.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.753 2.754 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.709 2.729 cp_fm_diag_elpa_base 48 14.0 1.867 2.561 2.039 2.724 wfi_extrapolate 11 7.9 0.001 0.001 2.691 2.691 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.673 2.674 jit_kernel_multiply 10 15.9 1.755 2.575 1.755 2.575 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.533 2.566 calculate_dm_sparse 110 9.5 0.001 0.001 2.485 2.535 calculate_first_density_matrix 1 7.0 0.000 0.000 2.481 2.482 fft3d_ps 1111 14.6 1.122 1.170 2.423 2.443 grid_collocate_task_list 110 9.6 2.329 2.426 2.329 2.426 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.377 2.394 rs_pw_transfer 902 11.9 0.010 0.010 2.207 2.328 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.773 1.996 potential_pw2rs 110 12.3 0.015 0.016 1.930 1.933 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.860 1.888 mp_alltoall_d11v 2046 13.8 1.729 1.887 1.729 1.887 cp_fm_cholesky_decompose 22 10.9 1.692 1.711 1.692 1.711 mp_allgather_i34 2055 14.4 0.552 1.678 0.552 1.678 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.506 1.629 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.578 1.583 dbcsr_complete_redistribute 325 12.2 0.558 0.580 1.477 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.441 1.457 qs_create_task_list 11 7.9 0.001 0.001 1.234 1.334 generate_qs_task_list 11 8.9 0.380 0.447 1.233 1.334 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.477 1.329 mp_irecv_dv 24056 15.7 0.453 1.287 0.453 1.287 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.834 1.208 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.170 1.201 mp_waitany 9240 13.8 1.059 1.200 1.059 1.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.112000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=783.636364, yerr=10.890135 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 1.328427E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 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 0.254 0.290 86.028 86.029 qs_mol_dyn_low 1 2.0 0.003 0.003 84.773 84.782 qs_forces 11 3.9 0.002 0.002 84.708 84.709 qs_energies 11 4.9 0.001 0.001 80.645 80.646 scf_env_do_scf 11 5.9 0.001 0.001 70.796 70.796 velocity_verlet 10 3.0 0.002 0.002 54.620 54.626 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.606 42.608 dbcsr_multiply_generic 2055 12.4 0.118 0.120 28.785 28.929 init_scf_loop 11 6.9 0.000 0.000 28.117 28.120 qs_scf_new_mos 99 7.5 0.001 0.001 26.266 26.348 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.265 26.348 prepare_preconditioner 11 7.9 0.000 0.000 26.187 26.198 make_preconditioner 11 8.9 0.000 0.000 26.187 26.198 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.414 25.645 ot_scf_mini 99 9.5 0.002 0.002 24.510 24.585 multiply_cannon 2055 13.4 0.344 0.365 21.480 22.184 multiply_cannon_loop 2055 14.4 0.340 0.343 19.640 20.016 cp_fm_upper_to_full 70 14.2 12.669 18.143 12.669 18.143 ot_mini 99 10.5 0.001 0.001 13.830 13.903 rebuild_ks_matrix 110 8.3 0.001 0.001 13.124 13.212 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.123 13.211 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.936 12.016 dbcsr_complete_redistribute 325 12.2 1.026 1.047 7.344 10.585 multiply_cannon_multrec 8220 15.4 4.370 4.568 9.577 9.687 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.321 9.561 qs_ot_get_derivative 99 11.5 0.001 0.001 9.133 9.207 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.758 8.956 mp_waitall_1 87304 16.6 7.758 8.834 7.758 8.834 mp_alltoall_i22 605 13.7 5.380 8.652 5.380 8.652 sum_up_and_integrate 110 10.3 0.151 0.153 6.624 6.639 integrate_v_rspace 110 11.3 0.004 0.004 6.473 6.488 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.417 6.454 calculate_rho_elec 110 8.6 0.227 0.227 6.416 6.454 make_m2s 4110 13.4 0.043 0.043 5.295 5.859 make_images 4110 14.4 0.881 0.932 5.109 5.674 init_scf_run 11 5.9 0.000 0.001 5.627 5.627 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.626 5.626 qs_ot_get_p 110 10.4 0.001 0.001 5.485 5.576 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.972 5.533 apply_single 110 13.6 0.000 0.000 4.971 5.533 dbcsr_mm_accdrv_process 11614 15.7 3.286 3.643 5.063 5.277 cp_fm_cholesky_invert 11 10.9 5.223 5.228 5.223 5.228 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.723 5.006 ot_diis_step 99 11.5 0.015 0.016 4.677 4.677 multiply_cannon_sync_h2d 8220 15.4 3.952 3.957 3.952 3.957 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.915 3.923 pw_transfer 1331 11.6 0.074 0.074 3.736 3.743 grid_integrate_task_list 110 12.3 3.656 3.715 3.656 3.715 hybrid_alltoall_any 4261 16.3 0.257 0.553 2.921 3.667 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.636 3.637 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.619 3.627 make_images_data 4110 15.4 0.041 0.044 2.923 3.592 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.862 3.541 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.477 3.477 wfi_extrapolate 11 7.9 0.001 0.001 3.209 3.209 calculate_dm_sparse 110 9.5 0.001 0.001 3.145 3.169 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.137 3.145 density_rs2pw 110 9.6 0.004 0.004 2.918 2.931 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.888 2.889 cp_fm_diag_elpa_base 48 14.0 2.368 2.557 2.886 2.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.853 2.859 fft3d_ps 1111 14.6 1.265 1.275 2.828 2.835 grid_collocate_task_list 110 9.6 2.626 2.648 2.626 2.648 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.374 2.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.334 2.364 calculate_first_density_matrix 1 7.0 0.000 0.000 2.314 2.315 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.173 2.230 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.101 2.175 rs_pw_transfer 902 11.9 0.011 0.011 2.097 2.133 mp_alltoall_d11v 2046 13.8 2.044 2.085 2.044 2.085 potential_pw2rs 110 12.3 0.022 0.022 2.062 2.067 cp_fm_cholesky_decompose 22 10.9 1.983 1.997 1.983 1.997 jit_kernel_multiply 10 15.3 1.573 1.986 1.573 1.986 qs_create_task_list 11 7.9 0.001 0.001 1.903 1.947 generate_qs_task_list 11 8.9 0.740 0.793 1.902 1.947 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.766 1.