=== This is the CP2K Performance-Test === Updating 5c163be6b..16f0f4032 Fast-forward src/emd/rt_hfx_utils.F | 9 +++++---- src/rt_propagation_velocity_gauge.F | 2 -- 2 files changed, 5 insertions(+), 6 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 16f0f4032911ae89ed9404c17c23edc8a547187d ################# 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/01 job id: 42038490 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/02 job id: 42038491 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/03 job id: 42038493 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/04 job id: 42038501 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/05 job id: 42038510 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/06 job id: 42038512 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/07 job id: 42038513 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/08 job id: 42038514 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/09 job id: 42038515 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/10 job id: 42038516 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/11 job id: 42038517 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/12 job id: 42038519 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/13 job id: 42038520 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/14 job id: 42038521 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/15 job id: 42038522 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/16 job id: 42038523 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/17 job id: 42038524 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/18 job id: 42038525 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/19 job id: 42038526 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/20 job id: 42038527 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/21 job id: 42038529 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/22 job id: 42038530 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/23 job id: 42038531 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/24 job id: 42038532 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/25 job id: 42038533 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/26 job id: 42038534 --- 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/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.036 133.677 133.678 farming_run 1 2.0 133.130 133.136 133.644 133.647 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.464902E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 115.317 115.318 qs_energies 1 2.0 0.000 0.000 115.083 115.086 mp2_main 1 3.0 0.000 0.000 113.130 113.133 mp2_gpw_main 1 4.0 0.036 0.044 112.236 112.240 mp2_ri_gpw_compute_in 1 5.0 0.176 0.179 93.453 93.941 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.377 55.868 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.169 41.682 47.080 get_2c_integrals 1 6.0 0.000 0.001 37.523 37.898 integrate_v_rspace 273 8.0 0.439 0.452 25.069 30.220 pw_transfer 6555 10.6 0.376 0.390 27.379 28.234 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 25.993 26.702 grid_integrate_task_list 273 9.0 20.883 26.513 20.883 26.513 fft_wrap_pw1pw2_100 2178 12.4 1.224 1.435 23.560 24.261 compute_2c_integrals 1 7.0 0.003 0.003 19.815 19.816 compute_2c_integrals_loop_lm 1 8.0 0.009 0.009 18.995 19.494 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.445 18.986 19.486 rpa_ri_compute_en 1 5.0 0.013 0.015 18.655 18.842 cp_fm_cholesky_decompose 12 8.2 17.745 18.098 17.745 18.098 cholesky_decomp 1 7.0 0.000 0.000 16.561 16.937 fft3d_s 5443 13.4 16.137 16.639 16.159 16.661 ao_to_mo_and_store_B_mult_1 272 7.0 10.857 15.581 10.857 15.581 calculate_wavefunction 272 8.0 5.405 5.561 12.526 13.185 rpa_num_int 1 6.0 0.000 0.006 10.585 10.595 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.548 10.563 calc_mat_Q 8 8.0 0.000 0.000 9.314 9.410 contract_S_to_Q 8 9.0 0.000 0.000 8.736 8.828 calc_potential_gpw 544 9.5 0.005 0.006 8.300 8.698 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.278 8.573 potential_pw2rs 545 10.0 0.108 0.110 7.711 8.481 parallel_gemm_fm 14 9.1 0.000 0.000 8.305 8.388 parallel_gemm_fm_cosma 14 10.1 8.305 8.388 8.305 8.388 collocate_single_gaussian 272 10.0 0.040 0.042 7.497 7.793 create_integ_mat 1 6.0 0.014 0.028 7.470 7.479 array2fm 1 7.0 0.000 0.000 6.594 7.018 pw_scatter_s 2720 13.7 4.413 4.586 4.413 4.586 pw_gather_s 2722 13.2 3.854 4.236 3.854 4.236 array2fm_buffer_send 1 8.0 2.946 3.145 2.946 3.145 pw_poisson_solve 545 10.5 1.134 1.189 2.225 2.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.239869, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 0.039 396.022 396.024 farming_run 1 2.0 395.228 395.235 395.976 395.979 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.220592E+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 0.014 0.042 209.234 209.234 qs_energies 1 2.0 0.000 0.000 209.019 209.023 scf_env_do_scf 1 3.0 0.000 0.000 106.221 106.221 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.234 105.242 rebuild_ks_matrix 4 6.0 0.000 0.000 105.232 105.241 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.232 105.240 hfx_ks_matrix 4 8.0 0.001 0.001 104.848 104.851 integrate_four_center 4 9.0 0.144 0.453 104.847 104.851 mp2_main 1 3.0 0.000 0.000 102.519 102.523 mp2_gpw_main 1 4.0 0.032 0.047 101.577 101.585 integrate_four_center_main 4 10.0 0.087 0.449 96.613 98.824 integrate_four_center_bin 266 11.0 96.525 98.817 96.525 98.817 init_scf_loop 1 4.0 0.000 0.000 91.857 91.858 mp2_ri_gpw_compute_in 1 5.0 0.065 0.091 74.913 75.931 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.558 55.578 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 42.110 47.321 integrate_v_rspace 95 8.0 0.398 0.570 28.484 33.532 pw_transfer 2240 10.6 0.148 0.167 29.875 30.365 ao_to_mo_and_store_B_mult_1 91 7.0 10.763 29.343 10.763 29.343 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.897 29.343 grid_integrate_task_list 95 9.0 23.760 29.000 23.760 29.000 mp2_ri_gpw_compute_en 1 5.0 0.057 0.068 