=== 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: 848fd4fc57d18f38d66960eafe7959570f56e303 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/01 job id: 41614827 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/02 job id: 41614836 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/03 job id: 41614843 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/04 job id: 41614851 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/05 job id: 41614852 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/06 job id: 41614853 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/07 job id: 41614856 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/08 job id: 41614858 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/09 job id: 41614859 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/10 job id: 41614860 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/11 job id: 41614862 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/12 job id: 41614863 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/13 job id: 41614864 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/14 job id: 41614865 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/15 job id: 41614866 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/16 job id: 41614867 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/17 job id: 41614868 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/18 job id: 41614869 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/19 job id: 41614870 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/20 job id: 41614871 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/21 job id: 41614872 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/22 job id: 41614873 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/23 job id: 41614874 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/24 job id: 41614876 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/25 job id: 41614877 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/26 job id: 41614878 --- 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/848fd4fc57d18f38d66960eafe7959570f56e303_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 0.021 0.035 135.406 135.406 farming_run 1 2.0 134.898 134.900 135.378 135.382 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.452425E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.025 117.381 117.381 qs_energies 1 2.0 0.000 0.000 116.996 116.997 mp2_main 1 3.0 0.000 0.000 115.024 115.026 mp2_gpw_main 1 4.0 0.021 0.046 114.131 114.133 mp2_ri_gpw_compute_in 1 5.0 0.172 0.186 95.247 95.443 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 56.258 56.456 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.171 42.564 47.732 get_2c_integrals 1 6.0 0.000 0.000 38.315 38.817 integrate_v_rspace 273 8.0 0.433 0.446 25.201 30.137 pw_transfer 6555 10.6 0.373 0.392 27.570 28.064 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.198 26.476 grid_integrate_task_list 273 9.0 20.960 26.426 20.960 26.426 fft_wrap_pw1pw2_100 2178 12.4 1.218 1.437 23.761 24.062 compute_2c_integrals 1 7.0 0.002 0.002 20.406 20.407 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 19.689 20.123 mp2_eri_2c_integrate_gpw 1 9.0 2.373 2.444 19.686 20.120 rpa_ri_compute_en 1 5.0 0.000 0.000 18.748 18.852 cp_fm_cholesky_decompose 12 8.2 17.909 18.392 17.909 18.392 cholesky_decomp 1 7.0 0.000 0.000 16.751 17.248 fft3d_s 5443 13.4 16.270 16.658 16.292 16.679 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.568 10.856 15.568 calculate_wavefunction 272 8.0 5.438 5.584 12.591 13.254 rpa_num_int 1 6.0 0.000 0.000 10.785 10.794 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.636 10.789 calc_potential_gpw 544 9.5 0.005 0.006 9.617 10.057 calc_mat_Q 8 8.0 0.000 0.000 9.431 9.633 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.962 9.243 contract_S_to_Q 8 9.0 0.000 0.000 8.852 9.054 parallel_gemm_fm 14 9.1 0.000 0.000 8.440 8.528 parallel_gemm_fm_cosma 14 10.1 8.440 8.528 8.440 8.528 potential_pw2rs 545 10.0 0.106 0.108 7.824 8.524 collocate_single_gaussian 272 10.0 0.039 0.042 7.537 7.774 create_integ_mat 1 6.0 0.014 0.027 7.672 7.681 array2fm 1 7.0 0.000 0.000 6.652 7.154 pw_scatter_s 2720 13.7 4.386 4.516 4.386 4.516 pw_gather_s 2722 13.2 3.959 4.388 3.959 4.388 pw_poisson_solve 545 10.5 2.489 2.533 3.569 3.767 array2fm_buffer_send 1 8.0 2.944 3.149 2.944 3.149 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.131488, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 0.031 0.038 401.060 401.062 farming_run 1 2.0 398.998 399.010 401.018 401.022 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223156E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.045 212.527 212.528 qs_energies 1 2.0 0.000 0.000 211.027 211.046 scf_env_do_scf 1 3.0 0.000 0.000 106.838 106.838 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.951 105.958 rebuild_ks_matrix 4 6.0 0.000 0.000 105.950 105.957 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.065 105.950 105.957 hfx_ks_matrix 4 8.0 0.001 0.001 105.571 105.574 integrate_four_center 4 9.0 0.145 0.513 105.570 105.573 mp2_main 1 3.0 0.000 0.001 103.897 103.916 mp2_gpw_main 1 4.0 0.047 0.105 103.028 103.048 integrate_four_center_main 4 10.0 0.102 0.588 96.877 99.611 integrate_four_center_bin 264 11.0 96.775 99.579 96.775 99.579 init_scf_loop 1 4.0 0.000 0.000 92.636 92.636 mp2_ri_gpw_compute_in 1 5.0 0.081 0.142 75.861 76.965 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.010 54.945 56.054 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.166 42.525 47.808 integrate_v_rspace 95 8.0 0.396 0.563 28.435 33.555 pw_transfer 2240 10.6 0.145 0.178 30.135 30.611 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 29.111 29.605 ao_to_mo_and_store_B_mult_1 91 7.0 10.741 29.377 10.741 29.377 grid_integrate_task_list 95 9.0 23.721 29.009 23.721 29.009 mp2_ri_gpw_compute_en 1 5.0 0.086 0.168 26.987 28.758 fft_wrap_pw1pw2_100 730 12.4 1.302 1.507 26.827 27.365 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.814 1.877 25.152 25.167 get_2c_integrals 1 6.0 0.006 0.037 20.814 20.846 compute_2c_integrals 1 7.0 0.008 0.024 19.785 19.799 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 19.363 19.646 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.888 19.361 19.645 fft3d_s 1823 13.4 18.513 18.923 18.526 18.936 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.200 14.200 calculate_wavefunction 91 8.0 2.041 2.089 9.823 10.123 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.584 8.871 9.479 potential_pw2rs 186 10.0 0.033 0.034 8.693 9.468 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.672 9.083 calc_potential_gpw 182 9.5 0.002 0.002 8.672 9.047 local_gemm 172 8.0 8.314 8.932 8.314 8.932 mp2_ri_gpw_compute_en_comm 22 7.0 0.515 0.542 8.082 8.709 collocate_single_gaussian 91 10.0 0.017 0.037 7.948 8.263 mp_sendrecv_dm3 2068 8.0 6.076 6.739 6.076 6.739 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.429 6.341 6.429 pw_gather_s 912 13.2 4.951 5.640 4.951 5.640 mp_sync 38 10.4 3.287 5.350 3.287 5.350 pw_scatter_s 910 13.7 3.986 4.317 3.986 4.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=103.043901, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 450.285568E+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 592243. 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 0.013 0.028 49.741 49.742 qs_mol_dyn_low 1 2.0 0.003 0.003 49.552 49.560 qs_forces 11 3.9 0.002 0.002 49.256 49.257 qs_energies 11 4.9 0.001 0.001 47.770 47.778 scf_env_do_scf 11 5.9 0.000 0.001 41.854 41.854 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 39.820 39.820 dbcsr_multiply_generic 2286 12.5 0.092 0.097 31.434 31.801 qs_scf_new_mos 108 7.5 0.000 0.001 29.591 29.853 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.591 29.853 ot_scf_mini 108 9.5 0.002 0.002 27.996 28.169 multiply_cannon 2286 13.5 0.185 0.194 24.868 26.202 multiply_cannon_loop 2286 14.5 1.515 1.582 24.163 25.514 velocity_verlet 10 3.0 0.001 0.002 24.778 24.781 ot_mini 108 10.5 0.001 0.001 17.616 17.865 qs_ot_get_derivative 108 11.5 0.001 0.001 14.806 14.977 mp_waitall_1 267858 16.1 7.572 13.260 7.572 13.260 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 5.251 11.705 multiply_cannon_multrec 54864 15.5 4.279 6.755 7.553 11.075 rebuild_ks_matrix 119 8.3 0.000 0.000 8.028 8.156 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.028 8.156 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.075 7.191 multiply_cannon_sync_h2d 54864 15.5 5.970 6.518 5.970 6.518 qs_ot_get_p 119 10.4 0.001 0.001 5.925 6.186 mp_sum_l 7207 12.9 4.480 6.072 4.480 6.072 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.020 5.462 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.917 5.026 init_scf_run 11 5.9 0.000 0.001 4.697 4.697 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.697 4.697 sum_up_and_integrate 119 10.3 0.008 0.009 4.664 4.673 integrate_v_rspace 119 11.3 0.002 0.002 4.656 4.666 dbcsr_mm_accdrv_process 76910 16.1 1.170 1.811 3.195 4.447 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.189 4.263 calculate_rho_elec 119 8.7 0.011 0.016 4.189 4.262 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.385 3.417 rs_pw_transfer 974 11.9 0.012 0.013 3.055 3.168 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.858 3.043 apply_single 119 13.6 0.000 0.000 2.857 3.042 calculate_dm_sparse 119 9.5 0.000 0.000 2.841 2.983 ot_diis_step 108 11.5 0.006 0.006 2.611 2.612 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.581 2.581 jit_kernel_multiply 13 15.8 1.963 2.549 1.963 2.549 calculate_first_density_matrix 1 7.0 0.001 0.002 2.533 2.541 density_rs2pw 119 9.7 0.003 0.004 2.321 2.461 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.366 2.367 cp_fm_redistribute_end 50 14.0 2.150 2.343 2.154 2.345 cp_fm_diag_elpa_base 50 14.0 0.189 2.290 0.190 2.302 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.244 2.300 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.255 2.257 acc_transpose_blocks 54864 15.5 0.226 0.244 1.761 2.200 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.419 2.151 grid_integrate_task_list 119 12.3 2.010 2.110 2.010 2.110 wfi_extrapolate 11 7.9 0.001 0.001 2.096 2.096 potential_pw2rs 119 12.3 0.004 0.004 2.033 2.054 init_scf_loop 11 6.9 0.000 0.000 2.014 2.015 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.966 2.011 make_m2s 4572 13.5 0.054 0.056 1.776 1.824 pw_transfer 1439 11.6 0.051 0.056 1.703 1.779 make_images 4572 14.5 0.133 0.139 1.694 1.741 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.629 1.705 mp_sum_d 4129 12.0 1.115 1.684 1.115 1.684 fft3d_ps 1201 14.6 0.359 0.463 1.404 1.476 mp_alltoall_d11v 2130 13.8 1.328 1.441 1.328 1.441 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.410 1.434 mp_waitany 12084 13.8 1.195 1.376 1.195 1.376 grid_collocate_task_list 119 9.7 1.288 1.348 1.288 1.348 fft_wrap_pw1pw2_140 487 13.2 0.084 0.096 1.255 1.330 acc_transpose_blocks_kernels 54864 16.5 0.251 0.383 0.780 1.062 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.998 1.014 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.596 1.008 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=49.742000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.000000, yerr=1.348400 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 487.444480E+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 985638. 