=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 0a7d28b76a9fad31202ff85cf948797d4075042b ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, 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 (26.10.2022) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/01 job id: 42675306 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/02 job id: 42675307 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/03 job id: 42675308 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/04 job id: 42675309 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/05 job id: 42675310 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/06 job id: 42675311 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/07 job id: 42675312 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/08 job id: 42675313 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/09 job id: 42675314 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/10 job id: 42675315 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/11 job id: 42675316 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/12 job id: 42675317 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/13 job id: 42675318 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/14 job id: 42675319 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/15 job id: 42675320 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/16 job id: 42675322 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/17 job id: 42675323 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/18 job id: 42675324 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/19 job id: 42675325 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/20 job id: 42675326 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/21 job id: 42675327 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/22 job id: 42675328 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/23 job id: 42675330 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/24 job id: 42675332 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/25 job id: 42675333 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/26 job id: 42675334 --- 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/0a7d28b76a9fad31202ff85cf948797d4075042b_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 0.020 0.034 144.014 144.015 farming_run 1 2.0 133.483 133.485 143.987 143.990 mp_bcast_i 22 4.2 8.296 8.858 8.296 8.858 read_input 2 3.0 0.005 0.011 8.852 8.852 section_vals_parse 40 7.0 0.000 0.000 8.846 8.847 parser_read_line 140 7.8 0.000 0.000 8.844 8.845 parser_read_line_low 2 6.0 0.553 8.838 8.844 8.845 broadcast_input_information 2 7.0 0.000 0.000 8.291 8.844 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460670E+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 0.025 115.321 115.321 qs_energies 1 2.0 0.000 0.000 115.125 115.126 mp2_main 1 3.0 0.000 0.000 113.232 113.233 mp2_gpw_main 1 4.0 0.019 0.025 112.418 112.420 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.402 93.912 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.495 56.006 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.811 47.270 get_2c_integrals 1 6.0 0.000 0.000 36.658 37.734 integrate_v_rspace 273 8.0 0.438 0.454 25.140 30.289 pw_transfer 6555 10.6 0.376 0.384 27.398 27.877 grid_integrate_task_list 273 9.0 20.943 26.579 20.943 26.579 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.056 26.541 fft_wrap_pw1pw2_100 2178 12.4 1.175 1.262 23.617 24.102 compute_2c_integrals 1 7.0 0.002 0.002 19.274 19.275 rpa_ri_compute_en 1 5.0 0.000 0.000 18.909 19.177 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.876 19.005 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.429 18.873 19.001 cp_fm_cholesky_decompose 12 8.2 17.322 18.379 17.322 18.379 cholesky_decomp 1 7.0 0.000 0.000 16.231 17.288 fft3d_s 5443 13.4 16.172 16.574 16.194 16.596 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.555 10.851 15.555 calculate_wavefunction 272 8.0 5.433 5.597 12.576 13.166 rpa_num_int 1 6.0 0.000 0.000 10.549 10.558 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.516 10.556 calc_mat_Q 8 8.0 0.000 0.000 9.369 9.460 contract_S_to_Q 8 9.0 0.000 0.000 8.789 8.877 calc_potential_gpw 544 9.5 0.005 0.006 8.211 8.620 parallel_gemm_fm 14 9.1 0.000 0.000 8.379 8.438 parallel_gemm_fm_cosma 14 10.1 8.379 8.438 8.379 8.438 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.200 8.399 potential_pw2rs 545 10.0 0.107 0.109 7.739 8.341 create_integ_mat 1 6.0 0.014 0.026 7.754 7.764 collocate_single_gaussian 272 10.0 0.040 0.042 7.464 7.698 array2fm 1 7.0 0.000 0.000 6.768 7.202 pw_scatter_s 2720 13.7 4.431 4.691 4.431 4.691 pw_gather_s 2722 13.2 3.907 4.418 3.907 4.418 array2fm_buffer_send 1 8.0 2.977 3.152 2.977 3.152 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.418816, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 0.027 0.038 400.415 400.416 farming_run 1 2.0 398.818 398.822 400.377 400.380 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 897827141120 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 249788822 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 1.226490E+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 3039998. 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 0.014 0.044 212.179 212.179 qs_energies 1 2.0 0.000 0.000 211.943 211.949 scf_env_do_scf 1 3.0 0.000 0.000 106.760 106.760 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.862 105.871 rebuild_ks_matrix 4 6.0 0.000 0.000 105.860 105.869 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.860 105.869 hfx_ks_matrix 4 8.0 0.001 0.001 105.486 105.489 integrate_four_center 4 9.0 0.144 0.453 105.485 105.488 mp2_main 1 3.0 0.000 0.000 104.894 104.899 mp2_gpw_main 1 4.0 0.034 0.071 104.047 104.054 integrate_four_center_main 4 10.0 0.099 0.554 96.683 99.854 integrate_four_center_bin 268 11.0 96.584 99.822 96.584 99.822 init_scf_loop 1 4.0 0.000 0.000 92.598 92.599 mp2_ri_gpw_compute_in 1 5.0 0.069 0.125 77.038 78.153 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.439 55.553 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.177 47.223 integrate_v_rspace 95 8.0 0.399 0.563 28.545 33.555 pw_transfer 2240 10.6 0.146 0.159 29.859 30.204 ao_to_mo_and_store_B_mult_1 91 7.0 10.576 29.453 10.576 29.453 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.892 29.285 grid_integrate_task_list 95 9.0 23.842 29.048 23.842 29.048 mp2_ri_gpw_compute_en 1 5.0 0.058 0.086 26.843 28.510 fft_wrap_pw1pw2_100 730 12.4 1.249 1.348 26.637 27.085 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.823 1.882 25.060 25.071 get_2c_integrals 1 6.0 0.001 0.011 22.505 22.535 compute_2c_integrals 1 7.0 0.029 0.218 21.489 21.497 compute_2c_integrals_loop_lm 1 8.0 0.002 0.010 19.058 21.098 mp2_eri_2c_integrate_gpw 1 9.0 1.752 1.836 19.056 21.088 fft3d_s 1823 13.4 18.412 18.777 18.426 18.791 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.159 14.159 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.217 1.746 8.418 10.215 calculate_wavefunction 91 8.0 2.027 2.051 9.747 10.026 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.589 8.810 9.215 potential_pw2rs 186 10.0 0.033 0.035 8.604 9.074 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.531 8.034 8.669 local_gemm 172 8.0 8.252 8.627 8.252 8.627 collocate_single_gaussian 91 10.0 0.054 0.354 7.888 8.413 calc_potential_gpw 182 9.5 0.002 0.002 7.893 8.072 mp_sendrecv_dm3 2068 8.0 6.067 6.729 6.067 6.729 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.459 6.348 6.459 mp_sync 38 10.4 3.480 6.228 3.480 6.228 pw_gather_s 912 13.2 4.924 5.183 4.924 5.183 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=104.037071, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.121152E+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 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 0.011 0.027 52.789 52.789 qs_mol_dyn_low 1 2.0 0.003 0.003 51.982 51.990 qs_forces 11 3.9 0.001 0.002 51.899 51.901 qs_energies 11 4.9 0.021 0.039 50.440 50.453 scf_env_do_scf 11 5.9 0.000 0.001 44.221 44.222 scf_env_do_scf_inner_loop 108 6.5 0.027 0.104 42.148 42.148 dbcsr_multiply_generic 2286 12.5 0.094 0.097 34.008 34.386 qs_scf_new_mos 108 7.5 0.000 0.001 32.372 32.669 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.372 32.669 ot_scf_mini 108 9.5 0.002 0.002 30.697 30.888 multiply_cannon 2286 13.5 0.180 0.190 26.495 28.062 multiply_cannon_loop 2286 14.5 1.523 1.598 25.804 27.368 velocity_verlet 10 3.0 0.001 0.002 26.129 26.129 ot_mini 108 10.5 0.001 0.001 19.655 19.918 qs_ot_get_derivative 108 11.5 0.001 0.001 16.725 16.930 mp_waitall_1 267858 16.1 8.863 14.797 8.863 14.797 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 6.215 13.075 multiply_cannon_multrec 54864 15.5 4.196 6.523 7.784 11.225 rebuild_ks_matrix 119 8.3 0.000 0.000 7.737 7.863 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.736 7.863 mp_sum_l 7207 12.9 5.548 7.000 5.548 7.000 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.838 6.952 multiply_cannon_sync_h2d 54864 15.5 5.901 6.894 5.901 6.894 qs_ot_get_p 119 10.4 0.001 0.001 6.350 6.632 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.943 6.345 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.314 5.415 init_scf_run 11 5.9 0.000 0.001 4.957 4.957 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.956 4.957 dbcsr_mm_accdrv_process 76910 16.1 1.214 1.876 3.511 4.883 sum_up_and_integrate 119 10.3 0.013 0.016 4.400 4.409 integrate_v_rspace 119 11.3 0.002 0.002 4.387 4.398 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.913 4.014 calculate_rho_elec 119 8.7 0.011 0.016 3.912 4.013 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.442 3.495 calculate_dm_sparse 119 9.5 0.000 0.000 3.017 3.151 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.856 3.144 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.903 3.092 apply_single 119 13.6 0.000 0.000 2.903 3.092 jit_kernel_multiply 13 15.8 2.234 2.926 2.234 2.926 calculate_first_density_matrix 1 7.0 0.000 0.000 2.827 2.833 rs_pw_transfer 974 11.9 0.012 0.013 2.673 2.778 ot_diis_step 108 11.5 0.006 0.006 2.681 2.682 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.644 2.644 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.523 2.579 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.503 2.506 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.439 2.439 cp_fm_redistribute_end 50 14.0 2.214 2.417 2.219 2.419 cp_fm_diag_elpa_base 50 14.0 0.198 2.382 0.199 2.391 acc_transpose_blocks 54864 15.5 0.233 0.250 1.816 2.256 density_rs2pw 119 9.7 0.004 0.004 2.053 2.140 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.074 2.124 grid_integrate_task_list 119 12.3 2.016 2.111 2.016 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.070 2.070 init_scf_loop 11 6.9 0.000 0.000 2.058 2.059 mp_sum_d 4125 12.0 1.279 1.916 1.279 1.916 potential_pw2rs 119 12.3 0.004 0.004 1.809 1.824 make_m2s 4572 13.5 0.054 0.057 1.653 1.696 pw_transfer 1439 11.6 0.051 0.056 1.557 1.630 make_images 4572 14.5 0.133 0.139 1.570 1.613 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.482 1.557 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.499 1.512 mp_alltoall_d11v 2130 13.8 1.235 1.474 1.235 1.474 grid_collocate_task_list 119 9.7 1.284 1.365 1.284 1.365 fft3d_ps 1201 14.6 0.350 0.454 1.258 1.323 mp_waitany 12084 13.8 1.201 1.303 1.201 1.303 fft_wrap_pw1pw2_140 487 13.2 0.081 0.095 1.130 1.205 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.691 1.096 acc_transpose_blocks_kernels 54864 16.5 0.254 0.385 0.807 1.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.789000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.727104E+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 1165979. 