26.507 28.170 fft_wrap_pw1pw2_100 730 12.4 1.258 1.410 26.637 27.114 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.927 24.833 24.845 get_2c_integrals 1 6.0 0.000 0.003 20.246 20.288 compute_2c_integrals 1 7.0 0.003 0.004 19.225 19.230 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.809 19.095 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.850 18.808 19.093 fft3d_s 1823 13.4 18.414 18.761 18.427 18.775 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.362 14.363 calculate_wavefunction 91 8.0 2.014 2.045 9.746 9.986 potential_pw2rs 186 10.0 0.033 0.035 8.599 9.221 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.583 8.712 8.967 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.204 8.493 local_gemm 172 8.0 8.157 8.418 8.157 8.418 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.525 7.885 8.301 calc_potential_gpw 182 9.5 0.002 0.002 7.927 8.131 collocate_single_gaussian 91 10.0 0.017 0.020 7.852 8.072 mp_sync 38 10.4 2.906 6.571 2.906 6.571 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.398 6.345 6.398 mp_sendrecv_dm3 2068 8.0 5.921 6.342 5.921 6.342 pw_gather_s 912 13.2 4.921 5.437 4.921 5.437 pw_scatter_s 910 13.7 3.938 4.233 3.938 4.233 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.563658, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 450.379776E+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 0.028 0.054 57.163 57.165 qs_mol_dyn_low 1 2.0 0.003 0.003 56.833 56.840 qs_forces 11 3.9 0.002 0.002 56.771 56.772 qs_energies 11 4.9 0.001 0.002 55.338 55.350 scf_env_do_scf 11 5.9 0.000 0.001 44.555 44.556 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.434 42.435 dbcsr_multiply_generic 2286 12.5 0.092 0.098 38.661 39.138 qs_scf_new_mos 108 7.5 0.000 0.001 32.380 32.683 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.380 32.683 multiply_cannon 2286 13.5 0.184 0.200 30.957 32.457 multiply_cannon_loop 2286 14.5 1.491 1.564 30.234 31.736 ot_scf_mini 108 9.5 0.002 0.002 30.739 30.927 velocity_verlet 10 3.0 0.001 0.002 25.956 25.957 ot_mini 108 10.5 0.001 0.001 19.813 20.060 mp_waitall_1 267858 16.1 10.612 19.966 10.612 19.966 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 7.752 18.073 qs_ot_get_derivative 108 11.5 0.001 0.001 16.839 17.006 multiply_cannon_multrec 54864 15.5 4.180 6.492 7.617 11.057 init_scf_run 11 5.9 0.000 0.001 9.546 9.546 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.546 9.546 rebuild_ks_matrix 119 8.3 0.000 0.000 7.837 7.986 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.837 7.986 multiply_cannon_sync_h2d 54864 15.5 5.780 7.729 5.780 7.729 mp_sum_l 7207 12.9 5.691 7.439 5.691 7.439 calculate_first_density_matrix 1 7.0 0.001 0.001 7.328 7.339 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.072 7.073 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.922 7.060 acc_transpose_blocks 54864 15.5 0.229 0.246 4.963 7.042 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.793 6.847 qs_ot_get_p 119 10.4 0.001 0.001 6.287 6.554 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.711 6.168 acc_transpose_blocks_kernels 54864 16.5 0.257 0.386 3.966 5.894 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.598 5.721 jit_kernel_transpose 5 15.5 3.709 5.513 3.709 5.513 dbcsr_mm_accdrv_process 76910 16.1 1.172 1.810 3.360 4.818 sum_up_and_integrate 119 10.3 0.012 0.014 4.493 4.500 integrate_v_rspace 119 11.3 0.002 0.002 4.481 4.489 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.152 4.280 calculate_rho_elec 119 8.7 0.011 0.016 4.152 4.280 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.443 3.490 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.943 3.449 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.909 3.087 apply_single 119 13.6 0.000 0.000 2.909 3.087 calculate_dm_sparse 119 9.5 0.000 0.000 2.928 3.086 jit_kernel_multiply 13 15.8 2.126 2.957 2.126 2.957 rs_pw_transfer 974 11.9 0.012 0.013 2.776 2.887 ot_diis_step 108 11.5 0.006 0.006 2.704 2.704 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.601 2.601 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.482 2.553 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.389 2.389 cp_fm_redistribute_end 50 14.0 2.170 2.367 2.174 2.368 cp_fm_diag_elpa_base 50 14.0 0.192 2.321 0.193 2.333 density_rs2pw 119 9.7 0.004 0.004 2.193 2.296 wfi_extrapolate 11 7.9 0.001 0.001 2.151 2.152 grid_integrate_task_list 119 12.3 2.022 2.121 2.022 2.121 init_scf_loop 11 6.9 0.000 0.000 2.101 2.102 mp_sum_d 4129 12.0 1.329 1.978 1.329 1.978 potential_pw2rs 119 12.3 0.004 0.004 1.847 1.864 make_m2s 4572 13.5 0.054 0.056 1.707 1.750 pw_transfer 1439 11.6 0.051 0.056 1.661 1.732 make_images 4572 14.5 0.132 0.138 1.625 1.667 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.586 1.658 mp_alltoall_d11v 2130 13.8 1.413 1.558 1.413 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.447 1.467 fft3d_ps 1201 14.6 0.358 0.460 1.363 1.427 grid_collocate_task_list 119 9.7 1.286 1.350 1.286 1.350 mp_waitany 12084 13.8 1.214 1.343 1.214 1.343 fft_wrap_pw1pw2_140 487 13.2 0.081 0.094 1.203 1.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.165000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 487.108608E+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 1032117. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 44.121 44.123 qs_mol_dyn_low 1 2.0 0.003 0.003 43.886 43.893 qs_forces 11 3.9 0.002 0.003 43.826 43.827 qs_energies 11 4.9 0.001 0.002 42.130 42.133 scf_env_do_scf 11 5.9 0.000 0.001 32.152 32.153 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.566 29.566 dbcsr_multiply_generic 2286 12.5 0.100 0.102 26.698 27.081 multiply_cannon 2286 13.5 0.209 0.219 21.395 23.000 multiply_cannon_loop 2286 14.5 0.904 0.979 20.199 21.661 qs_scf_new_mos 108 7.5 0.001 0.001 20.378 20.623 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.378 20.623 ot_scf_mini 108 9.5 0.002 0.003 19.468 19.639 velocity_verlet 10 3.0 0.001 0.002 18.728 18.729 mp_waitall_1 217478 16.2 8.039 16.338 8.039 16.338 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 5.875 14.428 ot_mini 108 10.5 0.001 0.001 12.155 12.391 qs_ot_get_derivative 108 11.5 0.001 0.001 9.667 9.843 multiply_cannon_multrec 27432 15.5 1.970 4.555 5.926 8.943 init_scf_run 11 5.9 0.000 0.001 8.749 8.749 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.749 8.749 rebuild_ks_matrix 119 8.3 0.000 0.000 7.245 7.390 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.245 7.389 calculate_first_density_matrix 1 7.0 0.000 0.001 7.202 7.204 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.687 6.690 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.398 6.