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.029000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1204.181818, yerr=50.551177 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/15/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 1.094965E+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 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 626.511872E+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 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 3236688. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.060 203.225 203.225 qs_mol_dyn_low 1 2.0 0.003 0.003 202.818 202.832 qs_forces 11 3.9 0.003 0.004 202.733 202.734 qs_energies 11 4.9 0.001 0.002 197.266 197.278 scf_env_do_scf 11 5.9 0.001 0.001 179.866 179.870 scf_env_do_scf_inner_loop 117 6.6 0.002 0.007 159.292 159.294 dbcsr_multiply_generic 2507 12.6 0.196 0.230 124.896 125.679 velocity_verlet 10 3.0 0.004 0.009 121.617 121.617 qs_scf_new_mos 117 7.6 0.001 0.001 120.754 121.041 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.754 121.040 ot_scf_mini 117 9.6 0.003 0.003 114.212 114.501 multiply_cannon 2507 13.6 0.239 0.246 101.148 103.291 multiply_cannon_loop 2507 14.6 2.109 2.176 98.977 100.912 ot_mini 117 10.6 0.001 0.001 65.426 65.708 multiply_cannon_multrec 60168 15.6 33.261 35.368 41.961 43.649 qs_ot_get_derivative 117 11.6 0.001 0.001 40.416 40.696 rebuild_ks_matrix 128 8.3 0.001 0.001 33.426 33.823 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 33.425 33.823 mp_waitall_1 291448 16.2 28.529 32.188 28.529 32.188 multiply_cannon_sync_h2d 60168 15.6 27.518 30.521 27.518 30.521 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.115 30.464 qs_ot_get_p 128 10.4 0.001 0.001 26.928 27.235 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.458 25.326 apply_single 128 13.6 0.001 0.001 24.458 25.326 ot_diis_step 117 11.6 0.007 0.008 24.676 24.678 init_scf_loop 11 6.9 0.000 0.000 20.499 20.500 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.140 20.237 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.751 18.972 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 15.677 17.936 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.554 17.556 prepare_preconditioner 11 7.9 0.000 0.000 15.691 15.737 make_preconditioner 11 8.9 0.000 0.000 15.691 15.737 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.913 15.087 make_m2s 5014 13.6 0.101 0.111 14.457 14.758 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.736 14.741 cp_fm_redistribute_end 83 14.4 11.643 14.670 11.655 14.672 make_images 5014 14.6 0.405 0.424 14.286 14.591 cp_fm_diag_elpa_base 83 14.4 2.975 14.376 3.007 14.489 sum_up_and_integrate 128 10.3 0.091 0.111 13.855 13.876 integrate_v_rspace 128 11.3 0.004 0.004 13.764 13.788 init_scf_run 11 5.9 0.000 0.001 13.314 13.315 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.314 13.315 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.522 12.615 calculate_rho_elec 128 8.7 0.045 0.063 12.521 12.614 mp_sum_l 7870 13.0 8.027 9.342 8.027 9.342 wfi_extrapolate 11 7.9 0.001 0.001 9.069 9.069 cp_fm_cholesky_invert 11 10.9 8.834 8.843 8.834 8.843 dbcsr_mm_accdrv_process 124484 16.2 3.288 3.456 8.258 8.831 calculate_dm_sparse 128 9.5 0.001 0.001 8.706 8.818 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.978 8.102 make_images_data 5014 15.6 0.073 0.079 7.027 8.045 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.784 7.870 multiply_cannon_metrocomm1 60168 15.6 0.087 0.090 5.817 7.828 grid_integrate_task_list 128 12.3 7.067 7.475 7.067 7.475 hybrid_alltoall_any 5200 16.5 0.293 2.271 6.059 7.413 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.376 7.383 pw_transfer 1547 11.6 0.075 0.105 6.763 7.032 density_rs2pw 128 9.7 0.006 0.007 6.374 6.975 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.560 6.808 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.482 6.559 rs_pw_transfer 1046 11.9 0.018 0.019 5.559 6.191 fft_wrap_pw1pw2_140 523 13.2 0.447 0.515 5.672 5.844 fft3d_ps 1291 14.7 2.090 2.559 5.350 5.536 grid_collocate_task_list 128 9.7 4.724 5.165 4.724 5.165 mp_alltoall_d11v 2415 14.1 4.164 5.058 4.164 5.058 cp_fm_cholesky_decompose 22 10.9 4.637 4.651 4.637 4.651 mp_sum_d 4464 12.1 3.719 4.526 3.719 4.526 potential_pw2rs 128 12.3 0.010 0.011 4.383 4.398 calculate_first_density_matrix 1 7.0 0.000 0.000 4.087 4.094 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.225000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.272727, yerr=6.675476 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 825.180160E+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 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 6687726. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 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 0.042 0.073 194.584 194.585 qs_mol_dyn_low 1 2.0 0.003 0.003 194.061 194.074 qs_forces 11 3.9 0.003 0.003 193.797 193.798 qs_energies 11 4.9 0.051 0.163 187.051 187.063 scf_env_do_scf 11 5.9 0.001 0.001 169.873 169.882 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.006 136.008 velocity_verlet 10 3.0 0.001 0.002 122.049 122.054 dbcsr_multiply_generic 2527 12.6 0.184 0.189 99.029 100.158 qs_scf_new_mos 118 7.6 0.001 0.001 96.408 97.037 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.407 97.037 ot_scf_mini 118 9.6 0.004 0.004 91.543 92.234 multiply_cannon 2527 13.6 0.483 0.531 78.534 83.325 multiply_cannon_loop 2527 14.6 1.264 1.305 75.074 77.441 ot_mini 118 10.6 0.001 0.001 50.552 51.165 mp_waitall_1 228564 16.4 25.727 40.541 25.727 40.541 multiply_cannon_multrec 30324 15.6 22.084 26.521 32.065 37.000 rebuild_ks_matrix 129 8.3 0.001 0.001 33.578 34.148 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 33.578 34.147 init_scf_loop 11 6.9 0.000 0.000 33.776 33.777 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.340 30.870 prepare_preconditioner 11 7.9 0.000 0.000 29.095 29.165 make_preconditioner 11 8.9 0.000 0.000 