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 0.034 0.056 39.055 39.058 qs_mol_dyn_low 1 2.0 0.006 0.013 38.685 38.702 qs_forces 11 3.9 0.004 0.016 38.599 38.602 qs_energies 11 4.9 0.001 0.002 36.910 36.915 scf_env_do_scf 11 5.9 0.000 0.001 31.659 31.660 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.033 29.033 dbcsr_multiply_generic 2286 12.5 0.101 0.105 21.609 21.957 qs_scf_new_mos 108 7.5 0.001 0.001 19.970 20.210 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.969 20.209 ot_scf_mini 108 9.5 0.003 0.003 19.055 19.230 velocity_verlet 10 3.0 0.002 0.002 18.322 18.324 multiply_cannon 2286 13.5 0.208 0.221 16.409 17.823 multiply_cannon_loop 2286 14.5 0.911 0.985 15.263 16.761 ot_mini 108 10.5 0.001 0.001 11.933 12.175 mp_waitall_1 217478 16.2 6.183 11.349 6.183 11.349 qs_ot_get_derivative 108 11.5 0.001 0.002 9.425 9.606 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.209 9.462 multiply_cannon_multrec 27432 15.5 1.968 4.471 5.858 8.608 rebuild_ks_matrix 119 8.3 0.000 0.000 7.307 7.453 qs_ks_build_kohn_sham_matrix 119 9.3 0.019 0.038 7.307 7.452 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.491 6.632 dbcsr_mm_accdrv_process 47894 16.0 3.078 5.149 3.819 5.643 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.628 4.473 qs_ot_get_p 119 10.4 0.001 0.001 4.174 4.410 sum_up_and_integrate 119 10.3 0.030 0.067 4.304 4.313 integrate_v_rspace 119 11.3 0.002 0.004 4.274 4.299 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.054 4.125 apply_single 119 13.6 0.000 0.000 3.054 4.125 mp_sum_l 7207 12.9 2.127 4.104 2.127 4.104 init_scf_run 11 5.9 0.000 0.001 3.960 3.960 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.960 3.960 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.635 3.671 calculate_rho_elec 119 8.7 0.021 0.024 3.634 3.671 rs_pw_transfer 974 11.9 0.010 0.011 2.740 2.882 multiply_cannon_sync_h2d 27432 15.5 2.191 2.857 2.191 2.857 make_m2s 4572 13.5 0.052 0.054 2.635 2.857 make_images 4572 14.5 0.201 0.237 2.546 2.765 qs_ot_p2m_diag 50 11.0 0.009 0.015 2.699 2.718 init_scf_loop 11 6.9 0.000 0.000 2.604 2.605 ot_diis_step 108 11.5 0.012 0.013 2.455 2.456 calculate_first_density_matrix 1 7.0 0.001 0.004 2.450 2.451 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.203 2.295 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.258 2.259 density_rs2pw 119 9.7 0.004 0.004 1.992 2.136 calculate_dm_sparse 119 9.5 0.001 0.001 2.054 2.127 potential_pw2rs 119 12.3 0.006 0.006 2.052 2.077 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.010 2.013 grid_integrate_task_list 119 12.3 1.825 1.923 1.825 1.923 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.913 1.914 cp_fm_redistribute_end 50 14.0 1.581 1.891 1.585 1.892 pw_transfer 1439 11.6 0.063 0.066 1.801 1.857 cp_fm_diag_elpa_base 50 14.0 0.299 1.826 0.306 1.856 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.796 1.837 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.712 1.770 jit_kernel_multiply 8 16.2 0.688 1.625 0.688 1.625 make_images_data 4572 15.5 0.045 0.051 1.202 1.613 prepare_preconditioner 11 7.9 0.000 0.000 1.576 1.602 make_preconditioner 11 8.9 0.000 0.000 1.576 1.602 acc_transpose_blocks 27432 15.5 0.109 0.114 1.223 1.534 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.477 1.530 fft3d_ps 1201 14.6 0.498 0.550 1.418 1.473 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.044 1.459 wfi_extrapolate 11 7.9 0.001 0.003 1.455 1.455 fft_wrap_pw1pw2_140 487 13.2 0.078 0.085 1.331 1.388 mp_alltoall_d11v 2130 13.8 1.215 1.386 1.215 1.386 mp_allgather_i34 2286 14.5 0.589 1.350 0.589 1.350 grid_collocate_task_list 119 9.7 1.241 1.333 1.241 1.333 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.258 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.202 1.249 mp_sum_d 4129 12.0 0.656 1.138 0.656 1.138 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.015 1.018 acc_transpose_blocks_kernels 27432 16.5 0.181 0.269 0.690 0.914 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.900 make_images_sizes 4572 15.5 0.005 0.005 0.584 0.840 mp_alltoall_i44 4572 16.5 0.579 0.836 0.579 0.836 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.809 0.833 mp_alltoall_z22v 1201 16.6 0.707 0.800 0.707 0.800 dbcsr_dot_sd 1205 11.9 0.071 0.086 0.436 0.795 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.058000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.796480E+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 0.019 0.039 32.711 32.712 qs_mol_dyn_low 1 2.0 0.003 0.005 32.426 32.435 qs_forces 11 3.9 0.002 0.003 32.360 32.360 qs_energies 11 4.9 0.002 0.005 30.753 30.756 scf_env_do_scf 11 5.9 0.001 0.001 25.873 25.873 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 23.293 23.294 dbcsr_multiply_generic 2286 12.5 0.094 0.100 16.873 16.978 velocity_verlet 10 3.0 0.001 0.002 15.311 15.313 qs_scf_new_mos 108 7.5 0.001 0.001 15.233 15.250 qs_scf_loop_do_ot 108 8.5 0.001 0.002 15.233 15.250 ot_scf_mini 108 9.5 0.002 0.003 14.501 14.516 multiply_cannon 2286 13.5 0.195 0.203 13.429 14.277 multiply_cannon_loop 2286 14.5 0.643 0.672 12.629 13.517 ot_mini 108 10.5 0.001 0.001 9.098 9.127 qs_ot_get_derivative 108 11.5 0.001 0.001 7.577 7.596 multiply_cannon_multrec 18288 15.5 1.978 3.066 6.839 7.275 rebuild_ks_matrix 119 8.3 0.000 0.000 6.445 6.463 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.444 6.462 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.667 5.683 dbcsr_mm_accdrv_process 38222 16.0 4.121 5.500 4.776 5.557 mp_waitall_1 169478 16.3 3.244 4.430 3.244 4.430 sum_up_and_integrate 119 10.3 0.018 0.020 4.050 4.060 integrate_v_rspace 119 11.3 0.002 0.003 4.032 4.045 init_scf_run 11 5.9 0.000 0.001 3.627 3.628 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.627 3.628 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.722 3.359 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.283 3.290 calculate_rho_elec 119 8.7 0.030 0.031 3.282 3.289 qs_ot_get_p 119 10.4 0.001 0.002 3.213 3.231 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.549 2.720 rs_pw_transfer 974 11.9 0.009 0.010 2.467 2.571 init_scf_loop 11 6.9 0.001 0.002 2.563 2.565 calculate_first_density_matrix 1 7.0 0.001 0.002 2.408 2.409 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.061 2.383 apply_single 119 13.6 0.000 0.000 2.060 2.383 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.080 2.087 jit_kernel_multiply 11 16.1 0.604 2.064 0.604 2.064 make_m2s 4572 13.5 0.044 0.045 1.860 2.012 density_rs2pw 119 9.7 0.003 0.004 1.882 1.971 make_images 4572 14.5 0.194 0.222 1.774 1.924 grid_integrate_task_list 119 12.3 1.806 1.906 1.806 1.906 potential_pw2rs 119 12.3 0.007 0.008 1.851 1.862 calculate_dm_sparse 119 9.5 0.001 0.001 1.843 1.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.821 1.823 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.813 1.814 pw_transfer 1439 11.6 0.063 0.065 1.749 1.764 prepare_preconditioner 11 7.9 0.000 0.000 1.750 1.753 make_preconditioner 11 8.9 0.000 0.001 1.750 1.753 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.608 1.690 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.659 1.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.662 1.672 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.620 1.629 multiply_cannon_sync_h2d 18288 15.5 1.398 1.589 1.398 1.589 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.573 1.574 mp_sum_l 7207 12.9 1.207 1.566 1.207 1.566 cp_fm_redistribute_end 50 14.0 1.169 1.552 1.170 1.552 cp_fm_diag_elpa_base 50 14.0 0.365 1.480 0.380 1.523 ot_diis_step 108 11.5 0.012 0.013 1.500 1.500 fft3d_ps 1201 14.6 0.506 0.523 1.345 1.361 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.305 1.320 acc_transpose_blocks 18288 15.5 0.076 0.077 1.283 1.307 grid_collocate_task_list 119 9.7 1.206 1.295 1.206 1.295 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.210 wfi_extrapolate 11 7.9 0.001 0.003 1.169 1.169 qs_energies_init_hamiltonians 11 5.9 0.004 0.017 1.008 1.008 make_images_data 4572 15.5 0.044 0.048 0.804 0.969 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.908 0.928 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.687 0.887 acc_transpose_blocks_kernels 18288 16.5 0.209 0.220 0.842 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.850 0.853 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.523 0.807 mp_alltoall_d11v 2130 13.8 0.682 0.784 0.682 0.784 cp_fm_cholesky_invert 11 10.9 0.757 0.761 0.757 0.761 mp_alltoall_z22v 1201 16.6 0.681 0.752 0.681 0.752 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.657 0.734 rs_pw_transfer_RS2PW_140 130 11.5 0.176 0.181 0.595 0.701 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.712000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.909091, yerr=2.574510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 551.038976E+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 0.110 0.270 36.558 36.559 qs_mol_dyn_low 1 2.0 0.003 0.005 36.067 36.080 qs_forces 11 3.9 0.002 0.007 35.725 35.727 qs_energies 11 4.9 0.002 0.005 34.017 34.023 scf_env_do_scf 11 5.9 0.000 0.001 28.525 28.526 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.985 24.986 dbcsr_multiply_generic 2286 12.5 0.100 0.105 18.740 18.882 velocity_verlet 10 3.0 0.002 0.002 18.038 18.052 qs_scf_new_mos 108 7.5 0.001 0.001 16.756 16.814 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.755 16.813 ot_scf_mini 108 9.5 0.002 0.003 15.815 15.864 multiply_cannon 2286 13.5 0.236 0.272 14.919 15.276 multiply_cannon_loop 2286 14.5 0.945 0.979 13.932 14.259 ot_mini 108 10.5 0.001 0.001 9.640 9.704 multiply_cannon_multrec 27432 15.5 2.329 2.996 8.587 8.922 qs_ot_get_derivative 108 11.5 0.001 0.001 7.650 7.702 dbcsr_mm_accdrv_process 47916 15.9 5.450 7.116 6.163 7.382 rebuild_ks_matrix 119 8.3 0.000 0.000 6.545 6.589 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 6.544 6.589 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.824 5.866 init_scf_run 11 5.9 0.000 0.001 4.056 4.056 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.056 4.056 sum_up_and_integrate 119 10.3 0.020 0.022 3.828 3.837 integrate_v_rspace 119 11.3 0.002 0.003 3.808 3.817 qs_ot_get_p 119 10.4 0.001 0.001 3.526 3.611 init_scf_loop 11 6.9 0.001 0.004 3.520 3.526 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.327 3.365 calculate_rho_elec 119 8.7 0.040 0.046 3.326 3.364 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.815 3.213 mp_waitall_1 145218 16.4 2.412 3.081 2.412 3.081 calculate_first_density_matrix 1 7.0 0.001 0.002 2.670 2.671 prepare_preconditioner 11 7.9 0.000 0.000 2.658 2.664 make_preconditioner 11 8.9 0.000 0.001 2.658 2.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.250 2.578 make_m2s 4572 13.5 0.054 0.055 2.348 2.477 make_images 4572 14.5 0.275 0.337 2.240 2.368 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.158 2.307 apply_single 119 13.6 0.000 0.000 2.158 2.307 rs_pw_transfer 974 11.9 0.009 0.009 2.161 2.269 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.237 2.245 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.153 2.154 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.070 2.100 calculate_dm_sparse 119 9.5 0.001 0.001 2.030 2.081 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.952 1.966 ot_diis_step 108 11.5 0.012 0.014 1.941 1.942 density_rs2pw 119 9.7 0.003 0.004 1.821 1.942 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.912 1.913 grid_integrate_task_list 119 12.3 1.833 1.903 1.833 1.903 pw_transfer 1439 11.6 0.063 0.066 1.796 1.829 acc_transpose_blocks 27432 15.5 0.112 0.115 1.680 1.795 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.706 1.741 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.674 1.675 cp_fm_redistribute_end 50 14.0 1.101 1.642 1.102 1.642 potential_pw2rs 119 12.3 0.009 0.009 1.631 1.635 cp_fm_diag_elpa_base 50 14.0 0.518 1.577 0.538 1.616 jit_kernel_multiply 8 16.1 0.652 1.484 0.652 1.484 multiply_cannon_metrocomm3 27432 15.5 0.038 0.038 0.871 1.481 fft3d_ps 1201 14.6 0.531 0.584 1.384 1.408 fft_wrap_pw1pw2_140 487 13.2 0.087 0.095 1.355 1.391 mp_sum_l 7207 12.9 1.111 1.391 1.111 1.391 wfi_extrapolate 11 7.9 0.001 0.003 1.338 1.339 grid_collocate_task_list 119 9.7 1.222 1.309 1.222 1.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.217 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.160 1.161 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 1.053 1.159 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.113 1.132 cp_fm_upper_to_full 72 13.5 0.808 1.123 0.808 1.123 dbcsr_complete_redistribute 329 12.2 0.122 0.149 0.814 1.087 multiply_cannon_sync_h2d 27432 15.5 0.987 1.050 0.987 1.050 make_images_data 4572 15.5 0.044 0.048 0.899 1.029 hybrid_alltoall_any 4725 16.4 0.062 0.153 0.774 0.967 jit_kernel_transpose 5 15.6 0.785 0.895 0.785 0.895 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.869 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.600 0.867 mp_alltoall_d11v 2130 13.8 0.741 0.859 0.741 0.859 cp_fm_cholesky_invert 11 10.9 0.819 0.822 0.819 0.822 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.813 0.818 multiply_cannon_metrocomm1 27432 15.5 0.032 0.034 0.350 0.782 mp_alltoall_z22v 1201 16.6 0.736 0.764 0.736 0.764 mp_alltoall_i22 627 13.8 0.443 0.736 0.443 0.736 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.559000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.363636, yerr=1.871933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 605.818880E+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 0.015 0.031 29.679 29.679 qs_mol_dyn_low 1 2.0 0.003 0.003 29.465 29.472 qs_forces 11 3.9 0.002 0.002 29.326 29.330 qs_energies 11 4.9 0.001 0.002 27.280 27.286 scf_env_do_scf 11 5.9 0.000 0.001 21.767 21.767 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.196 19.197 velocity_verlet 10 3.0 0.053 0.064 15.101 15.105 dbcsr_multiply_generic 2286 12.5 0.092 0.095 13.151 13.259 qs_scf_new_mos 108 7.5 0.001 0.001 11.608 11.629 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.607 11.628 multiply_cannon 2286 13.5 0.229 0.239 10.437 10.964 ot_scf_mini 108 9.5 0.002 0.002 10.931 10.955 multiply_cannon_loop 2286 14.5 0.334 0.347 9.478 9.697 multiply_cannon_multrec 9144 15.5 1.679 2.061 6.329 6.700 ot_mini 108 10.5 0.001 0.001 6.333 6.365 rebuild_ks_matrix 119 8.3 0.000 0.000 5.986 6.025 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.985 6.025 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.209 5.231 qs_ot_get_derivative 108 11.5 0.001 0.001 5.023 5.046 dbcsr_mm_accdrv_process 12550 15.8 3.553 4.199 4.548 4.680 init_scf_run 11 5.9 0.000 0.001 3.793 3.793 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.793 3.793 sum_up_and_integrate 119 10.3 0.022 0.023 3.623 3.630 integrate_v_rspace 119 11.3 0.003 0.003 3.600 3.608 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.215 3.224 calculate_rho_elec 119 8.7 0.059 0.061 3.215 3.223 qs_ot_get_p 119 10.4 0.001 0.001 2.839 2.879 calculate_first_density_matrix 1 7.0 0.000 0.000 2.551 2.552 init_scf_loop 11 6.9 0.000 0.000 2.550 2.551 mp_waitall_1 121218 16.5 1.863 2.391 1.863 2.391 make_m2s 4572 13.5 0.034 0.035 1.857 1.982 grid_integrate_task_list 119 12.3 1.845 1.927 1.845 1.927 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.890 1.920 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.893 1.895 make_images 4572 14.5 0.268 0.303 1.767 1.890 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.874 1.876 rs_pw_transfer 974 11.9 0.008 0.008 1.758 1.826 calculate_dm_sparse 119 9.5 0.000 0.000 1.803 1.821 prepare_preconditioner 11 7.9 0.000 0.000 1.791 1.795 make_preconditioner 11 8.9 0.000 0.000 1.791 1.795 pw_transfer 1439 11.6 0.063 0.066 1.731 1.741 density_rs2pw 119 9.7 0.003 0.004 1.651 1.708 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.699 1.