=== 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: f36b488847463310f25cb3cd7662924bf5e00627 ################# 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/01 job id: 42529009 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/02 job id: 42529010 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/03 job id: 42529011 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/04 job id: 42529012 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/05 job id: 42529013 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/06 job id: 42529014 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/07 job id: 42529015 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/08 job id: 42529016 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/09 job id: 42529017 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/10 job id: 42529018 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/11 job id: 42529019 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/12 job id: 42529020 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/13 job id: 42529021 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/14 job id: 42529022 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/15 job id: 42529023 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/16 job id: 42529024 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/17 job id: 42529025 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/18 job id: 42529026 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/19 job id: 42529027 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/20 job id: 42529028 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/21 job id: 42529029 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/22 job id: 42529030 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/23 job id: 42529031 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/24 job id: 42529032 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/25 job id: 42529033 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/26 job id: 42529034 --- 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/f36b488847463310f25cb3cd7662924bf5e00627_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 0.023 0.037 133.038 133.039 farming_run 1 2.0 132.570 132.573 133.010 133.015 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.449648E+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 0.074 0.306 115.758 115.758 qs_energies 1 2.0 0.005 0.037 114.985 115.206 mp2_main 1 3.0 0.013 0.105 113.016 113.159 mp2_gpw_main 1 4.0 0.022 0.051 112.153 112.205 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.195 93.613 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.523 55.940 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.168 41.833 47.162 get_2c_integrals 1 6.0 0.000 0.000 36.466 37.500 integrate_v_rspace 273 8.0 0.436 0.449 25.133 30.192 pw_transfer 6555 10.6 0.373 0.387 27.303 27.816 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.986 26.524 grid_integrate_task_list 273 9.0 20.946 26.491 20.946 26.491 fft_wrap_pw1pw2_100 2178 12.4 1.161 1.262 23.553 24.100 compute_2c_integrals 1 7.0 0.002 0.002 19.171 19.172 rpa_ri_compute_en 1 5.0 0.003 0.026 18.845 19.037 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.821 18.889 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.432 18.819 18.886 cp_fm_cholesky_decompose 12 8.2 17.253 18.270 17.253 18.270 cholesky_decomp 1 7.0 0.000 0.000 16.133 17.144 fft3d_s 5443 13.4 16.110 16.536 16.132 16.557 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.546 10.856 15.546 calculate_wavefunction 272 8.0 5.487 5.627 12.605 13.244 rpa_num_int 1 6.0 0.000 0.002 10.663 10.663 rpa_num_int_RPA_matrix_operati 8 7.0 0.002 0.017 10.569 10.623 calc_mat_Q 8 8.0 0.000 0.000 9.394 9.515 contract_S_to_Q 8 9.0 0.002 0.016 8.814 8.933 calc_potential_gpw 544 9.5 0.005 0.006 8.227 8.623 parallel_gemm_fm 14 9.1 0.000 0.000 8.390 8.487 parallel_gemm_fm_cosma 14 10.1 8.390 8.487 8.390 8.487 potential_pw2rs 545 10.0 0.108 0.110 7.694 8.416 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.183 8.402 collocate_single_gaussian 272 10.0 0.041 0.043 7.422 7.680 create_integ_mat 1 6.0 0.006 0.008 7.668 7.668 array2fm 1 7.0 0.000 0.000 6.659 7.138 pw_scatter_s 2720 13.7 4.439 4.652 4.439 4.652 pw_gather_s 2722 13.2 3.908 4.291 3.908 4.291 array2fm_buffer_send 1 8.0 2.959 3.134 2.959 3.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.232839, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2721.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.039 395.598 395.599 farming_run 1 2.0 394.789 394.801 395.555 395.559 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.225220E+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 14918599. 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 0.048 209.449 209.450 qs_energies 1 2.0 0.000 0.000 209.231 209.238 scf_env_do_scf 1 3.0 0.000 0.000 106.444 106.444 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.562 105.571 rebuild_ks_matrix 4 6.0 0.000 0.000 105.560 105.569 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.560 105.569 hfx_ks_matrix 4 8.0 0.001 0.001 105.187 105.190 integrate_four_center 4 9.0 0.144 0.451 105.186 105.190 mp2_main 1 3.0 0.000 0.000 102.499 102.506 mp2_gpw_main 1 4.0 0.039 0.087 101.658 101.667 integrate_four_center_main 4 10.0 0.110 0.581 96.506 99.211 integrate_four_center_bin 265 11.0 96.396 99.187 96.396 99.187 init_scf_loop 1 4.0 0.000 0.000 92.175 92.175 mp2_ri_gpw_compute_in 1 5.0 0.068 0.098 74.729 75.825 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.005 54.508 55.605 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.148 47.252 integrate_v_rspace 95 8.0 0.398 0.564 28.505 33.421 pw_transfer 2240 10.6 0.145 0.169 29.837 30.134 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.859 29.148 grid_integrate_task_list 95 9.0 23.794 28.921 23.794 28.921 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.737 28.449 ao_to_mo_and_store_B_mult_1 91 7.0 10.676 27.665 10.676 27.665 fft_wrap_pw1pw2_100 730 12.4 1.254 1.394 26.590 26.955 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.886 24.999 25.009 get_2c_integrals 1 6.0 0.000 0.004 20.117 20.157 compute_2c_integrals 1 7.0 0.003 0.003 19.102 19.112 compute_2c_integrals_loop_lm 1 8.0 0.004 0.028 18.789 18.984 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.843 18.785 18.983 fft3d_s 1823 13.4 18.409 18.829 18.422 18.843 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.268 14.268 calculate_wavefunction 91 8.0 2.025 2.058 9.749 9.953 potential_pw2rs 186 10.0 0.033 0.035 8.593 9.195 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.582 8.743 9.176 local_gemm 172 8.0 8.187 8.613 8.187 8.613 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.002 8.197 8.470 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.527 8.015 8.399 calc_potential_gpw 182 9.5 0.002 0.002 7.917 8.145 collocate_single_gaussian 91 10.0 0.017 0.022 7.839 8.075 mp2_ri_gpw_compute_en_ener 172 7.0 6.352 6.536 6.352 6.536 mp_sendrecv_dm3 2068 8.0 6.043 6.393 6.043 6.393 mp_sync 38 10.4 3.325 5.331 3.325 5.331 pw_gather_s 912 13.2 4.890 5.249 4.890 5.249 pw_scatter_s 910 13.7 3.941 4.190 3.941 4.190 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.663677, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 451.420160E+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 1430269. 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 0.043 0.073 50.792 50.803 qs_mol_dyn_low 1 2.0 0.003 0.006 50.495 50.503 qs_forces 11 3.9 0.002 0.004 50.432 50.434 qs_energies 11 4.9 0.001 0.002 48.971 48.996 scf_env_do_scf 11 5.9 0.000 0.002 42.863 42.864 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 40.797 40.798 dbcsr_multiply_generic 2286 12.5 0.094 0.099 32.677 33.094 qs_scf_new_mos 108 7.5 0.000 0.001 30.883 31.158 qs_scf_loop_do_ot 108 8.5 0.000 0.001 30.882 31.158 ot_scf_mini 108 9.5 0.002 0.002 29.299 29.502 multiply_cannon 2286 13.5 0.186 0.195 25.760 27.249 multiply_cannon_loop 2286 14.5 1.484 1.569 25.066 26.574 velocity_verlet 10 3.0 0.002 0.002 25.017 25.018 ot_mini 108 10.5 0.001 0.001 18.536 18.780 qs_ot_get_derivative 108 11.5 0.001 0.001 15.591 15.798 mp_waitall_1 267858 16.1 8.336 14.270 8.336 14.270 multiply_cannon_metrocomm3 54864 15.5 0.066 0.070 5.814 12.576 multiply_cannon_multrec 54864 15.5 4.215 6.475 7.735 11.184 rebuild_ks_matrix 119 8.3 0.000 0.000 7.810 7.951 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.809 7.951 multiply_cannon_sync_h2d 54864 15.5 5.904 7.041 5.904 7.041 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.895 7.027 qs_ot_get_p 119 10.4 0.001 0.001 6.280 6.574 mp_sum_l 7207 12.9 4.978 6.572 4.978 6.572 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.208 5.692 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.174 5.295 init_scf_run 11 5.9 0.000 0.001 4.836 4.836 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.836 4.836 dbcsr_mm_accdrv_process 76910 16.1 1.170 1.824 3.443 4.774 sum_up_and_integrate 119 10.3 0.012 0.015 4.461 4.468 integrate_v_rspace 119 11.3 0.002 0.002 4.449 4.457 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.038 4.122 calculate_rho_elec 119 8.7 0.011 0.016 4.038 4.122 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.457 3.493 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.884 3.078 apply_single 119 13.6 0.000 0.000 2.883 3.078 calculate_dm_sparse 119 9.5 0.000 0.000 2.926 3.054 rs_pw_transfer 974 11.9 0.011 0.012 2.776 2.887 jit_kernel_multiply 13 15.8 2.211 2.881 2.211 2.881 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.656 2.834 calculate_first_density_matrix 1 7.0 0.000 0.000 2.716 2.723 ot_diis_step 108 11.5 0.006 0.006 2.657 2.658 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.628 2.628 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.310 2.395 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.389 2.390 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.367 2.369 cp_fm_redistribute_end 50 14.0 2.166 2.365 2.171 2.367 cp_fm_diag_elpa_base 50 14.0 0.194 2.327 0.195 2.336 density_rs2pw 119 9.7 0.004 0.004 2.159 2.253 acc_transpose_blocks 54864 15.5 0.223 0.244 1.781 2.227 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.129 2.171 grid_integrate_task_list 119 12.3 2.013 2.110 2.013 2.110 wfi_extrapolate 11 7.9 0.001 0.001 2.056 2.056 init_scf_loop 11 6.9 0.000 0.000 2.046 2.047 mp_sum_d 4129 12.0 1.317 1.979 1.317 1.979 potential_pw2rs 119 12.3 0.004 0.004 1.839 1.854 pw_transfer 1439 11.6 0.051 0.056 1.653 1.727 make_m2s 4572 13.5 0.053 0.055 1.643 1.691 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.578 1.655 make_images 4572 14.5 0.132 0.138 1.562 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.505 1.528 mp_waitany 12084 13.8 1.272 1.465 1.272 1.465 mp_alltoall_d11v 2130 13.8 1.297 1.458 1.297 1.458 fft3d_ps 1201 14.6 0.359 0.461 1.359 1.430 grid_collocate_task_list 119 9.7 1.287 1.381 1.287 1.381 fft_wrap_pw1pw2_140 487 13.2 0.079 0.092 1.217 1.294 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.720 1.100 acc_transpose_blocks_kernels 54864 16.5 0.253 0.385 0.808 1.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=50.803000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 487.723008E+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 1462257. 