29.094 29.165 qs_ot_get_derivative 118 11.6 0.001 0.002 28.381 29.061 multiply_cannon_metrocomm3 30324 15.6 0.098 0.105 16.063 29.042 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.772 28.321 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.188 23.354 apply_single 129 13.6 0.001 0.001 22.187 23.353 qs_ot_get_p 129 10.4 0.001 0.001 22.492 23.323 multiply_cannon_sync_h2d 30324 15.6 19.411 22.139 19.411 22.139 ot_diis_step 118 11.6 0.014 0.015 21.998 22.000 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.531 17.575 cp_fm_cholesky_invert 11 10.9 17.036 17.050 17.036 17.050 make_m2s 5054 13.6 0.090 0.095 14.619 16.335 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.331 16.332 make_images 5054 14.6 1.175 1.377 14.420 16.138 sum_up_and_integrate 129 10.3 0.118 0.133 14.629 14.657 integrate_v_rspace 129 11.3 0.003 0.004 14.511 14.540 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.399 13.432 calculate_rho_elec 129 8.7 0.088 0.106 13.399 13.431 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.136 13.146 cp_fm_redistribute_end 83 14.4 7.687 13.062 7.700 13.063 cp_fm_diag_elpa_base 83 14.4 5.125 12.583 5.341 12.936 init_scf_run 11 5.9 0.000 0.001 11.878 11.879 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.877 11.879 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.139 11.605 make_images_data 5054 15.6 0.069 0.076 8.901 10.958 multiply_cannon_metrocomm4 27797 15.6 0.098 0.114 3.818 10.687 mp_irecv_dv 70031 16.3 3.622 10.289 3.622 10.289 hybrid_alltoall_any 5240 16.5 0.347 1.525 7.471 10.151 dbcsr_mm_accdrv_process 62734 16.2 4.615 5.908 9.434 10.053 wfi_extrapolate 11 7.9 0.001 0.001 8.449 8.449 pw_transfer 1559 11.6 0.086 0.112 8.031 8.106 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.805 7.872 grid_integrate_task_list 129 12.3 7.224 7.750 7.224 7.750 density_rs2pw 129 9.7 0.006 0.006 7.087 7.550 cp_fm_cholesky_decompose 22 10.9 7.141 7.248 7.141 7.248 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.456 7.233 calculate_dm_sparse 129 9.5 0.001 0.001 6.805 6.964 fft_wrap_pw1pw2_140 527 13.2 0.473 0.515 6.756 6.825 mp_sum_l 7930 13.1 4.088 6.471 4.088 6.471 rs_pw_transfer 1054 12.0 0.015 0.017 5.862 6.337 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.253 6.265 fft3d_ps 1301 14.7 2.800 2.961 6.162 6.222 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.338 5.489 grid_collocate_task_list 129 9.7 4.948 5.481 4.948 5.481 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.382 5.442 mp_allgather_i34 2527 14.6 2.022 5.391 2.022 5.391 potential_pw2rs 129 12.3 0.015 0.016 4.963 4.990 mp_alltoall_d11v 2423 14.1 4.336 4.840 4.336 4.840 mp_sum_d 4493 12.2 2.834 4.283 2.834 4.283 dbcsr_complete_redistribute 395 12.7 0.782 0.909 3.296 4.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.585000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.000000, yerr=2.593699 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 936.476672E+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 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 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 0.035 0.058 182.573 182.574 qs_mol_dyn_low 1 2.0 0.003 0.003 182.041 182.054 qs_forces 11 3.9 0.003 0.003 181.943 181.946 qs_energies 11 4.9 0.124 0.142 175.364 175.380 scf_env_do_scf 11 5.9 0.001 0.001 158.856 158.858 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 122.146 122.147 velocity_verlet 10 3.0 0.002 0.002 116.179 116.181 qs_scf_new_mos 117 7.6 0.001 0.001 84.844 85.199 qs_scf_loop_do_ot 117 8.6 0.001 0.001 84.843 85.198 dbcsr_multiply_generic 2507 12.6 0.179 0.185 83.374 84.586 ot_scf_mini 117 9.6 0.003 0.004 80.653 81.071 multiply_cannon 2507 13.6 0.500 0.525 62.750 67.753 multiply_cannon_loop 2507 14.6 0.860 0.887 59.441 62.120 ot_mini 117 10.6 0.001 0.001 43.766 44.194 init_scf_loop 11 6.9 0.000 0.000 36.605 36.606 mp_waitall_1 178456 16.5 26.542 36.180 26.542 36.180 prepare_preconditioner 11 7.9 0.000 0.000 32.276 32.325 make_preconditioner 11 8.9 0.000 0.000 32.276 32.325 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.874 31.283 rebuild_ks_matrix 128 8.3 0.001 0.001 30.754 31.264 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.753 31.263 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.824 28.303 multiply_cannon_multrec 20056 15.6 13.421 16.408 22.503 25.545 multiply_cannon_metrocomm3 20056 15.6 0.059 0.064 15.679 24.854 qs_ot_get_derivative 117 11.6 0.001 0.002 23.685 24.103 qs_ot_get_p 128 10.4 0.001 0.001 22.145 22.659 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.146 21.275 apply_single 128 13.6 0.001 0.001 20.146 21.274 ot_diis_step 117 11.6 0.018 0.020 19.977 19.977 qs_ot_p2m_diag 83 11.4 0.265 0.272 17.557 17.573 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.485 16.488 make_m2s 5014 13.6 0.082 0.085 15.228 16.293 make_images 5014 14.6 1.199 1.294 15.002 16.067 multiply_cannon_sync_h2d 20056 15.6 14.236 15.850 14.236 15.850 cp_fm_cholesky_invert 11 10.9 15.253 15.263 15.253 15.263 sum_up_and_integrate 128 10.3 0.135 0.148 14.495 14.521 integrate_v_rspace 128 11.3 0.004 0.004 14.360 14.384 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.407 13.430 calculate_rho_elec 128 8.7 0.132 0.147 13.406 13.429 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.121 13.124 cp_fm_redistribute_end 83 14.4 4.988 13.032 5.007 13.036 cp_fm_diag_elpa_base 83 14.4 7.595 12.354 8.001 12.896 init_scf_run 11 5.9 0.000 0.001 11.007 11.008 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.007 11.008 make_images_data 5014 15.6 0.062 0.069 9.410 10.895 hybrid_alltoall_any 5200 16.5 0.432 1.960 8.190 9.951 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.374 9.665 multiply_cannon_metrocomm4 17549 15.6 0.063 0.073 3.477 9.355 mp_irecv_dv 50230 16.2 3.353 9.103 3.353 9.103 dbcsr_mm_accdrv_process 41502 16.2 4.801 5.557 8.546 8.649 pw_transfer 1547 11.6 0.085 0.105 8.011 8.127 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.787 7.907 cp_fm_cholesky_decompose 22 10.9 7.781 7.831 7.781 7.831 grid_integrate_task_list 128 12.3 7.312 7.725 7.312 7.725 wfi_extrapolate 11 7.9 0.001 0.001 7.582 7.582 cp_fm_upper_to_full 105 14.5 5.786 7.509 5.786 7.509 density_rs2pw 128 9.7 0.006 0.006 6.808 7.004 fft_wrap_pw1pw2_140 523 13.2 0.477 0.531 6.748 6.870 dbcsr_complete_redistribute 395 12.7 1.179 1.223 4.860 6.675 calculate_dm_sparse 128 9.5 0.001 0.001 6.209 6.339 fft3d_ps 1291 14.7 2.696 2.908 6.053 6.134 rs_pw_transfer 1046 11.9 0.014 0.015 5.503 5.