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 0.018 0.035 39.327 39.328 qs_mol_dyn_low 1 2.0 0.003 0.003 39.116 39.123 qs_forces 11 3.9 0.002 0.002 38.904 38.905 qs_energies 11 4.9 0.002 0.005 37.209 37.212 scf_env_do_scf 11 5.9 0.001 0.001 31.560 31.560 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.004 29.005 dbcsr_multiply_generic 2286 12.5 0.100 0.104 22.005 22.359 qs_scf_new_mos 108 7.5 0.001 0.001 20.059 20.297 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.058 20.296 ot_scf_mini 108 9.5 0.002 0.003 19.158 19.329 velocity_verlet 10 3.0 0.001 0.002 18.389 18.393 multiply_cannon 2286 13.5 0.209 0.217 16.813 18.374 multiply_cannon_loop 2286 14.5 0.901 0.970 15.662 17.184 ot_mini 108 10.5 0.001 0.001 11.922 12.159 mp_waitall_1 217478 16.2 6.093 11.395 6.093 11.395 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.222 9.694 qs_ot_get_derivative 108 11.5 0.001 0.001 9.430 9.605 multiply_cannon_multrec 27432 15.5 1.964 4.349 6.041 9.085 rebuild_ks_matrix 119 8.3 0.000 0.000 7.180 7.312 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.180 7.312 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.335 6.453 dbcsr_mm_accdrv_process 47894 16.0 3.120 5.423 4.007 5.980 qs_ot_get_p 119 10.4 0.001 0.002 4.323 4.558 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.615 4.467 init_scf_run 11 5.9 0.000 0.001 4.419 4.420 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.419 4.419 sum_up_and_integrate 119 10.3 0.024 0.027 4.196 4.203 integrate_v_rspace 119 11.3 0.002 0.002 4.171 4.181 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.075 4.153 apply_single 119 13.6 0.000 0.000 3.075 4.153 mp_sum_l 7207 12.9 2.123 4.065 2.123 4.065 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.582 3.613 calculate_rho_elec 119 8.7 0.021 0.024 3.582 3.613 calculate_first_density_matrix 1 7.0 0.005 0.043 2.889 2.891 make_m2s 4572 13.5 0.053 0.056 2.618 2.873 multiply_cannon_sync_h2d 27432 15.5 2.179 2.822 2.179 2.822 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.770 2.789 make_images 4572 14.5 0.201 0.239 2.530 2.783 rs_pw_transfer 974 11.9 0.010 0.011 2.571 2.654 init_scf_loop 11 6.9 0.001 0.004 2.534 2.535 ot_diis_step 108 11.5 0.010 0.011 2.437 2.438 jit_kernel_multiply 11 16.2 0.835 2.415 0.835 2.415 calculate_dm_sparse 119 9.5 0.000 0.000 2.337 2.412 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.331 2.331 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.190 2.281 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.183 2.190 density_rs2pw 119 9.7 0.004 0.004 1.944 2.026 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.922 1.972 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.967 1.967 grid_integrate_task_list 119 12.3 1.838 1.955 1.838 1.955 potential_pw2rs 119 12.3 0.006 0.006 1.936 1.948 cp_fm_redistribute_end 50 14.0 1.615 1.937 1.619 1.939 cp_fm_diag_elpa_base 50 14.0 0.310 1.879 0.318 1.911 pw_transfer 1439 11.6 0.063 0.066 1.770 1.801 acc_transpose_blocks 27432 15.5 0.109 0.115 1.482 1.787 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.680 1.713 make_images_data 4572 15.5 0.045 0.052 1.180 1.637 prepare_preconditioner 11 7.9 0.000 0.000 1.592 1.620 make_preconditioner 11 8.9 0.000 0.001 1.592 1.620 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.550 1.557 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.489 1.549 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.023 1.514 wfi_extrapolate 11 7.9 0.001 0.001 1.477 1.477 fft3d_ps 1201 14.6 0.494 0.545 1.388 1.418 mp_allgather_i34 2286 14.5 0.590 1.353 0.590 1.353 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.302 1.333 grid_collocate_task_list 119 9.7 1.243 1.328 1.243 1.328 mp_alltoall_d11v 2130 13.8 1.209 1.307 1.209 1.307 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.185 1.237 acc_transpose_blocks_kernels 27432 16.5 0.181 0.269 0.946 1.159 mp_sum_d 4125 12.0 0.613 1.056 0.613 1.056 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.954 0.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.905 0.919 jit_kernel_transpose 5 15.5 0.765 0.893 0.765 0.893 make_images_sizes 4572 15.5 0.005 0.005 0.621 0.888 mp_alltoall_i44 4572 16.5 0.617 0.883 0.617 0.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.328000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.909091, yerr=1.831767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 523.075584E+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 0.010 0.027 37.173 37.174 qs_mol_dyn_low 1 2.0 0.003 0.003 36.937 36.944 qs_forces 11 3.9 0.002 0.002 36.853 36.854 qs_energies 11 4.9 0.001 0.002 35.317 35.319 scf_env_do_scf 11 5.9 0.000 0.001 25.855 25.856 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.241 23.241 dbcsr_multiply_generic 2286 12.5 0.093 0.096 21.369 21.457 multiply_cannon 2286 13.5 0.196 0.206 18.026 18.883 multiply_cannon_loop 2286 14.5 0.635 0.666 17.242 18.088 qs_scf_new_mos 108 7.5 0.001 0.001 15.276 15.297 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.276 15.297 velocity_verlet 10 3.0 0.002 0.002 15.234 15.236 ot_scf_mini 108 9.5 0.002 0.002 14.537 14.552 multiply_cannon_multrec 18288 15.5 2.022 3.454 11.041 11.740 dbcsr_mm_accdrv_process 38222 16.0 6.647 9.708 8.934 9.764 ot_mini 108 10.5 0.001 0.001 9.201 9.212 init_scf_run 11 5.9 0.000 0.001 8.026 8.026 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.026 8.026 qs_ot_get_derivative 108 11.5 0.001 0.001 7.700 7.714 calculate_first_density_matrix 1 7.0 0.000 0.000 6.840 6.840 rebuild_ks_matrix 119 8.3 0.000 0.000 6.273 6.291 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.273 6.290 jit_kernel_multiply 10 15.9 2.235 6.058 2.235 6.058 calculate_dm_sparse 119 9.5 0.000 0.000 6.000 6.013 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.552 5.568 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 5.356 5.360 mp_waitall_1 169478 16.3 3.341 4.276 3.341 4.276 sum_up_and_integrate 119 10.3 0.030 0.030 3.917 3.923 integrate_v_rspace 119 11.3 0.002 0.002 3.887 3.896 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.696 3.329 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.250 3.256 calculate_rho_elec 119 8.7 0.030 0.031 3.250 3.256 qs_ot_get_p 119 10.4 0.001 0.001 3.166 3.187 init_scf_loop 11 6.9 0.000 0.000 2.598 2.598 rs_pw_transfer 974 11.9 0.009 0.010 2.284 2.390 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.053 2.360 apply_single 119 13.6 0.000 0.000 2.053 2.360 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.454 2.183 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.077 2.084 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.069 2.071 density_rs2pw 119 9.7 0.004 0.004 1.847 1.962 make_m2s 4572 13.5 0.045 0.046 1.807 1.957 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.908 1.918 grid_integrate_task_list 119 12.3 1.801 1.905 1.801 1.905 make_images 4572 14.5 0.191 0.203 1.722 1.872 prepare_preconditioner 11 7.9 0.000 0.000 1.825 1.827 make_preconditioner 11 8.9 0.000 0.000 1.825 1.827 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.810 1.811 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.675 1.756 potential_pw2rs 119 12.3 0.007 0.008 1.706 1.715 pw_transfer 1439 11.6 0.063 0.066 1.695 1.708 acc_transpose_blocks 18288 15.5 0.075 0.077 1.599 1.637 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.604 1.619 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.583 1.584 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.564 1.570 cp_fm_redistribute_end 50 14.0 1.173 1.562 1.175 1.563 cp_fm_diag_elpa_base 50 14.0 0.371 1.506 0.386 1.546 multiply_cannon_sync_h2d 18288 15.5 1.372 1.525 1.372 1.525 ot_diis_step 108 11.5 0.011 0.011 1.478 1.478 mp_sum_l 7207 12.9 1.145 1.466 1.145 1.466 fft3d_ps 1201 14.6 0.505 0.521 1.290 1.306 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.276 1.291 grid_collocate_task_list 119 9.7 1.203 1.271 1.203 1.271 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 1.157 1.184 wfi_extrapolate 11 7.9 0.001 0.001 1.142 1.142 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.740 1.108 make_images_data 4572 15.5 0.045 0.049 0.819 0.977 jit_kernel_transpose 5 15.6 0.948 0.972 0.948 0.972 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.925 0.926 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.888 0.908 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.708 0.868 mp_alltoall_d11v 2130 13.8 0.688 0.845 0.688 0.845 cp_fm_cholesky_invert 11 10.9 0.796 0.802 0.796 0.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.790 0.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=37.174000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.545455, yerr=1.924183 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 554.205184E+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 0.010 0.030 35.429 35.430 qs_mol_dyn_low 1 2.0 0.003 0.003 35.268 35.275 qs_forces 11 3.9 0.002 0.002 35.194 35.195 qs_energies 11 4.9 0.001 0.001 33.486 33.491 scf_env_do_scf 11 5.9 0.000 0.001 27.878 27.880 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.382 24.383 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.817 18.924 velocity_verlet 10 3.0 0.001 0.002 17.866 17.868 qs_scf_new_mos 108 7.5 0.001 0.001 16.151 16.203 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.150 16.202 multiply_cannon 2286 13.5 0.230 0.273 15.095 15.615 ot_scf_mini 108 9.5 0.002 0.003 15.207 15.262 multiply_cannon_loop 2286 14.5 0.942 0.971 14.154 14.523 multiply_cannon_multrec 27432 15.5 2.331 3.056 8.981 9.394 ot_mini 108 10.5 0.001 0.001 9.246 9.313 dbcsr_mm_accdrv_process 47916 15.9 5.337 7.055 6.557 7.803 qs_ot_get_derivative 108 11.5 0.001 0.001 7.428 7.486 rebuild_ks_matrix 119 8.3 0.000 0.000 6.497 6.559 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.497 6.558 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.772 5.826 init_scf_run 11 5.9 0.000 0.001 4.296 4.296 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.296 4.296 sum_up_and_integrate 119 10.3 0.035 0.038 3.771 3.780 integrate_v_rspace 119 11.3 0.002 0.003 3.737 3.746 init_scf_loop 11 6.9 0.000 0.000 3.477 3.477 qs_ot_get_p 119 10.4 0.001 0.001 3.340 3.420 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.336 3.365 calculate_rho_elec 119 8.7 0.040 0.046 3.335 3.364 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.738 3.201 calculate_first_density_matrix 1 7.0 0.000 0.000 2.948 2.950 mp_waitall_1 145218 16.4 2.107 2.710 2.107 2.710 prepare_preconditioner 11 7.9 0.000 0.000 2.637 2.646 make_preconditioner 11 8.9 0.000 0.000 2.637 2.646 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.111 2.588 apply_single 119 13.6 0.000 0.000 2.111 2.588 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.226 2.565 calculate_dm_sparse 119 9.5 0.000 0.000 2.377 2.437 make_m2s 4572 13.5 0.054 0.056 2.223 2.324 jit_kernel_multiply 10 15.8 1.159 2.235 1.159 2.235 make_images 4572 14.5 0.273 0.332 2.115 2.215 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.195 2.196 rs_pw_transfer 974 11.9 0.009 0.009 2.014 2.115 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.043 2.052 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.014 2.025 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.920 1.950 grid_integrate_task_list 119 12.3 1.822 1.912 1.822 1.912 density_rs2pw 119 9.7 0.004 0.004 1.802 1.897 acc_transpose_blocks 27432 15.5 0.111 0.113 1.796 1.817 pw_transfer 1439 11.6 0.063 0.066 1.777 1.810 ot_diis_step 108 11.5 0.012 0.012 1.774 1.774 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.687 1.724 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.718 1.718 mp_sum_l 7207 12.9 1.072 1.584 1.072 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.553 1.564 potential_pw2rs 119 12.3 0.008 0.009 1.553 1.560 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.473 1.474 cp_fm_redistribute_end 50 14.0 0.970 1.450 0.972 1.450 cp_fm_diag_elpa_base 50 14.0 0.456 1.388 0.476 1.430 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.372 1.411 fft3d_ps 1201 14.6 0.533 0.581 1.369 1.395 wfi_extrapolate 11 7.9 0.001 0.001 1.302 1.302 grid_collocate_task_list 119 9.7 1.219 1.300 1.219 1.300 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.734 1.236 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 1.165 1.180 cp_fm_upper_to_full 72 13.5 0.820 1.140 0.820 1.140 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.110 1.127 dbcsr_complete_redistribute 329 12.2 0.123 0.152 0.798 1.078 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.061 1.062 multiply_cannon_sync_h2d 27432 15.5 0.991 1.058 0.991 1.058 make_images_data 4572 15.5 0.045 0.048 0.875 0.993 jit_kernel_transpose 5 15.6 0.897 0.908 0.897 0.908 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.759 0.899 mp_alltoall_d11v 2130 13.8 0.790 0.878 0.790 0.878 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.873 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.590 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.815 0.820 cp_fm_cholesky_invert 11 10.9 0.779 0.782 0.779 0.782 multiply_cannon_metrocomm1 27432 15.5 0.032 0.033 0.195 0.777 mp_alltoall_i22 627 13.8 0.441 0.752 0.441 0.752 mp_alltoall_z22v 1201 16.6 0.718 0.742 0.718 0.742 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.430000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.181818, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 595.169280E+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 0.010 0.027 28.813 28.813 qs_mol_dyn_low 1 2.0 0.003 0.003 28.633 28.640 qs_forces 11 3.9 0.001 0.002 28.570 28.571 qs_energies 11 4.9 0.001 0.001 26.865 26.868 scf_env_do_scf 11 5.9 0.000 0.001 21.059 21.059 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.537 18.538 velocity_verlet 10 3.0 0.001 0.002 14.285 14.288 dbcsr_multiply_generic 2286 12.5 0.098 0.100 13.133 13.217 multiply_cannon 2286 13.5 0.231 0.243 10.530 11.001 qs_scf_new_mos 108 7.5 0.001 0.001 10.954 10.981 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.954 10.980 ot_scf_mini 108 9.5 0.002 0.002 10.274 10.302 multiply_cannon_loop 2286 14.5 0.331 0.342 9.580 9.773 multiply_cannon_multrec 9144 15.5 1.578 1.826 6.787 7.004 rebuild_ks_matrix 119 8.3 0.000 0.000 5.814 5.838 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.814 5.837 ot_mini 108 10.5 0.001 0.001 5.665 5.699 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.183 5.205 dbcsr_mm_accdrv_process 12550 15.8 3.897 4.908 5.109 5.183 qs_ot_get_derivative 108 11.5 0.001 0.001 4.345 4.374 init_scf_run 11 5.9 0.000 0.001 4.298 4.298 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.297 4.298 sum_up_and_integrate 119 10.3 0.038 0.041 3.572 3.576 integrate_v_rspace 119 11.3 0.002 0.003 3.534 3.539 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.232 3.240 calculate_rho_elec 119 8.7 0.059 0.061 3.232 3.239 calculate_first_density_matrix 1 7.0 0.000 0.000 3.136 3.136 qs_ot_get_p 119 10.4 0.001 0.001 2.806 2.846 jit_kernel_multiply 10 15.6 1.174 2.788 1.174 2.788 calculate_dm_sparse 119 9.5 0.000 0.000 2.705 2.723 init_scf_loop 11 6.9 0.000 0.000 2.498 2.499 mp_waitall_1 121218 16.5 1.704 2.194 1.704 2.194 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.121 2.126 make_m2s 4572 13.5 0.035 0.036 1.790 1.981 grid_integrate_task_list 119 12.3 1.846 1.903 1.846 1.903 make_images 4572 14.5 0.268 0.301 1.702 1.891 rs_pw_transfer 974 11.9 0.008 0.008 1.736 1.872 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.852 1.855 prepare_preconditioner 11 7.9 0.000 0.000 1.768 1.771 make_preconditioner 11 8.9 0.000 0.000 1.768 1.771 density_rs2pw 119 9.7 0.003 0.004 1.648 1.730 pw_transfer 1439 11.6 0.064 0.066 1.697 1.