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 0.020 0.039 38.153 38.154 qs_mol_dyn_low 1 2.0 0.003 0.006 37.512 37.899 qs_forces 11 3.9 0.002 0.004 37.278 37.289 qs_energies 11 4.9 0.002 0.004 35.623 35.636 scf_env_do_scf 11 5.9 0.001 0.001 30.596 30.596 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.212 28.213 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.090 21.448 qs_scf_new_mos 108 7.5 0.001 0.001 19.481 19.717 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.480 19.717 ot_scf_mini 108 9.5 0.003 0.003 18.586 18.762 velocity_verlet 10 3.0 0.001 0.002 17.857 17.861 multiply_cannon 2286 13.5 0.207 0.216 16.172 17.795 multiply_cannon_loop 2286 14.5 0.897 0.972 15.084 16.505 ot_mini 108 10.5 0.001 0.001 11.559 11.799 mp_waitall_1 217478 16.2 5.776 10.922 5.776 10.922 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.103 9.372 qs_ot_get_derivative 108 11.5 0.001 0.001 9.108 9.288 multiply_cannon_multrec 27432 15.5 1.963 4.246 5.836 8.771 rebuild_ks_matrix 119 8.3 0.000 0.000 6.976 7.116 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.976 7.115 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.164 6.290 dbcsr_mm_accdrv_process 47894 16.0 2.922 4.807 3.804 5.615 qs_ot_get_p 119 10.4 0.001 0.001 4.150 4.384 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.518 4.370 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.042 4.118 apply_single 119 13.6 0.000 0.000 3.042 4.118 sum_up_and_integrate 119 10.3 0.024 0.027 4.020 4.026 integrate_v_rspace 119 11.3 0.002 0.002 3.996 4.003 mp_sum_l 7207 12.9 2.005 3.909 2.005 3.909 init_scf_run 11 5.9 0.000 0.001 3.750 3.751 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.750 3.750 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.463 3.492 calculate_rho_elec 119 8.7 0.021 0.024 3.462 3.491 multiply_cannon_sync_h2d 27432 15.5 2.207 2.827 2.207 2.827 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.749 2.768 make_m2s 4572 13.5 0.052 0.054 2.469 2.693 make_images 4572 14.5 0.216 0.313 2.381 2.603 rs_pw_transfer 974 11.9 0.010 0.011 2.398 2.502 ot_diis_step 108 11.5 0.011 0.014 2.402 2.403 init_scf_loop 11 6.9 0.000 0.001 2.363 2.363 calculate_first_density_matrix 1 7.0 0.000 0.002 2.289 2.291 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.248 2.249 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.079 2.170 calculate_dm_sparse 119 9.5 0.000 0.000 2.028 2.109 grid_integrate_task_list 119 12.3 1.838 1.953 1.838 1.953 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.937 1.937 density_rs2pw 119 9.7 0.004 0.004 1.832 1.928 cp_fm_redistribute_end 50 14.0 1.595 1.914 1.598 1.915 cp_fm_diag_elpa_base 50 14.0 0.308 1.864 0.316 1.896 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.851 1.853 potential_pw2rs 119 12.3 0.006 0.006 1.770 1.778 jit_kernel_multiply 9 16.1 0.830 1.763 0.830 1.763 pw_transfer 1439 11.6 0.063 0.067 1.696 1.728 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.659 1.701 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.607 1.641 make_images_data 4572 15.5 0.045 0.052 1.141 1.567 prepare_preconditioner 11 7.9 0.000 0.000 1.473 1.499 make_preconditioner 11 8.9 0.000 0.001 1.473 1.499 acc_transpose_blocks 27432 15.5 0.107 0.112 1.182 1.477 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.377 1.433 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.986 1.428 wfi_extrapolate 11 7.9 0.001 0.001 1.409 1.409 fft3d_ps 1201 14.6 0.497 0.551 1.314 1.346 grid_collocate_task_list 119 9.7 1.241 1.340 1.241 1.340 mp_alltoall_d11v 2130 13.8 1.143 1.339 1.143 1.339 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.273 1.305 mp_allgather_i34 2286 14.5 0.531 1.268 0.531 1.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.246 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.160 1.209 mp_sum_d 4129 12.0 0.584 1.028 0.584 1.028 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.004 1.016 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.881 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.653 0.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.154000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.000000, yerr=1.477098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 522.293248E+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 0.017 0.030 31.416 31.418 qs_mol_dyn_low 1 2.0 0.003 0.004 31.149 31.157 qs_forces 11 3.9 0.003 0.008 31.094 31.095 qs_energies 11 4.9 0.006 0.023 29.540 29.544 scf_env_do_scf 11 5.9 0.001 0.001 24.913 24.913 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.352 22.354 dbcsr_multiply_generic 2286 12.5 0.092 0.094 16.211 16.313 velocity_verlet 10 3.0 0.001 0.002 14.901 14.902 qs_scf_new_mos 108 7.5 0.001 0.001 14.555 14.572 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.554 14.572 ot_scf_mini 108 9.5 0.002 0.003 13.840 13.854 multiply_cannon 2286 13.5 0.194 0.200 13.015 13.802 multiply_cannon_loop 2286 14.5 0.637 0.669 12.233 13.068 ot_mini 108 10.5 0.001 0.001 8.621 8.635 qs_ot_get_derivative 108 11.5 0.001 0.001 7.126 7.138 multiply_cannon_multrec 18288 15.5 1.920 2.887 6.725 7.064 rebuild_ks_matrix 119 8.3 0.000 0.000 6.233 6.248 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.233 6.247 dbcsr_mm_accdrv_process 38222 16.0 3.937 5.296 4.719 5.640 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.500 5.512 mp_waitall_1 169478 16.3 2.814 4.075 2.814 4.075 sum_up_and_integrate 119 10.3 0.030 0.031 3.874 3.882 integrate_v_rspace 119 11.3 0.002 0.003 3.844 3.856 init_scf_run 11 5.9 0.000 0.001 3.436 3.436 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.436 3.436 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.677 3.281 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.161 3.173 calculate_rho_elec 119 8.7 0.030 0.031 3.161 3.172 qs_ot_get_p 119 10.4 0.001 0.003 3.088 3.109 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.384 2.559 init_scf_loop 11 6.9 0.001 0.003 2.543 2.544 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.019 2.350 apply_single 119 13.6 0.000 0.000 2.019 2.349 rs_pw_transfer 974 11.9 0.009 0.010 2.203 2.295 calculate_first_density_matrix 1 7.0 0.001 0.003 2.252 2.253 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.013 2.019 grid_integrate_task_list 119 12.3 1.810 1.897 1.810 1.897 make_m2s 4572 13.5 0.044 0.045 1.755 1.891 density_rs2pw 119 9.7 0.004 0.004 1.767 1.858 jit_kernel_multiply 10 16.1 0.731 1.856 0.731 1.856 calculate_dm_sparse 119 9.5 0.000 0.000 1.830 1.845 make_images 4572 14.5 0.191 0.203 1.670 1.805 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.748 1.749 prepare_preconditioner 11 7.9 0.000 0.000 1.733 1.734 make_preconditioner 11 8.9 0.000 0.002 1.733 1.734 potential_pw2rs 119 12.3 0.007 0.008 1.670 1.680 pw_transfer 1439 11.6 0.063 0.066 1.657 1.673 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.671 1.673 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.587 1.669 multiply_cannon_sync_h2d 18288 15.5 1.429 1.633 1.429 1.633 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.567 1.585 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.549 1.557 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.530 1.536 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.520 1.521 cp_fm_redistribute_end 50 14.0 1.127 1.500 1.128 1.500 cp_fm_diag_elpa_base 50 14.0 0.357 1.446 0.371 1.484 ot_diis_step 108 11.5 0.011 0.011 1.476 1.476 mp_sum_l 7207 12.9 1.061 1.421 1.061 1.421 grid_collocate_task_list 119 9.7 1.204 1.294 1.204 1.294 acc_transpose_blocks 18288 15.5 0.074 0.076 1.256 1.277 fft3d_ps 1201 14.6 0.502 0.519 1.254 1.276 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.242 1.260 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.215 wfi_extrapolate 11 7.9 0.001 0.001 1.137 1.137 make_images_data 4572 15.5 0.045 0.049 0.782 0.945 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.940 0.941 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.871 0.893 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.680 0.863 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.818 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.805 mp_alltoall_d11v 2130 13.8 0.652 0.787 0.652 0.787 cp_fm_cholesky_invert 11 10.9 0.716 0.720 0.716 0.720 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.712 mp_alltoall_z22v 1201 16.6 0.599 0.676 0.599 0.676 mp_waitany 9880 13.7 0.533 0.647 0.533 0.647 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.418000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=2.048987 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 554.381312E+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 0.031 0.066 35.041 35.043 qs_mol_dyn_low 1 2.0 0.003 0.005 34.754 34.761 qs_forces 11 3.9 0.002 0.005 34.415 34.423 qs_energies 11 4.9 0.001 0.002 32.711 32.725 scf_env_do_scf 11 5.9 0.000 0.002 27.582 27.583 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.155 24.155 dbcsr_multiply_generic 2286 12.5 0.122 0.131 18.199 18.294 velocity_verlet 10 3.0 0.002 0.002 17.981 17.995 qs_scf_new_mos 108 7.5 0.001 0.001 16.044 16.107 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.043 16.106 ot_scf_mini 108 9.5 0.002 0.003 14.945 15.004 multiply_cannon 2286 13.5 0.232 0.268 14.443 14.787 multiply_cannon_loop 2286 14.5 0.936 0.970 13.448 13.778 ot_mini 108 10.5 0.001 0.001 9.102 9.172 multiply_cannon_multrec 27432 15.5 2.340 3.026 8.596 8.964 qs_ot_get_derivative 108 11.5 0.001 0.003 7.299 7.354 dbcsr_mm_accdrv_process 47916 15.9 5.280 7.265 6.162 7.339 rebuild_ks_matrix 119 8.3 0.000 0.000 6.467 6.518 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.467 6.517 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.743 5.786 init_scf_run 11 5.9 0.000 0.001 3.788 3.789 scf_env_initial_rho_setup 11 6.9 0.002 0.013 3.788 3.788 sum_up_and_integrate 119 10.3 0.036 0.049 3.752 3.760 integrate_v_rspace 119 11.3 0.002 0.003 3.716 3.726 init_scf_loop 11 6.9 0.000 0.000 3.406 3.407 qs_ot_get_p 119 10.4 0.001 0.001 3.241 3.322 qs_rho_update_rho_low 119 7.7 0.003 0.018 3.281 3.316 calculate_rho_elec 119 8.7 0.040 0.046 3.278 3.315 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.738 3.134 mp_waitall_1 145218 16.4 2.181 2.704 2.181 2.704 prepare_preconditioner 11 7.9 0.000 0.000 2.561 2.570 make_preconditioner 11 8.9 0.000 0.000 2.561 2.570 make_m2s 4572 13.5 0.053 0.055 2.359 2.515 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.164 2.493 make_images 4572 14.5 0.277 0.334 2.252 2.407 calculate_first_density_matrix 1 7.0 0.000 0.002 2.387 2.389 calculate_dm_sparse 119 9.5 0.000 0.000 2.203 2.267 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.028 2.160 apply_single 119 13.6 0.000 0.000 2.027 2.159 rs_pw_transfer 974 11.9 0.009 0.009 1.982 2.102 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.959 1.968 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.949 1.950 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.901 1.931 grid_integrate_task_list 119 12.3 1.832 1.928 1.832 1.928 density_rs2pw 119 9.7 0.004 0.004 1.748 1.866 jit_kernel_multiply 10 16.0 0.822 1.827 0.822 1.827 pw_transfer 1439 11.6 0.063 0.065 1.759 1.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.767 1.782 ot_diis_step 108 11.5 0.013 0.018 1.762 1.765 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.669 1.700 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.646 1.646 potential_pw2rs 119 12.3 0.008 0.009 1.536 1.546 acc_transpose_blocks 27432 15.5 0.111 0.114 1.439 1.459 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.414 1.415 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.368 1.402 cp_fm_redistribute_end 50 14.0 0.930 1.390 0.931 1.390 fft3d_ps 1201 14.6 0.533 0.587 1.350 1.377 cp_fm_diag_elpa_base 50 14.0 0.437 1.332 0.457 1.373 wfi_extrapolate 11 7.9 0.003 0.020 1.325 1.325 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.749 1.315 grid_collocate_task_list 119 9.7 1.220 1.305 1.220 1.305 mp_sum_l 7207 12.9 0.959 1.277 0.959 1.277 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.222 1.233 make_images_data 4572 15.5 0.045 0.049 0.952 1.144 cp_fm_upper_to_full 72 13.5 0.814 1.127 0.814 1.127 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.096 1.116 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.083 1.091 dbcsr_complete_redistribute 329 12.2 0.122 0.156 0.790 1.062 multiply_cannon_sync_h2d 27432 15.5 1.000 1.053 1.000 1.053 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.785 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.870 mp_alltoall_d11v 2130 13.8 0.739 0.863 0.739 0.863 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.594 0.861 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.814 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.813 0.818 cp_fm_cholesky_invert 11 10.9 0.753 0.756 0.753 0.756 mp_alltoall_i22 627 13.8 0.450 0.751 0.450 0.751 mp_alltoall_z22v 1201 16.6 0.699 0.724 0.699 0.