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.551 5.556 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.718 5.447 grid_collocate_task_list 128 9.7 5.069 5.422 5.069 5.422 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.589 5.395 mp_alltoall_d11v 2415 14.1 4.599 5.134 4.599 5.134 mp_sum_l 7870 13.0 3.576 4.933 3.576 4.933 potential_pw2rs 128 12.3 0.021 0.023 4.813 4.838 mp_allgather_i34 2507 14.6 1.826 4.726 1.826 4.726 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.541 4.679 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.382 4.138 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.072 4.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.065 4.066 mp_alltoall_i22 716 14.1 2.025 4.015 2.025 4.015 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=182.574000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.636364, yerr=11.818881 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 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 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.142280E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57301. MP_Allreduce 11231 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.140 191.288 191.289 qs_mol_dyn_low 1 2.0 0.003 0.006 189.923 189.935 qs_forces 11 3.9 0.005 0.018 189.657 189.666 qs_energies 11 4.9 0.002 0.004 182.591 182.605 scf_env_do_scf 11 5.9 0.001 0.001 165.091 165.103 velocity_verlet 10 3.0 0.002 0.006 125.366 125.374 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.334 117.335 qs_scf_new_mos 116 7.6 0.001 0.001 81.462 81.786 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.461 81.785 dbcsr_multiply_generic 2485 12.5 0.186 0.190 79.372 80.138 ot_scf_mini 116 9.6 0.003 0.004 76.988 77.278 multiply_cannon 2485 13.5 0.552 0.582 54.730 57.299 multiply_cannon_loop 2485 14.5 1.170 1.201 51.024 52.864 init_scf_loop 11 6.9 0.000 0.001 47.614 47.617 prepare_preconditioner 11 7.9 0.000 0.000 43.516 43.544 make_preconditioner 11 8.9 0.000 0.001 43.516 43.544 ot_mini 116 10.6 0.001 0.002 42.247 42.540 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.988 42.106 multiply_cannon_multrec 29820 15.5 13.956 19.250 25.989 31.069 rebuild_ks_matrix 127 8.3 0.001 0.001 29.170 29.449 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.024 29.170 29.449 mp_waitall_1 152434 16.5 17.979 28.655 17.979 28.655 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.331 26.581 qs_ot_get_derivative 116 11.6 0.001 0.002 22.774 23.073 make_m2s 4970 13.5 0.097 0.102 20.304 21.408 qs_ot_get_p 127 10.4 0.001 0.002 21.015 21.333 make_images 4970 14.5 1.947 2.303 20.002 21.111 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.892 19.526 apply_single 127 13.6 0.001 0.001 18.892 19.526 ot_diis_step 116 11.6 0.018 0.026 19.363 19.369 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.789 16.842 cp_fm_cholesky_invert 11 10.9 16.821 16.831 16.821 16.831 cp_fm_upper_to_full 104 14.7 11.267 16.741 11.267 16.741 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.508 15.509 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.624 15.387 sum_up_and_integrate 127 10.3 0.141 0.155 14.231 14.260 integrate_v_rspace 127 11.3 0.004 0.004 14.090 14.124 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.396 13.430 calculate_rho_elec 127 8.7 0.174 0.189 13.396 13.429 dbcsr_complete_redistribute 393 12.7 1.522 1.672 9.382 13.208 make_images_data 4970 15.5 0.065 0.069 10.973 12.813 multiply_cannon_sync_h2d 29820 15.5 11.644 12.629 11.644 12.629 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.341 12.344 cp_fm_redistribute_end 82 14.4 2.120 12.241 2.139 12.247 dbcsr_mm_accdrv_process 61748 16.2 7.422 8.508 11.610 12.182 cp_fm_diag_elpa_base 82 14.4 9.470 11.619 10.065 12.091 hybrid_alltoall_any 5155 16.4 0.523 2.189 9.876 11.925 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.866 11.702 init_scf_run 11 5.9 0.000 0.001 11.199 11.201 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.199 11.201 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.506 10.238 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.372 9.579 mp_alltoall_i22 712 14.1 5.701 9.543 5.701 9.543 pw_transfer 1535 11.6 0.084 0.100 8.096 8.168 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.873 7.947 cp_fm_cholesky_decompose 22 10.9 7.855 7.943 7.855 7.943 wfi_extrapolate 11 7.9 0.001 0.001 7.807 7.807 grid_integrate_task_list 127 12.3 7.507 7.755 7.507 7.755 multiply_cannon_metrocomm4 24850 15.5 0.075 0.085 2.750 7.106 fft_wrap_pw1pw2_140 519 13.2 0.476 0.486 6.894 6.984 mp_irecv_dv 75445 16.2 2.607 6.848 2.607 6.848 density_rs2pw 127 9.7 0.005 0.006 6.468 6.798 calculate_dm_sparse 127 9.5 0.001 0.001 6.429 6.520 fft3d_ps 1281 14.7 2.776 2.856 6.117 6.181 grid_collocate_task_list 127 9.7 5.166 5.666 5.166 5.666 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.582 5.644 mp_alltoall_d11v 2401 14.1 5.160 5.571 5.160 5.571 rs_pw_transfer 1038 11.9 0.013 0.015 4.911 5.220 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.569 4.570 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.456 4.554 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.375 4.446 potential_pw2rs 127 12.3 0.023 0.023 4.423 4.441 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.167 4.224 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.289000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.090909, yerr=16.822358 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/19/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 5.865088E+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 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 