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.674 1.704 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.675 1.675 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.641 1.652 qs_energies_init_hamiltonians 11 5.9 0.027 0.030 1.497 1.501 jit_kernel_multiply 6 15.8 0.958 1.492 0.958 1.492 potential_pw2rs 119 12.3 0.010 0.011 1.399 1.401 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.400 1.400 cp_fm_redistribute_end 50 14.0 0.697 1.372 0.698 1.372 cp_fm_diag_elpa_base 50 14.0 0.629 1.299 0.673 1.356 grid_collocate_task_list 119 9.7 1.271 1.342 1.271 1.342 fft_wrap_pw1pw2_140 487 13.2 0.085 0.088 1.310 1.325 fft3d_ps 1201 14.6 0.539 0.552 1.305 1.317 ot_diis_step 108 11.5 0.012 0.013 1.290 1.291 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.172 1.250 apply_single 119 13.6 0.000 0.000 1.171 1.250 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.226 1.232 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.188 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.193 1.193 acc_transpose_blocks 9144 15.5 0.039 0.041 1.075 1.082 make_images_data 4572 15.5 0.038 0.042 0.855 1.029 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.794 1.018 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.925 0.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.937 mp_alltoall_d11v 2130 13.8 0.812 0.937 0.812 0.937 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.923 cp_fm_cholesky_invert 11 10.9 0.895 0.897 0.895 0.897 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.389 0.852 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.833 0.839 qs_create_task_list 11 7.9 0.001 0.001 0.794 0.818 generate_qs_task_list 11 8.9 0.190 0.212 0.793 0.817 multiply_cannon_sync_h2d 9144 15.5 0.703 0.784 0.703 0.784 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.731 0.740 jit_kernel_transpose 5 15.6 0.716 0.722 0.716 0.722 mp_allgather_i34 2286 14.5 0.291 0.693 0.291 0.693 mp_alltoall_z22v 1201 16.6 0.641 0.679 0.641 0.679 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.679000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.363636, yerr=6.718668 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 739.192832E+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 0.019 0.037 40.686 40.686 qs_mol_dyn_low 1 2.0 0.003 0.003 40.474 40.481 qs_forces 11 3.9 0.001 0.002 40.401 40.402 qs_energies 11 4.9 0.001 0.001 38.444 38.447 scf_env_do_scf 11 5.9 0.001 0.001 32.875 32.875 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.054 25.055 velocity_verlet 10 3.0 0.002 0.002 22.803 22.810 dbcsr_multiply_generic 2286 12.5 0.100 0.102 17.826 18.058 qs_scf_new_mos 108 7.5 0.001 0.001 16.195 16.297 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.195 16.296 ot_scf_mini 108 9.5 0.002 0.002 15.120 15.222 multiply_cannon 2286 13.5 0.300 0.306 13.923 14.930 multiply_cannon_loop 2286 14.5 0.344 0.350 12.651 13.653 ot_mini 108 10.5 0.001 0.001 9.138 9.256 multiply_cannon_multrec 9144 15.5 3.397 4.969 8.658 8.929 init_scf_loop 11 6.9 0.000 0.000 7.793 7.795 qs_ot_get_derivative 108 11.5 0.001 0.001 7.109 7.211 rebuild_ks_matrix 119 8.3 0.000 0.000 6.807 6.949 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.807 6.948 prepare_preconditioner 11 7.9 0.000 0.000 6.813 6.827 make_preconditioner 11 8.9 0.000 0.000 6.813 6.827 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.388 6.706 dbcsr_mm_accdrv_process 12550 15.8 4.194 5.917 5.135 6.641 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.169 6.298 cp_fm_upper_to_full 72 14.2 3.143 4.515 3.143 4.515 mp_waitall_1 97218 16.6 2.667 3.783 2.667 3.783 sum_up_and_integrate 119 10.3 0.038 0.039 3.636 3.644 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.636 3.644 calculate_rho_elec 119 8.7 0.118 0.121 3.636 3.643 integrate_v_rspace 119 11.3 0.003 0.003 3.598 3.606 init_scf_run 11 5.9 0.000 0.001 3.555 3.555 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.555 3.555 qs_ot_get_p 119 10.4 0.001 0.001 3.297 3.444 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.782 3.199 dbcsr_complete_redistribute 329 12.2 0.288 0.292 1.984 2.802 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.457 2.499 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.203 2.499 apply_single 119 13.6 0.000 0.000 2.203 2.499 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.662 2.465 make_m2s 4572 13.5 0.038 0.038 2.273 2.463 make_images 4572 14.5 0.354 0.389 2.152 2.342 mp_alltoall_i22 627 13.8 1.408 2.255 1.408 2.255 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.420 2.218 calculate_first_density_matrix 1 7.0 0.001 0.001 2.174 2.174 calculate_dm_sparse 119 9.5 0.000 0.000 2.147 2.160 pw_transfer 1439 11.6 0.067 0.067 2.016 2.019 grid_integrate_task_list 119 12.3 2.007 2.015 2.007 2.015 ot_diis_step 108 11.5 0.014 0.014 2.000 2.000 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.925 1.926 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.922 1.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.864 1.865 mp_sum_l 7207 12.9 1.092 1.862 1.092 1.862 density_rs2pw 119 9.7 0.003 0.003 1.790 1.809 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.708 1.760 jit_kernel_multiply 10 15.4 0.914 1.758 0.914 1.758 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.751 1.751 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.628 1.666 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.643 1.644 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.595 1.598 fft3d_ps 1201 14.6 0.568 0.577 1.556 1.560 cp_fm_cholesky_invert 11 10.9 1.483 1.487 1.483 1.487 grid_collocate_task_list 119 9.7 1.448 1.475 1.448 1.475 rs_pw_transfer 974 11.9 0.009 0.009 1.407 1.444 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.368 1.370 cp_fm_diag_elpa_base 50 14.0 1.224 1.277 1.366 1.368 wfi_extrapolate 11 7.9 0.001 0.001 1.318 1.319 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.043 1.296 make_images_data 4572 15.5 0.042 0.045 1.060 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.221 1.227 potential_pw2rs 119 12.3 0.014 0.014 1.218 1.222 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.162 1.182 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.091 1.107 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.991 1.046 multiply_cannon_sync_h2d 9144 15.5 1.039 1.041 1.039 1.041 acc_transpose_blocks 9144 15.5 0.039 0.040 0.991 1.001 mp_alltoall_d11v 2130 13.8 0.984 0.996 0.984 0.996 qs_create_task_list 11 7.9 0.006 0.007 0.951 0.962 generate_qs_task_list 11 8.9 0.373 0.392 0.945 0.956 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.861 0.873 mp_alltoall_z22v 1201 16.6 0.851 0.867 0.851 0.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.686000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=695.727273, yerr=13.672074 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 500.424704E+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 1483114. 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 0.044 0.081 83.163 83.163 qs_mol_dyn_low 1 2.0 0.003 0.003 82.823 82.833 qs_forces 11 3.9 0.002 0.002 82.743 82.744 qs_energies 11 4.9 0.001 0.002 79.907 79.922 scf_env_do_scf 11 5.9 0.000 0.001 70.927 70.929 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.412 65.412 dbcsr_multiply_generic 2055 12.4 0.105 0.109 52.667 53.002 qs_scf_new_mos 99 7.5 0.000 0.001 48.431 48.527 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.431 48.527 ot_scf_mini 99 9.5 0.002 0.002 46.037 46.110 multiply_cannon 2055 13.4 0.178 0.183 42.780 43.945 multiply_cannon_loop 2055 14.4 1.575 1.612 41.815 43.063 velocity_verlet 10 3.0 0.001 0.002 41.705 41.706 ot_mini 99 10.5 0.001 0.001 28.211 28.323 qs_ot_get_derivative 99 11.5 0.001 0.001 21.308 21.417 multiply_cannon_multrec 49320 15.4 12.104 12.896 17.119 17.761 rebuild_ks_matrix 110 8.3 0.000 0.001 14.540 14.662 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.539 14.662 mp_waitall_1 241148 16.1 11.871 13.052 11.871 13.052 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.769 12.868 multiply_cannon_sync_h2d 49320 15.4 9.926 10.490 9.926 10.490 qs_ot_get_p 110 10.4 0.001 0.001 9.308 9.407 multiply_cannon_metrocomm3 49320 15.4 0.076 0.081 7.105 8.488 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.177 7.674 apply_single 110 13.6 0.000 0.001 7.177 7.674 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.075 7.608 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.175 7.247 sum_up_and_integrate 110 10.3 0.023 0.028 7.111 7.124 integrate_v_rspace 110 11.3 0.002 0.003 7.088 7.107 init_scf_run 11 5.9 0.000 0.001 6.950 6.951 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.950 6.950 ot_diis_step 99 11.5 0.006 0.006 6.593 6.593 mp_sum_l 6514 12.8 5.543 6.515 5.543 6.515 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.191 6.333 calculate_rho_elec 110 8.6 0.020 0.024 6.190 6.333 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.019 6.061 init_scf_loop 11 6.9 0.000 0.000 5.487 5.488 dbcsr_mm_accdrv_process 87628 16.1 2.075 2.190 4.893 5.202 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.006 5.007 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.527 4.528 cp_fm_redistribute_end 48 14.0 3.941 4.499 3.945 4.500 cp_fm_diag_elpa_base 48 14.0 0.549 4.402 0.553 4.430 rs_pw_transfer 902 11.9 0.012 0.014 4.002 4.171 make_m2s 4110 13.4 0.060 0.065 3.910 4.040 wfi_extrapolate 11 7.9 0.001 0.001 3.981 3.981 make_images 4110 14.4 0.178 0.191 3.815 3.948 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.767 3.879 calculate_dm_sparse 110 9.5 0.000 0.001 3.755 3.856 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.749 3.753 density_rs2pw 110 9.6 0.004 0.004 3.278 3.451 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.382 3.436 grid_integrate_task_list 110 12.3 3.247 3.394 3.247 3.394 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.291 3.370 prepare_preconditioner 11 7.9 0.000 0.000 3.338 3.366 make_preconditioner 11 8.9 0.000 0.000 3.338 3.366 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.128 3.175 calculate_first_density_matrix 1 7.0 0.000 0.001 2.882 2.889 pw_transfer 1331 11.6 0.054 0.065 2.765 2.825 potential_pw2rs 110 12.3 0.006 0.006 2.737 2.756 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.673 2.737 jit_kernel_multiply 13 15.9 2.535 2.550 2.535 2.550 mp_alltoall_d11v 2046 13.8 2.056 2.429 2.056 2.429 acc_transpose_blocks 49320 15.4 0.215 0.226 2.250 2.350 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 2.257 2.321 fft3d_ps 1111 14.6 0.753 0.839 2.210 2.267 grid_collocate_task_list 110 9.6 2.087 2.207 2.087 2.207 mp_sum_d 3883 11.9 1.584 2.055 1.584 2.055 mp_waitany 14300 13.8 1.709 2.018 1.709 2.018 make_images_data 4110 15.4 0.043 0.046 1.782 1.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.909 1.936 cp_fm_cholesky_invert 11 10.9 1.840 1.844 1.840 1.844 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.552 1.833 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.635 1.665 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.163000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=1.827250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 585.596928E+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 3531006. 