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.043000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.363636, yerr=2.804954 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 601.112576E+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 0.014 0.030 27.058 27.058 qs_mol_dyn_low 1 2.0 0.003 0.003 26.848 26.855 qs_forces 11 3.9 0.002 0.002 26.789 26.792 qs_energies 11 4.9 0.001 0.001 25.109 25.114 scf_env_do_scf 11 5.9 0.001 0.001 20.537 20.537 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.102 18.103 velocity_verlet 10 3.0 0.002 0.002 13.938 13.940 dbcsr_multiply_generic 2286 12.5 0.091 0.094 11.890 11.969 qs_scf_new_mos 108 7.5 0.001 0.001 10.700 10.719 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.699 10.719 ot_scf_mini 108 9.5 0.002 0.002 10.031 10.059 multiply_cannon 2286 13.5 0.230 0.234 9.388 9.813 multiply_cannon_loop 2286 14.5 0.331 0.342 8.466 8.651 multiply_cannon_multrec 9144 15.5 1.579 1.829 5.725 5.974 rebuild_ks_matrix 119 8.3 0.000 0.000 5.701 5.719 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.701 5.719 ot_mini 108 10.5 0.001 0.001 5.553 5.587 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.080 5.097 qs_ot_get_derivative 108 11.5 0.001 0.001 4.260 4.290 dbcsr_mm_accdrv_process 12550 15.8 3.142 3.732 4.045 4.109 sum_up_and_integrate 119 10.3 0.038 0.042 3.479 3.484 integrate_v_rspace 119 11.3 0.003 0.003 3.442 3.446 init_scf_run 11 5.9 0.000 0.001 3.149 3.149 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.149 3.149 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.133 3.140 calculate_rho_elec 119 8.7 0.059 0.061 3.133 3.139 qs_ot_get_p 119 10.4 0.001 0.001 2.712 2.748 init_scf_loop 11 6.9 0.000 0.000 2.414 2.415 calculate_first_density_matrix 1 7.0 0.000 0.000 2.050 2.050 mp_waitall_1 121218 16.5 1.466 2.009 1.466 2.009 grid_integrate_task_list 119 12.3 1.846 1.923 1.846 1.923 make_m2s 4572 13.5 0.034 0.035 1.720 1.906 make_images 4572 14.5 0.269 0.304 1.631 1.814 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.803 1.806 calculate_dm_sparse 119 9.5 0.000 0.000 1.690 1.707 prepare_preconditioner 11 7.9 0.000 0.000 1.699 1.704 make_preconditioner 11 8.9 0.000 0.000 1.699 1.704 pw_transfer 1439 11.6 0.063 0.065 1.653 1.661 rs_pw_transfer 974 11.9 0.008 0.008 1.581 1.657 density_rs2pw 119 9.7 0.004 0.004 1.547 1.632 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.587 1.613 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.595 1.595 jit_kernel_multiply 9 15.8 0.865 1.587 0.865 1.587 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.562 1.570 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.507 1.509 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.486 1.507 grid_collocate_task_list 119 9.7 1.271 1.375 1.271 1.375 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.343 1.351 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.299 1.300 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.275 1.282 ot_diis_step 108 11.5 0.013 0.013 1.281 1.281 cp_fm_redistribute_end 50 14.0 0.645 1.280 0.646 1.281 potential_pw2rs 119 12.3 0.010 0.010 1.276 1.279 cp_fm_diag_elpa_base 50 14.0 0.590 1.210 0.634 1.268 fft3d_ps 1201 14.6 0.537 0.546 1.229 1.236 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.217 1.220 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.165 1.187 apply_single 119 13.6 0.000 0.000 1.165 1.187 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.132 1.147 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.113 1.119 wfi_extrapolate 11 7.9 0.001 0.001 1.055 1.055 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.764 1.033 make_images_data 4572 15.5 0.039 0.042 0.792 0.998 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.917 acc_transpose_blocks 9144 15.5 0.039 0.040 0.886 0.893 cp_fm_cholesky_invert 11 10.9 0.852 0.854 0.852 0.854 mp_alltoall_d11v 2130 13.8 0.780 0.848 0.780 0.848 multiply_cannon_sync_h2d 9144 15.5 0.711 0.805 0.711 0.805 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.742 0.744 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.673 0.725 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.710 0.720 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.310 0.667 mp_allgather_i34 2286 14.5 0.253 0.653 0.253 0.653 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.645 0.650 mp_alltoall_z22v 1201 16.6 0.566 0.595 0.566 0.595 qs_create_task_list 11 7.9 0.001 0.001 0.546 0.568 generate_qs_task_list 11 8.9 0.192 0.214 0.545 0.568 dbcsr_complete_redistribute 329 12.2 0.171 0.182 0.521 0.546 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.058000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.454545, yerr=5.449998 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 738.160640E+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 0.016 0.039 40.942 40.943 qs_mol_dyn_low 1 2.0 0.003 0.003 40.735 40.742 qs_forces 11 3.9 0.002 0.002 40.677 40.678 qs_energies 11 4.9 0.001 0.001 38.720 38.724 scf_env_do_scf 11 5.9 0.001 0.005 32.999 32.999 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.125 25.126 velocity_verlet 10 3.0 0.002 0.002 23.099 23.104 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.571 17.801 qs_scf_new_mos 108 7.5 0.001 0.001 16.168 16.285 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.168 16.284 ot_scf_mini 108 9.5 0.002 0.002 15.085 15.203 multiply_cannon 2286 13.5 0.305 0.311 13.588 14.576 multiply_cannon_loop 2286 14.5 0.345 0.351 12.304 13.289 ot_mini 108 10.5 0.001 0.001 8.875 9.012 multiply_cannon_multrec 9144 15.5 3.431 4.805 8.548 8.600 init_scf_loop 11 6.9 0.000 0.000 7.840 7.842 rebuild_ks_matrix 119 8.3 0.000 0.000 6.880 7.048 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.879 7.047 prepare_preconditioner 11 7.9 0.000 0.000 6.903 6.919 make_preconditioner 11 8.9 0.000 0.000 6.903 6.919 qs_ot_get_derivative 108 11.5 0.001 0.001 6.786 6.904 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.478 6.792 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.225 6.378 dbcsr_mm_accdrv_process 12550 15.8 4.032 5.619 4.994 6.345 cp_fm_upper_to_full 72 14.2 3.197 4.559 3.197 4.559 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.711 3.728 calculate_rho_elec 119 8.7 0.118 0.121 3.710 3.728 sum_up_and_integrate 119 10.3 0.064 0.066 3.705 3.712 init_scf_run 11 5.9 0.000 0.001 3.692 3.692 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.691 3.692 integrate_v_rspace 119 11.3 0.003 0.003 3.640 3.648 qs_ot_get_p 119 10.4 0.001 0.001 3.452 3.618 mp_waitall_1 97218 16.6 2.587 3.605 2.587 3.605 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.467 2.879 dbcsr_complete_redistribute 329 12.2 0.287 0.292 2.006 2.821 make_m2s 4572 13.5 0.037 0.038 2.310 2.479 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.673 2.476 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.216 2.435 apply_single 119 13.6 0.000 0.000 2.215 2.435 make_images 4572 14.5 0.355 0.384 2.190 2.360 mp_alltoall_i22 627 13.8 1.478 2.306 1.478 2.306 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 1.310 2.267 calculate_first_density_matrix 1 7.0 0.000 0.000 2.230 2.234 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.420 2.217 calculate_dm_sparse 119 9.5 0.000 0.000 2.189 2.209 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.098 2.099 ot_diis_step 108 11.5 0.014 0.015 2.060 2.061 pw_transfer 1439 11.6 0.066 0.066 2.055 2.060 grid_integrate_task_list 119 12.3 2.024 2.041 2.024 2.041 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.960 1.965 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.955 1.955 mp_sum_l 7207 12.9 1.134 1.889 1.134 1.889 density_rs2pw 119 9.7 0.003 0.003 1.806 1.823 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.820 1.821 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.716 1.776 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.757 1.757 jit_kernel_multiply 10 15.6 0.936 1.745 0.936 1.745 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.619 1.624 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.603 1.621 fft3d_ps 1201 14.6 0.567 0.577 1.595 1.599 cp_fm_cholesky_invert 11 10.9 1.519 1.522 1.519 1.522 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.507 1.507 cp_fm_diag_elpa_base 50 14.0 1.362 1.413 1.505 1.505 grid_collocate_task_list 119 9.7 1.462 1.475 1.462 1.475 rs_pw_transfer 974 11.9 0.009 0.009 1.419 1.457 wfi_extrapolate 11 7.9 0.001 0.001 1.397 1.397 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.139 1.379 make_images_data 4572 15.5 0.043 0.046 1.081 1.289 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.257 1.268 potential_pw2rs 119 12.3 0.014 0.014 1.235 1.239 mp_alltoall_d11v 2130 13.8 1.160 1.194 1.160 1.194 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.150 1.173 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.096 1.115 multiply_cannon_sync_h2d 9144 15.5 1.042 1.045 1.042 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.987 1.029 qs_create_task_list 11 7.9 0.001 0.001 0.946 0.955 generate_qs_task_list 11 8.9 0.377 0.396 0.946 0.955 acc_transpose_blocks 9144 15.5 0.038 0.039 0.907 0.927 mp_alltoall_z22v 1201 16.6 0.893 0.914 0.893 0.914 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.866 0.881 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.809 0.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.943000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=698.000000, yerr=12.120606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 503.439360E+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 1989866. 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 0.016 0.034 82.990 82.992 qs_mol_dyn_low 1 2.0 0.003 0.003 82.731 82.741 qs_forces 11 3.9 0.003 0.003 82.661 82.663 qs_energies 11 4.9 0.002 0.008 79.828 79.847 scf_env_do_scf 11 5.9 0.000 0.001 70.781 70.784 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.069 65.069 dbcsr_multiply_generic 2055 12.4 0.105 0.108 51.963 52.339 qs_scf_new_mos 99 7.5 0.000 0.001 48.105 48.220 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.105 48.219 ot_scf_mini 99 9.5 0.002 0.003 45.638 45.738 multiply_cannon 2055 13.4 0.180 0.188 42.495 43.509 multiply_cannon_loop 2055 14.4 1.535 1.577 41.512 42.558 velocity_verlet 10 3.0 0.002 0.002 41.167 41.168 ot_mini 99 10.5 0.001 0.001 27.558 27.643 qs_ot_get_derivative 99 11.5 0.001 0.001 20.708 20.803 multiply_cannon_multrec 49320 15.4 12.207 12.847 17.177 17.972 rebuild_ks_matrix 110 8.3 0.000 0.001 14.432 14.590 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.432 14.589 mp_waitall_1 241148 16.1 11.674 12.829 11.674 12.829 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.687 12.823 multiply_cannon_sync_h2d 49320 15.4 10.097 10.565 10.097 10.565 qs_ot_get_p 110 10.4 0.001 0.002 9.564 9.660 multiply_cannon_metrocomm3 49320 15.4 0.077 0.081 6.837 8.271 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.165 7.715 apply_single 110 13.6 0.000 0.001 7.165 7.715 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.649 7.186 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.971 7.039 sum_up_and_integrate 110 10.3 0.037 0.043 7.012 7.027 integrate_v_rspace 110 11.3 0.003 0.003 6.975 6.998 init_scf_run 11 5.9 0.000 0.001 6.978 6.979 scf_env_initial_rho_setup 11 6.9 0.017 0.036 6.978 6.979 ot_diis_step 99 11.5 0.006 0.006 6.554 6.554 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.211 6.345 calculate_rho_elec 110 8.6 0.020 0.024 6.211 6.345 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.222 6.247 mp_sum_l 6514 12.8 5.070 5.952 5.070 5.952 init_scf_loop 11 6.9 0.000 0.003 5.680 5.682 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.185 5.186 dbcsr_mm_accdrv_process 87628 16.1 1.981 2.114 4.848 5.171 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.603 4.605 cp_fm_redistribute_end 48 14.0 3.998 4.570 4.003 4.571 cp_fm_diag_elpa_base 48 14.0 0.562 4.498 0.565 4.523 wfi_extrapolate 11 7.9 0.001 0.001 4.033 4.033 make_m2s 4110 13.4 0.060 0.064 3.946 4.033 rs_pw_transfer 902 11.9 0.012 0.013 3.805 4.002 make_images 4110 14.4 0.179 0.193 3.851 3.939 calculate_dm_sparse 110 9.5 0.000 0.001 3.840 3.926 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.686 3.901 density_rs2pw 110 9.6 0.004 0.004 3.352 3.568 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.549 3.553 prepare_preconditioner 11 7.9 0.000 0.000 3.503 3.521 make_preconditioner 11 8.9 0.000 0.001 3.503 3.521 grid_integrate_task_list 110 12.3 3.256 3.393 3.256 3.393 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.285 3.333 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.269 3.313 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.179 3.226 pw_transfer 1331 11.6 0.053 0.066 2.957 3.026 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.870 2.942 calculate_first_density_matrix 1 7.0 0.000 0.004 2.840 2.863 potential_pw2rs 110 12.3 0.006 0.007 2.665 2.703 jit_kernel_multiply 13 15.9 2.592 2.629 2.592 2.629 mp_alltoall_d11v 2046 13.8 1.999 2.593 1.999 2.593 fft3d_ps 1111 14.6 0.750 0.838 2.410 2.475 fft_wrap_pw1pw2_140 451 13.1 0.167 0.187 2.336 2.408 grid_collocate_task_list 110 9.6 2.086 2.237 2.086 2.237 mp_sum_d 3883 11.9 1.706 2.220 1.706 2.220 acc_transpose_blocks 49320 15.4 0.204 0.217 2.092 2.184 cp_fm_cholesky_invert 11 10.9 1.930 1.934 1.930 1.934 mp_waitany 14300 13.8 1.770 1.919 1.770 1.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.886 1.906 make_images_data 4110 15.4 0.043 0.047 1.703 1.826 hybrid_alltoall_any 4261 16.3 0.082 0.483 1.480 1.700 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.645 1.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.992000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.818182, yerr=3.069444 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.468800E+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 1671445. 