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.471 6.512 acc_transpose_blocks 27432 15.5 0.109 0.113 4.442 6.336 dbcsr_mm_accdrv_process 47894 16.0 3.151 5.099 3.888 5.734 acc_transpose_blocks_kernels 27432 16.5 0.184 0.272 3.905 5.714 jit_kernel_transpose 5 15.5 3.721 5.446 3.721 5.446 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.746 4.583 qs_ot_get_p 119 10.4 0.001 0.001 4.318 4.553 sum_up_and_integrate 119 10.3 0.024 0.027 4.282 4.289 integrate_v_rspace 119 11.3 0.002 0.002 4.258 4.268 mp_sum_l 7207 12.9 2.199 4.220 2.199 4.220 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.069 4.154 apply_single 119 13.6 0.000 0.000 3.069 4.154 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.731 3.779 calculate_rho_elec 119 8.7 0.021 0.024 3.731 3.778 make_m2s 4572 13.5 0.052 0.054 2.662 2.910 rs_pw_transfer 974 11.9 0.010 0.011 2.752 2.847 make_images 4572 14.5 0.201 0.238 2.574 2.820 multiply_cannon_sync_h2d 27432 15.5 2.163 2.776 2.163 2.776 qs_ot_p2m_diag 50 11.0 0.008 0.013 2.738 2.757 init_scf_loop 11 6.9 0.000 0.000 2.559 2.560 ot_diis_step 108 11.5 0.010 0.011 2.435 2.436 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.274 2.364 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.288 2.288 calculate_dm_sparse 119 9.5 0.000 0.000 2.159 2.230 density_rs2pw 119 9.7 0.004 0.004 2.092 2.178 potential_pw2rs 119 12.3 0.006 0.007 2.033 2.044 grid_integrate_task_list 119 12.3 1.834 1.926 1.834 1.926 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.918 1.918 cp_fm_redistribute_end 50 14.0 1.581 1.892 1.585 1.893 cp_fm_diag_elpa_base 50 14.0 0.299 1.827 0.306 1.857 pw_transfer 1439 11.6 0.063 0.067 1.814 1.849 jit_kernel_multiply 9 16.1 0.684 1.811 0.684 1.811 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.725 1.759 make_images_data 4572 15.5 0.045 0.051 1.213 1.682 prepare_preconditioner 11 7.9 0.000 0.000 1.650 1.677 make_preconditioner 11 8.9 0.000 0.000 1.650 1.677 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.547 1.604 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.050 1.567 wfi_extrapolate 11 7.9 0.001 0.001 1.498 1.498 fft3d_ps 1201 14.6 0.496 0.547 1.428 1.463 mp_allgather_i34 2286 14.5 0.639 1.419 0.639 1.419 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.360 1.367 mp_alltoall_d11v 2130 13.8 1.247 1.361 1.247 1.361 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.259 1.306 grid_collocate_task_list 119 9.7 1.243 1.304 1.243 1.304 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.265 1.297 mp_sum_d 4129 12.0 0.596 1.038 0.596 1.038 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.939 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.905 0.919 make_images_sizes 4572 15.5 0.005 0.005 0.635 0.899 mp_alltoall_i44 4572 16.5 0.630 0.894 0.630 0.894 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=44.123000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.636364, yerr=1.298442 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 521.580544E+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 0.034 39.182 39.183 qs_mol_dyn_low 1 2.0 0.003 0.003 38.828 38.836 qs_forces 11 3.9 0.002 0.002 38.750 38.751 qs_energies 11 4.9 0.001 0.001 37.157 37.160 scf_env_do_scf 11 5.9 0.000 0.001 27.298 27.299 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.475 24.476 dbcsr_multiply_generic 2286 12.5 0.093 0.095 22.183 22.327 multiply_cannon 2286 13.5 0.195 0.200 18.370 19.176 multiply_cannon_loop 2286 14.5 0.640 0.669 17.472 18.321 velocity_verlet 10 3.0 0.010 0.026 16.295 16.296 qs_scf_new_mos 108 7.5 0.001 0.001 16.055 16.069 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.054 16.069 ot_scf_mini 108 9.5 0.002 0.003 15.293 15.302 ot_mini 108 10.5 0.001 0.001 9.440 9.460 init_scf_run 11 5.9 0.000 0.001 8.652 8.652 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.652 8.652 qs_ot_get_derivative 108 11.5 0.001 0.001 7.845 7.857 calculate_first_density_matrix 1 7.0 0.000 0.001 7.280 7.281 multiply_cannon_multrec 18288 15.5 1.929 2.913 6.806 7.156 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.608 6.609 rebuild_ks_matrix 119 8.3 0.000 0.000 6.564 6.582 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.564 6.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.405 6.420 acc_transpose_blocks 18288 15.5 0.076 0.078 6.165 6.185 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.802 5.818 acc_transpose_blocks_kernels 18288 16.5 0.212 0.222 5.720 5.732 dbcsr_mm_accdrv_process 38222 16.0 4.090 5.517 4.794 5.574 jit_kernel_transpose 5 15.6 5.508 5.514 5.508 5.514 mp_waitall_1 169478 16.3 3.694 4.662 3.694 4.662 sum_up_and_integrate 119 10.3 0.029 0.030 4.169 4.172 integrate_v_rspace 119 11.3 0.002 0.003 4.139 4.146 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.978 3.602 qs_ot_get_p 119 10.4 0.001 0.001 3.567 3.594 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.533 3.538 calculate_rho_elec 119 8.7 0.030 0.031 3.532 3.538 init_scf_loop 11 6.9 0.000 0.000 2.802 2.803 rs_pw_transfer 974 11.9 0.009 0.010 2.665 2.776 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.636 2.452 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.071 2.427 apply_single 119 13.6 0.000 0.000 2.071 2.427 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.299 2.306 density_rs2pw 119 9.7 0.004 0.004 2.119 2.231 make_m2s 4572 13.5 0.044 0.045 2.100 2.217 make_images 4572 14.5 0.192 0.205 2.015 2.131 calculate_dm_sparse 119 9.5 0.000 0.000 