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.647 1.683 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.652 1.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.631 1.632 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.606 1.615 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.522 1.544 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.428 1.441 potential_pw2rs 119 12.3 0.010 0.011 1.361 1.365 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.363 1.363 grid_collocate_task_list 119 9.7 1.272 1.356 1.272 1.356 cp_fm_redistribute_end 50 14.0 0.673 1.336 0.674 1.337 cp_fm_diag_elpa_base 50 14.0 0.617 1.264 0.661 1.323 ot_diis_step 108 11.5 0.012 0.012 1.302 1.302 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.292 1.300 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.286 1.287 fft3d_ps 1201 14.6 0.537 0.551 1.273 1.281 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.175 1.198 apply_single 119 13.6 0.000 0.000 1.175 1.198 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.163 1.179 wfi_extrapolate 11 7.9 0.001 0.001 1.116 1.117 hybrid_alltoall_any 4725 16.4 0.063 0.175 0.801 1.077 make_images_data 4572 15.5 0.040 0.042 0.840 1.060 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.917 mp_alltoall_d11v 2130 13.8 0.803 0.899 0.803 0.899 acc_transpose_blocks 9144 15.5 0.038 0.039 0.890 0.896 cp_fm_cholesky_invert 11 10.9 0.890 0.893 0.890 0.893 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.698 0.794 multiply_cannon_sync_h2d 9144 15.5 0.707 0.780 0.707 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.756 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.742 0.750 mp_allgather_i34 2286 14.5 0.282 0.720 0.282 0.720 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.647 0.650 mp_alltoall_z22v 1201 16.6 0.610 0.643 0.610 0.643 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.321 0.626 mp_waitany 5200 13.7 0.467 0.621 0.467 0.621 dbcsr_complete_redistribute 329 12.2 0.165 0.182 0.562 0.589 rs_pw_transfer_RS2PW_140 130 11.5 0.101 0.103 0.433 0.577 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.813000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.000000, yerr=3.411211 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 747.786240E+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 0.020 0.047 43.161 43.162 qs_mol_dyn_low 1 2.0 0.003 0.003 42.938 42.945 qs_forces 11 3.9 0.002 0.003 42.876 42.880 qs_energies 11 4.9 0.002 0.002 40.883 40.891 scf_env_do_scf 11 5.9 0.001 0.001 34.347 34.347 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.236 26.238 velocity_verlet 10 3.0 0.002 0.002 23.528 23.533 dbcsr_multiply_generic 2286 12.5 0.099 0.099 19.259 19.371 qs_scf_new_mos 108 7.5 0.001 0.001 17.108 17.191 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.107 17.191 ot_scf_mini 108 9.5 0.002 0.002 15.992 16.086 multiply_cannon 2286 13.5 0.308 0.328 14.800 15.907 multiply_cannon_loop 2286 14.5 0.345 0.350 13.496 14.675 ot_mini 108 10.5 0.001 0.001 9.789 9.905 multiply_cannon_multrec 9144 15.5 3.426 4.964 9.079 9.497 init_scf_loop 11 6.9 0.000 0.000 8.079 8.085 qs_ot_get_derivative 108 11.5 0.001 0.001 7.751 7.847 dbcsr_mm_accdrv_process 12550 15.8 4.352 6.422 5.530 7.157 prepare_preconditioner 11 7.9 0.000 0.000 7.126 7.139 make_preconditioner 11 8.9 0.000 0.000 7.126 7.138 rebuild_ks_matrix 119 8.3 0.000 0.000 6.994 7.132 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.994 7.131 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.617 7.003 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.302 6.424 cp_fm_upper_to_full 72 14.2 3.183 4.622 3.183 4.622 mp_waitall_1 97218 16.6 3.079 4.470 3.079 4.470 init_scf_run 11 5.9 0.000 0.001 4.459 4.459 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.458 4.459 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.945 4.083 calculate_rho_elec 119 8.7 0.118 0.121 3.945 4.083 sum_up_and_integrate 119 10.3 0.064 0.066 3.764 3.788 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.011 3.761 integrate_v_rspace 119 11.3 0.004 0.005 3.699 3.724 qs_ot_get_p 119 10.4 0.001 0.001 3.423 3.559 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.515 3.299 apply_single 119 13.6 0.000 0.000 2.515 3.299 calculate_first_density_matrix 1 7.0 0.000 0.000 2.832 2.975 dbcsr_complete_redistribute 329 12.2 0.290 0.294 2.118 2.951 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.657 2.867 calculate_dm_sparse 119 9.5 0.000 0.000 2.667 2.799 mp_sum_l 7207 12.9 1.452 2.667 1.452 2.667 make_m2s 4572 13.5 0.038 0.039 2.468 2.655 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.767 2.582 make_images 4572 14.5 0.351 0.380 2.347 2.533 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.503 2.312 mp_alltoall_i22 627 13.8 1.468 2.310 1.468 2.310 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.105 2.106 pw_transfer 1439 11.6 0.066 0.067 2.094 2.097 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.076 2.077 grid_integrate_task_list 119 12.3 2.014 2.043 2.014 2.043 ot_diis_step 108 11.5 0.014 0.015 2.006 2.006 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.999 2.002 density_rs2pw 119 9.7 0.004 0.004 1.890 1.914 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.826 1.864 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.704 1.839 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.796 1.796 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.790 1.794 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.743 1.794 jit_kernel_multiply 8 15.4 1.151 1.760 1.151 1.760 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.648 1.651 fft3d_ps 1201 14.6 0.565 0.577 1.634 1.639 cp_fm_cholesky_invert 11 10.9 1.614 1.618 1.614 1.618 rs_pw_transfer 974 11.9 0.009 0.009 1.510 1.551 mp_alltoall_d11v 2130 13.8 1.379 1.542 1.379 1.542 grid_collocate_task_list 119 9.7 1.454 1.500 1.454 1.500 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.485 1.485 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.224 1.485 cp_fm_diag_elpa_base 50 14.0 1.335 1.387 1.481 1.481 make_images_data 4572 15.5 0.043 0.046 1.186 1.431 wfi_extrapolate 11 7.9 0.001 0.001 1.421 1.421 acc_transpose_blocks 9144 15.5 0.038 0.038 1.219 1.358 potential_pw2rs 119 12.3 0.014 0.014 1.273 1.277 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.177 1.202 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.102 1.125 acc_transpose_blocks_kernels 9144 16.5 0.117 0.118 0.968 1.108 multiply_cannon_sync_h2d 9144 15.5 1.040 1.045 1.040 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.988 1.031 jit_kernel_transpose 5 15.6 0.850 0.990 0.850 0.990 qs_create_task_list 11 7.9 0.000 0.000 0.944 0.957 generate_qs_task_list 11 8.9 0.373 0.394 0.944 0.957 mp_alltoall_z22v 1201 16.6 0.932 0.951 0.932 0.951 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.922 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.162000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=703.909091, yerr=14.920727 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.501952E+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 1801931. 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 0.031 0.063 87.611 87.618 qs_mol_dyn_low 1 2.0 0.003 0.003 87.231 87.263 qs_forces 11 3.9 0.002 0.003 84.838 84.839 qs_energies 11 4.9 0.001 0.001 82.032 82.044 scf_env_do_scf 11 5.9 0.000 0.001 72.609 72.611 scf_env_do_scf_inner_loop 99 6.5 0.017 0.067 66.758 66.758 dbcsr_multiply_generic 2055 12.4 0.106 0.111 53.793 54.070 qs_scf_new_mos 99 7.5 0.000 0.001 49.631 49.775 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.631 49.775 ot_scf_mini 99 9.5 0.002 0.002 47.113 47.224 velocity_verlet 10 3.0 0.012 0.046 44.547 44.572 multiply_cannon 2055 13.4 0.174 0.182 43.465 44.267 multiply_cannon_loop 2055 14.4 1.572 1.608 42.491 43.357 ot_mini 99 10.5 0.001 0.001 28.789 28.901 qs_ot_get_derivative 99 11.5 0.001 0.001 21.894 22.008 multiply_cannon_multrec 49320 15.4 12.033 12.720 17.442 18.158 rebuild_ks_matrix 110 8.3 0.000 0.001 14.504 14.681 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.503 14.681 mp_waitall_1 241148 16.1 12.517 13.543 12.517 13.543 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.761 12.919 multiply_cannon_sync_h2d 49320 15.4 9.864 10.555 9.864 10.555 qs_ot_get_p 110 10.4 0.001 0.001 9.881 9.988 multiply_cannon_metrocomm3 49320 15.4 0.081 0.086 7.436 8.913 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.339 8.014 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.296 7.715 apply_single 110 13.6 0.000 0.001 7.296 7.715 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.412 7.479 sum_up_and_integrate 110 10.3 0.037 0.044 7.080 7.093 integrate_v_rspace 110 11.3 0.002 0.003 7.042 7.064 init_scf_run 11 5.9 0.000 0.001 6.967 6.967 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.967 6.967 mp_sum_l 6514 12.8 5.959 6.669 5.959 6.669 ot_diis_step 99 11.5 0.005 0.005 6.576 6.577 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.263 6.407 calculate_rho_elec 110 8.6 0.020 0.024 6.263 6.406 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.255 6.278 init_scf_loop 11 6.9 0.000 0.000 5.821 5.821 dbcsr_mm_accdrv_process 87628 16.1 2.098 2.207 5.288 5.645 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.207 5.207 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.700 4.702 cp_fm_redistribute_end 48 14.0 4.089 4.674 4.094 4.675 cp_fm_diag_elpa_base 48 14.0 0.575 4.609 0.579 4.635 rs_pw_transfer 902 11.9 0.012 0.013 3.932 4.355 make_m2s 4110 13.4 0.061 0.066 3.911 4.026 calculate_dm_sparse 110 9.5 0.000 0.001 3.879 4.013 wfi_extrapolate 11 7.9 0.001 0.001 3.998 3.998 make_images 4110 14.4 0.177 0.191 3.815 3.932 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.909 3.915 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.841 3.845 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.487 3.538 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.441 3.511 density_rs2pw 110 9.6 0.004 0.004 3.320 3.456 prepare_preconditioner 11 7.9 0.000 0.000 3.418 3.437 make_preconditioner 11 8.9 0.000 0.000 3.418 3.437 grid_integrate_task_list 110 12.3 3.230 3.387 3.230 3.387 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.194 3.241 jit_kernel_multiply 13 15.9 2.906 3.075 2.906 3.075 pw_transfer 1331 11.6 0.054 0.067 2.892 2.959 calculate_first_density_matrix 1 7.0 0.000 0.001 2.886 2.890 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.804 2.875 potential_pw2rs 110 12.3 0.006 0.007 2.681 2.708 mp_alltoall_d11v 2046 13.8 2.157 2.604 2.157 2.604 mp_sum_dm 438 4.9 2.403 2.468 2.403 2.468 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.342 2.414 fft3d_ps 1111 14.6 0.744 0.829 2.341 2.401 mp_waitany 14300 13.8 1.933 2.387 1.933 2.387 update_particle_set 20 4.0 0.000 0.000 2.339 2.375 md_output 10 3.0 0.000 0.000 0.028 2.362 md_write_output 11 3.9 0.021 1.938 0.026 2.340 mp_sum_d 3879 11.9 1.743 2.293 1.743 2.293 acc_transpose_blocks 49320 15.4 0.223 0.234 2.185 2.256 grid_collocate_task_list 110 9.6 2.088 2.215 2.088 2.215 cp_fm_cholesky_invert 11 10.9 1.901 1.905 1.901 1.905 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.883 1.899 make_images_data 4110 15.4 0.044 0.047 1.773 1.898 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.844 1.845 hybrid_alltoall_any 4261 16.3 0.084 0.485 1.538 1.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=87.618000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=1.928473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 586.522624E+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 3003626. 