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 0.026 0.073 70.435 70.436 qs_mol_dyn_low 1 2.0 0.003 0.005 70.048 70.061 qs_forces 11 3.9 0.002 0.004 69.957 69.959 qs_energies 11 4.9 0.002 0.009 66.605 66.609 scf_env_do_scf 11 5.9 0.001 0.001 57.743 57.747 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.794 49.796 dbcsr_multiply_generic 2055 12.4 0.112 0.116 38.431 38.668 velocity_verlet 10 3.0 0.002 0.003 36.978 36.980 qs_scf_new_mos 99 7.5 0.001 0.001 33.316 33.458 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.315 33.457 multiply_cannon 2055 13.4 0.221 0.244 31.247 32.908 ot_scf_mini 99 9.5 0.003 0.004 31.633 31.762 multiply_cannon_loop 2055 14.4 0.938 0.956 29.854 31.174 ot_mini 99 10.5 0.001 0.001 18.632 18.767 multiply_cannon_multrec 24660 15.4 7.632 9.539 13.836 15.905 rebuild_ks_matrix 110 8.3 0.000 0.000 14.083 14.295 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.082 14.295 qs_ot_get_derivative 99 11.5 0.001 0.002 12.774 12.908 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.431 12.613 mp_waitall_1 186928 16.3 8.237 10.827 8.237 10.827 multiply_cannon_sync_h2d 24660 15.4 7.060 8.335 7.060 8.335 init_scf_loop 11 6.9 0.000 0.001 7.913 7.916 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.149 7.845 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.570 7.331 apply_single 110 13.6 0.000 0.001 6.570 7.331 sum_up_and_integrate 110 10.3 0.029 0.032 6.858 6.871 integrate_v_rspace 110 11.3 0.002 0.003 6.830 6.841 init_scf_run 11 5.9 0.000 0.001 6.403 6.404 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.403 6.404 qs_ot_get_p 110 10.4 0.001 0.001 6.255 6.397 dbcsr_mm_accdrv_process 52282 16.1 4.704 5.628 6.044 6.328 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.108 6.115 calculate_rho_elec 110 8.6 0.039 0.047 6.107 6.114 prepare_preconditioner 11 7.9 0.000 0.000 5.791 5.810 make_preconditioner 11 8.9 0.000 0.001 5.791 5.810 ot_diis_step 99 11.5 0.011 0.013 5.805 5.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.365 5.531 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.749 5.499 make_m2s 4110 13.4 0.057 0.060 4.588 5.069 make_images 4110 14.4 0.404 0.459 4.478 4.956 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.225 4.246 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.772 3.772 wfi_extrapolate 11 7.9 0.001 0.001 3.683 3.683 pw_transfer 1331 11.6 0.066 0.071 3.451 3.594 rs_pw_transfer 902 11.9 0.012 0.013 3.435 3.585 density_rs2pw 110 9.6 0.004 0.004 3.345 3.497 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.340 3.482 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.422 3.424 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.342 3.408 grid_integrate_task_list 110 12.3 3.142 3.308 3.142 3.308 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.164 3.166 cp_fm_redistribute_end 48 14.0 2.360 3.133 2.362 3.134 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.021 3.088 cp_fm_diag_elpa_base 48 14.0 0.736 2.990 0.767 3.079 make_images_data 4110 15.4 0.047 0.050 2.518 3.036 calculate_dm_sparse 110 9.5 0.001 0.001 2.988 3.017 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.178 3.007 fft_wrap_pw1pw2_140 451 13.1 0.202 0.217 2.754 2.894 fft3d_ps 1111 14.6 1.072 1.270 2.696 2.825 cp_fm_cholesky_invert 11 10.9 2.721 2.728 2.721 2.728 potential_pw2rs 110 12.3 0.009 0.009 2.659 2.671 calculate_first_density_matrix 1 7.0 0.002 0.004 2.634 2.638 mp_sum_l 6514 12.8 2.011 2.600 2.011 2.600 grid_collocate_task_list 110 9.6 2.120 2.298 2.120 2.298 mp_alltoall_d11v 2046 13.8 1.972 2.286 1.972 2.286 jit_kernel_multiply 11 16.3 0.987 2.138 0.987 2.138 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.991 2.012 qs_energies_init_hamiltonians 11 5.9 0.010 0.024 1.912 1.913 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.820 1.847 mp_allgather_i34 2055 14.4 0.723 1.822 0.723 1.822 acc_transpose_blocks 24660 15.4 0.108 0.113 1.593 1.631 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.773 1.615 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.545 1.556 dbcsr_complete_redistribute 325 12.2 0.239 0.294 1.237 1.515 mp_irecv_dv 57340 16.2 0.650 1.501 0.650 1.501 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.474 mp_alltoall_z22v 1111 16.6 1.360 1.431 1.360 1.431 cp_fm_cholesky_decompose 22 10.9 1.403 1.411 1.403 1.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.436000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.545455, yerr=5.821022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 659.759104E+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 0.036 0.138 60.819 60.819 qs_mol_dyn_low 1 2.0 0.003 0.007 60.350 60.365 qs_forces 11 3.9 0.027 0.032 60.271 60.276 qs_energies 11 4.9 0.002 0.005 57.049 57.052 scf_env_do_scf 11 5.9 0.002 0.010 48.827 48.827 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.015 40.022 velocity_verlet 10 3.0 0.001 0.002 32.878 32.880 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.746 28.964 qs_scf_new_mos 99 7.5 0.001 0.001 25.083 25.178 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.082 25.178 ot_scf_mini 99 9.5 0.002 0.003 23.854 23.957 multiply_cannon 2055 13.4 0.212 0.223 22.206 23.342 multiply_cannon_loop 2055 14.4 0.622 0.637 20.957 21.956 ot_mini 99 10.5 0.001 0.001 13.761 13.864 rebuild_ks_matrix 110 8.3 0.000 0.000 12.382 12.504 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.025 12.382 12.503 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.929 11.038 multiply_cannon_multrec 16440 15.4 3.973 5.213 9.728 10.911 mp_waitall_1 146766 16.3 7.274 10.406 7.274 10.406 qs_ot_get_derivative 99 11.5 0.001 0.001 9.320 9.423 init_scf_loop 11 6.9 0.000 0.001 8.775 8.778 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.191 7.178 prepare_preconditioner 11 7.9 0.000 0.000 6.947 6.961 make_preconditioner 11 8.9 0.000 0.001 6.947 6.961 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.244 6.617 sum_up_and_integrate 110 10.3 0.034 0.038 6.583 6.600 integrate_v_rspace 110 11.3 0.002 0.003 6.549 6.563 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.840 5.848 calculate_rho_elec 110 8.6 0.058 0.058 5.840 5.848 dbcsr_mm_accdrv_process 34862 16.1 4.908 5.383 5.608 5.746 init_scf_run 11 5.9 0.000 0.001 5.650 5.650 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.650 5.650 qs_ot_get_p 110 10.4 0.001 0.001 5.287 5.421 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.966 5.390 apply_single 110 13.6 0.000 0.000 4.966 5.390 make_m2s 4110 13.4 0.050 0.051 4.393 4.777 make_images 4110 14.4 0.393 0.511 4.277 4.660 ot_diis_step 99 11.5 0.011 0.012 4.412 4.412 multiply_cannon_sync_h2d 16440 15.4 3.708 4.289 3.708 4.289 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.117 3.754 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.618 3.621 grid_integrate_task_list 110 12.3 3.210 3.404 3.210 3.404 pw_transfer 1331 11.6 0.066 0.073 3.354 3.363 rs_pw_transfer 902 11.9 0.010 0.011 3.086 3.326 density_rs2pw 110 9.6 0.004 0.004 3.094 3.309 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.280 3.281 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.243 3.253 wfi_extrapolate 11 7.9 0.001 0.003 2.994 2.994 make_images_data 4110 15.4 0.043 0.048 2.497 2.994 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.928 2.929 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.222 2.864 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.750 2.753 cp_fm_redistribute_end 48 14.0 1.713 2.729 1.715 2.730 fft_wrap_pw1pw2_140 451 13.1 0.214 0.219 2.714 2.728 cp_fm_cholesky_invert 11 10.9 2.711 2.718 2.711 2.718 cp_fm_diag_elpa_base 48 14.0 0.951 2.586 1.008 2.694 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.547 2.584 calculate_first_density_matrix 1 7.0 0.001 0.001 2.569 2.571 fft3d_ps 1111 14.6 1.065 1.081 2.554 2.568 calculate_dm_sparse 110 9.5 0.001 0.001 2.533 2.563 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.491 2.547 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.864 2.395 potential_pw2rs 110 12.3 0.011 0.011 2.379 2.392 grid_collocate_task_list 110 9.6 2.178 2.335 2.178 2.335 mp_irecv_dv 48980 15.7 0.795 2.272 0.795 2.272 mp_sum_l 6514 12.8 1.624 2.236 1.624 2.236 mp_alltoall_d11v 2046 13.8 1.883 2.168 1.883 2.168 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 2.118 2.118 dbcsr_complete_redistribute 325 12.2 0.343 0.383 1.543 2.015 cp_fm_upper_to_full 70 13.6 1.399 1.844 1.399 1.844 cp_fm_cholesky_decompose 22 10.9 1.663 1.687 1.663 1.687 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.655 1.669 mp_allgather_i34 2055 14.4 0.570 1.574 0.570 1.574 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.060 1.525 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.497 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.484 1.496 mp_waitany 17072 13.8 1.222 1.431 1.222 1.431 acc_transpose_blocks 16440 15.4 0.073 0.077 1.281 1.300 mp_alltoall_z22v 1111 16.6 1.275 1.299 1.275 1.299 qs_env_update_s_mstruct 11 6.9 0.001 0.007 1.207 1.287 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.272 1.282 rs_pw_transfer_RS2PW_140 121 11.5 0.261 0.268 1.008 1.254 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.819000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.454545, yerr=9.119047 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 739.803136E+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 0.038 0.147 66.496 66.497 qs_mol_dyn_low 1 2.0 0.003 0.004 66.107 66.117 qs_forces 11 3.9 0.002 0.004 66.035 66.036 qs_energies 11 4.9 0.002 0.005 62.629 62.633 scf_env_do_scf 11 5.9 0.001 0.001 54.048 54.051 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.940 41.943 velocity_verlet 10 3.0 0.001 0.002 37.736 37.738 dbcsr_multiply_generic 2055 12.4 0.115 0.119 30.377 30.584 qs_scf_new_mos 99 7.5 0.001 0.001 27.181 27.288 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.180 27.287 ot_scf_mini 99 9.5 0.003 0.003 25.518 25.653 multiply_cannon 2055 13.4 0.246 0.276 22.817 24.000 multiply_cannon_loop 2055 14.4 0.888 0.914 21.277 21.867 ot_mini 99 10.5 0.001 0.001 14.735 14.870 multiply_cannon_multrec 24660 15.4 4.252 6.886 12.739 13.878 rebuild_ks_matrix 110 8.3 0.000 0.000 12.193 12.295 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.192 12.294 init_scf_loop 11 6.9 0.000 0.002 12.068 12.072 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.814 10.904 qs_ot_get_derivative 99 11.5 0.001 0.001 10.422 10.550 prepare_preconditioner 11 7.9 0.000 0.000 10.280 10.302 make_preconditioner 11 8.9 0.000 0.001 10.280 10.302 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.493 9.961 dbcsr_mm_accdrv_process 52304 16.0 7.103 8.496 8.339 9.361 mp_waitall_1 126806 16.4 4.791 6.698 4.791 6.698 sum_up_and_integrate 110 10.3 0.038 0.039 6.486 6.500 integrate_v_rspace 110 11.3 0.003 0.003 6.449 6.461 make_m2s 4110 13.4 0.059 0.063 5.723 6.111 make_images 4110 14.4 0.583 0.714 5.582 5.966 qs_ot_get_p 110 10.4 0.001 0.001 5.649 5.799 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.785 5.795 calculate_rho_elec 110 8.6 0.077 0.081 5.784 5.795 init_scf_run 11 5.9 0.000 0.001 5.764 5.765 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.764 5.765 cp_fm_upper_to_full 70 13.8 3.325 4.728 3.325 4.728 ot_diis_step 99 11.5 0.011 0.011 4.210 4.219 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.096 4.195 apply_single 110 13.6 0.000 0.000 4.096 4.195 dbcsr_complete_redistribute 325 12.2 0.420 0.480 2.769 3.917 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.706 3.721 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.364 3.423 pw_transfer 1331 11.6 0.066 0.074 3.385 3.414 multiply_cannon_sync_h2d 24660 15.4 3.183 3.407 3.183 3.407 grid_integrate_task_list 110 12.3 3.283 3.384 3.283 3.384 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.188 3.328 make_images_data 4110 15.4 0.046 0.049 2.900 3.317 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.275 3.306 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.422 3.280 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.165 3.234 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.408 3.185 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.181 3.184 wfi_extrapolate 11 7.9 0.001 0.001 3.040 3.040 density_rs2pw 110 9.6 0.004 0.004 2.885 3.010 calculate_dm_sparse 110 9.5 0.001 0.001 2.962 2.989 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.930 2.932 cp_fm_cholesky_invert 11 10.9 2.896 2.904 2.896 2.904 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.778 2.900 mp_alltoall_i22 605 13.7 1.671 2.851 1.671 2.851 fft_wrap_pw1pw2_140 451 13.1 0.207 0.217 2.743 2.773 rs_pw_transfer 902 11.9 0.011 0.011 2.618 2.747 calculate_first_density_matrix 1 7.0 0.000 0.001 2.630 2.633 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.602 2.604 fft3d_ps 1111 14.6 1.067 1.103 2.572 2.594 cp_fm_redistribute_end 48 14.0 1.301 2.576 1.303 2.577 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.525 2.558 cp_fm_diag_elpa_base 48 14.0 1.192 2.432 1.271 2.546 grid_collocate_task_list 110 9.6 2.226 2.399 2.226 2.399 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.293 2.294 potential_pw2rs 110 12.3 0.013 0.013 2.262 2.274 mp_alltoall_d11v 2046 13.8 1.920 2.169 1.920 2.169 cp_fm_cholesky_decompose 22 10.9 1.796 1.850 1.796 1.850 jit_kernel_multiply 10 16.1 0.906 1.831 0.906 1.831 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.767 1.802 mp_sum_l 6514 12.8 1.165 1.791 1.165 1.791 mp_allgather_i34 2055 14.4 0.641 1.751 