2.007 2.019 prepare_preconditioner 11 7.9 0.000 0.000 1.997 2.000 make_preconditioner 11 8.9 0.000 0.000 1.997 2.000 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.997 1.998 potential_pw2rs 119 12.3 0.007 0.008 1.934 1.944 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.846 1.931 jit_kernel_multiply 10 16.0 0.651 1.929 0.651 1.929 grid_integrate_task_list 119 12.3 1.806 1.897 1.806 1.897 pw_transfer 1439 11.6 0.063 0.066 1.824 1.834 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.790 1.795 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.734 1.745 mp_sum_l 7207 12.9 1.327 1.727 1.327 1.727 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.704 1.706 multiply_cannon_sync_h2d 18288 15.5 1.343 1.687 1.343 1.687 cp_fm_redistribute_end 50 14.0 1.261 1.673 1.262 1.674 cp_fm_diag_elpa_base 50 14.0 0.394 1.597 0.408 1.635 ot_diis_step 108 11.5 0.011 0.011 1.555 1.555 fft3d_ps 1201 14.6 0.505 0.522 1.420 1.431 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.342 1.348 wfi_extrapolate 11 7.9 0.001 0.001 1.324 1.324 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.275 1.285 grid_collocate_task_list 119 9.7 1.204 1.277 1.204 1.277 make_images_data 4572 15.5 0.045 0.048 0.911 1.079 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.974 0.998 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.766 0.960 mp_alltoall_d11v 2130 13.8 0.796 0.949 0.796 0.949 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.942 0.943 cp_fm_cholesky_invert 11 10.9 0.909 0.914 0.909 0.914 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.835 0.839 mp_alltoall_z22v 1201 16.6 0.739 0.801 0.739 0.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.183000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.363636, yerr=2.057038 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 553.517056E+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 0.027 0.044 35.208 35.209 qs_mol_dyn_low 1 2.0 0.003 0.003 34.967 34.974 qs_forces 11 3.9 0.002 0.002 34.908 34.910 qs_energies 11 4.9 0.004 0.021 33.200 33.206 scf_env_do_scf 11 5.9 0.001 0.001 28.162 28.163 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.610 24.610 dbcsr_multiply_generic 2286 12.5 0.100 0.104 18.232 18.409 velocity_verlet 10 3.0 0.001 0.002 17.813 17.815 qs_scf_new_mos 108 7.5 0.001 0.001 16.328 16.376 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.328 16.376 ot_scf_mini 108 9.5 0.003 0.004 15.384 15.433 multiply_cannon 2286 13.5 0.231 0.259 14.524 14.838 multiply_cannon_loop 2286 14.5 0.937 0.971 13.572 13.940 ot_mini 108 10.5 0.001 0.001 9.480 9.543 multiply_cannon_multrec 27432 15.5 2.343 3.023 8.588 8.944 qs_ot_get_derivative 108 11.5 0.001 0.001 7.661 7.712 dbcsr_mm_accdrv_process 47916 15.9 5.274 7.306 6.151 7.363 rebuild_ks_matrix 119 8.3 0.000 0.000 6.552 6.601 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.552 6.600 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.831 5.873 sum_up_and_integrate 119 10.3 0.034 0.037 3.829 3.838 integrate_v_rspace 119 11.3 0.002 0.003 3.795 3.804 init_scf_run 11 5.9 0.000 0.001 3.651 3.653 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.651 3.653 init_scf_loop 11 6.9 0.000 0.002 3.531 3.531 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.346 3.377 calculate_rho_elec 119 8.7 0.040 0.046 3.346 3.376 qs_ot_get_p 119 10.4 0.001 0.001 3.293 3.369 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.689 3.149 mp_waitall_1 145218 16.4 2.336 3.048 2.336 3.048 prepare_preconditioner 11 7.9 0.000 0.000 2.678 2.687 make_preconditioner 11 8.9 0.000 0.000 2.678 2.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.274 2.614 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.118 2.565 apply_single 119 13.6 0.000 0.000 2.118 2.565 make_m2s 4572 13.5 0.054 0.056 2.241 2.361 calculate_first_density_matrix 1 7.0 0.001 0.003 2.281 2.285 make_images 4572 14.5 0.275 0.336 2.134 2.252 rs_pw_transfer 974 11.9 0.009 0.009 2.109 2.221 calculate_dm_sparse 119 9.5 0.000 0.000 2.063 2.116 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.011 2.020 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.958 1.983 density_rs2pw 119 9.7 0.003 0.004 1.835 1.939 grid_integrate_task_list 119 12.3 1.828 1.899 1.828 1.899 pw_transfer 1439 11.6 0.063 0.066 1.811 1.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.816 1.817 jit_kernel_multiply 10 15.9 0.816 1.817 0.816 1.817 ot_diis_step 108 11.5 0.012 0.012 1.779 1.779 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.722 1.754 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.691 1.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.621 1.636 potential_pw2rs 119 12.3 0.008 0.009 1.603 1.610 mp_sum_l 7207 12.9 1.050 1.561 1.050 1.561 acc_transpose_blocks 27432 15.5 0.113 0.116 1.477 1.513 multiply_cannon_metrocomm3 27432 15.5 0.037 0.039 0.845 1.464 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.451 1.452 fft3d_ps 1201 14.6 0.537 0.591 1.401 1.425 fft_wrap_pw1pw2_140 487 13.2 0.085 0.093 1.389 1.425 cp_fm_redistribute_end 50 14.0 0.956 1.424 0.957 1.425 cp_fm_diag_elpa_base 50 14.0 0.444 1.362 0.465 1.401 wfi_extrapolate 11 7.9 0.001 0.001 1.322 1.322 grid_collocate_task_list 119 9.7 1.221 1.314 1.221 1.314 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.240 1.250 cp_fm_upper_to_full 72 13.5 0.829 1.155 0.829 1.155 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.106 1.126 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.107 1.109 dbcsr_complete_redistribute 329 12.2 0.123 0.146 0.818 1.105 multiply_cannon_sync_h2d 27432 15.5 0.981 1.039 0.981 1.039 make_images_data 4572 15.5 0.045 0.049 0.890 1.031 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.772 0.937 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.598 0.875 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.869 acc_transpose_blocks_kernels 27432 16.5 0.272 0.278 0.844 0.866 mp_alltoall_d11v 2130 13.8 0.759 0.862 0.759 0.862 cp_fm_cholesky_invert 11 10.9 0.834 0.837 0.834 0.837 multiply_cannon_metrocomm1 27432 15.5 0.033 0.034 0.229 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.814 mp_alltoall_z22v 1201 16.6 0.746 0.771 0.746 0.771 mp_alltoall_i22 627 13.8 0.443 0.751 0.443 0.