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 0.013 0.031 68.535 68.536 qs_mol_dyn_low 1 2.0 0.003 0.003 68.216 68.225 qs_forces 11 3.9 0.002 0.002 68.120 68.121 qs_energies 11 4.9 0.002 0.006 64.834 64.839 scf_env_do_scf 11 5.9 0.001 0.002 56.419 56.422 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.693 48.694 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.588 37.812 velocity_verlet 10 3.0 0.001 0.002 36.056 36.058 qs_scf_new_mos 99 7.5 0.001 0.001 32.825 32.953 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.825 32.953 multiply_cannon 2055 13.4 0.221 0.244 30.926 32.191 ot_scf_mini 99 9.5 0.003 0.003 31.153 31.269 multiply_cannon_loop 2055 14.4 0.927 0.952 29.673 30.550 ot_mini 99 10.5 0.001 0.001 18.236 18.367 multiply_cannon_multrec 24660 15.4 7.621 9.030 13.782 15.370 rebuild_ks_matrix 110 8.3 0.000 0.001 13.650 13.742 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.649 13.742 qs_ot_get_derivative 99 11.5 0.001 0.001 12.390 12.514 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.048 12.130 mp_waitall_1 186928 16.3 7.962 10.228 7.962 10.228 multiply_cannon_sync_h2d 24660 15.4 7.029 7.977 7.029 7.977 multiply_cannon_metrocomm3 24660 15.4 0.070 0.071 5.204 7.925 init_scf_loop 11 6.9 0.001 0.004 7.686 7.687 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.522 7.244 apply_single 110 13.6 0.000 0.001 6.522 7.244 sum_up_and_integrate 110 10.3 0.052 0.059 6.532 6.543 integrate_v_rspace 110 11.3 0.002 0.002 6.479 6.493 qs_ot_get_p 110 10.4 0.001 0.001 6.281 6.427 dbcsr_mm_accdrv_process 52282 16.1 4.659 5.593 6.002 6.288 init_scf_run 11 5.9 0.000 0.001 6.052 6.052 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.051 6.052 ot_diis_step 99 11.5 0.010 0.010 5.799 5.799 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.780 5.790 calculate_rho_elec 110 8.6 0.039 0.047 5.780 5.790 prepare_preconditioner 11 7.9 0.000 0.000 5.645 5.669 make_preconditioner 11 8.9 0.000 0.001 5.645 5.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.230 5.395 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.655 5.389 make_m2s 4110 13.4 0.057 0.059 4.252 4.671 make_images 4110 14.4 0.402 0.448 4.147 4.562 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.368 4.387 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.936 3.936 wfi_extrapolate 11 7.9 0.001 0.001 3.497 3.497 pw_transfer 1331 11.6 0.066 0.070 3.294 3.432 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.379 3.381 cp_fm_redistribute_end 48 14.0 2.515 3.352 2.518 3.352 grid_integrate_task_list 110 12.3 3.149 3.346 3.149 3.346 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.187 3.326 cp_fm_diag_elpa_base 48 14.0 0.797 3.237 0.831 3.325 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.186 3.257 density_rs2pw 110 9.6 0.004 0.004 3.080 3.227 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.222 3.224 rs_pw_transfer 902 11.9 0.012 0.013 2.986 3.132 calculate_dm_sparse 110 9.5 0.001 0.001 2.975 3.008 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.884 2.911 make_images_data 4110 15.4 0.047 0.051 2.402 2.840 fft_wrap_pw1pw2_140 451 13.1 0.200 0.218 2.687 2.828 hybrid_alltoall_any 4261 16.3 0.104 0.446 2.083 2.703 fft3d_ps 1111 14.6 1.069 1.260 2.543 2.672 cp_fm_cholesky_invert 11 10.9 2.645 2.652 2.645 2.652 calculate_first_density_matrix 1 7.0 0.000 0.002 2.465 2.468 potential_pw2rs 110 12.3 0.008 0.009 2.395 2.411 mp_sum_l 6514 12.8 1.786 2.399 1.786 2.399 grid_collocate_task_list 110 9.6 2.120 2.277 2.120 2.277 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.967 1.987 mp_alltoall_d11v 2046 13.8 1.777 1.971 1.777 1.971 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.847 1.848 multiply_cannon_metrocomm4 22605 15.4 0.075 0.079 0.787 1.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.767 1.776 jit_kernel_multiply 9 16.4 0.995 1.740 0.995 1.740 mp_irecv_dv 57340 16.2 0.661 1.635 0.661 1.635 mp_allgather_i34 2055 14.4 0.581 1.619 0.581 1.619 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.549 acc_transpose_blocks 24660 15.4 0.110 0.115 1.515 1.529 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.478 dbcsr_complete_redistribute 325 12.2 0.250 0.315 1.176 1.441 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.536000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.272727, yerr=7.033568 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 659.726336E+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 0.047 0.098 60.839 60.840 qs_mol_dyn_low 1 2.0 0.003 0.003 60.042 60.051 qs_forces 11 3.9 0.002 0.002 59.977 59.978 qs_energies 11 4.9 0.003 0.011 56.802 56.806 scf_env_do_scf 11 5.9 0.001 0.001 48.340 48.340 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.688 39.688 velocity_verlet 10 3.0 0.002 0.002 32.623 32.624 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.895 29.122 qs_scf_new_mos 99 7.5 0.001 0.001 25.326 25.405 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.325 25.405 ot_scf_mini 99 9.5 0.002 0.003 24.074 24.178 multiply_cannon 2055 13.4 0.212 0.224 22.557 23.756 multiply_cannon_loop 2055 14.4 0.615 0.630 21.273 22.280 ot_mini 99 10.5 0.001 0.001 13.642 13.751 rebuild_ks_matrix 110 8.3 0.000 0.000 12.102 12.261 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.101 12.261 multiply_cannon_multrec 16440 15.4 3.951 5.165 10.079 11.155 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.667 10.806 mp_waitall_1 146766 16.3 7.137 9.793 7.137 9.793 qs_ot_get_derivative 99 11.5 0.001 0.001 9.177 9.280 init_scf_loop 11 6.9 0.001 0.003 8.608 8.609 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.256 7.036 prepare_preconditioner 11 7.9 0.000 0.000 6.873 6.889 make_preconditioner 11 8.9 0.000 0.001 6.873 6.889 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.223 6.565 sum_up_and_integrate 110 10.3 0.060 0.060 6.369 6.383 integrate_v_rspace 110 11.3 0.002 0.003 6.309 6.323 dbcsr_mm_accdrv_process 34862 16.1 4.817 5.742 5.984 6.209 init_scf_run 11 5.9 0.000 0.001 5.934 5.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.933 5.934 qs_ot_get_p 110 10.4 0.001 0.001 5.603 5.739 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.463 5.470 calculate_rho_elec 110 8.6 0.058 0.058 5.462 5.469 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.994 5.424 apply_single 110 13.6 0.000 0.000 4.994 5.423 make_m2s 4110 13.4 0.050 0.052 4.145 4.586 make_images 4110 14.4 0.395 0.511 4.032 4.475 multiply_cannon_sync_h2d 16440 15.4 3.711 4.463 3.711 4.463 ot_diis_step 99 11.5 0.011 0.011 4.431 4.431 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.952 3.956 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.168 3.827 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.596 3.597 grid_integrate_task_list 110 12.3 3.190 3.373 3.190 3.373 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.186 3.188 pw_transfer 1331 11.6 0.065 0.073 3.040 3.055 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.054 3.055 cp_fm_redistribute_end 48 14.0 1.893 3.022 1.895 3.022 rs_pw_transfer 902 11.9 0.010 0.011 2.846 3.013 wfi_extrapolate 11 7.9 0.001 0.001 2.999 2.999 cp_fm_diag_elpa_base 48 14.0 1.064 2.888 1.123 2.998 density_rs2pw 110 9.6 0.004 0.004 2.778 2.963 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.933 2.950 make_images_data 4110 15.4 0.044 0.048 2.363 2.906 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.847 2.891 calculate_first_density_matrix 1 7.0 0.001 0.002 2.850 2.851 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.085 2.848 cp_fm_cholesky_invert 11 10.9 2.793 2.799 2.793 2.799 calculate_dm_sparse 110 9.5 0.001 0.001 2.527 2.552 fft_wrap_pw1pw2_140 451 13.1 0.209 0.213 2.467 2.487 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.358 2.420 grid_collocate_task_list 110 9.6 2.176 2.349 2.176 2.349 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.855 2.349 fft3d_ps 1111 14.6 1.062 1.101 2.247 2.265 potential_pw2rs 110 12.3 0.011 0.011 2.243 2.252 mp_irecv_dv 48980 15.7 0.786 2.223 0.786 2.223 mp_sum_l 6514 12.8 1.585 2.205 1.585 2.205 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.039 2.040 mp_alltoall_d11v 2046 13.8 1.710 2.000 1.710 2.000 dbcsr_complete_redistribute 325 12.2 0.320 0.343 1.454 1.921 jit_kernel_multiply 8 16.3 0.783 1.843 0.783 1.843 cp_fm_upper_to_full 70 13.6 1.370 1.842 1.370 1.842 cp_fm_cholesky_decompose 22 10.9 1.647 1.663 1.647 1.663 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.636 1.655 mp_allgather_i34 2055 14.4 0.603 1.633 0.603 1.633 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.480 1.495 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.354 1.488 mp_waitany 17072 13.8 1.259 1.451 1.259 1.451 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.983 1.430 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.286 1.290 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.142 1.220 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.840000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.272727, yerr=8.528998 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 723.619840E+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 0.025 0.086 66.686 66.687 qs_mol_dyn_low 1 2.0 0.003 0.003 66.269 66.279 qs_forces 11 3.9 0.008 0.011 66.204 66.205 qs_energies 11 4.9 0.003 0.018 62.818 62.827 scf_env_do_scf 11 5.9 0.001 0.004 53.849 53.854 scf_env_do_scf_inner_loop 99 6.5 0.035 0.262 41.735 41.736 velocity_verlet 10 3.0 0.001 0.002 37.294 37.296 dbcsr_multiply_generic 2055 12.4 0.113 0.116 30.298 30.538 qs_scf_new_mos 99 7.5 0.001 0.001 26.819 26.923 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.818 26.922 ot_scf_mini 99 9.5 0.003 0.003 25.180 25.280 multiply_cannon 2055 13.4 0.243 0.266 23.108 24.295 multiply_cannon_loop 2055 14.4 0.890 0.912 21.743 22.343 ot_mini 99 10.5 0.001 0.001 14.471 14.568 multiply_cannon_multrec 24660 15.4 4.291 7.030 13.108 14.503 rebuild_ks_matrix 110 8.3 0.000 0.000 12.150 12.286 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.035 12.150 12.285 init_scf_loop 11 6.9 0.001 0.004 12.068 12.072 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.765 10.890 qs_ot_get_derivative 99 11.5 0.001 0.001 10.315 10.410 prepare_preconditioner 11 7.9 0.000 0.000 10.200 10.218 make_preconditioner 11 8.9 0.000 0.001 10.200 10.218 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.384 9.880 dbcsr_mm_accdrv_process 52304 16.0 7.119 8.505 8.671 9.637 mp_waitall_1 126806 16.4 4.700 6.781 4.700 6.781 sum_up_and_integrate 110 10.3 0.067 0.070 6.453 6.465 integrate_v_rspace 110 11.3 0.003 0.003 6.385 6.397 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.024 6.066 calculate_rho_elec 110 8.6 0.077 0.080 6.023 6.066 init_scf_run 11 5.9 0.000 0.001 5.953 5.954 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.953 5.953 qs_ot_get_p 110 10.4 0.001 0.001 5.691 5.831 make_m2s 4110 13.4 0.059 0.061 5.531 5.813 make_images 4110 14.4 0.580 0.701 5.390 5.668 cp_fm_upper_to_full 70 13.8 3.364 4.783 3.364 4.783 ot_diis_step 99 11.5 0.011 0.011 4.117 4.118 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.026 4.113 apply_single 110 13.6 0.000 0.000 4.026 4.113 dbcsr_complete_redistribute 325 12.2 0.416 0.456 2.733 3.903 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.883 3.898 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.427 3.428 grid_integrate_task_list 110 12.3 3.279 3.427 3.279 3.427 pw_transfer 1331 11.6 0.065 0.073 3.388 3.415 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.399 3.399 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.335 3.393 calculate_dm_sparse 110 9.5 0.001 0.001 3.345 3.374 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.186 3.336 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.283 3.312 multiply_cannon_sync_h2d 24660 15.4 3.165 3.309 3.165 3.309 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.061 3.112 make_images_data 4110 15.4 0.047 0.051 2.809 3.106 hybrid_alltoall_any 4261 16.3 0.121 0.461 2.366 3.059 density_rs2pw 110 9.6 0.004 0.004 2.891 3.020 wfi_extrapolate 11 7.9 0.001 0.001 3.007 3.007 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.807 2.942 mp_alltoall_i22 605 13.7 1.692 2.880 1.692 2.880 calculate_first_density_matrix 1 7.0 0.001 0.007 2.854 2.857 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.842 2.