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 0.020 0.045 68.471 68.479 qs_mol_dyn_low 1 2.0 0.003 0.003 68.251 68.260 qs_forces 11 3.9 0.006 0.016 68.183 68.187 qs_energies 11 4.9 0.004 0.005 64.886 64.897 scf_env_do_scf 11 5.9 0.000 0.001 56.395 56.399 scf_env_do_scf_inner_loop 99 6.5 0.002 0.010 48.626 48.627 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.845 37.994 velocity_verlet 10 3.0 0.001 0.002 36.184 36.185 qs_scf_new_mos 99 7.5 0.001 0.001 32.822 32.945 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.822 32.944 multiply_cannon 2055 13.4 0.221 0.246 31.117 32.412 ot_scf_mini 99 9.5 0.003 0.003 31.158 31.277 multiply_cannon_loop 2055 14.4 0.929 0.950 29.736 30.731 ot_mini 99 10.5 0.001 0.001 18.355 18.474 multiply_cannon_multrec 24660 15.4 7.627 9.556 13.803 15.699 rebuild_ks_matrix 110 8.3 0.000 0.001 13.597 13.741 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.596 13.741 qs_ot_get_derivative 99 11.5 0.001 0.001 12.484 12.600 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.993 12.121 mp_waitall_1 186928 16.3 7.915 10.757 7.915 10.757 multiply_cannon_sync_h2d 24660 15.4 7.067 8.185 7.067 8.185 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.215 7.740 init_scf_loop 11 6.9 0.000 0.000 7.731 7.732 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.592 7.290 apply_single 110 13.6 0.000 0.001 6.592 7.289 sum_up_and_integrate 110 10.3 0.052 0.060 6.511 6.521 integrate_v_rspace 110 11.3 0.002 0.003 6.458 6.471 dbcsr_mm_accdrv_process 52282 16.1 4.558 5.701 6.017 6.323 qs_ot_get_p 110 10.4 0.001 0.001 6.126 6.259 init_scf_run 11 5.9 0.000 0.001 6.132 6.132 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.131 6.132 ot_diis_step 99 11.5 0.010 0.010 5.823 5.824 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.738 5.782 calculate_rho_elec 110 8.6 0.039 0.047 5.737 5.781 prepare_preconditioner 11 7.9 0.000 0.000 5.706 5.725 make_preconditioner 11 8.9 0.000 0.000 5.706 5.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.276 5.433 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.677 5.425 make_m2s 4110 13.4 0.056 0.059 4.255 4.810 make_images 4110 14.4 0.398 0.444 4.147 4.698 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.191 4.212 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.752 3.752 wfi_extrapolate 11 7.9 0.001 0.001 3.530 3.530 pw_transfer 1331 11.6 0.065 0.069 3.289 3.428 density_rs2pw 110 9.6 0.004 0.004 3.044 3.320 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.183 3.320 grid_integrate_task_list 110 12.3 3.158 3.303 3.158 3.303 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.234 3.303 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.252 3.254 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.153 3.155 rs_pw_transfer 902 11.9 0.012 0.014 2.930 3.134 cp_fm_redistribute_end 48 14.0 2.346 3.124 2.348 3.125 cp_fm_diag_elpa_base 48 14.0 0.742 3.009 0.773 3.095 calculate_dm_sparse 110 9.5 0.001 0.001 3.018 3.059 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.892 2.938 make_images_data 4110 15.4 0.047 0.051 2.360 2.925 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.066 2.838 fft_wrap_pw1pw2_140 451 13.1 0.200 0.217 2.634 2.771 cp_fm_cholesky_invert 11 10.9 2.661 2.668 2.661 2.668 fft3d_ps 1111 14.6 1.072 1.265 2.541 2.665 calculate_first_density_matrix 1 7.0 0.000 0.000 2.513 2.517 mp_sum_l 6514 12.8 1.849 2.516 1.849 2.516 potential_pw2rs 110 12.3 0.008 0.009 2.383 2.402 grid_collocate_task_list 110 9.6 2.118 2.310 2.118 2.310 jit_kernel_multiply 11 16.3 1.111 2.118 1.111 2.118 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.969 1.990 mp_alltoall_d11v 2046 13.8 1.794 1.974 1.794 1.974 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.843 1.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.767 1.781 mp_allgather_i34 2055 14.4 0.709 1.730 0.709 1.730 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.590 1.603 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.778 1.589 acc_transpose_blocks 24660 15.4 0.106 0.110 1.510 1.551 dbcsr_complete_redistribute 325 12.2 0.248 0.308 1.229 1.494 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.473 mp_irecv_dv 57340 16.2 0.655 1.470 0.655 1.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.479000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.818182, yerr=6.520153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 659.955712E+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 0.046 0.073 59.345 59.346 qs_mol_dyn_low 1 2.0 0.003 0.003 58.962 58.971 qs_forces 11 3.9 0.003 0.006 58.896 58.899 qs_energies 11 4.9 0.004 0.008 55.764 55.770 scf_env_do_scf 11 5.9 0.001 0.001 47.759 47.759 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.287 39.287 velocity_verlet 10 3.0 0.002 0.002 31.849 31.851 dbcsr_multiply_generic 2055 12.4 0.109 0.111 28.188 28.440 qs_scf_new_mos 99 7.5 0.001 0.001 24.955 25.073 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.954 25.072 ot_scf_mini 99 9.5 0.003 0.003 23.745 23.867 multiply_cannon 2055 13.4 0.211 0.219 21.971 23.238 multiply_cannon_loop 2055 14.4 0.618 0.634 20.794 21.783 ot_mini 99 10.5 0.001 0.001 13.411 13.531 rebuild_ks_matrix 110 8.3 0.000 0.000 12.076 12.207 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.076 12.207 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.650 10.767 multiply_cannon_multrec 16440 15.4 4.005 5.000 9.811 10.676 mp_waitall_1 146766 16.3 6.802 9.861 6.802 9.861 qs_ot_get_derivative 99 11.5 0.001 0.001 9.015 9.136 init_scf_loop 11 6.9 0.001 0.003 8.438 8.440 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.105 7.013 prepare_preconditioner 11 7.9 0.000 0.000 6.692 6.709 make_preconditioner 11 8.9 0.000 0.001 6.691 6.709 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.053 6.409 sum_up_and_integrate 110 10.3 0.060 0.060 6.342 6.364 integrate_v_rspace 110 11.3 0.002 0.003 6.282 6.304 dbcsr_mm_accdrv_process 34862 16.1 4.639 5.423 5.659 5.783 qs_ot_get_p 110 10.4 0.001 0.002 5.564 5.704 init_scf_run 11 5.9 0.001 0.001 5.527 5.528 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.527 5.528 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.443 5.453 calculate_rho_elec 110 8.6 0.058 0.058 5.443 5.453 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.921 5.368 apply_single 110 13.6 0.000 0.000 4.921 5.368 make_m2s 4110 13.4 0.049 0.050 4.120 4.487 make_images 4110 14.4 0.391 0.508 4.005 4.371 ot_diis_step 99 11.5 0.010 0.011 4.368 4.368 multiply_cannon_sync_h2d 16440 15.4 3.710 4.306 3.710 4.306 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.948 3.952 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.065 3.694 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.611 3.612 grid_integrate_task_list 110 12.3 3.182 3.369 3.182 3.369 pw_transfer 1331 11.6 0.064 0.071 3.144 3.159 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.080 3.081 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.038 3.055 cp_fm_redistribute_end 48 14.0 1.911 3.050 1.913 3.050 cp_fm_diag_elpa_base 48 14.0 1.076 2.916 1.133 3.025 wfi_extrapolate 11 7.9 0.001 0.001 2.919 2.919 density_rs2pw 110 9.6 0.004 0.004 2.740 2.888 make_images_data 4110 15.4 0.044 0.047 2.372 2.846 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.842 2.844 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.178 2.810 rs_pw_transfer 902 11.9 0.010 0.011 2.633 2.766 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.600 2.619 cp_fm_cholesky_invert 11 10.9 2.610 2.617 2.610 2.617 calculate_first_density_matrix 1 7.0 0.001 0.003 2.525 2.528 calculate_dm_sparse 110 9.5 0.001 0.001 2.484 2.513 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.445 2.493 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.310 2.374 grid_collocate_task_list 110 9.6 2.170 2.371 2.170 2.371 fft3d_ps 1111 14.6 1.065 1.086 2.353 2.369 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.857 2.330 mp_irecv_dv 48980 15.7 0.788 2.204 0.788 2.204 potential_pw2rs 110 12.3 0.011 0.011 2.178 2.191 mp_sum_l 6514 12.8 1.494 2.098 1.494 2.098 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 2.041 2.044 mp_alltoall_d11v 2046 13.8 1.766 2.034 1.766 2.034 dbcsr_complete_redistribute 325 12.2 0.343 0.420 1.474 1.926 cp_fm_upper_to_full 70 13.6 1.366 1.831 1.366 1.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.645 mp_allgather_i34 2055 14.4 0.498 1.597 0.498 1.597 cp_fm_cholesky_decompose 22 10.9 1.562 1.580 1.562 1.580 jit_kernel_multiply 8 16.6 0.634 1.513 0.634 1.513 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.455 1.468 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.343 1.450 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.004 1.443 mp_waitany 17072 13.8 1.179 1.366 1.179 1.366 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.237 1.244 acc_transpose_blocks 16440 15.4 0.072 0.074 1.214 1.231 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.151 1.215 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.346000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.181818, yerr=8.199980 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 741.011456E+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 0.014 0.032 64.004 64.005 qs_mol_dyn_low 1 2.0 0.003 0.003 63.761 63.770 qs_forces 11 3.9 0.002 0.002 63.694 63.700 qs_energies 11 4.9 0.001 0.002 60.338 60.346 scf_env_do_scf 11 5.9 0.000 0.001 52.208 52.211 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.442 40.443 velocity_verlet 10 3.0 0.002 0.002 36.533 36.535 dbcsr_multiply_generic 2055 12.4 0.115 0.118 29.336 29.554 qs_scf_new_mos 99 7.5 0.001 0.001 26.140 26.249 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.140 26.248 ot_scf_mini 99 9.5 0.003 0.003 24.524 24.636 multiply_cannon 2055 13.4 0.244 0.262 22.580 23.381 multiply_cannon_loop 2055 14.4 0.880 0.903 21.050 21.648 ot_mini 99 10.5 0.001 0.001 14.163 14.283 multiply_cannon_multrec 24660 15.4 4.247 6.846 12.635 13.736 rebuild_ks_matrix 110 8.3 0.000 0.000 11.818 11.956 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.817 11.956 init_scf_loop 11 6.9 0.000 0.000 11.724 11.725 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.454 10.577 qs_ot_get_derivative 99 11.5 0.001 0.001 10.025 10.133 prepare_preconditioner 11 7.9 0.000 0.000 9.993 10.011 make_preconditioner 11 8.9 0.000 0.000 9.993 10.011 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.180 9.663 dbcsr_mm_accdrv_process 52304 16.0 6.855 8.220 8.241 9.187 mp_waitall_1 126806 16.4 4.366 6.628 4.366 6.628 sum_up_and_integrate 110 10.3 0.069 0.072 6.251 6.262 integrate_v_rspace 110 11.3 0.003 0.003 6.182 6.192 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.600 5.611 calculate_rho_elec 110 8.6 0.077 0.081 5.599 5.611 make_m2s 4110 13.4 0.059 0.061 5.161 5.603 qs_ot_get_p 110 10.4 0.001 0.001 5.434 5.581 make_images 4110 14.4 0.575 0.696 5.022 5.460 init_scf_run 11 5.9 0.000 0.001 5.402 5.403 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.402 5.403 cp_fm_upper_to_full 70 13.8 3.348 4.772 3.348 4.772 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.009 4.112 apply_single 110 13.6 0.000 0.000 4.008 4.112 ot_diis_step 99 11.5 0.011 0.012 4.101 4.102 dbcsr_complete_redistribute 325 12.2 0.430 0.481 2.726 3.889 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.623 3.637 multiply_cannon_sync_h2d 24660 15.4 3.152 3.437 3.152 3.437 grid_integrate_task_list 110 12.3 3.254 3.420 3.254 3.420 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.190 3.337 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.262 3.321 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.536 3.259 pw_transfer 1331 11.6 0.064 0.072 3.213 3.238 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.147 3.148 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.108 3.136 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.997 3.058 hybrid_alltoall_any 4261 16.3 0.120 0.453 2.235 2.967 make_images_data 4110 15.4 0.046 0.049 2.530 2.961 wfi_extrapolate 11 7.9 0.001 0.001 2.935 2.935 calculate_dm_sparse 110 9.5 0.001 0.001 2.898 2.933 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.804 2.927 density_rs2pw 110 9.6 0.004 0.004 2.740 2.900 mp_alltoall_i22 605 13.7 1.682 2.889 1.682 2.889 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.788 2.790 cp_fm_cholesky_invert 11 10.9 2.714 2.722 2.714 2.722 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.632 2.664 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.598 2.600 cp_fm_redistribute_end 48 14.0 1.292 2.571 1.293 2.571 cp_fm_diag_elpa_base 48 14.0 1.195 2.439 1.275 2.553 rs_pw_transfer 902 11.9 0.010 0.011 2.363 2.538 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.399 2.441 fft3d_ps 1111 14.6 1.064 1.101 2.410 2.427 calculate_first_density_matrix 1 7.0 0.000 0.000 2.380 2.383 grid_collocate_task_list 110 9.6 2.221 2.377 2.221 2.377 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.238 2.243 potential_pw2rs 110 12.3 0.012 0.013 2.053 2.065 mp_alltoall_d11v 2046 13.8 1.817 2.049 1.817 2.049 jit_kernel_multiply 10 15.8 