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.209000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.454545, yerr=3.367320 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 613.031936E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 28.471 28.472 qs_mol_dyn_low 1 2.0 0.003 0.003 28.273 28.281 qs_forces 11 3.9 0.002 0.002 28.067 28.069 qs_energies 11 4.9 0.001 0.001 26.341 26.345 scf_env_do_scf 11 5.9 0.000 0.001 21.644 21.644 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.971 18.971 velocity_verlet 10 3.0 0.002 0.002 14.823 14.835 dbcsr_multiply_generic 2286 12.5 0.091 0.093 12.407 12.485 qs_scf_new_mos 108 7.5 0.001 0.001 11.327 11.346 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.327 11.346 ot_scf_mini 108 9.5 0.002 0.002 10.629 10.654 multiply_cannon 2286 13.5 0.229 0.235 9.617 10.126 multiply_cannon_loop 2286 14.5 0.331 0.342 8.634 8.827 multiply_cannon_multrec 9144 15.5 1.577 1.814 5.748 5.976 ot_mini 108 10.5 0.001 0.001 5.924 5.953 rebuild_ks_matrix 119 8.3 0.000 0.000 5.850 5.873 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.850 5.872 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.202 5.223 qs_ot_get_derivative 108 11.5 0.001 0.001 4.582 4.608 dbcsr_mm_accdrv_process 12550 15.8 3.202 4.089 4.068 4.129 sum_up_and_integrate 119 10.3 0.037 0.040 3.584 3.590 integrate_v_rspace 119 11.3 0.003 0.003 3.547 3.553 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.281 3.289 calculate_rho_elec 119 8.7 0.060 0.061 3.280 3.289 init_scf_run 11 5.9 0.000 0.001 3.254 3.254 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.254 3.254 qs_ot_get_p 119 10.4 0.001 0.001 2.872 2.906 init_scf_loop 11 6.9 0.000 0.000 2.650 2.652 mp_waitall_1 121218 16.5 1.906 2.455 1.906 2.455 make_m2s 4572 13.5 0.034 0.035 1.939 2.126 calculate_first_density_matrix 1 7.0 0.001 0.001 2.054 2.056 make_images 4572 14.5 0.270 0.304 1.850 2.036 grid_integrate_task_list 119 12.3 1.845 1.921 1.845 1.921 prepare_preconditioner 11 7.9 0.000 0.000 1.909 1.912 make_preconditioner 11 8.9 0.000 0.000 1.909 1.912 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.887 1.889 rs_pw_transfer 974 11.9 0.008 0.008 1.756 1.843 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.795 1.822 density_rs2pw 119 9.7 0.003 0.004 1.694 1.778 calculate_dm_sparse 119 9.5 0.000 0.000 1.732 1.750 pw_transfer 1439 11.6 0.063 0.065 1.726 1.734 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.675 1.676 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.636 1.643 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.606 1.621 jit_kernel_multiply 9 15.8 0.829 1.617 0.829 1.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.534 1.536 potential_pw2rs 119 12.3 0.010 0.010 1.374 1.377 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.362 1.369 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.361 1.362 cp_fm_redistribute_end 50 14.0 0.676 1.335 0.678 1.337 grid_collocate_task_list 119 9.7 1.272 1.333 1.272 1.333 ot_diis_step 108 11.5 0.012 0.013 1.327 1.327 cp_fm_diag_elpa_base 50 14.0 0.613 1.259 0.657 1.316 fft3d_ps 1201 14.6 0.537 0.550 1.301 1.309 fft_wrap_pw1pw2_140 487 13.2 0.083 0.088 1.293 1.301 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.223 1.236 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.198 1.222 apply_single 119 13.6 0.000 0.000 1.198 1.222 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.217 1.219 make_images_data 4572 15.5 0.039 0.042 0.936 1.195 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.880 1.187 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.136 1.140 wfi_extrapolate 11 7.9 0.001 0.001 1.138 1.138 cp_fm_cholesky_invert 11 10.9 0.999 1.003 0.999 1.003 mp_alltoall_d11v 2130 13.8 0.836 0.934 0.836 0.934 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.919 acc_transpose_blocks 9144 15.5 0.039 0.040 0.896 0.904 mp_allgather_i34 2286 14.5 0.315 0.805 0.315 0.805 multiply_cannon_sync_h2d 9144 15.5 0.705 0.786 0.705 0.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.774 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.766 0.773 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.675 0.723 mp_alltoall_z22v 1201 16.6 0.638 0.672 0.638 0.672 acc_transpose_blocks_kernels 9144 16.5 0.118 0.122 0.653 0.657 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.335 0.630 dbcsr_complete_redistribute 329 12.2 0.159 0.170 0.579 0.617 mp_waitany 5200 13.7 0.477 0.578 0.477 0.578 qs_create_task_list 11 7.9 0.001 0.001 0.548 0.572 generate_qs_task_list 11 8.9 0.191 0.213 0.548 0.571 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.472000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=580.181818, yerr=6.087516 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 752.943104E+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 0.017 0.036 42.718 42.718 qs_mol_dyn_low 1 2.0 0.003 0.003 42.486 42.499 qs_forces 11 3.9 0.002 0.002 42.425 42.425 qs_energies 11 4.9 0.002 0.002 40.445 40.448 scf_env_do_scf 11 5.9 0.001 0.001 34.581 34.582 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.249 26.251 velocity_verlet 10 3.0 0.002 0.002 23.894 23.900 dbcsr_multiply_generic 2286 12.5 0.101 0.103 18.229 18.345 qs_scf_new_mos 108 7.5 0.001 0.001 17.017 17.109 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.017 17.108 ot_scf_mini 108 9.5 0.002 0.002 15.899 15.995 multiply_cannon 2286 13.5 0.304 0.312 13.813 14.754 multiply_cannon_loop 2286 14.5 0.342 0.350 12.490 13.416 ot_mini 108 10.5 0.001 0.001 9.394 9.511 multiply_cannon_multrec 9144 15.5 3.487 4.928 8.627 8.780 init_scf_loop 11 6.9 0.000 0.000 8.300 8.303 qs_ot_get_derivative 108 11.5 0.001 0.001 7.307 7.404 prepare_preconditioner 11 7.9 0.000 0.000 7.373 7.386 make_preconditioner 11 8.9 0.000 0.000 7.373 7.386 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.867 7.261 rebuild_ks_matrix 119 8.3 0.000 