1.472152E+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 5.532237E+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 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_ISendRecv 7680 122880. MP_Wait 19962 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 0.026 0.049 176.204 176.204 qs_mol_dyn_low 1 2.0 0.003 0.003 175.727 175.740 qs_forces 11 3.9 0.003 0.003 175.640 175.642 qs_energies 11 4.9 0.001 0.002 168.306 168.316 scf_env_do_scf 11 5.9 0.001 0.001 151.006 151.014 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.606 114.607 velocity_verlet 10 3.0 0.007 0.009 114.585 114.589 qs_scf_new_mos 117 7.6 0.001 0.001 78.208 78.340 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.207 78.339 dbcsr_multiply_generic 2507 12.6 0.179 0.185 76.055 76.577 ot_scf_mini 117 9.6 0.003 0.004 73.767 73.890 multiply_cannon 2507 13.6 0.592 0.649 55.585 60.079 multiply_cannon_loop 2507 14.6 0.449 0.460 50.868 51.963 ot_mini 117 10.6 0.001 0.001 40.965 41.097 init_scf_loop 11 6.9 0.000 0.000 36.243 36.244 mp_waitall_1 129618 16.6 27.652 35.791 27.652 35.791 prepare_preconditioner 11 7.9 0.000 0.000 32.360 32.391 make_preconditioner 11 8.9 0.000 0.000 32.360 32.391 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.164 30.429 rebuild_ks_matrix 128 8.3 0.001 0.001 29.096 29.224 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.096 29.223 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.595 26.713 multiply_cannon_multrec 10028 15.6 10.382 14.367 17.669 22.055 qs_ot_get_derivative 117 11.6 0.002 0.002 20.668 20.779 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 13.052 20.708 ot_diis_step 117 11.6 0.020 0.020 20.220 20.221 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.734 20.148 apply_single 128 13.6 0.001 0.001 19.734 20.148 qs_ot_get_p 128 10.4 0.001 0.001 19.620 19.766 make_m2s 5014 13.6 0.067 0.072 16.708 18.841 make_images 5014 14.6 2.314 2.673 16.408 18.543 cp_fm_cholesky_invert 11 10.9 18.506 18.512 18.506 18.512 qs_ot_p2m_diag 83 11.4 0.495 0.501 15.655 15.674 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.484 14.485 sum_up_and_integrate 128 10.3 0.181 0.190 14.314 14.362 integrate_v_rspace 128 11.3 0.004 0.004 14.134 14.191 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.420 13.454 calculate_rho_elec 128 8.7 0.258 0.269 13.419 13.453 make_images_data 5014 15.6 0.054 0.060 10.251 12.689 hybrid_alltoall_any 5200 16.5 0.845 3.819 9.967 12.439 multiply_cannon_sync_h2d 10028 15.6 11.501 12.300 11.501 12.300 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.366 11.377 cp_fm_diag_elpa_base 83 14.4 11.118 11.186 11.358 11.369 init_scf_run 11 5.9 0.000 0.001 10.517 10.517 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.516 10.517 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.244 8.333 cp_fm_cholesky_decompose 22 10.9 8.104 8.227 8.104 8.227 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.716 8.190 grid_integrate_task_list 128 12.3 7.684 8.153 7.684 8.153 pw_transfer 1547 11.6 0.083 0.090 8.065 8.079 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.844 7.866 mp_allgather_i34 2507 14.6 2.908 7.609 2.908 7.609 dbcsr_mm_accdrv_process 20762 16.1 2.579 3.435 6.923 7.554 wfi_extrapolate 11 7.9 0.001 0.001 7.386 7.386 fft_wrap_pw1pw2_140 523 13.2 0.504 0.524 6.842 6.869 density_rs2pw 128 9.7 0.005 0.006 6.182 6.601 calculate_dm_sparse 128 9.5 0.001 0.001 6.096 6.202 fft3d_ps 1291 14.7 2.727 2.798 6.000 6.034 dbcsr_complete_redistribute 395 12.7 2.128 2.167 5.439 5.825 mp_alltoall_d11v 2415 14.1 5.085 5.814 5.085 5.814 grid_collocate_task_list 128 9.7 5.484 5.772 5.484 5.772 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.870 5.424 mp_irecv_dv 28860 15.9 1.833 5.336 1.833 5.336 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.248 5.253 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.230 5.231 rs_pw_transfer 1046 11.9 0.013 0.013 4.277 4.700 potential_pw2rs 128 12.3 0.027 0.028 4.279 4.299 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.102 4.138 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.539 3.873 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.555 3.843 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.811 3.842 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.784 3.818 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.661 3.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.204000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1371.272727, yerr=43.245790 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 11.786060E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.040432E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144969E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318398E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592338480 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.057 295.192 295.192 qs_mol_dyn_low 1 2.0 0.003 0.003 294.528 294.557 qs_forces 11 3.9 0.003 0.004 294.420 294.423 qs_energies 11 4.9 0.002 0.002 285.807 285.815 scf_env_do_scf 11 5.9 0.001 0.001 262.781 262.796 velocity_verlet 10 3.0 0.017 0.020 211.855 211.863 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.517 136.519 init_scf_loop 11 6.9 0.000 0.000 125.992 125.995 prepare_preconditioner 11 7.9 0.000 0.000 120.944 120.971 make_preconditioner 11 8.9 0.000 0.000 120.944 120.971 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.407 118.093 qs_scf_new_mos 118 7.6 0.001 0.001 93.897 94.028 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.896 94.027 ot_scf_mini 118 9.6 0.004 0.004 88.962 89.029 dbcsr_multiply_generic 2529 12.6 0.214 0.220 85.347 85.955 cp_fm_upper_to_full 106 14.8 53.278 76.805 53.278 76.805 multiply_cannon 2529 13.6 0.716 0.776 59.955 60.891 multiply_cannon_loop 2529 14.6 0.477 0.487 56.065 57.673 ot_mini 118 10.6 0.001 0.001 45.694 45.755 dbcsr_complete_redistribute 397 12.7 4.397 4.459 30.695 43.759 copy_fm_to_dbcsr 210 11.7 0.001 0.001 27.228 40.329 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.492 37.469 mp_alltoall_i22 720 14.1 22.270 35.606 22.270 35.606 rebuild_ks_matrix 129 8.3 