1.059 1.814 1.059 1.814 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.687 1.722 cp_fm_cholesky_decompose 22 10.9 1.662 1.710 1.662 1.710 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.603 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.607 1.621 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.518 1.537 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.858 1.511 acc_transpose_blocks 24660 15.4 0.104 0.105 1.489 1.511 mp_sum_l 6514 12.8 0.906 1.501 0.906 1.501 mp_irecv_dv 62702 16.1 0.759 1.433 0.759 1.433 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.185 1.282 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.005000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.000000, yerr=14.628739 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 833.421312E+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 0.014 0.030 55.018 55.018 qs_mol_dyn_low 1 2.0 0.003 0.003 54.758 54.767 qs_forces 11 3.9 0.002 0.002 54.690 54.695 qs_energies 11 4.9 0.013 0.015 51.081 51.093 scf_env_do_scf 11 5.9 0.000 0.001 42.880 42.880 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.000 35.001 velocity_verlet 10 3.0 0.002 0.002 30.811 30.814 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.732 22.889 qs_scf_new_mos 99 7.5 0.001 0.001 20.949 20.997 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.948 20.997 ot_scf_mini 99 9.5 0.002 0.002 19.700 19.733 multiply_cannon 2055 13.4 0.248 0.259 17.211 18.408 multiply_cannon_loop 2055 14.4 0.323 0.336 15.825 16.087 rebuild_ks_matrix 110 8.3 0.000 0.000 11.390 11.440 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.389 11.440 ot_mini 99 10.5 0.001 0.001 10.357 10.383 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.150 10.194 multiply_cannon_multrec 8220 15.4 3.196 4.623 7.405 8.693 mp_waitall_1 106626 16.5 6.131 7.856 6.131 7.856 init_scf_loop 11 6.9 0.000 0.000 7.829 7.830 qs_ot_get_derivative 99 11.5 0.001 0.001 6.504 6.537 prepare_preconditioner 11 7.9 0.000 0.000 6.211 6.217 make_preconditioner 11 8.9 0.000 0.000 6.211 6.217 sum_up_and_integrate 110 10.3 0.080 0.081 6.162 6.175 integrate_v_rspace 110 11.3 0.003 0.003 6.082 6.094 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.781 5.855 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.607 5.626 calculate_rho_elec 110 8.6 0.115 0.115 5.607 5.625 qs_ot_get_p 110 10.4 0.001 0.001 5.413 5.437 init_scf_run 11 5.9 0.000 0.001 5.061 5.062 scf_env_initial_rho_setup 11 6.9 0.003 0.003 5.061 5.061 dbcsr_mm_accdrv_process 17442 15.9 2.779 3.612 4.078 4.964 make_m2s 4110 13.4 0.038 0.040 4.247 4.491 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.049 4.373 make_images 4110 14.4 0.635 0.689 4.117 4.360 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.996 4.000 ot_diis_step 99 11.5 0.012 0.014 3.830 3.830 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.749 3.814 apply_single 110 13.6 0.000 0.000 3.749 3.814 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.681 3.682 grid_integrate_task_list 110 12.3 3.355 3.517 3.355 3.517 pw_transfer 1331 11.6 0.065 0.070 3.246 3.260 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.139 3.157 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.113 3.114 multiply_cannon_sync_h2d 8220 15.4 2.901 3.084 2.901 3.084 cp_fm_redistribute_end 48 14.0 0.785 3.069 0.790 3.070 cp_fm_cholesky_invert 11 10.9 3.030 3.034 3.030 3.034 cp_fm_diag_elpa_base 48 14.0 2.091 2.873 2.271 3.034 make_images_data 4110 15.4 0.038 0.043 2.391 2.789 density_rs2pw 110 9.6 0.004 0.004 2.659 2.778 wfi_extrapolate 11 7.9 0.001 0.001 2.724 2.724 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.303 2.703 fft_wrap_pw1pw2_140 451 13.1 0.212 0.222 2.656 2.676 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 2.666 2.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.535 2.537 calculate_dm_sparse 110 9.5 0.001 0.001 2.483 2.529 grid_collocate_task_list 110 9.6 2.331 2.468 2.331 2.468 fft3d_ps 1111 14.6 1.115 1.150 2.397 2.417 rs_pw_transfer 902 11.9 0.010 0.010 2.194 2.340 calculate_first_density_matrix 1 7.0 0.000 0.000 2.246 2.248 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.164 2.180 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.778 1.991 potential_pw2rs 110 12.3 0.015 0.016 1.932 1.937 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.918 1.935 mp_alltoall_d11v 2046 13.8 1.656 1.902 1.656 1.902 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.833 1.846 cp_fm_cholesky_decompose 22 10.9 1.713 1.726 1.713 1.726 mp_allgather_i34 2055 14.4 0.570 1.662 0.570 1.662 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.506 1.627 dbcsr_complete_redistribute 325 12.2 0.555 0.584 1.504 1.606 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.573 1.579 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.440 1.452 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.474 1.424 mp_irecv_dv 24056 15.7 0.450 1.384 0.450 1.384 jit_kernel_multiply 8 15.9 0.993 1.373 0.993 1.373 qs_create_task_list 11 7.9 0.002 0.003 1.228 1.327 generate_qs_task_list 11 8.9 0.380 0.449 1.225 1.325 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.178 1.194 mp_waitany 9240 13.8 1.055 1.192 1.055 1.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.018000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=790.727273, yerr=9.126295 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 1.347850E+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 0.036 0.041 86.041 86.041 qs_mol_dyn_low 1 2.0 0.003 0.003 85.400 85.408 qs_forces 11 3.9 0.002 0.003 85.316 85.322 qs_energies 11 4.9 0.001 0.001 81.276 81.282 scf_env_do_scf 11 5.9 0.001 0.001 70.912 70.912 velocity_verlet 10 3.0 0.005 0.023 54.431 54.437 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.691 42.692 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.744 28.824 init_scf_loop 11 6.9 0.000 0.000 28.147 28.149 qs_scf_new_mos 99 7.5 0.001 0.001 26.435 26.495 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.434 26.494 prepare_preconditioner 11 7.9 0.000 0.000 26.229 26.235 make_preconditioner 11 8.9 0.000 0.000 26.229 26.235 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.483 25.699 ot_scf_mini 99 9.5 0.002 0.002 24.670 24.723 multiply_cannon 2055 13.4 0.349 0.379 21.687 22.446 multiply_cannon_loop 2055 14.4 0.342 0.346 19.823 20.137 cp_fm_upper_to_full 70 14.2 12.671 18.130 12.671 18.130 ot_mini 99 10.5 0.001 0.001 13.645 13.695 rebuild_ks_matrix 110 8.3 0.000 0.001 13.037 13.068 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.036 13.067 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.864 11.891 dbcsr_complete_redistribute 325 12.2 1.028 1.060 7.328 10.424 multiply_cannon_multrec 8220 15.4 4.391 4.610 9.778 9.932 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.299 9.394 qs_ot_get_derivative 99 11.5 0.001 0.001 9.014 9.069 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.730 8.788 mp_waitall_1 87304 16.6 7.671 8.649 7.671 8.649 mp_alltoall_i22 605 13.7 5.347 8.463 5.347 8.463 sum_up_and_integrate 110 10.3 0.150 0.153 6.570 6.583 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.403 6.436 calculate_rho_elec 110 8.6 0.227 0.227 6.403 6.435 integrate_v_rspace 110 11.3 0.004 0.004 6.419 6.433 init_scf_run 11 5.9 0.000 0.001 6.137 6.137 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.137 6.137 qs_ot_get_p 110 10.4 0.001 0.001 5.871 5.926 make_m2s 4110 13.4 0.042 0.043 5.275 5.831 make_images 4110 14.4 0.881 0.930 5.088 5.642 dbcsr_mm_accdrv_process 11614 15.7 3.246 3.617 5.244 5.492 cp_fm_cholesky_invert 11 10.9 5.311 5.315 5.311 5.315 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.756 5.233 apply_single 110 13.6 0.000 0.000 4.756 5.233 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.657 5.039 ot_diis_step 99 11.5 0.015 0.015 4.612 4.612 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.321 4.328 multiply_cannon_sync_h2d 8220 15.4 3.948 3.952 3.948 3.952 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.876 3.876 grid_integrate_task_list 110 12.3 3.660 3.717 3.660 3.717 pw_transfer 1331 11.6 0.073 0.073 3.684 3.690 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.870 3.663 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.635 3.641 make_images_data 4110 15.4 0.042 0.044 2.891 3.588 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.569 3.575 calculate_dm_sparse 110 9.5 0.001 0.001 3.260 3.285 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.832 3.271 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.259 3.259 cp_fm_diag_elpa_base 48 14.0 2.724 2.917 3.257 3.257 wfi_extrapolate 11 7.9 0.001 0.001 3.195 3.195 fft_wrap_pw1pw2_140 451 13.1 0.214 0.216 3.088 3.092 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.974 2.978 density_rs2pw 110 9.6 0.004 0.004 2.906 2.925 calculate_first_density_matrix 1 7.0 0.000 0.000 2.838 2.839 fft3d_ps 1111 14.6 1.265 1.271 2.778 2.785 grid_collocate_task_list 110 9.6 2.629 2.649 2.629 2.649 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.472 2.483 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.395 2.426 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.170 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.180 rs_pw_transfer 902 11.9 0.010 0.011 2.073 2.114 mp_alltoall_d11v 2046 13.8 2.013 2.072 2.013 2.072 potential_pw2rs 110 12.3 0.021 0.021 2.015 2.021 cp_fm_cholesky_decompose 22 10.9 1.992 2.010 1.992 2.010 jit_kernel_multiply 10 15.3 1.798 1.971 1.798 1.971 qs_create_task_list 11 7.9 0.000 0.001 1.903 1.946 generate_qs_task_list 11 8.9 0.745 0.799 1.903 1.945 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.791 1.806 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.757 1.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.777 1.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.041000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1206.818182, yerr=61.179049 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.837824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 1601627. 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 0.031 0.126 201.279 201.281 qs_mol_dyn_low 1 2.0 0.003 0.005 200.748 200.762 qs_forces 11 3.9 0.005 0.020 200.636 200.648 qs_energies 11 4.9 0.005 0.026 195.088 195.121 scf_env_do_scf 11 5.9 0.001 0.002 178.469 178.474 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 157.640 157.644 dbcsr_multiply_generic 2507 12.6 0.194 0.197 122.869 123.591 velocity_verlet 10 3.0 0.001 0.002 121.364 121.365 qs_scf_new_mos 117 7.6 0.001 0.001 119.096 119.248 qs_scf_loop_do_ot 117 8.6 0.001 0.001 119.095 119.247 ot_scf_mini 117 9.6 0.003 0.004 112.573 112.772 multiply_cannon 2507 13.6 0.239 0.247 100.471 102.285 multiply_cannon_loop 2507 14.6 2.031 2.092 98.241 99.918 ot_mini 117 10.6 0.001 0.002 64.597 64.814 multiply_cannon_multrec 60168 15.6 33.732 35.849 41.771 44.094 qs_ot_get_derivative 117 11.6 0.001 0.002 39.820 40.008 rebuild_ks_matrix 128 8.3 0.001 0.001 33.219 33.431 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 33.219 33.431 mp_waitall_1 291448 16.2 27.721 31.266 27.721 31.266 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.872 30.093 multiply_cannon_sync_h2d 60168 15.6 28.013 29.694 28.013 29.694 qs_ot_get_p 128 10.4 0.001 0.002 26.306 26.626 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.328 25.139 apply_single 128 13.6 0.001 0.001 24.327 25.139 ot_diis_step 117 11.6 0.008 0.010 24.440 24.441 init_scf_loop 11 6.9 0.002 0.008 20.750 20.766 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.616 19.658 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.198 18.306 multiply_cannon_metrocomm3 60168 15.6 0.112 0.116 15.134 17.283 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.239 17.240 prepare_preconditioner 11 7.9 0.000 0.000 16.057 16.102 make_preconditioner 11 8.9 0.001 0.003 16.057 16.102 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.268 15.413 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.413 14.417 cp_fm_redistribute_end 83 14.4 11.359 14.342 11.370 14.344 make_m2s 5014 13.6 0.102 0.113 13.925 14.258 cp_fm_diag_elpa_base 83 14.4 2.932 14.116 2.964 14.237 make_images 5014 14.6 0.405 0.424 13.746 14.088 sum_up_and_integrate 128 10.3 0.091 0.108 13.894 13.911 integrate_v_rspace 128 11.3 0.003 0.004 13.802 13.822 qs_rho_update_rho_low 128 7.7 0.001 0.002 12.552 12.649 calculate_rho_elec 128 8.7 0.045 0.063 12.552 12.649 init_scf_run 11 5.9 0.000 0.001 12.387 12.388 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.387 12.388 cp_fm_cholesky_invert 11 10.9 9.173 9.181 9.173 9.181 wfi_extrapolate 11 7.9 0.001 0.001 9.128 9.128 mp_sum_l 7870 13.0 7.234 8.708 7.234 8.708 calculate_dm_sparse 128 9.5 0.001 0.001 8.329 8.431 dbcsr_mm_accdrv_process 124484 16.2 3.017 3.199 7.583 8.118 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.892 8.006 make_images_data 5014 15.6 0.069 0.075 6.895 7.738 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.679 7.732 grid_integrate_task_list 128 12.3 7.063 7.452 7.063 7.452 hybrid_alltoall_any 5200 16.5 0.291 2.269 6.006 7.308 multiply_cannon_metrocomm1 60168 15.6 0.086 0.091 5.575 7.275 pw_transfer 1547 11.6 0.074 0.104 6.846 7.123 density_rs2pw 128 9.7 0.006 0.007 6.421 6.910 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.644 6.