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.835 2.836 cp_fm_redistribute_end 48 14.0 1.412 2.809 1.413 2.810 cp_fm_cholesky_invert 11 10.9 2.791 2.800 2.791 2.800 cp_fm_diag_elpa_base 48 14.0 1.311 2.676 1.393 2.788 rs_pw_transfer 902 11.9 0.010 0.011 2.631 2.771 fft_wrap_pw1pw2_140 451 13.1 0.201 0.211 2.723 2.755 jit_kernel_multiply 13 15.6 1.220 2.752 1.220 2.752 fft3d_ps 1111 14.6 1.064 1.096 2.585 2.601 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.470 2.503 qs_energies_init_hamiltonians 11 5.9 0.004 0.010 2.475 2.482 grid_collocate_task_list 110 9.6 2.222 2.394 2.222 2.394 mp_alltoall_d11v 2046 13.8 2.164 2.317 2.164 2.317 potential_pw2rs 110 12.3 0.012 0.013 2.193 2.204 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.951 1.967 cp_fm_cholesky_decompose 22 10.9 1.722 1.758 1.722 1.758 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.708 1.748 acc_transpose_blocks 24660 15.4 0.104 0.107 1.705 1.727 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.602 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.648 mp_sum_l 6514 12.8 1.011 1.561 1.011 1.561 multiply_cannon_metrocomm4 20550 15.4 0.057 0.062 0.856 1.549 mp_allgather_i34 2055 14.4 0.463 1.516 0.463 1.516 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.386 1.485 mp_irecv_dv 62702 16.1 0.755 1.472 0.755 1.472 mp_waitany 13376 13.8 1.190 1.422 1.190 1.422 mp_alltoall_z22v 1111 16.6 1.305 1.368 1.305 1.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.687000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.636364, yerr=7.474474 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 823.644160E+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 0.032 0.341 56.573 56.574 qs_mol_dyn_low 1 2.0 0.003 0.003 56.007 56.016 qs_forces 11 3.9 0.002 0.002 55.937 55.938 qs_energies 11 4.9 0.001 0.001 52.310 52.315 scf_env_do_scf 11 5.9 0.000 0.001 43.403 43.404 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.334 35.335 velocity_verlet 10 3.0 0.002 0.002 31.104 31.107 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.695 23.810 qs_scf_new_mos 99 7.5 0.001 0.001 21.177 21.226 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.176 21.225 ot_scf_mini 99 9.5 0.002 0.002 19.919 19.949 multiply_cannon 2055 13.4 0.246 0.258 18.031 19.386 multiply_cannon_loop 2055 14.4 0.325 0.336 16.634 16.931 rebuild_ks_matrix 110 8.3 0.000 0.000 11.489 11.515 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.488 11.515 ot_mini 99 10.5 0.001 0.001 10.742 10.770 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.239 10.264 multiply_cannon_multrec 8220 15.4 3.382 4.732 8.203 9.195 init_scf_loop 11 6.9 0.000 0.000 8.010 8.011 mp_waitall_1 106626 16.5 6.140 7.821 6.140 7.821 qs_ot_get_derivative 99 11.5 0.001 0.001 6.906 6.936 prepare_preconditioner 11 7.9 0.000 0.000 6.373 6.378 make_preconditioner 11 8.9 0.000 0.000 6.373 6.378 sum_up_and_integrate 110 10.3 0.080 0.081 6.218 6.229 integrate_v_rspace 110 11.3 0.003 0.003 6.137 6.150 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.945 6.030 dbcsr_mm_accdrv_process 17442 15.9 3.100 4.239 4.691 5.812 init_scf_run 11 5.9 0.000 0.001 5.779 5.779 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.779 5.779 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.640 5.654 calculate_rho_elec 110 8.6 0.115 0.115 5.640 5.653 qs_ot_get_p 110 10.4 0.001 0.001 5.259 5.293 make_m2s 4110 13.4 0.039 0.040 4.260 4.501 make_images 4110 14.4 0.639 0.698 4.130 4.369 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.938 4.308 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.880 3.884 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.737 3.811 apply_single 110 13.6 0.000 0.000 3.737 3.811 ot_diis_step 99 11.5 0.012 0.012 3.810 3.811 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.572 3.572 grid_integrate_task_list 110 12.3 3.357 3.505 3.357 3.505 pw_transfer 1331 11.6 0.065 0.071 3.288 3.304 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.181 3.201 cp_fm_cholesky_invert 11 10.9 3.164 3.168 3.164 3.168 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.033 3.034 multiply_cannon_sync_h2d 8220 15.4 2.910 3.028 2.910 3.028 cp_fm_redistribute_end 48 14.0 0.765 3.002 0.770 3.002 cp_fm_diag_elpa_base 48 14.0 2.047 2.807 2.224 2.971 calculate_first_density_matrix 1 7.0 0.000 0.000 2.926 2.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.851 2.852 density_rs2pw 110 9.6 0.004 0.004 2.701 2.832 make_images_data 4110 15.4 0.039 0.044 2.421 2.815 calculate_dm_sparse 110 9.5 0.001 0.001 2.720 2.764 wfi_extrapolate 11 7.9 0.001 0.001 2.761 2.761 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.720 2.745 hybrid_alltoall_any 4261 16.3 0.201 0.864 2.331 2.703 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.646 2.646 grid_collocate_task_list 110 9.6 2.326 2.497 2.326 2.497 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.455 2.466 fft3d_ps 1111 14.6 1.117 1.141 2.438 2.453 rs_pw_transfer 902 11.9 0.010 0.010 2.231 2.379 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.963 1.987 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.986 jit_kernel_multiply 8 15.8 1.284 1.985 1.284 1.985 potential_pw2rs 110 12.3 0.015 0.016 1.954 1.959 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.855 1.870 mp_alltoall_d11v 2046 13.8 1.682 1.869 1.682 1.869 cp_fm_cholesky_decompose 22 10.9 1.757 1.769 1.757 1.769 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.666 1.676 dbcsr_complete_redistribute 325 12.2 0.571 0.623 1.537 1.634 mp_allgather_i34 2055 14.4 0.563 1.631 0.563 1.631 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.495 1.614 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.580 1.584 qs_create_task_list 11 7.9 0.000 0.000 1.225 1.326 generate_qs_task_list 11 8.9 0.379 0.447 1.224 1.325 mp_waitany 9240 13.8 1.088 1.254 1.088 1.254 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.478 1.240 mp_sum_l 6514 12.8 1.003 1.208 1.003 1.208 mp_irecv_dv 24056 15.7 0.454 1.198 0.454 1.198 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.175 1.194 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.574000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.090909, yerr=8.479436 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 1.365709E+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 0.040 0.055 87.066 87.066 qs_mol_dyn_low 1 2.0 0.003 0.003 86.152 86.160 qs_forces 11 3.9 0.006 0.007 86.056 86.057 qs_energies 11 4.9 0.001 0.001 81.999 82.002 scf_env_do_scf 11 5.9 0.001 0.001 72.068 72.069 velocity_verlet 10 3.0 0.037 0.038 55.339 55.345 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.449 43.450 dbcsr_multiply_generic 2055 12.4 0.118 0.120 29.001 29.053 init_scf_loop 11 6.9 0.000 0.000 28.546 28.548 qs_scf_new_mos 99 7.5 0.001 0.001 26.792 26.823 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.792 26.822 prepare_preconditioner 11 7.9 0.000 0.000 26.562 26.568 make_preconditioner 11 8.9 0.000 0.000 26.562 26.568 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.741 26.028 ot_scf_mini 99 9.5 0.002 0.002 25.019 25.048 multiply_cannon 2055 13.4 0.353 0.388 21.647 22.387 multiply_cannon_loop 2055 14.4 0.341 0.344 19.817 20.210 cp_fm_upper_to_full 70 14.2 12.727 18.256 12.727 18.256 ot_mini 99 10.5 0.001 0.001 13.878 13.901 rebuild_ks_matrix 110 8.3 0.001 0.001 13.283 13.336 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.282 13.335 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.085 12.133 dbcsr_complete_redistribute 325 12.2 1.045 1.082 7.470 10.702 multiply_cannon_multrec 8220 15.4 4.380 4.576 9.605 9.705 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.394 9.635 qs_ot_get_derivative 99 11.5 0.001 0.001 9.210 9.246 mp_waitall_1 87304 16.6 8.191 9.060 8.191 9.060 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.805 9.008 mp_alltoall_i22 605 13.7 5.423 8.678 5.423 8.678 sum_up_and_integrate 110 10.3 0.151 0.152 6.691 6.705 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.628 6.663 calculate_rho_elec 110 8.6 0.227 0.227 6.628 6.662 integrate_v_rspace 110 11.3 0.004 0.004 6.539 6.554 make_m2s 4110 13.4 0.043 0.044 5.550 6.082 qs_ot_get_p 110 10.4 0.001 0.001 5.892 5.917 make_images 4110 14.4 0.878 0.916 5.361 5.893 init_scf_run 11 5.9 0.000 0.001 5.666 5.666 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.665 5.666 cp_fm_cholesky_invert 11 10.9 5.417 5.421 5.417 5.421 dbcsr_mm_accdrv_process 11614 15.7 3.243 3.640 5.083 5.357 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.865 5.277 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.780 5.227 apply_single 110 13.6 0.000 0.000 4.780 5.226 ot_diis_step 99 11.5 0.015 0.016 4.638 4.639 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.259 4.267 multiply_cannon_sync_h2d 8220 15.4 3.949 3.956 3.949 3.956 pw_transfer 1331 11.6 0.073 0.074 3.906 3.917 hybrid_alltoall_any 4261 16.3 0.257 0.554 3.120 3.909 make_images_data 4110 15.4 0.042 0.045 3.134 3.819 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.809 3.811 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.790 3.801 grid_integrate_task_list 110 12.3 3.664 3.736 3.664 3.736 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.662 3.663 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.912 3.367 wfi_extrapolate 11 7.9 0.001 0.001 3.285 3.285 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.209 3.265 calculate_dm_sparse 110 9.5 0.001 0.001 3.163 3.189 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.186 3.186 cp_fm_diag_elpa_base 48 14.0 2.655 2.847 3.184 3.184 density_rs2pw 110 9.6 0.004 0.004 3.117 3.138 fft3d_ps 1111 14.6 1.274 1.309 2.996 3.008 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.906 2.911 grid_collocate_task_list 110 9.6 2.632 2.658 2.632 2.658 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.463 2.480 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.379 2.395 calculate_first_density_matrix 1 7.0 0.000 0.000 2.274 2.274 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.183 2.243 rs_pw_transfer 902 11.9 0.011 0.011 2.188 2.223 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.183 mp_alltoall_d11v 2046 13.8 2.108 2.153 2.108 2.153 potential_pw2rs 110 12.3 0.021 0.021 2.103 2.108 cp_fm_cholesky_decompose 22 10.9 2.035 2.047 2.035 2.047 qs_create_task_list 11 7.9 0.000 0.000 1.900 1.944 generate_qs_task_list 11 8.9 0.739 0.793 1.900 1.944 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.837 1.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.802 1.807 jit_kernel_multiply 10 15.4 1.638 1.766 1.638 1.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.066000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1209.909091, yerr=73.900984 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 631.812096E+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 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 2633587. 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 0.014 0.052 205.742 205.744 qs_mol_dyn_low 1 2.0 0.003 0.003 205.259 205.273 qs_forces 11 3.9 0.004 0.009 205.173 205.178 qs_energies 11 4.9 0.001 0.002 199.707 199.721 scf_env_do_scf 11 5.9 0.001 0.001 182.984 182.988 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 161.738 161.740 dbcsr_multiply_generic 2507 12.6 0.175 0.180 124.758 125.425 qs_scf_new_mos 117 7.6 0.001 0.001 122.893 123.153 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.892 123.152 velocity_verlet 10 3.0 0.001 0.002 122.636 122.637 ot_scf_mini 117 9.6 0.003 0.003 116.218 116.408 multiply_cannon 2507 13.6 0.236 0.246 101.078 102.989 multiply_cannon_loop 2507 14.6 2.141 2.212 98.794 100.604 ot_mini 117 10.6 0.001 0.001 65.939 66.121 multiply_cannon_multrec 60168 15.6 33.000 35.130 41.336 42.950 qs_ot_get_derivative 117 11.6 0.001 0.001 41.117 41.302 rebuild_ks_matrix 128 8.3 0.001 0.001 33.306 33.584 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.305 33.583 mp_waitall_1 291448 16.2 29.351 32.379 29.351 32.379 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.988 30.271 qs_ot_get_p 128 10.4 0.001 0.001 28.553 28.835 multiply_cannon_sync_h2d 60168 15.6 27.070 28.705 27.070 28.705 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.201 24.675 apply_single 128 13.6 0.001 0.001 24.200 24.675 ot_diis_step 117 11.6 0.007 0.008 24.436 24.437 qs_ot_p2m_diag 83 11.4 0.077 0.091 21.379 21.456 init_scf_loop 11 6.9 0.000 0.000 21.168 21.169 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.200 19.370 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.634 18.636 multiply_cannon_metrocomm3 60168 15.6 0.119 0.125 15.860 17.959 prepare_preconditioner 11 7.9 0.000 0.000 16.256 16.310 make_preconditioner 11 8.9 0.000 0.000 16.256 16.309 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.760 15.766 cp_fm_redistribute_end 83 14.4 12.420 15.687 12.435 15.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.488 15.664 cp_fm_diag_elpa_base 83 14.4 3.211 15.459 3.244 15.584 make_m2s 5014 13.6 0.102 0.112 13.952 14.234 make_images 5014 14.6 0.405 0.426 13.778 14.063 sum_up_and_integrate 128 10.3 0.090 0.109 13.977 13.994 integrate_v_rspace 128 11.3 0.003 0.004 13.887 13.903 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.689 12.795 calculate_rho_elec 128 8.7 0.045 0.063 12.688 12.794 init_scf_run 11 5.9 0.000 0.001 12.571 12.571 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.571 12.571 mp_sum_l 7870 13.0 8.479 9.696 8.479 9.696 cp_fm_cholesky_invert 11 10.9 9.365 9.372 9.365 9.372 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 6.425 9.205 wfi_extrapolate 11 7.9 0.001 0.001 9.028 9.028 calculate_dm_sparse 128 9.5 0.001 0.001 8.470 8.571 dbcsr_mm_accdrv_process 124484 16.2 3.300 3.466 7.900 8.465 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.941 8.033 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.866 7.991 make_images_data 5014 15.6 0.070 0.077 6.907 7.898 grid_integrate_task_list 128 12.3 7.110 7.503 7.110 7.503 hybrid_alltoall_any 5200 16.5 0.295 2.280 6.005 7.444 pw_transfer 1547 11.6 0.075 0.105 6.889 7.162 density_rs2pw 128 9.7 0.006 0.007 6.494 7.046 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.685 6.