0.641 1.751 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.603 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.637 multiply_cannon_metrocomm4 20550 15.4 0.056 0.060 0.837 1.591 acc_transpose_blocks 24660 15.4 0.106 0.108 1.561 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.562 mp_irecv_dv 62702 16.1 0.739 1.514 0.739 1.514 mp_alltoall_z22v 1111 16.6 1.289 1.352 1.289 1.352 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.241 1.342 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.497000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.363636, yerr=13.479093 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 833.732608E+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 0.053 0.062 54.490 54.490 qs_mol_dyn_low 1 2.0 0.003 0.003 54.222 54.232 qs_forces 11 3.9 0.002 0.002 53.679 53.680 qs_energies 11 4.9 0.002 0.003 50.055 50.058 scf_env_do_scf 11 5.9 0.000 0.001 41.815 41.815 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.167 34.168 velocity_verlet 10 3.0 0.023 0.047 30.734 30.764 dbcsr_multiply_generic 2055 12.4 0.104 0.105 22.543 22.760 qs_scf_new_mos 99 7.5 0.001 0.001 20.206 20.245 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.205 20.244 ot_scf_mini 99 9.5 0.002 0.002 18.974 19.000 multiply_cannon 2055 13.4 0.248 0.259 17.241 18.659 multiply_cannon_loop 2055 14.4 0.324 0.336 15.908 16.365 rebuild_ks_matrix 110 8.3 0.000 0.000 11.380 11.417 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.379 11.416 ot_mini 99 10.5 0.001 0.001 10.275 10.299 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.114 10.146 multiply_cannon_multrec 8220 15.4 3.192 4.401 7.440 8.420 init_scf_loop 11 6.9 0.000 0.000 7.596 7.597 mp_waitall_1 106626 16.5 5.840 7.370 5.840 7.370 qs_ot_get_derivative 99 11.5 0.001 0.001 6.501 6.525 sum_up_and_integrate 110 10.3 0.048 0.049 6.143 6.155 integrate_v_rspace 110 11.3 0.003 0.003 6.095 6.107 prepare_preconditioner 11 7.9 0.000 0.000 5.973 5.975 make_preconditioner 11 8.9 0.000 0.000 5.973 5.975 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.555 5.628 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.552 5.561 calculate_rho_elec 110 8.6 0.115 0.116 5.551 5.561 init_scf_run 11 5.9 0.000 0.001 5.137 5.137 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.137 5.137 dbcsr_mm_accdrv_process 17442 15.9 2.834 3.847 4.118 5.030 qs_ot_get_p 110 10.4 0.001 0.001 4.853 4.883 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.970 4.330 make_m2s 4110 13.4 0.039 0.040 4.046 4.278 make_images 4110 14.4 0.642 0.696 3.916 4.147 ot_diis_step 99 11.5 0.012 0.012 3.753 3.753 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.667 3.705 apply_single 110 13.6 0.000 0.000 3.667 3.705 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.527 3.530 grid_integrate_task_list 110 12.3 3.375 3.499 3.375 3.499 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.211 3.211 pw_transfer 1331 11.6 0.066 0.071 3.165 3.178 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.054 3.071 multiply_cannon_sync_h2d 8220 15.4 2.915 3.024 2.915 3.024 cp_fm_cholesky_invert 11 10.9 2.875 2.878 2.875 2.878 density_rs2pw 110 9.6 0.004 0.004 2.614 2.719 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.693 2.693 cp_fm_redistribute_end 48 14.0 0.680 2.663 0.684 2.663 wfi_extrapolate 11 7.9 0.001 0.001 2.660 2.660 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.653 2.654 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 2.622 2.640 cp_fm_diag_elpa_base 48 14.0 1.794 2.473 1.973 2.635 make_images_data 4110 15.4 0.038 0.043 2.247 2.621 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.186 2.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.541 2.542 calculate_dm_sparse 110 9.5 0.001 0.001 2.487 2.523 grid_collocate_task_list 110 9.6 2.331 2.498 2.331 2.498 calculate_first_density_matrix 1 7.0 0.001 0.001 2.392 2.393 rs_pw_transfer 902 11.9 0.010 0.010 2.238 2.374 fft3d_ps 1111 14.6 1.117 1.144 2.309 2.330 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.182 2.205 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.769 1.992 potential_pw2rs 110 12.3 0.015 0.015 1.952 1.958 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.906 1.921 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.801 1.815 mp_alltoall_d11v 2046 13.8 1.571 1.721 1.571 1.721 cp_fm_cholesky_decompose 22 10.9 1.668 1.681 1.668 1.681 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.499 1.617 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.598 1.603 dbcsr_complete_redistribute 325 12.2 0.564 0.588 1.452 1.533 mp_allgather_i34 2055 14.4 0.518 1.487 0.518 1.487 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.455 1.466 jit_kernel_multiply 8 15.8 0.976 1.413 0.976 1.413 qs_create_task_list 11 7.9 0.001 0.001 1.226 1.326 generate_qs_task_list 11 8.9 0.381 0.447 1.225 1.325 mp_waitany 9240 13.8 1.024 1.156 1.024 1.156 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.115 1.136 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.483 1.110 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.490000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=786.636364, yerr=10.636752 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 1.187938E+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 0.054 0.186 85.891 85.892 qs_mol_dyn_low 1 2.0 0.003 0.003 85.352 85.366 qs_forces 11 3.9 0.045 0.089 85.095 85.266 qs_energies 11 4.9 0.001 0.002 81.006 81.219 scf_env_do_scf 11 5.9 0.001 0.001 70.999 70.999 velocity_verlet 10 3.0 0.002 0.002 54.242 54.248 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.889 42.890 dbcsr_multiply_generic 2055 12.4 0.119 0.121 29.191 29.279 init_scf_loop 11 6.9 0.000 0.000 28.038 28.040 qs_scf_new_mos 99 7.5 0.001 0.001 26.503 26.569 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.502 26.568 prepare_preconditioner 11 7.9 0.000 0.000 26.077 26.094 make_preconditioner 11 8.9 0.000 0.000 26.077 26.093 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.438 25.553 ot_scf_mini 99 9.5 0.002 0.002 24.778 24.835 multiply_cannon 2055 13.4 0.347 0.372 21.859 22.541 multiply_cannon_loop 2055 14.4 0.342 0.345 20.013 20.410 cp_fm_upper_to_full 70 14.2 12.579 17.933 12.579 17.933 ot_mini 99 10.5 0.001 0.001 14.063 14.119 rebuild_ks_matrix 110 8.3 0.001 0.001 13.241 13.290 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.241 13.289 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.058 12.102 dbcsr_complete_redistribute 325 12.2 1.015 1.037 7.205 10.314 multiply_cannon_multrec 8220 15.4 4.363 4.554 9.521 9.594 qs_ot_get_derivative 99 11.5 0.001 0.001 9.494 9.553 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.197 9.310 mp_waitall_1 87304 16.6 8.063 8.812 8.063 8.812 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.624 8.716 mp_alltoall_i22 605 13.7 5.253 8.351 5.253 8.351 sum_up_and_integrate 110 10.3 0.090 0.092 6.551 6.568 integrate_v_rspace 110 11.3 0.003 0.003 6.461 6.478 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.357 6.395 calculate_rho_elec 110 8.6 0.226 0.227 6.357 6.394 make_m2s 4110 13.4 0.043 0.044 5.376 5.840 make_images 4110 14.4 0.882 0.929 5.189 5.653 init_scf_run 11 5.9 0.000 0.001 5.647 5.647 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.646 5.647 qs_ot_get_p 110 10.4 0.001 0.001 5.490 5.559 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.011 5.414 cp_fm_cholesky_invert 11 10.9 5.319 5.324 5.319 5.324 dbcsr_mm_accdrv_process 11614 15.7 3.326 3.619 5.014 5.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.674 5.155 apply_single 110 13.6 0.000 0.000 4.674 5.155 ot_diis_step 99 11.5 0.015 0.015 4.545 4.547 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.990 3.998 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.774 3.988 multiply_cannon_sync_h2d 8220 15.4 3.950 3.952 3.950 3.952 pw_transfer 1331 11.6 0.073 0.073 3.779 3.782 grid_integrate_task_list 110 12.3 3.658 3.717 3.658 3.717 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.655 3.658 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.553 3.553 make_images_data 4110 15.4 0.041 0.044 2.914 3.545 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.862 3.535 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.833 3.286 fft_wrap_pw1pw2_140 451 13.1 0.230 0.231 3.177 3.181 wfi_extrapolate 11 7.9 0.001 0.001 3.174 3.175 calculate_dm_sparse 110 9.5 0.001 0.001 3.119 3.150 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.017 3.017 cp_fm_diag_elpa_base 48 14.0 2.487 2.672 3.015 3.015 density_rs2pw 110 9.6 0.004 0.004 2.986 3.005 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.911 2.916 fft3d_ps 1111 14.6 1.267 1.275 2.851 2.856 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.707 2.739 grid_collocate_task_list 110 9.6 2.626 2.658 2.626 2.658 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.294 2.543 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.431 2.446 rs_pw_transfer 902 11.9 0.011 0.011 2.222 2.438 calculate_first_density_matrix 1 7.0 0.000 0.001 2.368 2.370 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.104 2.195 potential_pw2rs 110 12.3 0.021 0.022 2.052 2.056 cp_fm_cholesky_decompose 22 10.9 2.021 2.042 2.021 2.042 qs_create_task_list 11 7.9 0.009 0.009 1.911 1.958 generate_qs_task_list 11 8.9 0.739 0.794 1.902 1.949 mp_alltoall_d11v 2046 13.8 1.791 1.840 1.791 1.840 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.790 1.795 jit_kernel_multiply 10 15.2 1.486 1.731 1.486 1.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.892000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1119.454545, yerr=10.192365 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 626.360320E+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 2963290. 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 0.077 0.151 204.833 204.834 qs_mol_dyn_low 1 2.0 0.006 0.020 204.213 204.228 qs_forces 11 3.9 0.022 0.105 204.049 204.052 qs_energies 11 4.9 0.007 0.025 198.390 198.410 scf_env_do_scf 11 5.9 0.001 0.001 181.204 181.208 scf_env_do_scf_inner_loop 117 6.6 0.003 0.015 160.457 160.459 dbcsr_multiply_generic 2507 12.6 0.175 0.179 124.724 125.239 velocity_verlet 10 3.0 0.002 0.003 122.560 122.573 qs_scf_new_mos 117 7.6 0.001 0.001 121.202 121.394 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.201 121.394 ot_scf_mini 117 9.6 0.003 0.004 114.603 114.774 multiply_cannon 2507 13.6 0.237 0.246 101.244 103.560 multiply_cannon_loop 2507 14.6 2.125 2.202 99.043 100.968 ot_mini 117 10.6 0.001 0.002 65.740 65.916 multiply_cannon_multrec 60168 15.6 33.087 35.667 41.787 43.343 qs_ot_get_derivative 117 11.6 0.001 0.002 41.062 41.244 rebuild_ks_matrix 128 8.3 0.001 0.001 33.658 34.053 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 33.657 34.052 mp_waitall_1 291448 16.2 28.674 31.595 28.674 31.595 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.279 30.649 multiply_cannon_sync_h2d 60168 15.6 27.220 29.905 27.220 29.905 qs_ot_get_p 128 10.4 0.001 0.001 26.970 27.141 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.230 24.772 apply_single 128 13.6 0.001 0.001 24.230 24.772 ot_diis_step 117 11.6 0.008 0.010 24.324 24.325 init_scf_loop 11 6.9 0.000 0.000 20.668 20.669 qs_ot_p2m_diag 83 11.4 0.077 0.092 20.265 20.377 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.958 19.095 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 15.700 17.798 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.584 17.585 prepare_preconditioner 11 7.9 0.000 0.000 16.002 16.032 make_preconditioner 11 8.9 0.000 0.000 16.002 16.032 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.235 15.395 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.753 14.758 cp_fm_redistribute_end 83 14.4 11.652 14.682 11.663 14.684 cp_fm_diag_elpa_base 83 14.4 2.980 14.389 3.010 14.514 sum_up_and_integrate 128 10.3 0.053 0.065 14.317 14.335 integrate_v_rspace 128 11.3 0.003 0.005 14.263 14.282 make_m2s 5014 13.6 0.103 0.111 13.801 14.219 make_images 5014 14.6 0.402 0.420 13.619 14.046 qs_rho_update_rho_low 128 7.7 0.001 0.002 12.866 12.956 calculate_rho_elec 128 8.7 0.045 0.063 12.866 12.955 init_scf_run 11 5.9 0.000 0.001 12.765 12.766 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.765 12.766 mp_sum_l 7870 13.0 8.460 9.488 8.460 9.488 cp_fm_cholesky_invert 11 10.9 9.122 9.130 9.122 9.130 wfi_extrapolate 11 7.9 0.001 0.003 9.033 9.033 calculate_dm_sparse 128 9.5 0.001 0.004 8.803 8.936 dbcsr_mm_accdrv_process 124484 16.2 3.298 3.467 8.263 8.880 multiply_cannon_metrocomm1 60168 15.6 0.086 0.091 6.146 8.549 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.147 8.279 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.914 8.010 make_images_data 5014 15.6 0.069 0.076 6.759 7.697 grid_integrate_task_list 128 12.3 7.090 7.507 7.090 7.507 density_rs2pw 128 9.7 0.006 0.006 6.662 7.069 pw_transfer 1547 11.6 0.075 0.104 6.754 7.041 hybrid_alltoall_any 5200 16.5 0.289 2.251 5.910 7.014 rs_pw_transfer 1046 11.9 0.017 0.019 6.423 6.860 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 6.540 6.