0.001 0.001 34.261 34.350 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.260 34.349 cp_fm_cholesky_invert 11 10.9 33.878 33.884 33.878 33.884 mp_waitall_1 106386 16.7 29.309 33.384 29.309 33.384 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.004 32.091 qs_ot_get_p 129 10.4 0.001 0.001 27.425 27.478 qs_ot_get_derivative 118 11.6 0.002 0.002 24.893 24.961 qs_ot_p2m_diag 84 11.4 0.889 0.894 23.210 23.242 make_m2s 5058 13.6 0.077 0.081 20.624 21.806 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.444 21.445 make_images 5058 14.6 3.808 3.923 20.148 21.332 ot_diis_step 118 11.6 0.022 0.023 20.766 20.767 multiply_cannon_metrocomm3 10116 15.6 0.023 0.023 19.161 20.457 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.944 20.194 apply_single 129 13.6 0.001 0.001 19.944 20.193 multiply_cannon_multrec 10116 15.6 10.645 12.469 18.481 18.515 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.155 18.156 cp_fm_diag_elpa_base 84 14.4 13.705 15.330 18.151 18.152 sum_up_and_integrate 129 10.3 0.323 0.326 15.843 15.939 multiply_cannon_sync_h2d 10116 15.6 15.777 15.789 15.777 15.789 integrate_v_rspace 129 11.3 0.004 0.004 15.520 15.617 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.218 15.238 calculate_rho_elec 129 8.7 0.487 0.488 15.218 15.237 make_images_data 5058 15.6 0.062 0.067 11.222 13.303 hybrid_alltoall_any 5245 16.5 1.314 3.048 11.269 13.293 init_scf_run 11 5.9 0.000 0.001 13.017 13.018 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.017 13.017 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.755 9.812 dbcsr_mm_accdrv_process 20934 16.1 3.844 5.719 7.597 9.508 cp_fm_cholesky_decompose 22 10.9 9.056 9.074 9.056 9.074 grid_integrate_task_list 129 12.3 8.614 8.774 8.614 8.774 wfi_extrapolate 11 7.9 0.001 0.001 8.745 8.745 pw_transfer 1559 11.6 0.092 0.093 8.590 8.604 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.351 8.364 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 7.967 7.967 fft_wrap_pw1pw2_140 527 13.2 0.553 0.557 7.398 7.418 mp_alltoall_d11v 2429 14.1 7.019 7.180 7.019 7.180 calculate_dm_sparse 129 9.5 0.001 0.001 7.036 7.120 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.827 6.896 grid_collocate_task_list 129 9.7 6.388 6.418 6.388 6.418 fft3d_ps 1301 14.7 2.809 2.844 6.362 6.377 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.153 6.234 density_rs2pw 129 9.7 0.005 0.006 6.189 6.221 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.192000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2689.909091, yerr=190.974275 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 1.260130E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 26567447. MP_ISendRecv 570 19200. MP_Wait 1302 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 0.017 0.078 85.234 85.235 qs_energies 1 2.0 0.000 0.000 84.723 84.730 ls_scf 1 3.0 0.000 0.000 83.804 83.811 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.818 72.972 multiply_cannon 111 7.7 0.018 0.021 55.993 57.215 multiply_cannon_loop 111 8.7 0.213 0.226 52.567 53.913 ls_scf_main 1 4.0 0.000 0.000 52.417 52.418 density_matrix_trs4 2 5.0 0.002 0.003 46.903 47.003 ls_scf_init_scf 1 4.0 0.000 0.000 28.328 28.329 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.215 27.260 mp_waitall_1 11316 10.9 22.434 25.877 22.434 25.877 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.100 25.121 multiply_cannon_multrec 2664 9.7 8.217 9.001 15.620 17.422 multiply_cannon_sync_h2d 2664 9.7 13.701 15.563 13.701 15.563 make_m2s 222 7.7 0.009 0.011 13.215 13.751 make_images 222 8.7 0.101 0.112 13.193 13.732 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.598 12.587 make_images_data 222 9.7 0.005 0.006 7.725 8.319 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.602 7.021 8.020 hybrid_alltoall_any 227 10.6 0.219 1.850 6.599 7.824 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.434 7.593 dbcsr_mm_accdrv_process_sort 4760 11.4 6.305 7.207 6.305 7.207 calculate_norms 4752 9.8 5.538 6.195 5.538 6.195 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.043 5.178 mp_sum_l 807 5.4 3.090 4.301 3.090 4.301 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.075 3.607 mp_irecv_dv 6231 10.9 2.058 3.577 2.058 3.577 make_images_sizes 222 9.7 0.000 0.000 0.763 3.490 mp_alltoall_i44 222 10.7 0.762 3.490 0.762 3.490 arnoldi_extremal 4 6.8 0.000 0.000 3.255 3.291 arnoldi_normal_ev 4 7.8 0.001 0.002 3.255 3.291 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.303 3.274 build_subspace 16 8.4 0.009 0.012 3.158 3.160 ls_scf_post 1 4.0 0.000 0.000 3.059 3.066 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.906 dbcsr_special_finalize 555 9.7 0.005 0.006 2.360 2.839 dbcsr_merge_single_wm 555 10.7 0.457 0.586 2.351 2.830 make_images_pack 222 9.7 2.218 2.645 2.220 2.647 dbcsr_sort_data 658 11.4 2.151 2.580 2.151 2.580 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.332 2.570 dbcsr_matrix_vector_mult_local 304 10.0 2.055 2.449 2.057 2.450 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.265 2.354 buffer_matrices_ensure_size 222 8.7 1.755 2.047 1.755 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.803 1.804 rebuild_ks_matrix 3 7.3 0.000 0.000 1.793 1.795 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.793 1.795 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.235000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 2.090050E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 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 0.082 91.644 91.644 qs_energies 1 2.0 0.000 0.000 90.979 91.115 ls_scf 1 3.0 0.000 0.000 89.641 89.707 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.403 75.818 multiply_cannon 111 7.7 0.031 0.044 53.545 56.943 ls_scf_main 1 4.0 0.000 0.000 55.015 55.032 multiply_cannon_loop 111 8.7 0.117 0.124 50.235 53.302 density_matrix_trs4 2 5.0 0.002 0.003 49.312 49.493 ls_scf_init_scf 1 4.0 0.000 0.000 30.888 30.894 mp_waitall_1 9246 10.9 21.159 30.327 21.159 30.327 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.725 29.818 multiply_cannon_multrec 1332 9.7 13.279 17.078 22.604 27.