0.000 7.002 7.144 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.002 7.143 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.338 6.466 dbcsr_mm_accdrv_process 12550 15.8 3.925 5.266 5.014 6.437 cp_fm_upper_to_full 72 14.2 3.346 4.794 3.346 4.794 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.910 3.971 calculate_rho_elec 119 8.7 0.118 0.121 3.910 3.971 mp_waitall_1 97218 16.6 2.976 3.841 2.976 3.841 init_scf_run 11 5.9 0.000 0.001 3.823 3.823 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.823 3.823 sum_up_and_integrate 119 10.3 0.064 0.065 3.816 3.821 qs_ot_get_p 119 10.4 0.001 0.001 3.652 3.791 integrate_v_rspace 119 11.3 0.003 0.003 3.752 3.759 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.586 3.094 dbcsr_complete_redistribute 329 12.2 0.286 0.293 2.122 2.976 make_m2s 4572 13.5 0.038 0.038 2.581 2.775 make_images 4572 14.5 0.356 0.386 2.461 2.655 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.765 2.624 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.302 2.541 apply_single 119 13.6 0.000 0.000 2.302 2.541 mp_alltoall_i22 627 13.8 1.553 2.396 1.553 2.396 calculate_first_density_matrix 1 7.0 0.001 0.001 2.313 2.376 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.500 2.350 calculate_dm_sparse 119 9.5 0.000 0.000 2.242 2.303 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.227 2.228 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 1.423 2.219 pw_transfer 1439 11.6 0.066 0.067 2.129 2.133 grid_integrate_task_list 119 12.3 2.039 2.058 2.039 2.058 ot_diis_step 108 11.5 0.014 0.014 2.054 2.055 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.034 2.038 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.999 1.999 mp_sum_l 7207 12.9 1.300 1.994 1.300 1.994 density_rs2pw 119 9.7 0.003 0.003 1.912 1.953 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.938 1.939 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.849 1.898 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.755 1.755 jit_kernel_multiply 10 15.7 1.061 1.727 1.061 1.727 cp_fm_cholesky_invert 11 10.9 1.706 1.710 1.706 1.710 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.688 fft3d_ps 1201 14.6 0.569 0.579 1.667 1.671 fft_wrap_pw1pw2_140 487 13.2 0.088 0.089 1.655 1.660 rs_pw_transfer 974 11.9 0.009 0.009 1.544 1.604 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.601 1.603 cp_fm_diag_elpa_base 50 14.0 1.447 1.505 1.599 1.601 hybrid_alltoall_any 4725 16.4 0.088 0.150 1.266 1.554 make_images_data 4572 15.5 0.043 0.045 1.249 1.521 grid_collocate_task_list 119 9.7 1.471 1.499 1.471 1.499 wfi_extrapolate 11 7.9 0.001 0.001 1.393 1.393 mp_alltoall_d11v 2130 13.8 1.297 1.390 1.297 1.390 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.280 1.339 potential_pw2rs 119 12.3 0.014 0.015 1.312 1.314 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.202 1.226 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.093 1.106 multiply_cannon_sync_h2d 9144 15.5 1.040 1.042 1.040 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.991 1.040 mp_alltoall_z22v 1201 16.6 0.964 0.983 0.964 0.983 qs_create_task_list 11 7.9 0.001 0.001 0.942 0.953 generate_qs_task_list 11 8.9 0.373 0.392 0.941 0.952 acc_transpose_blocks 9144 15.5 0.039 0.039 0.898 0.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.878 0.892 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.825 0.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.718000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=709.545455, yerr=12.999046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 500.219904E+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 2451464. 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 0.013 0.031 82.509 82.510 qs_mol_dyn_low 1 2.0 0.003 0.003 82.272 82.283 qs_forces 11 3.9 0.002 0.003 82.196 82.197 qs_energies 11 4.9 0.001 0.002 79.339 79.357 scf_env_do_scf 11 5.9 0.000 0.001 70.573 70.575 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.974 64.974 dbcsr_multiply_generic 2055 12.4 0.115 0.118 52.271 52.556 qs_scf_new_mos 99 7.5 0.000 0.001 48.089 48.203 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.088 48.203 ot_scf_mini 99 9.5 0.002 0.002 45.676 45.796 multiply_cannon 2055 13.4 0.205 0.222 42.643 43.617 multiply_cannon_loop 2055 14.4 1.550 1.587 41.623 42.695 velocity_verlet 10 3.0 0.001 0.002 41.701 41.701 ot_mini 99 10.5 0.001 0.001 27.800 27.914 qs_ot_get_derivative 99 11.5 0.001 0.001 20.979 21.105 multiply_cannon_multrec 49320 15.4 12.121 12.907 17.136 17.834 rebuild_ks_matrix 110 8.3 0.000 0.001 14.427 14.626 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.426 14.625 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.667 12.844 mp_waitall_1 241148 16.1 11.877 12.749 11.877 12.749 multiply_cannon_sync_h2d 49320 15.4 9.914 10.740 9.914 10.740 qs_ot_get_p 110 10.4 0.001 0.001 9.323 9.444 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.051 8.339 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.206 7.697 apply_single 110 13.6 0.000 0.001 7.206 7.697 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.033 7.588 sum_up_and_integrate 110 10.3 0.037 0.044 7.016 7.031 integrate_v_rspace 110 11.3 0.003 0.003 6.979 7.003 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.935 7.002 init_scf_run 11 5.9 0.000 0.001 6.754 6.754 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.753 6.754 ot_diis_step 99 11.5 0.005 0.006 6.557 6.557 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.149 6.281 calculate_rho_elec 110 8.6 0.020 0.024 6.149 6.280 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.084 6.115 mp_sum_l 6514 12.8 5.213 5.933 5.213 5.933 init_scf_loop 11 6.9 0.000 0.000 5.573 5.573 dbcsr_mm_accdrv_process 87628 16.1 2.070 2.157 4.894 5.196 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.080 5.080 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.590 4.591 cp_fm_redistribute_end 48 14.0 3.999 4.566 4.003 4.567 cp_fm_diag_elpa_base 48 14.0 0.557 4.470 0.561 4.493 make_m2s 4110 13.4 0.061 0.065 3.947 4.050 wfi_extrapolate 