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.672 6.680 rs_pw_transfer 1046 11.9 0.017 0.019 5.667 6.285 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.821 5.963 fft_wrap_pw1pw2_140 523 13.2 0.446 0.510 5.750 5.945 fft3d_ps 1291 14.7 2.092 2.563 5.474 5.667 mp_alltoall_d11v 2415 14.1 4.228 5.354 4.228 5.354 grid_collocate_task_list 128 9.7 4.730 5.206 4.730 5.206 cp_fm_cholesky_decompose 22 10.9 4.694 4.707 4.694 4.707 mp_sum_d 4469 12.2 3.883 4.495 3.883 4.495 potential_pw2rs 128 12.3 0.009 0.011 4.470 4.494 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.744000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.181818, yerr=8.386333 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 827.940864E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56957. MP_Allreduce 11297 945. MP_Sync 170 MP_Alltoall 1969 4910959. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.078 190.489 190.490 qs_mol_dyn_low 1 2.0 0.005 0.019 189.720 189.733 qs_forces 11 3.9 0.013 0.032 189.586 189.590 qs_energies 11 4.9 0.004 0.018 182.868 182.877 scf_env_do_scf 11 5.9 0.001 0.001 166.292 166.302 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.527 133.529 velocity_verlet 10 3.0 0.003 0.013 118.599 118.603 dbcsr_multiply_generic 2507 12.6 0.183 0.189 97.599 98.663 qs_scf_new_mos 117 7.6 0.001 0.001 95.007 95.431 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.006 95.430 ot_scf_mini 117 9.6 0.004 0.005 90.223 90.786 multiply_cannon 2507 13.6 0.481 0.535 77.641 82.151 multiply_cannon_loop 2507 14.6 1.250 1.289 74.350 76.901 ot_mini 117 10.6 0.001 0.004 50.307 50.795 mp_waitall_1 226760 16.4 24.575 39.166 24.575 39.166 multiply_cannon_multrec 30084 15.6 22.127 26.811 32.133 37.345 rebuild_ks_matrix 128 8.3 0.001 0.001 32.758 33.221 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.031 32.757 33.220 init_scf_loop 11 6.9 0.000 0.000 32.677 32.678 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.531 29.951 qs_ot_get_derivative 117 11.6 0.001 0.002 28.039 28.571 prepare_preconditioner 11 7.9 0.000 0.000 28.293 28.345 make_preconditioner 11 8.9 0.000 0.000 28.293 28.345 multiply_cannon_metrocomm3 30084 15.6 0.093 0.098 15.464 28.182 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.990 27.513 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 21.978 23.088 apply_single 128 13.6 0.001 0.001 21.978 23.087 qs_ot_get_p 128 10.4 0.001 0.001 21.708 22.227 ot_diis_step 117 11.6 0.060 0.387 22.100 22.103 multiply_cannon_sync_h2d 30084 15.6 19.344 21.623 19.344 21.623 qs_ot_p2m_diag 83 11.4 0.219 0.467 16.882 16.913 cp_fm_cholesky_invert 11 10.9 16.543 16.555 16.543 16.555 make_m2s 5014 13.6 0.092 0.098 14.262 15.884 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.717 15.750 make_images 5014 14.6 1.170 1.355 14.058 15.678 sum_up_and_integrate 128 10.3 0.116 0.134 14.318 14.348 integrate_v_rspace 128 11.3 0.004 0.004 14.202 14.236 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.840 12.878 calculate_rho_elec 128 8.7 0.088 0.105 12.840 12.877 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.413 12.421 cp_fm_redistribute_end 83 14.4 7.222 12.353 7.235 12.357 cp_fm_diag_elpa_base 83 14.4 4.890 11.889 5.105 12.261 init_scf_run 11 5.9 0.000 0.001 11.792 11.793 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.792 11.793 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.885 11.247 make_images_data 5014 15.6 0.066 0.074 8.670 10.680 multiply_cannon_metrocomm4 27577 15.6 0.099 0.114 3.723 10.464 mp_irecv_dv 69486 16.3 3.525 10.074 3.525 10.074 dbcsr_mm_accdrv_process 62242 16.2 4.728 5.767 9.460 10.066 hybrid_alltoall_any 5200 16.5 0.344 1.505 7.399 9.853 wfi_extrapolate 11 7.9 0.003 0.016 8.300 8.300 pw_transfer 1547 11.6 0.086 0.103 7.732 7.791 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.507 7.561 grid_integrate_task_list 128 12.3 7.160 7.542 7.160 7.542 density_rs2pw 128 9.7 0.006 0.006 6.662 7.302 cp_fm_cholesky_decompose 22 10.9 6.914 7.003 6.914 7.003 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.153 6.984 calculate_dm_sparse 128 9.5 0.001 0.001 6.449 6.630 fft_wrap_pw1pw2_140 523 13.2 0.470 0.523 6.566 6.610 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.141 6.150 rs_pw_transfer 1046 11.9 0.014 0.017 5.499 6.062 fft3d_ps 1291 14.7 2.767 2.919 5.877 5.912 mp_sum_l 7870 13.0 3.956 5.516 3.956 5.516 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.275 5.409 grid_collocate_task_list 128 9.7 4.907 5.380 4.907 5.380 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.293 5.351 mp_alltoall_d11v 2415 14.1 4.104 5.160 4.104 5.160 mp_allgather_i34 2507 14.6 1.854 4.861 1.854 4.861 potential_pw2rs 128 12.3 0.015 0.017 4.764 4.784 mp_sum_d 4469 12.2 2.956 4.383 2.956 4.383 dbcsr_complete_redistribute 395 12.7 0.788 0.875 3.110 3.967 dbcsr_dot_sd 1318 12.0 0.777 0.854 2.756 3.908 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.490000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.454545, yerr=1.233151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/17/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 2.928533E+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 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 934.567936E+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 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. 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 0.050 0.135 175.745 175.745 qs_mol_dyn_low 1 2.0 0.003 0.003 175.083 175.096 qs_forces 11 3.9 0.003 0.003 174.986 174.994 qs_energies 11 4.9 0.002 0.005 168.522 168.535 scf_env_do_scf 11 5.9 0.001 0.001 152.916 152.916 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 118.112 118.112 velocity_verlet 10 3.0 0.002 0.002 111.622 111.624 dbcsr_multiply_generic 2507 12.6 0.178 0.185 81.881 83.357 qs_scf_new_mos 117 7.6 0.001 0.001 81.803 82.147 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.803 82.146 ot_scf_mini 117 9.6 0.003 0.004 77.676 78.070 multiply_cannon 2507 13.6 0.527 0.743 61.610 66.495 multiply_cannon_loop 2507 14.6 0.857 0.881 58.399 61.663 ot_mini 117 10.6 0.001 0.001 43.141 43.529 init_scf_loop 11 6.9 0.000 0.002 34.705 34.705 mp_waitall_1 178456 16.5 24.855 33.518 24.855 33.518 prepare_preconditioner 11 7.9 0.000 0.000 30.705 30.748 make_preconditioner 11 8.9 0.000 0.001 30.705 30.748 rebuild_ks_matrix 128 8.3 0.001 0.001 29.858 30.381 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.857 30.381 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.340 29.712 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.981 27.465 multiply_cannon_multrec 20056 15.6 13.543 16.629 22.303 25.319 multiply_cannon_metrocomm3 20056 15.6 0.061 0.066 14.895 24.692 qs_ot_get_derivative 117 11.6 0.001 0.002 23.469 23.865 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.230 21.509 apply_single 128 13.6 0.001 0.001 20.229 21.509 qs_ot_get_p 128 10.4 0.001 0.001 20.116 20.585 ot_diis_step 117 11.6 0.018 0.018 19.566 19.566 multiply_cannon_sync_h2d 20056 15.6 14.204 15.877 14.204 15.877 qs_ot_p2m_diag 83 11.4 0.266 0.273 15.776 15.787 make_m2s 5014 13.6 0.083 0.087 14.662 15.574 make_images 5014 14.6 1.195 1.294 14.432 15.343 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.767 14.769 cp_fm_cholesky_invert 11 10.9 14.337 14.346 14.337 14.346 sum_up_and_integrate 128 10.3 0.133 0.144 14.061 14.085 integrate_v_rspace 128 11.3 0.004 0.004 13.928 13.953 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.907 12.932 calculate_rho_elec 128 8.7 0.131 0.147 12.906 12.932 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.599 11.600 cp_fm_redistribute_end 83 14.4 4.355 11.536 4.370 11.538 cp_fm_diag_elpa_base 83 14.4 6.745 10.970 7.147 11.449 init_scf_run 11 5.9 0.000 0.001 10.584 10.585 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.584 10.585 make_images_data 5014 15.6 0.062 0.069 8.954 10.249 hybrid_alltoall_any 5200 16.5 0.433 1.970 7.766 9.432 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.042 9.332 multiply_cannon_metrocomm4 17549 15.6 0.060 0.070 3.404 9.236 mp_irecv_dv 50230 16.2 3.283 8.988 3.283 8.988 dbcsr_mm_accdrv_process 41502 16.2 4.444 5.271 8.215 8.358 grid_integrate_task_list 128 12.3 7.309 7.769 7.309 7.769 pw_transfer 1547 11.6 0.085 0.104 7.598 7.709 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.375 7.493 cp_fm_upper_to_full 105 14.5 5.685 7.342 5.685 7.342 cp_fm_cholesky_decompose 22 10.9 7.272 7.306 7.272 7.306 wfi_extrapolate 11 7.9 0.001 0.001 7.285 7.285 density_rs2pw 128 9.7 0.006 0.006 6.434 6.747 fft_wrap_pw1pw2_140 523 13.2 0.478 0.530 6.482 6.610 dbcsr_complete_redistribute 395 12.7 1.180 1.212 4.591 6.328 calculate_dm_sparse 128 9.5 0.001 0.001 5.717 5.851 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.716 5.824 fft3d_ps 1291 14.7 2.694 2.904 5.642 5.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.669 5.676 grid_collocate_task_list 128 9.7 5.076 5.520 5.076 5.520 rs_pw_transfer 1046 11.9 0.014 0.014 5.101 5.482 mp_sum_l 7870 13.0 3.798 5.339 3.798 5.339 copy_fm_to_dbcsr 209 11.7 0.002 0.003 3.419 5.158 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.600 4.913 mp_alltoall_d11v 2415 14.1 4.167 4.783 4.167 4.783 potential_pw2rs 128 12.3 0.020 0.022 4.441 4.460 mp_allgather_i34 2507 14.6 1.700 4.441 1.700 4.441 mp_alltoall_i22 716 14.1 2.148 4.093 2.148 4.093 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.345 4.046 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.011 4.041 qs_energies_init_hamiltonians 11 5.9 0.001 0.007 3.754 3.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.745000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=887.636364, yerr=8.657868 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 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 1.164689E+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 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57274. MP_Allreduce 11236 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 0.014 0.037 188.954 188.954 qs_mol_dyn_low 1 2.0 0.003 0.003 188.597 188.610 qs_forces 11 3.9 0.003 0.003 188.487 188.492 qs_energies 11 4.9 0.001 0.002 181.443 181.453 scf_env_do_scf 11 5.9 0.001 0.001 164.808 164.819 velocity_verlet 10 3.0 0.002 0.002 124.067 124.070 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.137 117.138 qs_scf_new_mos 116 7.6 0.001 0.001 81.926 82.207 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.925 82.206 dbcsr_multiply_generic 2485 12.5 0.185 0.190 80.066 81.033 ot_scf_mini 116 9.6 0.003 0.003 77.438 77.713 multiply_cannon 2485 13.5 0.553 0.586 55.327 58.397 multiply_cannon_loop 2485 14.5 1.177 1.202 51.641 53.734 init_scf_loop 11 6.9 0.000 0.000 47.536 47.537 ot_mini 116 10.6 0.001 0.001 43.199 43.478 prepare_preconditioner 11 7.9 0.000 0.000 43.095 43.120 make_preconditioner 11 8.9 0.000 0.000 43.095 43.120 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.664 41.785 multiply_cannon_multrec 29820 15.5 14.052 18.971 26.100 30.711 rebuild_ks_matrix 127 8.3 0.001 0.001 28.842 29.080 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 28.842 29.080 mp_waitall_1 152434 16.5 17.800 26.864 17.800 26.864 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.037 26.243 qs_ot_get_derivative 116 11.6 0.001 0.002 23.648 23.917 make_m2s 4970 13.5 0.095 0.099 20.296 21.324 qs_ot_get_p 127 10.4 0.001 0.001 20.875 21.221 make_images 4970 14.5 1.960 2.281 19.996 21.024 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.912 19.574 apply_single 127 13.6 0.001 0.001 18.912 19.574 ot_diis_step 116 11.6 0.017 0.018 19.435 19.436 cp_fm_upper_to_full 104 14.7 11.315 16.750 11.315 16.750 cp_fm_cholesky_invert 11 10.9 16.575 16.585 16.575 16.585 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.332 16.382 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.562 15.653 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.067 15.069 sum_up_and_integrate 127 10.3 0.141 0.152 14.069 14.092 integrate_v_rspace 127 11.3 0.004 0.004 13.928 13.955 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.024 13.065 calculate_rho_elec 127 8.7 0.174 0.189 13.023 13.065 dbcsr_complete_redistribute 393 12.7 1.522 1.639 9.069 12.875 make_images_data 4970 15.5 0.065 0.068 10.904 12.640 multiply_cannon_sync_h2d 29820 15.5 11.664 12.610 11.664 12.610 dbcsr_mm_accdrv_process 61748 16.2 7.304 8.420 11.622 12.171 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.970 11.972 cp_fm_redistribute_end 82 14.4 2.039 11.885 2.054 11.892 cp_fm_diag_elpa_base 82 14.4 9.210 11.288 9.801 11.771 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.663 11.473 hybrid_alltoall_any 5155 16.4 0.524 2.205 9.691 11.471 init_scf_run 11 5.9 0.000 0.001 10.750 10.752 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.750 10.752 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.409 10.117 mp_alltoall_i22 712 14.1 5.720 9.641 5.720 9.641 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.438 9.635 pw_transfer 1535 11.6 0.085 0.099 7.785 7.857 grid_integrate_task_list 127 12.3 7.498 7.779 7.498 7.779 cp_fm_cholesky_decompose 22 10.9 7.671 7.776 7.671 7.776 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.561 7.639 wfi_extrapolate 11 7.9 0.001 0.001 7.532 7.532 density_rs2pw 127 9.7 0.005 0.006 6.245 6.781 fft_wrap_pw1pw2_140 519 13.2 0.477 0.483 6.677 6.779 calculate_dm_sparse 127 9.5 0.001 0.001 6.146 6.221 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.711 6.198 mp_irecv_dv 75445 16.2 2.567 5.930 2.567 5.930 fft3d_ps 1281 14.7 2.768 2.832 5.814 5.