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.664 6.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.863 5.954 fft_wrap_pw1pw2_140 523 13.2 0.437 0.494 5.552 5.734 fft3d_ps 1291 14.7 2.093 2.583 5.341 5.562 mp_alltoall_d11v 2415 14.1 4.196 5.467 4.196 5.467 grid_collocate_task_list 128 9.7 4.727 5.134 4.727 5.134 potential_pw2rs 128 12.3 0.009 0.011 4.862 4.894 cp_fm_cholesky_decompose 22 10.9 4.698 4.714 4.698 4.714 mp_sum_d 4459 12.1 3.522 4.168 3.522 4.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.834000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.181818, yerr=6.603280 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 826.474496E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57228. MP_Allreduce 11247 947. MP_Sync 168 MP_Alltoall 1955 4824147. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.113 190.626 190.628 qs_mol_dyn_low 1 2.0 0.005 0.019 190.068 190.083 qs_forces 11 3.9 0.005 0.018 189.929 189.931 qs_energies 11 4.9 0.006 0.037 183.225 183.237 scf_env_do_scf 11 5.9 0.002 0.013 166.434 166.444 scf_env_do_scf_inner_loop 116 6.6 0.004 0.013 132.996 133.000 velocity_verlet 10 3.0 0.001 0.003 118.808 118.813 dbcsr_multiply_generic 2485 12.5 0.183 0.192 97.459 98.720 qs_scf_new_mos 116 7.6 0.001 0.001 94.709 95.232 qs_scf_loop_do_ot 116 8.6 0.001 0.002 94.709 95.232 ot_scf_mini 116 9.6 0.004 0.005 89.939 90.466 multiply_cannon 2485 13.5 0.472 0.524 77.480 81.870 multiply_cannon_loop 2485 14.5 1.254 1.295 74.244 77.172 ot_mini 116 10.6 0.001 0.002 50.051 50.548 mp_waitall_1 224796 16.4 24.642 38.099 24.642 38.099 multiply_cannon_multrec 29820 15.5 21.867 26.778 31.702 36.776 init_scf_loop 11 6.9 0.004 0.031 33.349 33.352 rebuild_ks_matrix 127 8.3 0.001 0.001 32.504 33.152 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 32.504 33.152 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.279 29.867 prepare_preconditioner 11 7.9 0.000 0.000 28.953 29.036 make_preconditioner 11 8.9 0.000 0.002 28.953 29.036 qs_ot_get_derivative 116 11.6 0.001 0.002 28.242 28.778 multiply_cannon_metrocomm3 29820 15.5 0.095 0.101 15.840 28.428 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.594 28.116 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 22.052 23.263 apply_single 127 13.6 0.001 0.001 22.052 23.262 qs_ot_get_p 127 10.4 0.001 0.001 21.821 22.409 ot_diis_step 116 11.6 0.015 0.016 21.641 21.644 multiply_cannon_sync_h2d 29820 15.5 19.208 21.580 19.208 21.580 qs_ot_p2m_diag 82 11.4 0.185 0.214 17.096 17.130 cp_fm_cholesky_invert 11 10.9 16.979 16.992 16.979 16.992 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.957 15.958 make_m2s 4970 13.5 0.089 0.094 13.976 15.542 make_images 4970 14.5 1.146 1.326 13.767 15.334 sum_up_and_integrate 127 10.3 0.061 0.071 14.202 14.232 integrate_v_rspace 127 11.3 0.003 0.003 14.141 14.172 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.900 12.908 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.863 12.898 calculate_rho_elec 127 8.7 0.087 0.104 12.862 12.898 cp_fm_redistribute_end 82 14.4 7.542 12.833 7.554 12.835 cp_fm_diag_elpa_base 82 14.4 5.047 12.314 5.263 12.731 init_scf_run 11 5.9 0.000 0.001 11.841 11.842 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.840 11.842 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 10.877 11.269 multiply_cannon_metrocomm4 27335 15.5 0.095 0.112 3.735 10.563 dbcsr_mm_accdrv_process 61726 16.2 4.758 5.945 9.298 10.375 mp_irecv_dv 68888 16.3 3.543 10.191 3.543 10.191 make_images_data 4970 15.5 0.065 0.072 8.336 10.134 hybrid_alltoall_any 5155 16.4 0.339 1.483 7.173 9.413 wfi_extrapolate 11 7.9 0.002 0.013 8.338 8.338 pw_transfer 1535 11.6 0.082 0.092 7.584 7.639 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.541 7.509 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.357 7.413 grid_integrate_task_list 127 12.3 7.077 7.404 7.077 7.404 density_rs2pw 127 9.7 0.006 0.006 6.737 7.194 cp_fm_cholesky_decompose 22 10.9 7.043 7.117 7.043 7.117 calculate_dm_sparse 127 9.5 0.001 0.003 6.726 6.898 fft_wrap_pw1pw2_140 519 13.2 0.455 0.488 6.431 6.500 rs_pw_transfer 1038 11.9 0.014 0.015 5.864 6.399 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.176 6.186 mp_sum_l 7804 13.0 4.239 6.141 4.239 6.141 fft3d_ps 1281 14.7 2.753 2.906 5.753 5.790 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.315 5.474 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.309 5.380 grid_collocate_task_list 127 9.7 4.871 5.216 4.871 5.216 mp_alltoall_d11v 2401 14.1 4.064 4.921 4.064 4.921 potential_pw2rs 127 12.3 0.016 0.018 4.837 4.855 mp_allgather_i34 2485 14.5 1.823 4.659 1.823 4.659 dbcsr_complete_redistribute 393 12.7 0.775 0.865 3.193 4.028 mp_sum_d 4445 12.1 2.707 3.950 2.707 3.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.628000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.000000, yerr=1.758098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 938.545152E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56898. MP_Allreduce 11306 983. MP_Sync 170 MP_Alltoall 1724 9394172. MP_ISendRecv 15996 75008. MP_Wait 29984 MP_comm_split 83 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.117 176.781 176.782 qs_mol_dyn_low 1 2.0 0.003 0.005 176.193 176.207 qs_forces 11 3.9 0.005 0.027 176.086 176.091 qs_energies 11 4.9 0.003 0.020 169.587 169.600 scf_env_do_scf 11 5.9 0.001 0.004 153.592 153.593 scf_env_do_scf_inner_loop 118 6.6 0.006 0.030 118.574 118.575 velocity_verlet 10 3.0 0.002 0.002 113.062 113.064 dbcsr_multiply_generic 2527 12.6 0.183 0.187 82.550 83.643 qs_scf_new_mos 118 7.6 0.001 0.001 81.643 81.970 qs_scf_loop_do_ot 118 8.6 0.001 0.002 81.642 81.969 ot_scf_mini 118 9.6 0.004 0.004 77.455 77.841 multiply_cannon 2527 13.6 0.501 0.522 62.598 67.594 multiply_cannon_loop 2527 14.6 0.871 0.896 59.565 62.600 ot_mini 118 10.6 0.001 0.002 43.101 43.470 init_scf_loop 11 6.9 0.000 0.001 34.919 34.920 mp_waitall_1 179876 16.5 25.263 34.036 25.263 34.036 prepare_preconditioner 11 7.9 0.000 0.000 30.930 30.976 make_preconditioner 11 8.9 0.000 0.001 30.930 30.976 rebuild_ks_matrix 129 8.3 0.001 0.001 30.309 30.824 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.020 30.308 30.824 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.625 30.004 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.400 27.872 multiply_cannon_multrec 20216 15.6 13.596 16.707 22.651 25.923 multiply_cannon_metrocomm3 20216 15.6 0.058 0.062 15.300 24.580 qs_ot_get_derivative 118 11.6 0.002 0.002 23.242 23.626 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 20.141 21.088 apply_single 129 13.6 0.001 0.001 20.141 21.088 qs_ot_get_p 129 10.4 0.001 0.001 19.708 20.142 ot_diis_step 118 11.6 0.021 0.040 19.753 19.753 multiply_cannon_sync_h2d 20216 15.6 14.412 16.480 14.412 16.480 make_m2s 5054 13.6 0.082 0.089 14.627 15.512 qs_ot_p2m_diag 83 11.4 0.266 0.273 15.385 15.394 make_images 5054 14.6 1.189 1.267 14.391 15.275 cp_fm_cholesky_invert 11 10.9 14.433 14.442 14.433 14.442 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.370 14.372 sum_up_and_integrate 129 10.3 0.071 0.076 14.266 14.290 integrate_v_rspace 129 11.3 0.003 0.003 14.195 14.221 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.165 13.197 calculate_rho_elec 129 8.7 0.133 0.147 13.164 13.197 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.363 11.364 cp_fm_redistribute_end 83 14.4 4.288 11.302 4.303 11.305 cp_fm_diag_elpa_base 83 14.4 6.578 10.709 6.984 11.190 init_scf_run 11 5.9 0.000 0.001 10.934 10.935 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.934 10.935 make_images_data 5054 15.6 0.061 0.068 8.944 10.239 hybrid_alltoall_any 5240 16.5 0.435 1.981 7.752 9.372 multiply_cannon_metrocomm4 17689 15.6 0.064 0.075 3.470 9.262 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.971 9.239 mp_irecv_dv 50625 16.2 3.345 9.007 3.345 9.007 dbcsr_mm_accdrv_process 41830 16.2 4.572 5.789 8.503 8.798 grid_integrate_task_list 129 12.3 7.359 7.828 7.359 7.828 pw_transfer 1559 11.6 0.083 0.102 7.711 7.819 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.484 7.598 cp_fm_upper_to_full 105 14.5 5.744 7.456 5.744 7.456 cp_fm_cholesky_decompose 22 10.9 7.404 7.439 7.404 7.439 wfi_extrapolate 11 7.9 0.001 0.002 7.317 7.318 density_rs2pw 129 9.7 0.006 0.006 6.548 6.879 fft_wrap_pw1pw2_140 527 13.2 0.475 0.517 6.548 6.672 calculate_dm_sparse 129 9.5 0.001 0.002 6.050 6.176 dbcsr_complete_redistribute 395 12.7 1.167 1.195 4.409 6.169 fft3d_ps 1301 14.7 2.721 2.928 5.747 5.841 rs_pw_transfer 1054 12.0 0.013 0.014 5.344 5.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.603 5.611 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.703 5.580 grid_collocate_task_list 129 9.7 5.112 5.525 5.112 5.525 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.231 4.992 mp_alltoall_d11v 2423 14.1 4.171 4.872 4.171 4.872 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.710 4.835 mp_sum_l 7930 13.1 3.524 4.796 3.524 4.796 potential_pw2rs 129 12.3 0.021 0.023 4.648 4.665 mp_allgather_i34 2527 14.6 1.554 4.397 1.554 4.397 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.049 4.081 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.286 4.023 mp_alltoall_i22 718 14.1 1.934 3.893 1.934 3.893 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.771 3.773 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.782000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.090909, yerr=8.670268 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.144816E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4129 56599. MP_Allreduce 11361 1063. MP_Sync 172 MP_Alltoall 1724 12509605. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.217 189.515 189.516 qs_mol_dyn_low 1 2.0 0.003 0.005 188.873 188.933 qs_forces 11 3.9 0.003 0.004 188.733 188.739 qs_energies 11 4.9 0.002 0.007 181.669 181.685 scf_env_do_scf 11 5.9 0.001 0.001 164.889 164.901 velocity_verlet 10 3.0 0.003 0.011 125.134 125.138 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.107 118.109 qs_scf_new_mos 118 7.6 0.001 0.001 82.163 82.481 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.163 82.481 dbcsr_multiply_generic 2529 12.6 0.191 0.197 80.722 81.470 ot_scf_mini 118 9.6 0.003 0.004 77.521 77.799 multiply_cannon 2529 13.6 0.563 0.594 55.432 58.503 multiply_cannon_loop 2529 14.6 1.195 1.232 51.312 53.086 init_scf_loop 11 6.9 0.001 0.005 46.656 46.662 ot_mini 118 10.6 0.001 0.002 43.209 43.482 prepare_preconditioner 11 7.9 0.000 0.000 42.406 42.425 make_preconditioner 11 8.9 0.000 0.001 42.406 42.425 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.922 40.961 multiply_cannon_multrec 30348 15.6 14.220 19.492 26.231 31.067 rebuild_ks_matrix 129 8.3 0.001 0.001 29.327 29.579 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 29.326 29.578 mp_waitall_1 155106 16.5 17.425 27.282 17.425 27.282 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.483 26.711 qs_ot_get_derivative 118 11.6 0.002 0.002 23.320 23.602 make_m2s 5058 13.6 0.095 0.099 20.749 22.044 make_images 5058 14.6 1.973 2.264 20.440 21.734 qs_ot_get_p 129 10.4 0.005 0.031 20.337 20.633 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.343 19.794 apply_single 129 13.6 0.001 0.001 19.343 19.794 ot_diis_step 118 11.6 0.018 0.021 19.763 19.765 cp_fm_upper_to_full 106 14.7 11.199 16.601 11.199 16.601 qs_ot_p2m_diag 84 11.4 0.347 0.396 16.066 16.118 cp_fm_cholesky_invert 11 10.9 16.092 16.100 16.092 16.100 multiply_cannon_metrocomm3 30348 15.6 0.047 0.050 6.465 15.298 cp_dbcsr_syevd 84 12.4 0.005 0.005 14.773 14.774 sum_up_and_integrate 129 10.3 0.080 0.087 14.178 14.217 integrate_v_rspace 129 11.3 0.003 0.003 14.097 14.139 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.313 13.343 calculate_rho_elec 129 8.7 0.176 0.191 13.313 13.342 dbcsr_complete_redistribute 397 12.7 1.540 1.669 9.204 13.006 multiply_cannon_sync_h2d 30348 15.6 11.834 12.721 11.834 12.721 make_images_data 5058 15.6 0.063 0.068 10.925 12.537 dbcsr_mm_accdrv_process 62780 16.2 7.484 8.425 11.583 12.133 hybrid_alltoall_any 5245 16.5 0.529 2.222 9.761 11.775 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.842 11.665 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.547 11.549 cp_fm_redistribute_end 84 14.4 1.997 11.480 2.011 11.485 cp_fm_diag_elpa_base 84 14.4 8.852 10.846 9.445 11.371 init_scf_run 11 5.9 0.000 0.001 10.820 10.821 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.819 10.821 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.462 10.143 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.736 9.945 mp_alltoall_i22 720 14.1 5.625 9.299 5.625 9.299 grid_integrate_task_list 129 12.3 7.596 7.951 7.596 7.951 pw_transfer 1559 11.6 0.083 0.100 7.825 7.902 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.595 7.675 cp_fm_cholesky_decompose 22 10.9 7.577 7.659 7.577 7.659 wfi_extrapolate 11 7.9 0.002 0.012 7.567 7.568 multiply_cannon_metrocomm4 