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.707 6.716 rs_pw_transfer 1046 11.9 0.017 0.019 5.574 6.109 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.844 5.969 fft_wrap_pw1pw2_140 523 13.2 0.444 0.507 5.697 5.886 fft3d_ps 1291 14.7 2.096 2.569 5.439 5.628 mp_alltoall_d11v 2415 14.1 4.163 5.405 4.163 5.405 grid_collocate_task_list 128 9.7 4.722 5.092 4.722 5.092 cp_fm_cholesky_decompose 22 10.9 4.663 4.679 4.663 4.679 potential_pw2rs 128 12.3 0.009 0.010 4.424 4.465 mp_sum_d 4459 12.1 3.525 4.274 3.525 4.274 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=201.281000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.272727, yerr=6.743224 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 831.688704E+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 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 5122298. 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 0.016 0.032 187.672 187.674 qs_mol_dyn_low 1 2.0 0.003 0.003 187.307 187.320 qs_forces 11 3.9 0.003 0.004 187.227 187.228 qs_energies 11 4.9 0.002 0.010 180.547 180.555 scf_env_do_scf 11 5.9 0.001 0.001 164.368 164.378 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.862 131.865 velocity_verlet 10 3.0 0.001 0.002 117.964 117.965 dbcsr_multiply_generic 2507 12.6 0.184 0.191 97.198 98.386 qs_scf_new_mos 117 7.6 0.001 0.001 93.830 94.396 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.830 94.395 ot_scf_mini 117 9.6 0.004 0.005 89.039 89.703 multiply_cannon 2507 13.6 0.476 0.525 77.519 81.675 multiply_cannon_loop 2507 14.6 1.255 1.289 74.112 76.557 ot_mini 117 10.6 0.001 0.001 49.528 50.123 mp_waitall_1 226760 16.4 24.379 37.825 24.379 37.825 multiply_cannon_multrec 30084 15.6 22.050 26.008 31.759 36.186 rebuild_ks_matrix 128 8.3 0.001 0.001 32.461 32.972 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.460 32.972 init_scf_loop 11 6.9 0.001 0.006 32.419 32.421 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.244 29.715 qs_ot_get_derivative 117 11.6 0.001 0.002 27.518 28.166 prepare_preconditioner 11 7.9 0.000 0.000 28.071 28.153 make_preconditioner 11 8.9 0.000 0.001 28.071 28.153 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.616 27.424 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.798 27.333 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.084 23.242 apply_single 128 13.6 0.001 0.001 22.084 23.242 qs_ot_get_p 128 10.4 0.001 0.002 21.310 22.066 ot_diis_step 117 11.6 0.014 0.015 21.836 21.838 multiply_cannon_sync_h2d 30084 15.6 19.295 21.763 19.295 21.763 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.596 16.628 cp_fm_cholesky_invert 11 10.9 16.478 16.490 16.478 16.490 make_m2s 5014 13.6 0.088 0.094 14.024 15.747 make_images 5014 14.6 1.153 1.346 13.817 15.531 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.468 15.469 sum_up_and_integrate 128 10.3 0.117 0.131 14.074 14.106 integrate_v_rspace 128 11.3 0.003 0.004 13.957 13.991 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.593 12.627 calculate_rho_elec 128 8.7 0.088 0.105 12.593 12.627 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.413 12.425 cp_fm_redistribute_end 83 14.4 7.219 12.362 7.229 12.362 cp_fm_diag_elpa_base 83 14.4 4.897 11.908 5.113 12.278 init_scf_run 11 5.9 0.000 0.001 11.455 11.457 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.455 11.456 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.794 11.283 make_images_data 5014 15.6 0.067 0.073 8.478 10.436 multiply_cannon_metrocomm4 27577 15.6 0.097 0.114 3.745 10.421 mp_irecv_dv 69486 16.3 3.551 10.042 3.551 10.042 hybrid_alltoall_any 5200 16.5 0.340 1.474 7.218 9.765 dbcsr_mm_accdrv_process 62242 16.2 4.569 5.555 9.167 9.744 wfi_extrapolate 11 7.9 0.001 0.001 8.246 8.246 pw_transfer 1547 11.6 0.084 0.096 7.517 7.591 grid_integrate_task_list 128 12.3 7.166 7.572 7.166 7.572 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.294 7.363 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.187 6.924 density_rs2pw 128 9.7 0.006 0.006 6.465 6.923 cp_fm_cholesky_decompose 22 10.9 6.808 6.883 6.808 6.883 calculate_dm_sparse 128 9.5 0.001 0.001 6.437 6.570 fft_wrap_pw1pw2_140 523 13.2 0.469 0.520 6.412 6.498 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.234 6.246 rs_pw_transfer 1046 11.9 0.014 0.016 5.277 5.849 mp_sum_l 7870 13.0 3.875 5.816 3.875 5.816 fft3d_ps 1291 14.7 2.771 2.951 5.667 5.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.393 5.522 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.326 5.398 grid_collocate_task_list 128 9.7 4.913 5.267 4.913 5.267 mp_allgather_i34 2507 14.6 1.979 5.065 1.979 5.065 mp_alltoall_d11v 2415 14.1 3.887 4.744 3.887 4.744 potential_pw2rs 128 12.3 0.015 0.017 4.576 4.593 mp_sum_d 4464 12.1 2.643 3.967 2.643 3.967 dbcsr_complete_redistribute 395 12.7 0.767 0.859 3.010 3.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.674000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.363636, yerr=6.064393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 969.388032E+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 0.048 0.150 174.795 174.797 qs_mol_dyn_low 1 2.0 0.003 0.004 174.102 174.115 qs_forces 11 3.9 0.014 0.055 174.006 174.009 qs_energies 11 4.9 0.003 0.008 167.461 167.480 scf_env_do_scf 11 5.9 0.001 0.002 152.236 152.238 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.426 117.427 velocity_verlet 10 3.0 0.001 0.002 111.611 111.613 dbcsr_multiply_generic 2507 12.6 0.182 0.185 80.947 82.028 qs_scf_new_mos 117 7.6 0.001 0.001 81.321 81.733 qs_scf_loop_do_ot 117 8.6 0.001 0.002 81.320 81.732 ot_scf_mini 117 9.6 0.004 0.004 77.191 77.677 multiply_cannon 2507 13.6 0.498 0.513 61.396 65.001 multiply_cannon_loop 2507 14.6 0.860 0.885 58.314 61.071 ot_mini 117 10.6 0.001 0.001 42.247 42.711 init_scf_loop 11 6.9 0.001 0.004 34.711 34.715 mp_waitall_1 178456 16.5 24.766 34.115 24.766 34.115 prepare_preconditioner 11 7.9 0.000 0.000 30.747 30.798 make_preconditioner 11 8.9 0.000 0.002 30.747 30.798 rebuild_ks_matrix 128 8.3 0.001 0.001 29.823 30.307 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 29.823 30.307 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.392 29.734 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.892 27.323 multiply_cannon_multrec 20056 15.6 13.509 16.880 22.184 25.525 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 14.966 23.945 qs_ot_get_derivative 117 11.6 0.001 0.002 22.536 23.025 qs_ot_get_p 128 10.4 0.001 0.002 20.376 20.904 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.801 20.895 apply_single 128 13.6 0.001 0.001 19.801 20.895 ot_diis_step 117 11.6 0.018 0.018 19.604 19.604 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.167 16.174 multiply_cannon_sync_h2d 20056 15.6 14.300 15.914 14.300 15.914 make_m2s 5014 13.6 0.080 0.085 14.542 15.400 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.162 15.163 make_images 5014 14.6 1.172 1.259 14.310 15.159 cp_fm_cholesky_invert 11 10.9 14.337 14.346 14.337 14.346 sum_up_and_integrate 128 10.3 0.134 0.146 14.026 14.062 integrate_v_rspace 128 11.3 0.004 0.004 13.892 13.932 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.810 12.861 calculate_rho_elec 128 8.7 0.132 0.147 12.809 12.860 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.015 12.016 cp_fm_redistribute_end 83 14.4 4.511 11.950 4.525 11.952 cp_fm_diag_elpa_base 83 14.4 7.002 11.381 7.407 11.865 init_scf_run 11 5.9 0.000 0.001 10.208 10.209 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.208 10.209 make_images_data 5014 15.6 0.061 0.068 8.938 10.131 hybrid_alltoall_any 5200 16.5 0.430 1.952 7.827 9.419 multiply_cannon_metrocomm4 17549 15.6 0.063 0.074 3.460 9.297 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.822 9.192 mp_irecv_dv 50230 16.2 3.335 9.043 3.335 9.043 dbcsr_mm_accdrv_process 41502 16.2 4.452 5.228 8.129 8.267 pw_transfer 1547 11.6 0.084 0.103 7.623 7.740 grid_integrate_task_list 128 12.3 7.334 7.633 7.334 7.633 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.400 7.523 cp_fm_upper_to_full 105 14.5 5.707 7.401 5.707 7.401 cp_fm_cholesky_decompose 22 10.9 7.305 7.398 7.305 7.398 wfi_extrapolate 11 7.9 0.001 0.001 7.290 7.290 fft_wrap_pw1pw2_140 523 13.2 0.478 0.528 6.517 6.647 density_rs2pw 128 9.7 0.006 0.006 6.337 6.627 dbcsr_complete_redistribute 395 12.7 1.166 1.195 4.539 6.325 calculate_dm_sparse 128 9.5 0.001 0.001 5.718 5.811 fft3d_ps 1291 14.7 2.690 2.894 5.667 5.740 grid_collocate_task_list 128 9.7 5.061 5.536 5.061 5.536 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.345 5.352 rs_pw_transfer 1046 11.9 0.013 0.014 4.969 5.343 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.486 5.173 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.379 5.158 mp_sum_l 7870 13.0 3.195 4.670 3.195 4.670 mp_alltoall_d11v 2415 14.1 4.140 4.617 4.140 4.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.466 4.591 mp_allgather_i34 2507 14.6 1.606 4.582 1.606 4.582 potential_pw2rs 128 12.3 0.020 0.022 4.445 4.466 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.336 4.085 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.021 4.054 mp_alltoall_i22 716 14.1 1.888 3.791 1.888 3.791 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.746 3.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.797000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=912.545455, yerr=17.839505 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 1.139356E+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 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57136. MP_Allreduce 11261 1067. MP_Sync 168 MP_Alltoall 1700 12496371. 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 0.048 0.067 184.946 184.948 qs_mol_dyn_low 1 2.0 0.003 0.004 184.486 184.498 qs_forces 11 3.9 0.005 0.020 184.386 184.391 qs_energies 11 4.9 0.003 0.007 177.399 177.413 scf_env_do_scf 11 5.9 0.001 0.003 160.804 160.814 velocity_verlet 10 3.0 0.002 0.002 122.362 122.365 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 114.234 114.236 qs_scf_new_mos 116 7.6 0.001 0.001 79.150 79.440 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.149 79.439 dbcsr_multiply_generic 2485 12.5 0.189 0.195 78.400 79.103 ot_scf_mini 116 9.6 0.003 0.004 74.691 74.983 multiply_cannon 2485 13.5 0.551 0.576 53.938 57.611 multiply_cannon_loop 2485 14.5 1.170 1.196 50.326 51.885 init_scf_loop 11 6.9 0.001 0.002 46.441 46.442 prepare_preconditioner 11 7.9 0.000 0.000 42.298 42.324 make_preconditioner 11 8.9 0.000 0.002 42.298 42.324 ot_mini 116 10.6 0.001 0.001 41.519 41.807 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.800 40.877 multiply_cannon_multrec 29820 15.5 14.182 19.644 25.958 30.766 rebuild_ks_matrix 127 8.3 0.001 0.001 28.840 29.023 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 28.840 29.022 mp_waitall_1 152434 16.5 17.230 27.029 17.230 27.029 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.072 26.236 qs_ot_get_derivative 116 11.6 0.001 0.002 22.184 22.478 make_m2s 4970 13.5 0.093 0.098 20.263 21.238 make_images 4970 14.5 1.938 2.279 19.960 20.935 qs_ot_get_p 127 10.4 0.001 0.003 19.501 19.824 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.754 19.372 apply_single 127 13.6 0.001 0.001 18.753 19.371 ot_diis_step 116 11.6 0.018 0.021 19.212 19.213 cp_fm_upper_to_full 104 14.7 11.208 16.622 11.208 16.622 cp_fm_cholesky_invert 11 10.9 16.069 16.077 16.069 16.077 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.508 15.559 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.176 14.527 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.253 14.254 sum_up_and_integrate 127 10.3 0.140 0.154 13.911 13.968 integrate_v_rspace 127 11.3 0.004 0.004 13.770 13.833 dbcsr_complete_redistribute 393 12.7 1.506 1.622 9.086 12.967 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.904 12.935 calculate_rho_elec 127 8.7 0.174 0.189 12.903 12.934 make_images_data 4970 15.5 0.063 0.068 10.987 12.580 multiply_cannon_sync_h2d 29820 15.5 11.673 12.490 11.673 12.490 dbcsr_mm_accdrv_process 61748 16.2 7.317 8.259 11.353 11.873 hybrid_alltoall_any 5155 16.4 0.511 2.053 9.611 11.859 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.710 11.500 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.192 11.194 cp_fm_redistribute_end 82 14.4 1.903 11.121 1.917 11.127 cp_fm_diag_elpa_base 82 14.4 8.600 10.536 9.178 11.024 init_scf_run 11 5.9 0.000 0.001 10.658 10.659 scf_env_initial_rho_setup 11 6.9 0.002 0.002 10.657 10.659 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.474 10.145 mp_alltoall_i22 712 14.1 5.670 9.442 5.670 9.442 qs_ot_get_derivative_diag 76 12.4 0.003 0.006 9.128 9.345 pw_transfer 1535 11.6 0.084 0.098 7.804 7.888 grid_integrate_task_list 127 12.3 7.509 7.779 7.509 7.779 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.580 7.668 wfi_extrapolate 11 7.9 0.001 0.001 7.590 7.590 cp_fm_cholesky_decompose 22 10.9 7.463 7.587 7.463 7.587 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.679 7.127 mp_irecv_dv 75445 16.2 2.536 6.862 2.536 6.862 fft_wrap_pw1pw2_140 519 13.2 0.478 0.489 6.706 6.795 density_rs2pw 127 9.7 0.006 0.006 6.145 6.533 calculate_dm_sparse 127 9.5 0.001 0.001 6.124 6.204 fft3d_ps 1281 14.7 2.772 2.847 5.824 5.