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.602 5.641 grid_collocate_task_list 127 9.7 5.167 5.585 5.167 5.585 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.276 5.358 rs_pw_transfer 1038 11.9 0.013 0.015 4.719 5.247 mp_alltoall_d11v 2401 14.1 4.674 5.052 4.674 5.052 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.527 4.625 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.390 4.392 potential_pw2rs 127 12.3 0.022 0.023 4.355 4.371 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.191 4.251 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.954000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1084.636364, yerr=24.929985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.494696E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58095. MP_Allreduce 11022 1174. MP_Sync 87 MP_Alltoall 1712 18838222. 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 0.064 0.067 174.473 174.473 qs_mol_dyn_low 1 2.0 0.003 0.003 173.337 173.349 qs_forces 11 3.9 0.003 0.003 173.240 173.248 qs_energies 11 4.9 0.001 0.002 165.901 165.909 scf_env_do_scf 11 5.9 0.001 0.001 147.017 147.030 velocity_verlet 10 3.0 0.004 0.004 112.439 112.443 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 110.964 110.966 dbcsr_multiply_generic 2507 12.6 0.180 0.184 75.651 76.122 qs_scf_new_mos 117 7.6 0.001 0.001 75.722 75.855 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.722 75.854 ot_scf_mini 117 9.6 0.004 0.004 71.350 71.454 multiply_cannon 2507 13.6 0.587 0.629 55.858 60.066 multiply_cannon_loop 2507 14.6 0.449 0.458 50.852 51.770 ot_mini 117 10.6 0.001 0.001 39.644 39.741 init_scf_loop 11 6.9 0.000 0.000 35.899 35.902 mp_waitall_1 129618 16.6 25.771 33.564 25.771 33.564 prepare_preconditioner 11 7.9 0.000 0.000 32.020 32.045 make_preconditioner 11 8.9 0.000 0.000 32.020 32.045 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.782 30.058 rebuild_ks_matrix 128 8.3 0.001 0.001 28.360 28.462 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.360 28.462 qs_ks_update_qs_env 128 7.6 0.001 0.001 25.850 25.944 multiply_cannon_multrec 10028 15.6 10.451 14.918 19.113 22.024 ot_diis_step 117 11.6 0.020 0.021 19.868 19.868 qs_ot_get_derivative 117 11.6 0.002 0.002 19.710 19.815 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.397 19.717 apply_single 128 13.6 0.001 0.001 19.396 19.717 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.146 19.356 make_m2s 5014 13.6 0.068 0.073 16.144 18.884 qs_ot_get_p 128 10.4 0.001 0.001 18.701 18.795 make_images 5014 14.6 2.334 2.878 15.838 18.580 cp_fm_cholesky_invert 11 10.9 18.238 18.244 18.238 18.244 qs_ot_p2m_diag 83 11.4 0.496 0.501 14.843 14.860 sum_up_and_integrate 128 10.3 0.182 0.192 13.909 13.954 integrate_v_rspace 128 11.3 0.004 0.005 13.726 13.779 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.698 13.699 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.966 12.997 calculate_rho_elec 128 8.7 0.259 0.270 12.966 12.996 make_images_data 5014 15.6 0.055 0.061 9.790 12.347 multiply_cannon_sync_h2d 10028 15.6 11.635 12.199 11.635 12.199 init_scf_run 11 5.9 0.000 0.001 12.094 12.094 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.093 12.094 hybrid_alltoall_any 5200 16.5 0.815 3.578 9.548 11.965 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.659 10.660 cp_fm_diag_elpa_base 83 14.4 10.401 10.486 10.651 10.651 dbcsr_mm_accdrv_process 20762 16.1 2.659 3.477 8.294 8.945 cp_fm_cholesky_decompose 22 10.9 8.012 8.156 8.012 8.156 grid_integrate_task_list 128 12.3 7.774 8.037 7.774 8.037 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.856 7.926 pw_transfer 1547 11.6 0.084 0.091 7.503 7.526 mp_allgather_i34 2507 14.6 3.205 7.507 3.205 7.507 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.281 7.311 wfi_extrapolate 11 7.9 0.001 0.001 7.294 7.294 multiply_cannon_metrocomm1 10028 15.6 0.028 0.028 4.339 7.272 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.617 6.633 fft_wrap_pw1pw2_140 523 13.2 0.505 0.524 6.401 6.446 calculate_dm_sparse 128 9.5 0.001 0.001 6.043 6.134 density_rs2pw 128 9.7 0.005 0.006 5.831 6.040 dbcsr_complete_redistribute 395 12.7 2.173 2.332 5.365 5.703 grid_collocate_task_list 128 9.7 5.486 5.671 5.486 5.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.495 5.525 fft3d_ps 1291 14.7 2.728 2.815 5.440 5.484 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.263 5.265 mp_alltoall_d11v 2415 14.1 4.654 5.089 4.654 5.089 calculate_first_density_matrix 1 7.0 0.000 0.000 4.602 4.603 rs_pw_transfer 1046 11.9 0.013 0.013 4.119 4.356 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.852 4.096 potential_pw2rs 128 12.3 0.026 0.026 4.007 4.023 mp_irecv_dv 28860 15.9 1.815 4.011 1.815 4.011 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.575 3.908 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.557 3.878 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.685 3.718 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.534 3.633 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.473000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1386.909091, yerr=42.354466 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 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 1986255912960 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 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 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 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 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 3.021480E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - 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 0.072 289.667 289.667 qs_mol_dyn_low 1 2.0 0.003 0.003 289.126 289.138 qs_forces 11 3.9 0.003 0.003 288.326 288.328 qs_energies 11 4.9 0.002 0.002 279.766 279.777 scf_env_do_scf 11 5.9 0.001 0.001 257.238 257.246 velocity_verlet 10 3.0 0.048 0.049 209.203 209.294 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.449 132.452 init_scf_loop 11 6.9 0.000 0.000 124.529 124.532 prepare_preconditioner 11 7.9 0.000 0.000 119.953 119.984 make_preconditioner 11 8.9 0.000 0.000 119.953 119.984 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.452 117.112 qs_scf_new_mos 118 7.6 0.001 0.001 90.503 90.634 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.503 90.633 ot_scf_mini 118 9.6 0.004 0.004 85.691 85.725 dbcsr_multiply_generic 2529 12.6 0.213 0.219 83.817 84.353 cp_fm_upper_to_full 106 14.8 53.466 76.963 53.466 76.963 multiply_cannon 2529 13.6 0.711 0.765 59.602 60.319 multiply_cannon_loop 2529 14.6 0.477 0.492 55.945 57.292 ot_mini 118 10.6 0.001 0.001 44.426 44.456 dbcsr_complete_redistribute 397 12.7 4.009 4.042 30.209 43.464 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.799 40.047 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.454 37.560 mp_alltoall_i22 720 14.1 22.281 35.772 22.281 35.772 rebuild_ks_matrix 129 8.3 0.001 0.001 33.403 33.439 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.403 33.439 cp_fm_cholesky_invert 11 10.9 33.054 33.060 33.054 33.060 mp_waitall_1 106386 16.7 27.971 31.811 27.971 31.811 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.221 31.263 qs_ot_get_p 129 10.4 0.001 0.001 25.767 25.775 qs_ot_get_derivative 118 11.6 0.002 0.002 24.182 24.216 qs_ot_p2m_diag 84 11.4 0.890 0.895 21.756 21.785 make_m2s 5058 13.6 0.078 0.080 19.722 20.706 make_images 5058 14.6 3.767 3.950 19.242 20.228 ot_diis_step 118 11.6 0.022 0.023 20.218 20.219 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.029 20.031 multiply_cannon_metrocomm3 10116 15.6 0.024 0.025 18.678 19.780 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.417 19.619 apply_single 129 13.6 0.001 0.001 19.416 19.619 multiply_cannon_multrec 10116 15.6 10.650 12.472 18.971 19.050 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.835 16.836 cp_fm_diag_elpa_base 84 14.4 12.418 14.047 16.832 16.832 multiply_cannon_sync_h2d 10116 15.6 15.777 15.788 15.777 15.788 sum_up_and_integrate 129 10.3 0.324 0.326 15.609 15.696 integrate_v_rspace 129 11.3 0.004 0.004 15.285 15.372 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.873 14.900 calculate_rho_elec 129 8.7 0.487 0.488 14.872 14.899 init_scf_run 11 5.9 0.000 0.001 12.775 12.775 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.775 12.775 hybrid_alltoall_any 5245 16.5 1.311 3.059 10.687 12.569 make_images_data 5058 15.6 0.061 0.065 10.478 12.169 dbcsr_mm_accdrv_process 20934 16.1 3.920 5.938 8.086 9.985 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.479 9.510 cp_fm_cholesky_decompose 22 10.9 8.729 8.789 8.729 8.789 grid_integrate_task_list 129 12.3 8.599 8.780 8.599 8.780 wfi_extrapolate 11 7.9 0.001 0.001 8.542 8.542 pw_transfer 1559 11.6 0.091 0.093 8.234 8.242 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 7.999 8.007 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.760 7.761 fft_wrap_pw1pw2_140 527 13.2 0.545 0.547 7.115 7.127 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.045 7.113 mp_alltoall_d11v 2429 14.1 6.791 6.914 6.791 6.914 calculate_dm_sparse 129 9.5 0.001 0.001 6.645 6.766 grid_collocate_task_list 129 9.7 6.388 6.420 6.388 6.420 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.068 6.109 fft3d_ps 1301 14.7 2.778 2.784 6.022 6.029 density_rs2pw 129 9.7 0.005 0.005 5.912 5.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.667000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2665.272727, yerr=184.898051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 1.261462E+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 12225489. 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 0.030 0.041 85.566 85.567 qs_energies 1 2.0 0.000 0.000 84.544 84.553 ls_scf 1 3.0 0.000 0.001 83.641 83.651 dbcsr_multiply_generic 111 6.7 0.016 0.017 72.788 72.937 multiply_cannon 111 7.7 0.017 0.020 56.040 57.110 multiply_cannon_loop 111 8.7 0.213 0.224 52.575 53.923 ls_scf_main 1 4.0 0.000 0.000 52.319 52.319 density_matrix_trs4 2 5.0 0.002 0.003 46.864 46.949 ls_scf_init_scf 1 4.0 0.000 0.002 28.278 28.279 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.158 27.224 mp_waitall_1 11316 10.9 22.674 26.240 22.674 26.240 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.047 25.068 multiply_cannon_multrec 2664 9.7 8.187 8.831 15.619 17.193 multiply_cannon_sync_h2d 2664 9.7 13.463 15.505 13.463 15.505 make_m2s 222 7.7 0.009 0.012 13.115 13.698 make_images 222 8.7 0.102 0.111 13.093 13.677 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.877 12.699 make_images_data 222 9.7 0.005 0.006 7.688 8.210 dbcsr_mm_accdrv_process 4760 10.4 0.520 0.617 7.047 8.050 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.441 8.007 hybrid_alltoall_any 227 10.6 0.218 1.847 6.557 7.809 dbcsr_mm_accdrv_process_sort 4760 11.4 6.328 7.263 6.328 7.263 calculate_norms 4752 9.8 5.511 6.144 5.511 6.144 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.066 5.205 mp_sum_l 807 5.4 3.090 4.321 3.090 4.321 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.074 3.545 mp_irecv_dv 6231 10.9 2.056 3.514 2.056 3.514 make_images_sizes 222 9.7 0.000 0.000 0.719 3.481 mp_alltoall_i44 222 10.7 0.719 3.481 0.719 3.481 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.313 3.302 arnoldi_extremal 4 6.8 0.000 0.000 3.201 3.232 arnoldi_normal_ev 4 7.8 0.001 0.003 3.201 3.232 build_subspace 16 8.4 0.009 0.012 3.108 3.110 ls_scf_post 1 4.0 0.000 0.000 3.045 3.055 ls_scf_store_result 1 5.0 0.000 0.000 2.870 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.343 2.734 dbcsr_merge_single_wm 555 10.7 0.460 0.581 2.334 2.725 make_images_pack 222 9.7 2.216 2.632 2.218 2.633 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.310 2.552 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.456 2.067 2.458 dbcsr_sort_data 658 11.4 2.132 2.449 2.132 2.449 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.203 2.283 buffer_matrices_ensure_size 222 8.7 1.760 2.060 1.760 2.060 