11 7.9 0.001 0.001 3.994 3.994 rs_pw_transfer 902 11.9 0.012 0.014 3.718 3.961 make_images 4110 14.4 0.177 0.190 3.852 3.958 calculate_dm_sparse 110 9.5 0.000 0.001 3.775 3.887 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.730 3.664 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.588 3.592 prepare_preconditioner 11 7.9 0.000 0.000 3.428 3.447 make_preconditioner 11 8.9 0.000 0.000 3.428 3.447 grid_integrate_task_list 110 12.3 3.239 3.400 3.239 3.400 density_rs2pw 110 9.6 0.004 0.004 3.228 3.389 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.292 3.331 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.254 3.303 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.217 3.273 pw_transfer 1331 11.6 0.053 0.062 2.868 2.925 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.781 2.840 calculate_first_density_matrix 1 7.0 0.001 0.001 2.673 2.679 potential_pw2rs 110 12.3 0.006 0.007 2.615 2.639 jit_kernel_multiply 13 15.9 2.539 2.577 2.539 2.577 mp_alltoall_d11v 2046 13.8 2.104 2.480 2.104 2.480 fft_wrap_pw1pw2_140 451 13.1 0.176 0.195 2.334 2.398 fft3d_ps 1111 14.6 0.751 0.836 2.311 2.364 acc_transpose_blocks 49320 15.4 0.218 0.226 2.175 2.242 mp_waitany 14300 13.8 1.838 2.232 1.838 2.232 grid_collocate_task_list 110 9.6 2.088 2.205 2.088 2.205 mp_sum_d 3883 11.9 1.417 1.984 1.417 1.984 make_images_data 4110 15.4 0.043 0.046 1.789 1.923 cp_fm_cholesky_invert 11 10.9 1.918 1.922 1.918 1.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.899 1.921 hybrid_alltoall_any 4261 16.3 0.081 0.481 1.534 1.814 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.638 1.660 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.510000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.090909, yerr=2.678349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 588.865536E+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 2050137. 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 0.018 0.036 68.631 68.632 qs_mol_dyn_low 1 2.0 0.003 0.003 68.391 68.402 qs_forces 11 3.9 0.002 0.003 68.190 68.192 qs_energies 11 4.9 0.001 0.002 64.889 64.897 scf_env_do_scf 11 5.9 0.000 0.001 56.447 56.451 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.660 48.660 dbcsr_multiply_generic 2055 12.4 0.114 0.117 37.764 37.956 velocity_verlet 10 3.0 0.001 0.002 36.203 36.207 qs_scf_new_mos 99 7.5 0.001 0.001 32.929 33.062 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.929 33.062 multiply_cannon 2055 13.4 0.222 0.242 31.087 32.290 ot_scf_mini 99 9.5 0.003 0.003 31.259 31.381 multiply_cannon_loop 2055 14.4 0.928 0.949 29.697 30.516 ot_mini 99 10.5 0.001 0.001 18.380 18.506 multiply_cannon_multrec 24660 15.4 7.591 9.335 13.815 15.367 rebuild_ks_matrix 110 8.3 0.000 0.001 13.553 13.634 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.552 13.634 qs_ot_get_derivative 99 11.5 0.001 0.001 12.555 12.675 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.949 12.032 mp_waitall_1 186928 16.3 7.849 10.588 7.849 10.588 multiply_cannon_sync_h2d 24660 15.4 6.993 8.035 6.993 8.035 init_scf_loop 11 6.9 0.000 0.001 7.749 7.750 multiply_cannon_metrocomm3 24660 15.4 0.070 0.071 5.178 7.661 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.530 7.277 apply_single 110 13.6 0.000 0.001 6.530 7.277 sum_up_and_integrate 110 10.3 0.052 0.058 6.507 6.520 integrate_v_rspace 110 11.3 0.002 0.003 6.455 6.467 qs_ot_get_p 110 10.4 0.001 0.001 6.249 6.407 dbcsr_mm_accdrv_process 52282 16.1 4.644 5.663 6.066 6.376 init_scf_run 11 5.9 0.000 0.001 6.095 6.095 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.094 6.095 ot_diis_step 99 11.5 0.010 0.010 5.770 5.770 prepare_preconditioner 11 7.9 0.000 0.000 5.735 5.751 make_preconditioner 11 8.9 0.000 0.000 5.735 5.751 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.735 5.744 calculate_rho_elec 110 8.6 0.039 0.047 5.734 5.744 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.295 5.451 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.675 5.371 make_m2s 4110 13.4 0.057 0.059 4.179 4.738 make_images 4110 14.4 0.402 0.447 4.071 4.626 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.317 4.337 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.877 3.877 wfi_extrapolate 11 7.9 0.001 0.001 3.529 3.529 pw_transfer 1331 11.6 0.065 0.071 3.283 3.423 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.266 3.334 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.176 3.315 grid_integrate_task_list 110 12.3 3.161 3.311 3.161 3.311 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.291 3.293 cp_fm_redistribute_end 48 14.0 2.449 3.257 2.451 3.258 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.253 3.254 cp_fm_diag_elpa_base 48 14.0 0.770 3.128 0.802 3.215 density_rs2pw 110 9.6 0.004 0.004 3.054 3.197 rs_pw_transfer 902 11.9 0.012 0.014 2.950 3.112 calculate_dm_sparse 110 9.5 0.001 0.001 2.970 3.002 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.921 2.956 make_images_data 4110 15.4 0.047 0.052 2.307 2.857 hybrid_alltoall_any 4261 16.3 0.102 0.449 1.993 2.810 fft_wrap_pw1pw2_140 451 13.1 0.201 0.220 2.644 2.781 cp_fm_cholesky_invert 11 10.9 2.706 2.713 2.706 2.713 fft3d_ps 1111 14.6 1.074 1.267 2.531 2.654 mp_sum_l 6514 12.8 1.873 2.504 1.873 2.504 calculate_first_density_matrix 1 7.0 0.001 0.001 2.481 2.484 potential_pw2rs 110 12.3 0.008 0.009 2.396 2.406 grid_collocate_task_list 110 9.6 2.121 2.272 2.121 2.272 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.978 1.994 jit_kernel_multiply 10 16.4 1.066 1.966 1.066 1.966 mp_alltoall_d11v 2046 13.8 1.726 1.943 1.726 1.943 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.831 1.