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.282 27.305 multiply_cannon_metrocomm3 1332 9.7 0.007 0.009 11.724 20.855 make_m2s 222 7.7 0.007 0.009 15.297 15.975 make_images 222 8.7 1.585 1.979 15.267 15.946 dbcsr_mm_accdrv_process 4041 10.4 0.306 0.448 8.923 10.534 dbcsr_mm_accdrv_process_sort 4041 11.4 8.514 10.086 8.514 10.086 make_images_data 222 9.7 0.004 0.005 8.790 9.611 hybrid_alltoall_any 227 10.6 0.523 2.479 8.226 9.338 mp_sum_l 807 5.4 5.278 8.410 5.278 8.410 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.234 7.645 mp_irecv_dv 3311 11.0 3.215 7.595 3.215 7.595 calculate_norms 2376 9.8 6.005 6.882 6.005 6.882 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.088 6.656 multiply_cannon_sync_h2d 1332 9.7 4.737 5.871 4.737 5.871 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.086 5.289 arnoldi_extremal 4 6.8 0.000 0.000 4.732 4.765 arnoldi_normal_ev 4 7.8 0.001 0.004 4.732 4.765 build_subspace 16 8.4 0.014 0.021 4.456 4.462 ls_scf_post 1 4.0 0.000 0.000 3.737 3.804 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.193 3.427 ls_scf_store_result 1 5.0 0.000 0.000 3.261 3.379 dbcsr_matrix_vector_mult_local 304 10.0 2.775 3.243 2.777 3.244 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.272 2.801 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.539 2.640 make_images_pack 222 9.7 2.031 2.375 2.034 2.377 mp_allgather_i34 111 8.7 1.003 2.296 1.003 2.296 dbcsr_sort_data 436 11.2 1.830 2.034 1.830 2.034 dbcsr_data_new 4174 10.1 1.615 1.841 1.615 1.841 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.829 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.644000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 2.702184E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 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 0.079 0.163 95.493 95.494 qs_energies 1 2.0 0.000 0.000 94.359 94.369 ls_scf 1 3.0 0.000 0.001 92.639 92.651 dbcsr_multiply_generic 111 6.7 0.017 0.021 77.397 77.668 ls_scf_main 1 4.0 0.001 0.012 58.254 58.259 multiply_cannon 111 7.7 0.043 0.101 53.272 57.432 multiply_cannon_loop 111 8.7 0.100 0.108 49.737 53.424 density_matrix_trs4 2 5.0 0.002 0.003 52.374 52.530 mp_waitall_1 7374 11.0 24.361 34.154 24.361 34.154 ls_scf_init_scf 1 4.0 0.000 0.001 30.810 30.814 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.631 29.700 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.236 27.251 multiply_cannon_multrec 888 9.7 12.632 15.314 21.265 24.526 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.216 23.262 make_m2s 222 7.7 0.007 0.008 17.370 18.618 make_images 222 8.7 1.978 2.312 17.331 18.578 hybrid_alltoall_any 227 10.6 0.620 2.869 9.536 11.006 make_images_data 222 9.7 0.004 0.005 9.914 10.939 dbcsr_mm_accdrv_process 3754 10.4 0.247 0.418 8.139 9.317 dbcsr_mm_accdrv_process_sort 3754 11.4 7.767 8.906 7.767 8.906 mp_sum_l 807 5.4 5.175 8.787 5.175 8.787 multiply_cannon_sync_h2d 888 9.7 6.019 7.453 6.019 7.453 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.494 7.116 mp_irecv_dv 2335 11.1 2.479 7.073 2.479 7.073 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.004 6.952 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.854 6.860 arnoldi_extremal 4 6.8 0.000 0.001 5.086 5.104 arnoldi_normal_ev 4 7.8 0.001 0.007 5.086 5.104 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.885 5.089 build_subspace 16 8.4 0.014 0.020 4.775 4.786 calculate_norms 1584 9.8 4.264 4.622 4.264 4.622 mp_allgather_i34 111 8.7 1.367 3.888 1.367 3.888 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.442 3.767 dbcsr_matrix_vector_mult_local 304 10.0 3.031 3.604 3.033 3.606 ls_scf_post 1 4.0 0.001 0.009 3.575 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.308 3.397 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.765 2.858 make_images_sizes 222 9.7 0.000 0.000 1.151 2.456 mp_alltoall_i44 222 10.7 1.150 2.455 1.150 2.455 dbcsr_sort_data 325 11.1 1.862 2.165 1.862 2.165 make_images_pack 222 9.7 1.813 2.120 1.816 2.123 dbcsr_data_release 9322 10.9 1.348 1.959 1.348 1.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.494000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2181.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 3.391762E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 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 0.033 0.059 97.435 97.435 qs_energies 1 2.0 0.000 0.000 96.725 96.730 ls_scf 1 3.0 0.000 0.000 95.049 95.055 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.550 78.812 ls_scf_main 1 4.0 0.000 0.000 58.871 58.880 multiply_cannon 111 7.7 0.047 0.125 51.753 55.987 density_matrix_trs4 2 5.0 0.002 0.003 52.794 52.899 multiply_cannon_loop 111 8.7 0.115 0.127 46.752 49.693 ls_scf_init_scf 1 4.0 0.000 0.000 32.936 32.967 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.423 31.478 mp_waitall_1 6438 11.0 22.774 29.668 22.774 29.668 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.932 28.946 multiply_cannon_multrec 1332 9.7 14.264 17.139 22.237 24.749 make_m2s 222 7.7 0.008 0.009 21.230 22.555 make_images 222 8.7 3.149 3.618 21.178 22.506 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.266 15.995 make_images_data 222 9.7 0.004 0.005 11.841 13.419 hybrid_alltoall_any 227 10.6 0.798 3.821 11.255 12.916 dbcsr_mm_accdrv_process 3641 10.4 0.204 0.379 7.597 9.121 dbcsr_mm_accdrv_process_sort 3641 11.4 7.224 8.708 7.224 8.708 mp_sum_l 807 5.4 3.967 7.172 3.967 7.172 multiply_cannon_sync_h2d 1332 9.7 5.528 6.177 5.528 6.177 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.093 6.028 mp_irecv_dv 3229 10.9 2.070 5.948 2.070 5.948 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.045 5.588 arnoldi_extremal 4 6.8 0.000 0.000 5.244 5.257 arnoldi_normal_ev 4 7.8 0.001 0.005 5.244 5.257 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.501 5.152 build_subspace 16 8.4 0.015 0.021 4.910 4.920 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.540 4.726 calculate_norms 2376 9.8 4.184 4.537 4.184 4.537 mp_allgather_i34 111 8.7 2.113 4.379 2.113 4.379 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.603 3.905 dbcsr_matrix_vector_mult_local 304 10.0 3.201 3.719 3.203 3.721 dbcsr_sort_data 658 11.4 3.088 3.403 3.088 3.403 ls_scf_post 1 4.0 0.000 0.000 3.242 3.247 dbcsr_special_finalize 555 9.7 0.006 0.007 2.830 3.200 dbcsr_merge_single_wm 555 10.7 0.534 0.660 2.821 3.192 ls_scf_store_result 1 5.0 0.000 0.000 2.985 3.042 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.919 2.948 dbcsr_data_release 10477 10.7 1.610 2.439 1.610 2.439 dbcsr_finalize 304 7.8 0.050 0.062 1.816 1.