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.472 5.510 grid_collocate_task_list 127 9.7 5.163 5.503 5.163 5.503 mp_alltoall_d11v 2401 14.1 4.668 5.149 4.668 5.149 rs_pw_transfer 1038 11.9 0.013 0.015 4.490 4.962 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.449 4.451 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.392 4.436 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.311 4.385 potential_pw2rs 127 12.3 0.022 0.023 4.203 4.253 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.123 4.190 mp_allgather_i34 2485 14.5 1.722 4.011 1.722 4.011 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=184.948000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.000000, yerr=17.304755 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.505247E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828189. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.047 170.686 170.686 qs_mol_dyn_low 1 2.0 0.003 0.003 170.282 170.293 qs_forces 11 3.9 0.003 0.003 170.185 170.190 qs_energies 11 4.9 0.001 0.002 162.885 162.894 scf_env_do_scf 11 5.9 0.001 0.006 145.699 145.713 velocity_verlet 10 3.0 0.002 0.002 110.995 110.998 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 110.119 110.120 qs_scf_new_mos 116 7.6 0.001 0.001 75.167 75.300 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.167 75.299 dbcsr_multiply_generic 2485 12.5 0.180 0.184 73.676 74.140 ot_scf_mini 116 9.6 0.004 0.004 70.793 70.904 multiply_cannon 2485 13.5 0.578 0.604 54.240 59.669 multiply_cannon_loop 2485 14.5 0.446 0.458 49.656 50.553 ot_mini 116 10.6 0.001 0.001 39.339 39.465 init_scf_loop 11 6.9 0.000 0.000 35.418 35.419 mp_waitall_1 128490 16.6 25.760 32.964 25.760 32.964 prepare_preconditioner 11 7.9 0.000 0.000 31.567 31.594 make_preconditioner 11 8.9 0.000 0.000 31.567 31.594 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.420 29.697 rebuild_ks_matrix 127 8.3 0.001 0.001 28.130 28.269 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.129 28.269 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.631 25.763 multiply_cannon_multrec 9940 15.5 10.332 14.775 17.832 21.855 ot_diis_step 116 11.6 0.019 0.020 19.851 19.852 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.407 19.724 apply_single 127 13.6 0.001 0.001 19.406 19.724 multiply_cannon_metrocomm3 9940 15.5 0.022 0.024 12.501 19.626 qs_ot_get_derivative 116 11.6 0.001 0.002 19.413 19.536 qs_ot_get_p 127 10.4 0.001 0.001 18.637 18.781 cp_fm_cholesky_invert 11 10.9 18.028 18.036 18.028 18.036 make_m2s 4970 13.5 0.065 0.069 15.749 17.954 make_images 4970 14.5 2.305 2.667 15.446 17.643 qs_ot_p2m_diag 82 11.4 0.489 0.496 14.876 14.890 sum_up_and_integrate 127 10.3 0.180 0.189 13.842 13.892 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.754 13.755 integrate_v_rspace 127 11.3 0.004 0.004 13.662 13.718 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.929 12.967 calculate_rho_elec 127 8.7 0.256 0.267 12.929 12.967 multiply_cannon_sync_h2d 9940 15.5 11.444 12.206 11.444 12.206 make_images_data 4970 15.5 0.052 0.059 9.530 11.769 hybrid_alltoall_any 5155 16.4 0.830 3.775 9.360 11.408 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.762 10.764 cp_fm_diag_elpa_base 82 14.4 10.525 10.594 10.755 10.755 init_scf_run 11 5.9 0.000 0.001 10.399 10.399 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.399 10.399 grid_integrate_task_list 127 12.3 7.743 8.102 7.743 8.102 cp_fm_cholesky_decompose 22 10.9 7.854 7.975 7.854 7.975 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.725 7.815 dbcsr_mm_accdrv_process 20590 16.1 2.652 3.522 7.133 7.788 pw_transfer 1535 11.6 0.082 0.089 7.610 7.623 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.256 7.423 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.392 7.404 mp_allgather_i34 2485 14.5 2.803 7.304 2.803 7.304 wfi_extrapolate 11 7.9 0.001 0.001 7.272 7.272 fft_wrap_pw1pw2_140 519 13.2 0.501 0.518 6.490 6.508 calculate_dm_sparse 127 9.5 0.001 0.001 6.091 6.168 density_rs2pw 127 9.7 0.005 0.006 5.813 6.137 dbcsr_complete_redistribute 393 12.7 2.138 2.171 5.259 5.692 grid_collocate_task_list 127 9.7 5.445 5.634 5.445 5.634 fft3d_ps 1281 14.7 2.709 2.773 5.563 5.607 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.329 5.344 mp_alltoall_d11v 2401 14.1 4.577 5.226 4.577 5.226 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.216 5.217 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.846 4.845 mp_irecv_dv 28618 15.9 1.809 4.762 1.809 4.762 rs_pw_transfer 1038 11.9 0.012 0.013 4.007 4.343 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.222 4.268 potential_pw2rs 127 12.3 0.026 0.026 4.017 4.037 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.566 3.854 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.449 3.789 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.688 3.738 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.558 3.610 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.391 3.423 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.686000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1394.000000, yerr=49.248904 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 1973537472512 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 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.038724E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 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 0.045 291.906 291.907 qs_mol_dyn_low 1 2.0 0.003 0.003 291.393 291.404 qs_forces 11 3.9 0.003 0.003 291.292 291.297 qs_energies 11 4.9 0.002 0.002 282.680 282.687 scf_env_do_scf 11 5.9 0.001 0.002 260.964 260.973 velocity_verlet 10 3.0 0.002 0.004 211.827 211.835 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.484 135.486 init_scf_loop 11 6.9 0.000 0.000 125.224 125.227 prepare_preconditioner 11 7.9 0.000 0.000 120.521 120.549 make_preconditioner 11 8.9 0.000 0.000 120.521 120.549 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.070 117.646 qs_scf_new_mos 118 7.6 0.001 0.001 93.200 93.339 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.199 93.338 ot_scf_mini 118 9.6 0.004 0.004 88.349 88.462 dbcsr_multiply_generic 2534 12.6 0.215 0.221 84.069 84.885 cp_fm_upper_to_full 106 14.8 53.175 76.626 53.175 76.626 multiply_cannon 2534 13.6 0.709 0.755 58.995 60.128 multiply_cannon_loop 2534 14.6 0.479 0.487 55.109 57.034 ot_mini 118 10.6 0.001 0.001 45.100 45.222 dbcsr_complete_redistribute 397 12.7 3.988 4.039 30.108 43.038 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.738 39.554 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.404 37.097 mp_alltoall_i22 720 14.1 22.241 35.289 22.241 35.289 rebuild_ks_matrix 129 8.3 0.001 0.001 33.773 33.906 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.772 33.906 cp_fm_cholesky_invert 11 10.9 33.718 33.724 33.718 33.724 mp_waitall_1 106586 16.7 28.434 32.251 28.434 32.251 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.549 31.667 qs_ot_get_p 129 10.4 0.001 0.001 27.488 27.618 qs_ot_get_derivative 118 11.6 0.002 0.002 24.496 24.612 qs_ot_p2m_diag 84 11.4 0.890 0.897 23.312 23.339 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.552 21.555 make_m2s 5068 13.6 0.076 0.079 20.099 21.211 make_images 5068 14.6 3.806 3.931 19.617 20.729 multiply_cannon_metrocomm3 10136 15.6 0.023 0.024 18.810 20.659 ot_diis_step 118 11.6 0.022 0.023 20.559 20.559 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.656 19.962 apply_single 129 13.6 0.001 0.001 19.656 19.962 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.261 18.262 cp_fm_diag_elpa_base 84 14.4 13.805 15.418 18.257 18.257 multiply_cannon_multrec 10136 15.6 10.535 12.394 17.948 18.033 sum_up_and_integrate 129 10.3 0.324 0.326 15.801 15.900 multiply_cannon_sync_h2d 10136 15.6 15.783 15.793 15.783 15.793 integrate_v_rspace 129 11.3 0.004 0.004 15.476 15.576 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.054 15.068 calculate_rho_elec 129 8.7 0.486 0.487 15.053 15.067 hybrid_alltoall_any 5255 16.5 1.313 3.071 10.858 12.872 make_images_data 5068 15.6 0.061 0.066 10.741 12.627 init_scf_run 11 5.9 0.000 0.001 11.720 11.720 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.720 11.720 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.589 9.682 dbcsr_mm_accdrv_process 20954 16.1 3.763 5.573 7.174 9.011 cp_fm_cholesky_decompose 22 10.9 8.972 9.008 8.972 9.008 grid_integrate_task_list 129 12.3 8.600 8.764 8.600 8.764 wfi_extrapolate 11 7.9 0.001 0.001 8.683 8.683 pw_transfer 1559 11.6 0.091 0.092 8.341 8.349 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.106 8.113 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.990 7.990 fft_wrap_pw1pw2_140 527 13.2 0.545 0.548 7.164 7.176 mp_alltoall_d11v 2429 14.1 7.017 7.161 7.017 7.161 calculate_dm_sparse 129 9.5 0.001 0.001 6.688 6.730 grid_collocate_task_list 129 9.7 6.386 6.423 6.386 6.423 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.284 6.328 fft3d_ps 1301 14.7 2.789 2.822 6.131 6.142 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.052 6.142 density_rs2pw 129 9.7 0.005 0.005 6.053 6.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.907000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2678.000000, yerr=184.133351 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 1.261191E+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 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 84.797 84.797 qs_energies 1 2.0 0.000 0.000 84.383 84.391 ls_scf 1 3.0 0.000 0.000 83.495 83.503 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.599 72.759 multiply_cannon 111 7.7 0.017 0.020 55.943 57.088 multiply_cannon_loop 111 8.7 0.208 0.221 52.500 53.872 ls_scf_main 1 4.0 0.000 0.000 52.240 52.241 density_matrix_trs4 2 5.0 0.002 0.003 46.807 46.870 ls_scf_init_scf 1 4.0 0.000 0.000 28.210 28.211 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.151 27.204 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.038 25.059 mp_waitall_1 11316 10.9 22.282 24.863 22.282 24.863 multiply_cannon_multrec 2664 9.7 8.184 8.827 15.481 17.119 multiply_cannon_sync_h2d 2664 9.7 13.772 15.600 13.772 15.600 make_m2s 222 7.7 0.008 0.011 12.945 13.513 make_images 222 8.7 0.100 0.109 12.923 13.493 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.731 12.489 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.354 8.346 make_images_data 222 9.7 0.004 0.005 7.518 8.083 dbcsr_mm_accdrv_process 4760 10.4 0.498 0.601 6.917 7.903 hybrid_alltoall_any 227 10.6 0.216 1.838 6.440 7.635 dbcsr_mm_accdrv_process_sort 4760 11.4 6.219 7.152 6.219 7.152 calculate_norms 4752 9.8 5.524 6.132 5.524 6.132 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.055 5.207 mp_sum_l 807 5.4 3.183 4.645 3.183 4.645 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.372 3.647 make_images_sizes 222 9.7 0.000 0.000 0.671 3.410 mp_alltoall_i44 222 10.7 0.671 3.410 0.671 3.410 arnoldi_extremal 4 6.8 0.000 0.000 3.324 3.354 arnoldi_normal_ev 4 7.8 0.001 0.003 3.324 3.354 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.059 3.301 mp_irecv_dv 6231 10.9 2.041 3.273 2.041 3.273 build_subspace 16 8.4 0.009 0.012 3.216 3.218 ls_scf_post 1 4.0 0.000 0.000 3.045 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.898 dbcsr_special_finalize 555 9.7 0.005 0.006 2.399 2.832 dbcsr_merge_single_wm 555 10.7 0.464 0.630 2.391 2.824 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.335 2.638 make_images_pack 222 9.7 2.209 2.629 2.211 2.631 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.525 2.067 2.527 dbcsr_sort_data 658 11.4 2.185 2.515 2.185 2.515 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.192 2.281 buffer_matrices_ensure_size 222 8.7 1.764 2.204 1.764 2.204 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.717 1.719 rebuild_ks_matrix 3 7.3 0.000 0.000 1.708 1.709 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.708 1.