25290 15.6 0.076 0.086 2.821 6.893 fft_wrap_pw1pw2_140 527 13.2 0.476 0.485 6.733 6.821 mp_irecv_dv 76751 16.2 2.676 6.630 2.676 6.630 density_rs2pw 129 9.7 0.005 0.006 6.327 6.616 calculate_dm_sparse 129 9.5 0.001 0.001 6.325 6.413 fft3d_ps 1301 14.7 2.821 2.902 5.817 5.918 grid_collocate_task_list 129 9.7 5.244 5.614 5.244 5.614 mp_alltoall_d11v 2429 14.1 5.131 5.570 5.131 5.570 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.514 5.559 rs_pw_transfer 1054 12.0 0.013 0.014 4.868 5.167 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.439 4.603 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.412 4.509 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.407 4.458 potential_pw2rs 129 12.3 0.024 0.024 4.435 4.457 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.422 4.425 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.516000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.818182, yerr=19.474077 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.504895E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.046 169.715 169.715 qs_mol_dyn_low 1 2.0 0.003 0.003 169.225 169.238 qs_forces 11 3.9 0.003 0.003 169.119 169.124 qs_energies 11 4.9 0.001 0.002 161.862 161.868 scf_env_do_scf 11 5.9 0.001 0.001 144.895 144.909 velocity_verlet 10 3.0 0.015 0.016 111.604 111.608 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 109.499 109.501 dbcsr_multiply_generic 2527 12.6 0.182 0.187 74.023 74.450 qs_scf_new_mos 118 7.6 0.001 0.001 73.906 73.991 qs_scf_loop_do_ot 118 8.6 0.001 0.001 73.905 73.991 ot_scf_mini 118 9.6 0.004 0.004 69.465 69.522 multiply_cannon 2527 13.6 0.594 0.621 54.607 59.246 multiply_cannon_loop 2527 14.6 0.454 0.467 49.989 50.783 ot_mini 118 10.6 0.001 0.001 39.533 39.602 init_scf_loop 11 6.9 0.000 0.000 35.246 35.248 mp_waitall_1 130650 16.6 25.631 33.184 25.631 33.184 prepare_preconditioner 11 7.9 0.000 0.000 31.396 31.425 make_preconditioner 11 8.9 0.000 0.000 31.396 31.425 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.239 29.500 rebuild_ks_matrix 129 8.3 0.001 0.001 28.579 28.624 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.578 28.624 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.129 26.164 multiply_cannon_multrec 10108 15.6 10.522 14.431 17.919 21.192 ot_diis_step 118 11.6 0.020 0.022 19.917 19.917 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.529 19.787 apply_single 129 13.6 0.001 0.001 19.529 19.787 qs_ot_get_derivative 118 11.6 0.002 0.002 19.548 19.602 multiply_cannon_metrocomm3 10108 15.6 0.022 0.024 12.287 19.386 make_m2s 5054 13.6 0.067 0.071 15.873 18.212 make_images 5054 14.6 2.322 2.695 15.563 17.904 cp_fm_cholesky_invert 11 10.9 17.861 17.867 17.861 17.867 qs_ot_get_p 129 10.4 0.001 0.001 17.072 17.192 sum_up_and_integrate 129 10.3 0.103 0.110 14.026 14.073 integrate_v_rspace 129 11.3 0.003 0.004 13.923 13.975 qs_ot_p2m_diag 83 11.4 0.496 0.503 13.386 13.402 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.075 13.106 calculate_rho_elec 129 8.7 0.259 0.270 13.074 13.106 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.248 12.249 multiply_cannon_sync_h2d 10108 15.6 11.753 12.160 11.753 12.160 hybrid_alltoall_any 5240 16.5 0.851 3.861 9.437 11.966 make_images_data 5054 15.6 0.052 0.060 9.611 11.895 init_scf_run 11 5.9 0.000 0.001 10.206 10.206 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.205 10.206 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.283 9.287 cp_fm_diag_elpa_base 83 14.4 9.042 9.110 9.273 9.277 grid_integrate_task_list 129 12.3 7.835 8.166 7.835 8.166 cp_fm_cholesky_decompose 22 10.9 7.865 7.961 7.865 7.961 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.791 7.827 dbcsr_mm_accdrv_process 20926 16.1 2.715 3.524 7.026 7.655 pw_transfer 1559 11.6 0.084 0.090 7.602 7.621 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.374 7.399 multiply_cannon_metrocomm1 10108 15.6 0.029 0.030 4.390 7.238 mp_allgather_i34 2527 14.6 2.817 7.204 2.817 7.204 wfi_extrapolate 11 7.9 0.001 0.001 7.166 7.166 fft_wrap_pw1pw2_140 527 13.2 0.497 0.516 6.512 6.543 calculate_dm_sparse 129 9.5 0.001 0.001 6.149 6.224 density_rs2pw 129 9.7 0.005 0.006 5.907 6.190 grid_collocate_task_list 129 9.7 5.523 5.785 5.523 5.785 dbcsr_complete_redistribute 395 12.7 2.134 2.233 5.188 5.581 fft3d_ps 1301 14.7 2.748 2.819 5.526 5.543 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.227 5.229 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.143 5.153 mp_alltoall_d11v 2423 14.1 4.548 5.089 4.548 5.089 rs_pw_transfer 1054 12.0 0.013 0.013 4.234 4.546 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.853 4.361 mp_irecv_dv 29086 15.9 1.815 4.279 1.815 4.279 potential_pw2rs 129 12.3 0.027 0.028 4.122 4.131 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.085 4.117 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.570 3.858 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.417 3.753 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.682 3.711 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.467 3.530 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.497 3.524 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.715000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1400.545455, yerr=43.361826 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430451062784 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986258751488 0.0% 0.0% 100.0% flops 32 x 32 x 32 1986422374400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1991997444096 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 11613050155008 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 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 94.256601E+12 0.0% 0.0% 100.0% flops max/rank 11.789007E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806737984 0.0% 0.0% 100.0% number of processed stacks 1984336 0.0% 0.0% 100.0% average stack size 0.0 0.0 3430.2 marketing flops 145.674553E+12 ------------------------------------------------------------------------------- # multiplications 2540 max memory usage/rank 2.278687E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101600 MPI messages size (bytes): total size 1.145338E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.273011E+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 46088 35802578944 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592076688 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4035 58676. MP_Allreduce 11084 1511. MP_Sync 88 MP_Alltoall 1724 36993736. 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 0.064 0.085 290.827 290.827 qs_mol_dyn_low 1 2.0 0.003 0.003 290.119 290.132 qs_forces 11 3.9 0.003 0.003 290.013 290.018 qs_energies 11 4.9 0.002 0.002 281.429 281.436 scf_env_do_scf 11 5.9 0.001 0.001 259.366 259.375 velocity_verlet 10 3.0 0.014 0.016 209.471 209.479 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.046 136.047 init_scf_loop 11 6.9 0.000 0.000 123.058 123.060 prepare_preconditioner 11 7.9 0.000 0.000 118.362 118.391 make_preconditioner 11 8.9 0.000 0.000 118.362 118.391 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.428 115.504 qs_scf_new_mos 118 7.6 0.001 0.001 93.359 93.450 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.359 93.449 ot_scf_mini 118 9.6 0.004 0.004 88.531 88.576 dbcsr_multiply_generic 2540 12.6 0.216 0.223 84.969 85.545 cp_fm_upper_to_full 106 14.8 52.468 74.377 52.468 74.377 multiply_cannon 2540 13.6 0.695 0.724 59.614 60.731 multiply_cannon_loop 2540 14.6 0.481 0.489 55.754 57.150 ot_mini 118 10.6 0.001 0.001 45.697 45.737 dbcsr_complete_redistribute 397 12.7 3.996 4.067 28.753 40.980 copy_fm_to_dbcsr 210 11.7 0.001 0.001 25.260 37.522 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 22.891 35.060 rebuild_ks_matrix 129 8.3 0.001 0.001 34.120 34.176 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.119 34.175 cp_fm_cholesky_invert 11 10.9 33.771 33.779 33.771 33.779 mp_waitall_1 106826 16.7 29.310 33.060 29.310 33.060 mp_alltoall_i22 720 14.1 20.710 32.916 20.710 32.916 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.836 31.891 qs_ot_get_p 129 10.4 0.001 0.001 27.038 27.116 qs_ot_get_derivative 118 11.6 0.002 0.002 24.848 24.895 qs_ot_p2m_diag 84 11.4 0.890 0.894 22.837 22.868 make_m2s 5080 13.6 0.078 0.080 20.538 21.707 make_images 5080 14.6 3.826 4.048 20.059 21.231 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.058 21.060 ot_diis_step 118 11.6 0.023 0.024 20.804 20.804 multiply_cannon_metrocomm3 10160 15.6 0.023 0.024 19.243 20.490 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.018 20.213 apply_single 129 13.6 0.001 0.001 20.018 20.212 multiply_cannon_multrec 10160 15.6 10.586 12.362 18.128 18.248 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.789 17.790 cp_fm_diag_elpa_base 84 14.4 13.323 14.921 17.785 17.786 sum_up_and_integrate 129 10.3 0.195 0.197 15.745 15.838 multiply_cannon_sync_h2d 10160 15.6 15.782 15.798 15.782 15.798 integrate_v_rspace 129 11.3 0.004 0.004 15.549 15.643 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.109 15.126 calculate_rho_elec 129 8.7 0.486 0.487 15.109 15.125 hybrid_alltoall_any 5267 16.5 1.313 3.062 11.236 13.220 make_images_data 5080 15.6 0.059 0.064 11.164 13.159 init_scf_run 11 5.9 0.000 0.001 12.031 12.031 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.030 12.031 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.724 9.766 dbcsr_mm_accdrv_process 20978 16.1 3.909 5.863 7.301 9.197 cp_fm_cholesky_decompose 22 10.9 9.053 9.126 9.053 9.126 wfi_extrapolate 11 7.9 0.001 0.001 8.797 8.797 grid_integrate_task_list 129 12.3 8.601 8.786 8.601 8.786 pw_transfer 1559 11.6 0.094 0.095 8.448 8.454 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.206 8.213 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 7.989 7.992 fft_wrap_pw1pw2_140 527 13.2 0.528 0.532 7.236 7.247 mp_alltoall_d11v 2429 14.1 6.924 7.052 6.924 7.052 calculate_dm_sparse 129 9.5 0.001 0.001 6.719 6.791 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.483 6.584 grid_collocate_task_list 129 9.7 6.387 6.421 6.387 6.421 fft3d_ps 1301 14.7 2.783 2.793 6.252 6.269 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.134 6.265 density_rs2pw 129 9.7 0.005 0.005 6.138 6.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.827000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2142.636364, yerr=53.743960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 1.260405E+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 12512919. 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 0.060 0.143 85.311 85.312 qs_energies 1 2.0 0.000 0.001 84.663 84.670 ls_scf 1 3.0 0.001 0.012 83.748 83.756 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.645 72.801 multiply_cannon 111 7.7 0.017 0.020 56.055 57.075 multiply_cannon_loop 111 8.7 0.212 0.225 52.591 53.767 ls_scf_main 1 4.0 0.000 0.002 52.330 52.331 density_matrix_trs4 2 5.0 0.002 0.025 46.817 46.898 ls_scf_init_scf 1 4.0 0.000 0.001 28.350 28.351 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.104 27.143 mp_waitall_1 11316 10.9 22.486 25.405 22.486 25.405 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.953 24.968 multiply_cannon_multrec 2664 9.7 8.161 8.843 15.463 17.129 multiply_cannon_sync_h2d 2664 9.7 13.760 16.100 13.760 16.100 make_m2s 222 7.7 0.008 0.011 13.014 13.662 make_images 222 8.7 0.098 0.107 12.992 13.643 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.714 12.774 make_images_data 222 9.7 0.004 0.005 7.642 8.302 hybrid_alltoall_any 227 10.6 0.215 1.841 6.593 7.987 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.628 6.921 7.906 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.447 7.377 dbcsr_mm_accdrv_process_sort 4760 11.4 6.207 7.101 6.207 7.101 calculate_norms 4752 9.8 5.515 6.097 5.515 6.097 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.185 mp_sum_l 807 5.4 3.133 4.361 3.133 4.361 make_images_sizes 222 9.7 0.000 0.000 0.668 3.371 mp_alltoall_i44 222 10.7 0.667 3.370 0.667 3.370 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.045 3.274 mp_irecv_dv 6231 10.9 2.027 3.261 2.027 3.261 arnoldi_extremal 4 6.8 0.000 0.000 3.228 3.255 arnoldi_normal_ev 4 7.8 0.001 0.005 3.228 3.255 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.242 3.204 build_subspace 16 8.4 0.009 0.012 3.132 3.135 ls_scf_post 1 4.0 0.000 0.002 3.067 3.075 ls_scf_store_result 1 5.0 0.000 0.000 2.875 2.919 dbcsr_special_finalize 555 9.7 0.005 0.006 2.353 2.808 dbcsr_merge_single_wm 555 10.7 0.458 0.590 2.345 2.800 make_images_pack 222 9.7 2.207 2.632 2.208 2.633 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.311 2.575 dbcsr_sort_data 658 11.4 2.146 2.535 2.146 2.535 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.479 2.067 2.481 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.273 2.353 buffer_matrices_ensure_size 222 8.7 1.752 2.052 1.752 2.052 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.836 1.839 rebuild_ks_matrix 3 7.3 0.000 0.000 1.827 1.830 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.827 1.