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.435000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 4.625289E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 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 0.054 99.195 99.196 qs_energies 1 2.0 0.000 0.000 98.414 98.420 ls_scf 1 3.0 0.000 0.000 96.500 96.505 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.103 78.313 ls_scf_main 1 4.0 0.000 0.000 62.204 62.205 multiply_cannon 111 7.7 0.096 0.160 55.444 61.047 density_matrix_trs4 2 5.0 0.002 0.003 55.240 55.332 multiply_cannon_loop 111 8.7 0.069 0.075 51.054 52.695 mp_waitall_1 5481 11.0 26.697 32.644 26.697 32.644 ls_scf_init_scf 1 4.0 0.000 0.000 30.713 30.719 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.546 29.588 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.381 27.389 multiply_cannon_multrec 444 9.7 14.074 16.362 21.172 23.199 make_m2s 222 7.7 0.006 0.006 17.892 20.275 make_images 222 8.7 3.728 4.425 17.828 20.213 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.080 16.940 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.356 14.526 make_images_data 222 9.7 0.003 0.004 10.063 12.450 hybrid_alltoall_any 227 10.6 0.790 3.761 9.913 12.318 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.340 6.804 7.933 dbcsr_mm_accdrv_process_sort 3003 11.4 6.490 7.588 6.490 7.588 multiply_cannon_sync_h2d 444 9.7 6.573 7.582 6.573 7.582 mp_allgather_i34 111 8.7 2.606 7.009 2.606 7.009 arnoldi_extremal 4 6.8 0.000 0.000 5.792 5.800 arnoldi_normal_ev 4 7.8 0.001 0.004 5.792 5.799 build_subspace 16 8.4 0.015 0.020 5.400 5.408 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.700 4.681 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.525 4.666 mp_irecv_dv 1241 11.2 1.680 4.643 1.680 4.643 mp_sum_l 807 5.4 2.813 4.638 2.813 4.638 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.178 4.353 dbcsr_matrix_vector_mult_local 304 10.0 3.708 4.143 3.710 4.145 calculate_norms 792 9.8 3.539 3.679 3.539 3.679 ls_scf_post 1 4.0 0.000 0.000 3.582 3.587 make_images_sizes 222 9.7 0.000 0.000 1.161 3.531 mp_alltoall_i44 222 10.7 1.160 3.531 1.160 3.531 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.385 3.485 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.920 3.461 ls_scf_store_result 1 5.0 0.000 0.000 3.362 3.406 dbcsr_finalize 304 7.8 0.062 0.078 2.204 2.265 dbcsr_merge_all 275 8.9 0.474 0.521 2.050 2.096 dbcsr_data_release 10123 10.8 1.334 1.989 1.334 1.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.196000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3654.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b41f84f0a640ce8588a35f300e725818d42616be_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 8.724824E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.098 109.078 109.078 qs_energies 1 2.0 0.000 0.000 107.167 107.181 ls_scf 1 3.0 0.000 0.000 104.061 104.093 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.853 77.984 ls_scf_main 1 4.0 0.000 0.000 65.843 65.844 density_matrix_trs4 2 5.0 0.002 0.003 57.000 57.044 multiply_cannon 111 7.7 0.140 0.237 50.284 52.399 multiply_cannon_loop 111 8.7 0.068 0.070 46.747 47.814 ls_scf_init_scf 1 4.0 0.000 0.000 34.534 34.554 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.808 32.846 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.966 29.983 mp_waitall_1 4569 11.1 22.383 26.064 22.383 26.064 make_m2s 222 7.7 0.007 0.007 23.938 24.951 make_images 222 8.7 4.589 4.976 23.828 24.838 multiply_cannon_multrec 444 9.7 17.901 18.667 22.637 23.456 make_images_data 222 9.7 0.004 0.004 13.152 15.624 hybrid_alltoall_any 227 10.6 1.663 3.633 12.943 15.604 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.638 10.989 multiply_cannon_sync_h2d 444 9.7 8.845 8.896 8.845 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.390 7.410 arnoldi_normal_ev 4 7.8 0.002 0.008 7.389 7.410 build_subspace 16 8.4 0.026 0.037 6.837 6.850 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.492 5.655 dbcsr_matrix_vector_mult_local 304 10.0 5.062 5.382 5.064 5.384 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.018 5.276 dbcsr_mm_accdrv_process 1814 10.4 0.228 0.652 4.566 4.898 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.799 4.886 dbcsr_mm_accdrv_process_sort 1814 11.4 4.184 4.323 4.184 4.323 ls_scf_post 1 4.0 0.000 0.000 3.684 3.698 mp_allgather_i34 111 8.7 1.155 3.625 1.155 3.625 make_images_sizes 222 9.7 0.000 0.000 1.432 3.493 mp_alltoall_i44 222 10.7 1.432 3.492 1.432 3.492 ls_scf_store_result 1 5.0 0.000 0.000 3.430 3.443 calculate_norms 792 9.8 3.242 3.275 3.242 3.275 dbcsr_finalize 304 7.8 0.082 0.090 3.095 3.137 qs_energies_init_hamiltonians 1 3.0 0.021 0.040 3.075 3.094 dbcsr_merge_all 275 8.9 0.895 0.921 2.878 2.915 dbcsr_complete_redistribute 5 7.6 1.454 1.479 2.790 2.890 dbcsr_data_release 12724 10.6 2.340 2.889 2.340 2.889 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.583 2.585 matrix_ls_to_qs 2 6.0 0.000 0.000 2.437 2.542 rebuild_ks_matrix 3 7.3 0.000 0.000 2.518 2.520 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.518 2.520 dbcsr_sort_data 325 11.1 2.442 2.501 2.442 2.501 dbcsr_new_transposed 4 7.5 0.243 0.251 2.271 2.282 dbcsr_frobenius_norm 74 6.6 2.059 2.132 2.196 2.234 dbcsr_add_d 103 6.2 0.000 0.000 2.138 2.209 dbcsr_add_anytype 103 7.2 0.861 0.893 2.138 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.078000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6856.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b41f84f0a640ce8588a35f300e725818d42616be Summary: empty Status: OK