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.797000, 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/f36b488847463310f25cb3cd7662924bf5e00627_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 2.093507E+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 0.036 0.076 91.135 91.137 qs_energies 1 2.0 0.000 0.000 90.454 90.458 ls_scf 1 3.0 0.000 0.000 89.147 89.151 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.344 75.721 multiply_cannon 111 7.7 0.028 0.046 53.391 56.930 ls_scf_main 1 4.0 0.000 0.000 54.851 54.851 multiply_cannon_loop 111 8.7 0.116 0.124 50.123 52.912 density_matrix_trs4 2 5.0 0.002 0.003 49.197 49.400 ls_scf_init_scf 1 4.0 0.000 0.000 30.717 30.718 mp_waitall_1 9246 10.9 21.508 30.573 21.508 30.573 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.588 29.686 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.166 27.180 multiply_cannon_multrec 1332 9.7 13.068 16.421 22.264 26.739 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.038 21.292 make_m2s 222 7.7 0.006 0.007 15.462 16.176 make_images 222 8.7 1.570 1.918 15.432 16.147 dbcsr_mm_accdrv_process 4041 10.4 0.291 0.463 8.799 10.392 make_images_data 222 9.7 0.004 0.004 8.873 9.943 dbcsr_mm_accdrv_process_sort 4041 11.4 8.392 9.929 8.392 9.929 hybrid_alltoall_any 227 10.6 0.521 2.428 8.295 9.457 mp_sum_l 807 5.4 5.202 7.984 5.202 7.984 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.221 7.888 mp_irecv_dv 3311 11.0 3.202 7.818 3.202 7.818 calculate_norms 2376 9.8 6.001 6.900 6.001 6.900 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.005 6.541 multiply_cannon_sync_h2d 1332 9.7 4.811 5.838 4.811 5.838 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.093 5.296 arnoldi_extremal 4 6.8 0.000 0.000 4.651 4.674 arnoldi_normal_ev 4 7.8 0.001 0.005 4.650 4.673 build_subspace 16 8.4 0.014 0.021 4.395 4.398 ls_scf_post 1 4.0 0.000 0.000 3.579 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.260 3.402 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.129 3.381 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.234 2.744 3.236 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.221 2.627 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.516 2.604 mp_allgather_i34 111 8.7 0.965 2.510 0.965 2.510 make_images_pack 222 9.7 2.029 2.435 2.031 2.437 dbcsr_sort_data 436 11.2 1.823 2.054 1.823 2.054 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.137000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1766.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 2.704695E+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 0.038 0.053 93.702 93.702 qs_energies 1 2.0 0.000 0.000 93.046 93.061 ls_scf 1 3.0 0.000 0.001 91.649 91.666 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.486 76.810 ls_scf_main 1 4.0 0.000 0.000 57.549 57.555 multiply_cannon 111 7.7 0.040 0.093 52.987 56.802 multiply_cannon_loop 111 8.7 0.100 0.119 49.441 52.406 density_matrix_trs4 2 5.0 0.002 0.003 51.645 51.783 mp_waitall_1 7374 11.0 24.035 33.519 24.035 33.519 ls_scf_init_scf 1 4.0 0.000 0.001 30.536 30.537 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.406 29.472 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.001 27.013 multiply_cannon_multrec 888 9.7 12.592 15.205 21.103 24.199 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.152 22.466 make_m2s 222 7.7 0.006 0.006 16.994 18.238 make_images 222 8.7 1.974 2.314 16.956 18.199 make_images_data 222 9.7 0.003 0.004 9.763 10.685 hybrid_alltoall_any 227 10.6 0.619 2.857 9.423 10.603 dbcsr_mm_accdrv_process 3754 10.4 0.251 0.414 8.041 9.243 dbcsr_mm_accdrv_process_sort 3754 11.4 7.666 8.832 7.666 8.832 mp_sum_l 807 5.4 4.949 7.541 4.949 7.541 multiply_cannon_sync_h2d 888 9.7 6.126 7.215 6.126 7.215 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.455 6.934 mp_irecv_dv 2335 11.1 2.440 6.877 2.440 6.877 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.737 6.531 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.741 5.782 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.991 5.177 arnoldi_extremal 4 6.8 0.000 0.000 5.132 5.148 arnoldi_normal_ev 4 7.8 0.001 0.005 5.132 5.148 build_subspace 16 8.4 0.014 0.020 4.829 4.836 calculate_norms 1584 9.8 4.255 4.572 4.255 4.572 mp_allgather_i34 111 8.7 1.388 3.836 1.388 3.836 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.461 3.805 dbcsr_matrix_vector_mult_local 304 10.0 3.043 3.621 3.045 3.623 ls_scf_post 1 4.0 0.002 0.017 3.564 3.579 ls_scf_store_result 1 5.0 0.000 0.000 3.312 3.389 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.738 2.844 make_images_pack 222 9.7 1.824 2.133 1.827 2.136 dbcsr_sort_data 325 11.1 1.890 2.104 1.890 2.104 make_images_sizes 222 9.7 0.000 0.000 0.894 1.988 mp_alltoall_i44 222 10.7 0.894 1.988 0.894 1.988 dbcsr_data_release 9322 10.9 1.300 1.966 1.300 1.966 dbcsr_finalize 304 7.8 0.026 0.032 1.606 1.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.702000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 3.348263E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.049 97.001 97.002 qs_energies 1 2.0 0.000 0.000 96.406 96.411 ls_scf 1 3.0 0.000 0.000 94.759 94.764 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.741 79.056 ls_scf_main 1 4.0 0.000 0.000 58.957 58.958 multiply_cannon 111 7.7 0.045 0.154 51.766 55.729 density_matrix_trs4 2 5.0 0.002 0.003 52.908 53.055 multiply_cannon_loop 111 8.7 0.115 0.129 46.653 50.721 ls_scf_init_scf 1 4.0 0.000 0.000 32.533 32.535 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.400 31.481 mp_waitall_1 6438 11.0 22.653 29.432 22.653 29.432 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.948 28.962 multiply_cannon_multrec 1332 9.7 14.215 17.488 22.060 24.495 make_m2s 222 7.7 0.007 0.008 20.885 22.389 make_images 222 8.7 3.135 3.590 20.835 22.341 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.346 17.357 make_images_data 222 9.7 0.004 0.004 11.602 13.262 hybrid_alltoall_any 227 10.6 0.794 3.762 10.815 12.730 dbcsr_mm_accdrv_process 3641 10.4 0.223 0.408 7.483 9.021 dbcsr_mm_accdrv_process_sort 3641 11.4 7.105 8.604 7.105 8.604 mp_sum_l 807 5.4 4.526 8.169 4.526 8.169 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.518 6.510 multiply_cannon_sync_h2d 1332 9.7 5.471 6.136 5.471 6.136 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.091 5.987 mp_irecv_dv 3229 10.9 2.068 5.910 2.068 5.910 arnoldi_extremal 4 6.8 0.000 0.000 5.236 5.249 arnoldi_normal_ev 4 7.8 0.001 0.005 5.236 5.249 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.527 4.976 build_subspace 16 8.4 0.014 0.021 4.899 4.908 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.582 4.808 mp_allgather_i34 111 8.7 2.238 4.699 2.238 4.699 calculate_norms 2376 9.8 4.221 4.601 4.221 4.601 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.606 3.885 dbcsr_matrix_vector_mult_local 304 10.0 3.197 3.694 3.199 3.696 dbcsr_sort_data 658 11.4 3.017 3.385 3.017 3.385 ls_scf_post 1 4.0 0.000 0.000 3.269 3.275 dbcsr_special_finalize 555 9.7 0.006 0.007 2.766 3.220 dbcsr_merge_single_wm 555 10.7 0.531 0.662 2.757 3.212 ls_scf_store_result 1 5.0 0.000 0.000 3.001 3.075 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.886 2.927 dbcsr_data_release 10477 10.7 1.576 2.439 1.576 2.439 dbcsr_finalize 304 7.8 0.049 0.061 1.791 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.002000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 4.624089E+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 0.040 0.057 98.106 98.107 qs_energies 1 2.0 0.000 0.000 97.298 97.303 ls_scf 1 3.0 0.000 0.000 95.394 95.399 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.218 77.434 ls_scf_main 1 4.0 0.000 0.000 61.395 61.395 multiply_cannon 111 7.7 0.071 0.131 54.840 59.744 density_matrix_trs4 2 5.0 0.002 0.003 54.436 54.510 multiply_cannon_loop 111 8.7 0.069 0.079 50.279 51.594 mp_waitall_1 5481 11.0 25.961 31.754 25.961 31.754 ls_scf_init_scf 1 4.0 0.000 0.000 30.423 30.426 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.253 29.291 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.095 27.104 multiply_cannon_multrec 444 9.7 14.079 16.402 21.105 23.763 make_m2s 222 7.7 0.004 0.005 17.659 20.230 make_images 222 8.7 3.728 4.430 17.597 20.171 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.799 16.245 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.074 14.587 make_images_data 222 9.7 0.003 0.004 9.889 12.385 hybrid_alltoall_any 227 10.6 0.792 3.784 9.678 12.249 multiply_cannon_sync_h2d 444 9.7 6.561 8.100 6.561 8.100 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.344 6.733 7.854 dbcsr_mm_accdrv_process_sort 3003 11.4 6.418 7.509 6.418 7.509 mp_allgather_i34 111 8.7 2.789 6.977 2.789 6.977 arnoldi_extremal 4 6.8 0.000 0.000 5.795 5.805 arnoldi_normal_ev 4 7.8 0.002 0.005 5.795 5.805 build_subspace 16 8.4 0.015 0.020 5.412 5.422 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.576 4.713 mp_sum_l 807 5.4 2.768 4.516 2.768 4.516 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.182 4.385 dbcsr_matrix_vector_mult_local 304 10.0 3.729 4.204 3.732 4.206 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.566 3.814 mp_irecv_dv 1241 11.2 1.547 3.792 1.547 3.792 calculate_norms 792 9.8 3.539 3.666 3.539 3.666 ls_scf_post 1 4.0 0.000 0.000 3.577 3.581 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.357 3.447 make_images_sizes 222 9.7 0.000 0.000 1.109 3.442 mp_alltoall_i44 222 10.7 1.108 3.441 1.108 3.441 ls_scf_store_result 1 5.0 0.000 0.000 3.341 3.406 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.871 3.217 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.241 dbcsr_merge_all 275 8.9 0.473 0.519 2.047 2.088 dbcsr_data_release 10123 10.8 1.321 1.963 1.321 1.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.107000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3631.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f36b488847463310f25cb3cd7662924bf5e00627_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 8.768528E+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 0.110 0.167 107.846 107.860 qs_energies 1 2.0 0.000 0.000 106.340 106.352 ls_scf 1 3.0 0.000 0.000 103.427 103.438 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.635 77.775 ls_scf_main 1 4.0 0.000 0.000 65.816 65.817 density_matrix_trs4 2 5.0 0.002 0.003 56.993 57.052 multiply_cannon 111 7.7 0.126 0.312 49.869 51.722 multiply_cannon_loop 111 8.7 0.068 0.072 46.403 47.463 ls_scf_init_scf 1 4.0 0.000 0.000 33.942 33.943 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.596 32.619 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.802 29.810 mp_waitall_1 4569 11.1 22.215 26.130 22.215 26.130 make_m2s 222 7.7 0.005 0.005 23.953 24.941 make_images 222 8.7 4.587 4.973 23.847 24.832 multiply_cannon_multrec 444 9.7 17.874 18.713 22.492 23.127 hybrid_alltoall_any 227 10.6 1.661 3.638 12.953 15.669 make_images_data 222 9.7 0.003 0.003 13.174 15.624 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.404 11.123 multiply_cannon_sync_h2d 444 9.7 8.823 8.861 8.823 8.861 arnoldi_extremal 4 6.8 0.000 0.000 7.391 7.403 arnoldi_normal_ev 4 7.8 0.003 0.009 7.391 7.403 build_subspace 16 8.4 0.026 0.037 6.843 6.854 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.500 5.649 dbcsr_matrix_vector_mult_local 304 10.0 5.090 5.393 5.093 5.396 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.997 5.251 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.779 4.872 dbcsr_mm_accdrv_process 1814 10.4 0.243 0.346 4.442 4.578 dbcsr_mm_accdrv_process_sort 1814 11.4 4.131 4.253 4.131 4.253 ls_scf_post 1 4.0 0.000 0.000 3.669 3.680 make_images_sizes 222 9.7 0.000 0.000 1.457 3.534 mp_alltoall_i44 222 10.7 1.457 3.533 1.457 3.533 mp_allgather_i34 111 8.7 1.125 3.526 1.125 3.526 ls_scf_store_result 1 5.0 0.000 0.000 3.401 3.417 calculate_norms 792 9.8 3.263 3.299 3.263 3.299 dbcsr_finalize 304 7.8 0.082 0.089 3.080 3.127 dbcsr_merge_all 275 8.9 0.891 0.915 2.867 2.907 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.882 2.882 dbcsr_complete_redistribute 5 7.6 1.432 1.466 2.762 2.872 dbcsr_data_release 12724 10.6 2.329 2.848 2.329 2.848 matrix_ls_to_qs 2 6.0 0.000 0.000 2.422 2.542 dbcsr_sort_data 325 11.1 2.441 2.504 2.441 2.504 mp_sum_l 807 5.4 1.507 2.348 1.507 2.348 dbcsr_new_transposed 4 7.5 0.242 0.250 2.267 2.277 dbcsr_frobenius_norm 74 6.6 2.058 2.139 2.202 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.205 dbcsr_add_anytype 103 7.2 0.860 0.892 2.131 2.205 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.197 2.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.860000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6816.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: f36b488847463310f25cb3cd7662924bf5e00627 Summary: empty Status: OK