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.312000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 2.112258E+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 0.100 0.182 91.310 91.311 qs_energies 1 2.0 0.000 0.000 90.568 90.574 ls_scf 1 3.0 0.001 0.012 89.226 89.234 dbcsr_multiply_generic 111 6.7 0.016 0.022 75.160 75.448 multiply_cannon 111 7.7 0.027 0.043 53.147 57.479 ls_scf_main 1 4.0 0.000 0.001 54.868 54.876 multiply_cannon_loop 111 8.7 0.116 0.122 49.916 53.410 density_matrix_trs4 2 5.0 0.002 0.003 49.181 49.364 ls_scf_init_scf 1 4.0 0.000 0.001 30.788 30.789 mp_waitall_1 9246 10.9 21.147 29.719 21.147 29.719 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.527 29.648 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.115 27.131 multiply_cannon_multrec 1332 9.7 12.975 16.500 22.175 26.371 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.922 21.603 make_m2s 222 7.7 0.006 0.008 15.176 15.755 make_images 222 8.7 1.570 1.910 15.146 15.727 dbcsr_mm_accdrv_process 4041 10.4 0.260 0.422 8.806 10.305 dbcsr_mm_accdrv_process_sort 4041 11.4 8.402 9.876 8.402 9.876 make_images_data 222 9.7 0.005 0.016 8.617 9.463 mp_sum_l 807 5.4 5.575 9.006 5.575 9.006 hybrid_alltoall_any 227 10.6 0.524 2.457 8.034 8.927 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.244 7.800 mp_irecv_dv 3311 11.0 3.224 7.738 3.224 7.738 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.255 7.085 calculate_norms 2376 9.8 6.000 6.601 6.000 6.601 multiply_cannon_sync_h2d 1332 9.7 4.736 6.025 4.736 6.025 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.234 arnoldi_extremal 4 6.8 0.000 0.002 4.712 4.733 arnoldi_normal_ev 4 7.8 0.004 0.029 4.711 4.733 build_subspace 16 8.4 0.014 0.021 4.430 4.435 ls_scf_post 1 4.0 0.001 0.007 3.569 3.576 ls_scf_store_result 1 5.0 0.000 0.000 3.272 3.387 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.157 3.378 dbcsr_matrix_vector_mult_local 304 10.0 2.752 3.224 2.754 3.226 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.234 3.067 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.527 2.601 make_images_pack 222 9.7 2.016 2.381 2.018 2.383 mp_allgather_i34 111 8.7 0.933 2.351 0.933 2.351 dbcsr_sort_data 436 11.2 1.826 2.094 1.826 2.094 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.914 1.916 rebuild_ks_matrix 3 7.3 0.000 0.000 1.901 1.903 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.901 1.903 dbcsr_data_new 4174 10.1 1.614 1.834 1.614 1.834 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.311000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1765.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 2.748514E+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 0.052 0.072 95.176 95.176 qs_energies 1 2.0 0.000 0.000 94.417 94.432 ls_scf 1 3.0 0.000 0.001 93.007 93.025 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.737 78.056 ls_scf_main 1 4.0 0.000 0.010 58.143 58.147 multiply_cannon 111 7.7 0.039 0.070 53.830 57.770 multiply_cannon_loop 111 8.7 0.100 0.104 50.199 54.318 density_matrix_trs4 2 5.0 0.003 0.028 52.242 52.439 mp_waitall_1 7374 11.0 24.586 34.448 24.586 34.448 ls_scf_init_scf 1 4.0 0.000 0.002 31.313 31.316 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.100 30.205 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.357 27.371 multiply_cannon_multrec 888 9.7 12.566 15.040 21.219 24.505 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.761 23.702 make_m2s 222 7.7 0.006 0.007 17.025 18.338 make_images 222 8.7 1.967 2.291 16.987 18.296 hybrid_alltoall_any 227 10.6 0.619 2.853 9.205 10.804 make_images_data 222 9.7 0.003 0.004 9.649 10.737 dbcsr_mm_accdrv_process 3754 10.4 0.274 0.674 8.188 9.386 mp_sum_l 807 5.4 5.323 8.968 5.323 8.968 dbcsr_mm_accdrv_process_sort 3754 11.4 7.665 8.815 7.665 8.815 multiply_cannon_sync_h2d 888 9.7 6.055 7.174 6.055 7.174 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.099 7.160 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.447 6.702 mp_irecv_dv 2335 11.1 2.430 6.643 2.430 6.643 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.789 6.641 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.251 5.456 arnoldi_extremal 4 6.8 0.000 0.001 5.139 5.157 arnoldi_normal_ev 4 7.8 0.001 0.007 5.139 5.157 build_subspace 16 8.4 0.014 0.020 4.830 4.838 calculate_norms 1584 9.8 4.266 4.587 4.266 4.587 mp_allgather_i34 111 8.7 1.474 3.900 1.474 3.900 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.461 3.793 dbcsr_matrix_vector_mult_local 304 10.0 3.044 3.624 3.046 3.626 ls_scf_post 1 4.0 0.002 0.015 3.552 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.292 3.371 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.740 2.843 dbcsr_sort_data 325 11.1 1.871 2.205 1.871 2.205 make_images_sizes 222 9.7 0.000 0.000 1.085 2.194 mp_alltoall_i44 222 10.7 1.084 2.193 1.084 2.193 make_images_pack 222 9.7 1.811 2.128 1.814 2.130 dbcsr_data_release 9322 10.9 1.308 2.013 1.308 2.013 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.176000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2166.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 3.310981E+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 0.062 0.095 97.749 97.749 qs_energies 1 2.0 0.000 0.000 97.027 97.035 ls_scf 1 3.0 0.000 0.001 95.386 95.393 dbcsr_multiply_generic 111 6.7 0.017 0.022 79.080 79.348 ls_scf_main 1 4.0 0.000 0.001 59.312 59.313 multiply_cannon 111 7.7 0.050 0.141 51.688 55.578 density_matrix_trs4 2 5.0 0.002 0.003 53.229 53.426 multiply_cannon_loop 111 8.7 0.115 0.125 46.741 50.628 ls_scf_init_scf 1 4.0 0.000 0.001 32.843 32.845 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.612 31.675 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.136 29.150 mp_waitall_1 6438 11.0 22.906 28.448 22.906 28.448 multiply_cannon_multrec 1332 9.7 14.171 17.150 21.996 24.795 make_m2s 222 7.7 0.007 0.008 21.139 22.474 make_images 222 8.7 3.132 3.580 21.088 22.425 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.160 17.741 make_images_data 222 9.7 0.004 0.004 11.790 13.376 hybrid_alltoall_any 227 10.6 0.798 3.777 11.157 12.832 mp_sum_l 807 5.4 4.702 9.054 4.702 9.054 dbcsr_mm_accdrv_process 3641 10.4 0.186 0.372 7.461 8.994 dbcsr_mm_accdrv_process_sort 3641 11.4 7.088 8.572 7.088 8.572 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.630 7.258 multiply_cannon_sync_h2d 1332 9.7 5.514 6.183 5.514 6.183 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.102 6.025 mp_irecv_dv 3229 10.9 2.078 5.944 2.078 5.944 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.786 5.522 arnoldi_extremal 4 6.8 0.001 0.003 5.317 5.333 arnoldi_normal_ev 4 7.8 0.002 0.008 5.317 5.331 build_subspace 16 8.4 0.015 0.023 4.963 4.972 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.593 4.835 calculate_norms 2376 9.8 4.176 4.570 4.176 4.570 mp_allgather_i34 111 8.7 2.070 4.139 2.070 4.139 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.632 3.919 dbcsr_matrix_vector_mult_local 304 10.0 3.229 3.707 3.231 3.709 dbcsr_sort_data 658 11.4 3.081 3.455 3.081 3.455 ls_scf_post 1 4.0 0.000 0.001 3.231 3.240 dbcsr_special_finalize 555 9.7 0.006 0.007 2.825 3.191 dbcsr_merge_single_wm 555 10.7 0.535 0.657 2.817 3.182 ls_scf_store_result 1 5.0 0.000 0.000 2.982 3.033 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.916 2.970 dbcsr_data_release 10477 10.7 1.596 2.414 1.596 2.414 dbcsr_finalize 304 7.8 0.049 0.061 1.798 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.749000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2714.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 4.637196E+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 0.145 0.156 99.413 99.413 qs_energies 1 2.0 0.000 0.000 98.469 98.475 ls_scf 1 3.0 0.000 0.000 96.543 96.548 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.230 78.438 ls_scf_main 1 4.0 0.000 0.000 62.160 62.160 multiply_cannon 111 7.7 0.102 0.193 55.176 59.858 density_matrix_trs4 2 5.0 0.002 0.003 55.139 55.208 multiply_cannon_loop 111 8.7 0.069 0.076 50.774 53.214 mp_waitall_1 5481 11.0 26.878 32.484 26.878 32.484 ls_scf_init_scf 1 4.0 0.000 0.000 30.758 30.763 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.561 29.609 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.401 27.408 multiply_cannon_multrec 444 9.7 13.969 16.310 20.983 23.929 make_m2s 222 7.7 0.004 0.005 17.971 20.364 make_images 222 8.7 3.712 4.397 17.909 20.305 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.166 16.191 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.222 15.479 make_images_data 222 9.7 0.003 0.004 10.288 12.561 hybrid_alltoall_any 227 10.6 0.789 3.773 10.024 12.452 multiply_cannon_sync_h2d 444 9.7 6.516 8.111 6.516 8.111 dbcsr_mm_accdrv_process 3003 10.4 0.167 0.342 6.724 7.855 dbcsr_mm_accdrv_process_sort 3003 11.4 6.402 7.507 6.402 7.507 mp_allgather_i34 111 8.7 2.640 6.967 2.640 6.967 arnoldi_extremal 4 6.8 0.000 0.000 5.862 5.870 arnoldi_normal_ev 4 7.8 0.002 0.005 5.862 5.870 build_subspace 16 8.4 0.015 0.020 5.457 5.469 mp_sum_l 807 5.4 3.124 4.888 3.124 4.888 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.593 4.785 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.194 4.388 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.690 4.301 mp_irecv_dv 1241 11.2 1.669 4.277 1.669 4.277 dbcsr_matrix_vector_mult_local 304 10.0 3.712 4.169 3.714 4.171 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.241 3.728 calculate_norms 792 9.8 3.520 3.644 3.520 3.644 make_images_sizes 222 9.7 0.000 0.000 1.046 3.642 mp_alltoall_i44 222 10.7 1.046 3.641 1.046 3.641 ls_scf_post 1 4.0 0.000 0.000 3.625 3.631 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.409 3.509 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.449 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.259 dbcsr_merge_all 275 8.9 0.474 0.524 2.049 2.105 dbcsr_data_release 10123 10.8 1.332 2.010 1.332 2.010 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.413000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3595.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/848fd4fc57d18f38d66960eafe7959570f56e303_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 8.729641E+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 0.081 0.099 107.855 107.856 qs_energies 1 2.0 0.000 0.000 106.390 106.404 ls_scf 1 3.0 0.000 0.000 103.467 103.481 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.705 77.818 ls_scf_main 1 4.0 0.000 0.000 65.912 65.913 density_matrix_trs4 2 5.0 0.002 0.004 57.015 57.068 multiply_cannon 111 7.7 0.146 0.257 50.161 51.929 multiply_cannon_loop 111 8.7 0.067 0.070 46.681 47.806 ls_scf_init_scf 1 4.0 0.000 0.000 33.886 33.887 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.548 32.560 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.713 29.724 mp_waitall_1 4569 11.1 22.447 26.270 22.447 26.270 make_m2s 222 7.7 0.005 0.005 23.908 24.928 make_images 222 8.7 4.575 4.947 23.802 24.821 multiply_cannon_multrec 444 9.7 17.893 18.606 22.495 23.031 hybrid_alltoall_any 227 10.6 1.658 3.620 12.883 15.633 make_images_data 222 9.7 0.003 0.003 13.142 15.589 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.723 11.275 multiply_cannon_sync_h2d 444 9.7 8.843 8.892 8.843 8.892 arnoldi_extremal 4 6.8 0.000 0.000 7.391 7.405 arnoldi_normal_ev 4 7.8 0.003 0.009 7.391 7.405 build_subspace 16 8.4 0.026 0.036 6.831 6.843 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.476 5.623 dbcsr_matrix_vector_mult_local 304 10.0 5.056 5.333 5.059 5.335 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.067 5.303 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.831 4.927 dbcsr_mm_accdrv_process 1814 10.4 0.255 0.316 4.424 4.540 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.247 4.120 4.247 ls_scf_post 1 4.0 0.000 0.000 3.668 3.682 make_images_sizes 222 9.7 0.000 0.000 1.450 3.605 mp_alltoall_i44 222 10.7 1.450 3.604 1.450 3.604 mp_allgather_i34 111 8.7 1.138 3.547 1.138 3.547 ls_scf_store_result 1 5.0 0.000 0.000 3.420 3.428 calculate_norms 792 9.8 3.244 3.281 3.244 3.281 dbcsr_finalize 304 7.8 0.082 0.089 3.080 3.136 dbcsr_merge_all 275 8.9 0.891 0.919 2.866 2.912 dbcsr_complete_redistribute 5 7.6 1.434 1.473 2.769 2.911 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.893 2.893 dbcsr_data_release 12724 10.6 2.334 2.847 2.334 2.847 matrix_ls_to_qs 2 6.0 0.000 0.000 2.413 2.562 dbcsr_sort_data 325 11.1 2.439 2.493 2.439 2.493 dbcsr_new_transposed 4 7.5 0.244 0.251 2.242 2.251 dbcsr_frobenius_norm 74 6.6 2.056 2.129 2.190 2.234 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.221 2.223 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.205 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.856000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6919.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 848fd4fc57d18f38d66960eafe7959570f56e303 Summary: empty Status: OK