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.772 1.773 rebuild_ks_matrix 3 7.3 0.000 0.000 1.760 1.761 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.760 1.761 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.567000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 2.110816E+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 0.109 0.166 92.285 92.288 qs_energies 1 2.0 0.001 0.012 91.041 91.050 ls_scf 1 3.0 0.000 0.000 89.727 89.730 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.576 75.905 multiply_cannon 111 7.7 0.029 0.043 53.508 57.151 ls_scf_main 1 4.0 0.000 0.000 55.088 55.095 multiply_cannon_loop 111 8.7 0.117 0.126 50.199 53.016 density_matrix_trs4 2 5.0 0.002 0.003 49.407 49.620 ls_scf_init_scf 1 4.0 0.000 0.001 31.102 31.104 mp_waitall_1 9246 10.9 21.283 29.886 21.283 29.886 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.583 29.694 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.179 27.193 multiply_cannon_multrec 1332 9.7 13.175 16.327 22.537 26.924 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.878 20.748 make_m2s 222 7.7 0.008 0.008 15.375 16.101 make_images 222 8.7 1.587 1.960 15.345 16.072 dbcsr_mm_accdrv_process 4041 10.4 0.285 0.454 8.959 10.556 dbcsr_mm_accdrv_process_sort 4041 11.4 8.551 10.104 8.551 10.104 make_images_data 222 9.7 0.004 0.005 8.829 9.809 hybrid_alltoall_any 227 10.6 0.522 2.443 8.251 9.215 mp_sum_l 807 5.4 5.393 8.295 5.393 8.295 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.251 7.982 mp_irecv_dv 3311 11.0 3.231 7.928 3.231 7.928 calculate_norms 2376 9.8 6.006 6.798 6.006 6.798 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.121 6.645 multiply_cannon_sync_h2d 1332 9.7 4.755 5.869 4.755 5.869 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.052 5.261 arnoldi_extremal 4 6.8 0.000 0.000 4.588 4.609 arnoldi_normal_ev 4 7.8 0.001 0.004 4.588 4.609 build_subspace 16 8.4 0.014 0.021 4.338 4.340 ls_scf_post 1 4.0 0.000 0.000 3.536 3.541 ls_scf_store_result 1 5.0 0.000 0.000 3.225 3.358 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.122 3.354 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.218 2.741 3.220 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.203 2.863 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.523 2.609 mp_allgather_i34 111 8.7 0.994 2.401 0.994 2.401 make_images_pack 222 9.7 2.029 2.371 2.032 2.373 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.191 2.193 rebuild_ks_matrix 3 7.3 0.000 0.000 2.177 2.179 qs_ks_build_kohn_sham_matrix 3 8.3 0.047 0.376 2.177 2.179 dbcsr_sort_data 436 11.2 1.841 2.064 1.841 2.064 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.288000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 2.716754E+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 0.029 0.053 95.899 95.901 qs_energies 1 2.0 0.000 0.000 94.915 94.921 ls_scf 1 3.0 0.000 0.000 93.524 93.530 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.837 78.142 ls_scf_main 1 4.0 0.000 0.000 58.332 58.337 multiply_cannon 111 7.7 0.036 0.073 53.338 57.214 multiply_cannon_loop 111 8.7 0.101 0.119 49.765 54.257 density_matrix_trs4 2 5.0 0.002 0.003 52.478 52.669 mp_waitall_1 7374 11.0 24.255 34.414 24.255 34.414 ls_scf_init_scf 1 4.0 0.000 0.001 31.378 31.380 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.241 30.319 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.827 27.841 multiply_cannon_multrec 888 9.7 12.622 15.415 21.281 24.677 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.361 23.661 make_m2s 222 7.7 0.007 0.008 17.202 18.423 make_images 222 8.7 1.978 2.304 17.163 18.383 hybrid_alltoall_any 227 10.6 0.622 2.842 9.386 10.835 make_images_data 222 9.7 0.004 0.005 9.761 10.827 mp_sum_l 807 5.4 5.728 9.905 5.728 9.905 dbcsr_mm_accdrv_process 3754 10.4 0.259 0.415 8.179 9.393 dbcsr_mm_accdrv_process_sort 3754 11.4 7.809 8.978 7.809 8.978 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.522 7.977 multiply_cannon_sync_h2d 888 9.7 5.952 7.329 5.952 7.329 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.480 6.950 mp_irecv_dv 2335 11.1 2.464 6.891 2.464 6.891 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.758 6.857 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.941 5.139 arnoldi_extremal 4 6.8 0.000 0.000 5.047 5.057 arnoldi_normal_ev 4 7.8 0.002 0.009 5.047 5.057 build_subspace 16 8.4 0.014 0.020 4.735 4.745 calculate_norms 1584 9.8 4.315 4.725 4.315 4.725 mp_allgather_i34 111 8.7 1.398 3.892 1.398 3.892 ls_scf_post 1 4.0 0.000 0.000 3.814 3.820 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.414 3.749 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.592 3.021 3.594 ls_scf_store_result 1 5.0 0.000 0.000 3.365 3.477 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.723 2.802 make_images_sizes 222 9.7 0.000 0.000 1.116 2.232 mp_alltoall_i44 222 10.7 1.116 2.232 1.116 2.232 make_images_pack 222 9.7 1.816 2.122 1.819 2.124 dbcsr_sort_data 325 11.1 1.862 2.107 1.862 2.107 dbcsr_data_release 9322 10.9 1.333 1.930 1.333 1.930 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.901000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2211.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 3.340202E+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 0.099 0.126 97.681 97.682 qs_energies 1 2.0 0.000 0.000 96.300 96.305 ls_scf 1 3.0 0.000 0.000 94.629 94.636 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.606 78.832 ls_scf_main 1 4.0 0.000 0.000 58.846 58.847 multiply_cannon 111 7.7 0.065 0.142 51.836 55.688 density_matrix_trs4 2 5.0 0.002 0.003 52.817 52.939 multiply_cannon_loop 111 8.7 0.115 0.126 46.805 49.389 ls_scf_init_scf 1 4.0 0.000 0.001 32.547 32.550 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.417 31.478 mp_waitall_1 6438 11.0 22.800 29.914 22.800 29.914 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.955 28.969 multiply_cannon_multrec 1332 9.7 14.347 17.593 22.349 25.397 make_m2s 222 7.7 0.008 0.009 21.265 22.632 make_images 222 8.7 3.151 3.635 21.214 22.584 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.151 16.945 make_images_data 222 9.7 0.004 0.005 11.900 13.488 hybrid_alltoall_any 227 10.6 0.799 3.787 11.250 13.164 dbcsr_mm_accdrv_process 3641 10.4 0.214 0.416 7.637 9.169 dbcsr_mm_accdrv_process_sort 3641 11.4 7.262 8.748 7.262 8.748 mp_sum_l 807 5.4 3.912 7.104 3.912 7.104 multiply_cannon_sync_h2d 1332 9.7 5.539 6.256 5.539 6.256 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.065 5.981 mp_irecv_dv 3229 10.9 2.043 5.910 2.043 5.910 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.020 5.537 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.582 5.370 arnoldi_extremal 4 6.8 0.000 0.000 5.185 5.201 arnoldi_normal_ev 4 7.8 0.001 0.005 5.185 5.201 build_subspace 16 8.4 0.014 0.021 4.845 4.854 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.530 4.721 calculate_norms 2376 9.8 4.182 4.540 4.182 4.540 mp_allgather_i34 111 8.7 2.128 4.421 2.128 4.421 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.566 3.875 dbcsr_matrix_vector_mult_local 304 10.0 3.196 3.699 3.198 3.701 dbcsr_sort_data 658 11.4 3.072 3.473 3.072 3.473 dbcsr_special_finalize 555 9.7 0.006 0.008 2.828 3.270 dbcsr_merge_single_wm 555 10.7 0.538 0.690 2.820 3.262 ls_scf_post 1 4.0 0.000 0.000 3.236 3.240 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.038 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.914 2.973 dbcsr_data_release 10477 10.7 1.599 2.419 1.599 2.419 dbcsr_finalize 304 7.8 0.050 0.062 1.808 1.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.682000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 4.616876E+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 0.077 0.092 99.918 99.919 qs_energies 1 2.0 0.000 0.000 99.103 99.106 ls_scf 1 3.0 0.000 0.000 97.177 97.182 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.338 78.569 ls_scf_main 1 4.0 0.000 0.000 62.343 62.343 multiply_cannon 111 7.7 0.099 0.216 55.666 60.609 density_matrix_trs4 2 5.0 0.002 0.003 55.301 55.406 multiply_cannon_loop 111 8.7 0.070 0.079 51.100 52.779 mp_waitall_1 5481 11.0 26.748 31.216 26.748 31.216 ls_scf_init_scf 1 4.0 0.000 0.001 30.908 30.913 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.737 29.769 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.551 27.557 multiply_cannon_multrec 444 9.7 14.206 16.704 21.335 24.933 make_m2s 222 7.7 0.006 0.007 17.856 20.355 make_images 222 8.7 3.735 4.439 17.793 20.293 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.838 16.579 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.615 15.432 make_images_data 222 9.7 0.003 0.004 10.095 12.534 hybrid_alltoall_any 227 10.6 0.790 3.779 9.791 12.363 multiply_cannon_sync_h2d 444 9.7 6.572 8.731 6.572 8.731 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.347 6.838 7.964 dbcsr_mm_accdrv_process_sort 3003 11.4 6.522 7.621 6.522 7.621 mp_allgather_i34 111 8.7 2.788 7.014 2.788 7.014 arnoldi_extremal 4 6.8 0.000 0.000 5.850 5.857 arnoldi_normal_ev 4 7.8 0.002 0.005 5.850 5.857 build_subspace 16 8.4 0.015 0.020 5.465 5.474 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.636 4.790 mp_sum_l 807 5.4 2.837 4.575 2.837 4.575 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.184 4.390 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.574 4.188 dbcsr_matrix_vector_mult_local 304 10.0 3.704 4.168 3.706 4.170 mp_irecv_dv 1241 11.2 1.553 4.155 1.553 4.155 ls_scf_post 1 4.0 0.000 0.000 3.926 3.930 calculate_norms 792 9.8 3.524 3.668 3.524 3.668 make_images_sizes 222 9.7 0.000 0.000 1.089 3.588 mp_alltoall_i44 222 10.7 1.088 3.587 1.088 3.587 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.935 3.540 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.400 3.508 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.444 dbcsr_finalize 304 7.8 0.062 0.078 2.213 2.325 dbcsr_merge_all 275 8.9 0.476 0.528 2.058 2.155 dbcsr_data_release 10123 10.8 1.344 2.004 1.344 2.004 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.919000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3614.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0a7d28b76a9fad31202ff85cf948797d4075042b_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 8.732987E+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 0.102 108.934 108.935 qs_energies 1 2.0 0.000 0.000 107.370 107.391 ls_scf 1 3.0 0.000 0.000 104.051 104.072 dbcsr_multiply_generic 111 6.7 0.023 0.027 78.036 78.156 ls_scf_main 1 4.0 0.000 0.000 66.305 66.305 density_matrix_trs4 2 5.0 0.002 0.003 57.371 57.448 multiply_cannon 111 7.7 0.121 0.162 50.280 52.016 multiply_cannon_loop 111 8.7 0.068 0.070 46.850 47.712 ls_scf_init_scf 1 4.0 0.001 0.001 34.064 34.064 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.672 32.684 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.842 29.853 mp_waitall_1 4569 11.1 22.685 26.498 22.685 26.498 make_m2s 222 7.7 0.007 0.007 24.059 24.976 make_images 222 8.7 4.600 4.981 23.949 24.866 multiply_cannon_multrec 444 9.7 17.882 18.696 22.548 23.171 hybrid_alltoall_any 227 10.6 1.665 3.626 12.961 15.902 make_images_data 222 9.7 0.003 0.004 13.244 15.625 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.838 11.500 multiply_cannon_sync_h2d 444 9.7 8.845 8.878 8.845 8.878 arnoldi_extremal 4 6.8 0.000 0.000 7.436 7.448 arnoldi_normal_ev 4 7.8 0.003 0.009 7.436 7.448 build_subspace 16 8.4 0.026 0.037 6.890 6.900 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.523 5.702 dbcsr_matrix_vector_mult_local 304 10.0 5.073 5.408 5.076 5.410 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.069 5.339 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.852 4.944 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.304 4.503 4.638 dbcsr_mm_accdrv_process_sort 1814 11.4 4.204 4.340 4.204 4.340 ls_scf_post 1 4.0 0.000 0.000 3.683 3.704 make_images_sizes 222 9.7 0.000 0.000 1.450 3.648 mp_alltoall_i44 222 10.7 1.450 3.648 1.450 3.648 ls_scf_store_result 1 5.0 0.000 0.000 3.430 3.441 mp_allgather_i34 111 8.7 1.041 3.290 1.041 3.290 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 3.288 3.288 calculate_norms 792 9.8 3.242 3.278 3.242 3.278 dbcsr_finalize 304 7.8 0.082 0.090 3.093 3.137 dbcsr_complete_redistribute 5 7.6 1.445 1.481 2.799 2.935 dbcsr_merge_all 275 8.9 0.895 0.921 2.877 2.914 dbcsr_data_release 12724 10.6 2.329 2.876 2.329 2.876 matrix_ls_to_qs 2 6.0 0.000 0.000 2.440 2.593 dbcsr_sort_data 325 11.1 2.440 2.504 2.440 2.504 dbcsr_new_transposed 4 7.5 0.243 0.252 2.314 2.332 dbcsr_frobenius_norm 74 6.6 2.055 2.138 2.192 2.224 dbcsr_add_d 103 6.2 0.000 0.000 2.138 2.214 dbcsr_add_anytype 103 7.2 0.860 0.894 2.138 2.214 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.195 2.196 mp_sum_l 807 5.4 1.367 2.183 1.367 2.183 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.935000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6882.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 0a7d28b76a9fad31202ff85cf948797d4075042b Summary: empty Status: OK