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.768 1.778 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.795 1.736 mp_allgather_i34 2055 14.4 0.719 1.670 0.719 1.670 mp_irecv_dv 57340 16.2 0.671 1.622 0.671 1.622 acc_transpose_blocks 24660 15.4 0.109 0.111 1.548 1.567 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.541 1.554 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.366 1.473 dbcsr_complete_redistribute 325 12.2 0.240 0.293 1.189 1.461 cp_fm_cholesky_decompose 22 10.9 1.380 1.386 1.380 1.386 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.632000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.363636, yerr=6.664325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 658.329600E+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 0.029 0.077 61.466 61.467 qs_mol_dyn_low 1 2.0 0.003 0.003 61.017 61.026 qs_forces 11 3.9 0.002 0.002 60.833 60.833 qs_energies 11 4.9 0.007 0.017 57.671 57.675 scf_env_do_scf 11 5.9 0.001 0.002 49.366 49.367 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 40.431 40.432 velocity_verlet 10 3.0 0.003 0.006 33.142 33.147 dbcsr_multiply_generic 2055 12.4 0.109 0.114 29.145 29.463 qs_scf_new_mos 99 7.5 0.001 0.001 25.811 25.894 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.811 25.893 ot_scf_mini 99 9.5 0.003 0.003 24.563 24.665 multiply_cannon 2055 13.4 0.216 0.231 22.578 23.779 multiply_cannon_loop 2055 14.4 0.620 0.633 21.352 22.502 ot_mini 99 10.5 0.001 0.001 13.965 14.068 rebuild_ks_matrix 110 8.3 0.000 0.000 12.294 12.446 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.021 12.294 12.446 multiply_cannon_multrec 16440 15.4 3.973 5.403 9.924 11.245 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.873 11.010 mp_waitall_1 146766 16.3 7.483 10.846 7.483 10.846 qs_ot_get_derivative 99 11.5 0.001 0.001 9.445 9.547 init_scf_loop 11 6.9 0.001 0.006 8.900 8.902 multiply_cannon_metrocomm3 16440 15.4 0.042 0.043 4.436 7.626 prepare_preconditioner 11 7.9 0.000 0.000 7.084 7.099 make_preconditioner 11 8.9 0.000 0.001 7.084 7.099 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.433 6.766 sum_up_and_integrate 110 10.3 0.059 0.060 6.523 6.538 integrate_v_rspace 110 11.3 0.002 0.003 6.464 6.478 dbcsr_mm_accdrv_process 34862 16.1 4.880 5.556 5.809 5.961 qs_ot_get_p 110 10.4 0.001 0.001 5.744 5.882 init_scf_run 11 5.9 0.001 0.004 5.833 5.833 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.832 5.833 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.565 5.577 calculate_rho_elec 110 8.6 0.058 0.058 5.565 5.576 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.016 5.491 apply_single 110 13.6 0.000 0.000 5.016 5.490 make_m2s 4110 13.4 0.050 0.052 4.311 4.679 make_images 4110 14.4 0.397 0.518 4.196 4.564 multiply_cannon_sync_h2d 16440 15.4 3.670 4.487 3.670 4.487 ot_diis_step 99 11.5 0.010 0.011 4.485 4.485 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.048 4.055 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.308 3.963 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.685 3.685 grid_integrate_task_list 110 12.3 3.184 3.351 3.184 3.351 pw_transfer 1331 11.6 0.064 0.071 3.158 3.167 rs_pw_transfer 902 11.9 0.011 0.011 2.903 3.140 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.111 3.112 density_rs2pw 110 9.6 0.004 0.004 2.877 3.081 cp_fm_redistribute_end 48 14.0 1.935 3.080 1.938 3.081 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.052 3.064 cp_fm_diag_elpa_base 48 14.0 1.080 2.933 1.139 3.042 wfi_extrapolate 11 7.9 0.001 0.001 3.015 3.015 make_images_data 4110 15.4 0.044 0.048 2.468 2.938 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.927 2.929 cp_fm_cholesky_invert 11 10.9 2.892 2.899 2.892 2.899 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.212 2.837 calculate_dm_sparse 110 9.5 0.001 0.001 2.721 2.756 calculate_first_density_matrix 1 7.0 0.001 0.004 2.736 2.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.570 2.629 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.543 2.554 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.443 2.497 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.865 2.468 fft3d_ps 1111 14.6 1.061 1.072 2.364 2.375 potential_pw2rs 110 12.3 0.010 0.010 2.361 2.368 grid_collocate_task_list 110 9.6 2.173 2.367 2.173 2.367 mp_irecv_dv 48980 15.7 0.795 2.343 0.795 2.343 mp_sum_l 6514 12.8 1.683 2.239 1.683 2.239 mp_alltoall_d11v 2046 13.8 1.800 2.121 1.800 2.121 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 2.004 2.004 dbcsr_complete_redistribute 325 12.2 0.322 0.348 1.483 1.922 cp_fm_upper_to_full 70 13.6 1.370 1.839 1.370 1.839 cp_fm_cholesky_decompose 22 10.9 1.725 1.746 1.725 1.746 jit_kernel_multiply 8 16.5 0.534 1.742 0.534 1.742 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.659 1.674 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.624 1.640 mp_allgather_i34 2055 14.4 0.545 1.633 0.545 1.633 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.356 1.474 copy_fm_to_dbcsr 174 11.2 0.001 0.002 0.987 1.436 mp_waitany 17072 13.8 1.218 1.392 1.218 1.392 acc_transpose_blocks 16440 15.4 0.073 0.076 1.336 1.362 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.310 1.316 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.467000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.727273, yerr=8.236183 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 720.728064E+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 0.013 0.028 65.787 65.788 qs_mol_dyn_low 1 2.0 0.003 0.003 65.565 65.574 qs_forces 11 3.9 0.002 0.002 65.497 65.498 qs_energies 11 4.9 0.001 0.002 62.133 62.136 scf_env_do_scf 11 5.9 0.000 0.001 53.817 53.820 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.655 41.656 velocity_verlet 10 3.0 0.004 0.006 37.367 37.369 dbcsr_multiply_generic 2055 12.4 0.115 0.118 30.012 30.309 qs_scf_new_mos 99 7.5 0.001 0.001 27.311 27.419 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.310 27.419 ot_scf_mini 99 9.5 0.003 0.003 25.664 25.763 multiply_cannon 2055 13.4 0.241 0.260 22.703 23.942 multiply_cannon_loop 2055 14.4 0.880 0.913 21.283 21.984 ot_mini 99 10.5 0.001 0.001 14.649 14.769 multiply_cannon_multrec 24660 15.4 4.214 6.793 12.622 13.755 init_scf_loop 11 6.9 0.000 0.000 12.119 12.119 rebuild_ks_matrix 110 8.3 0.000 0.000 11.907 12.015 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.906 12.015 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.542 10.637 qs_ot_get_derivative 99 11.5 0.001 0.001 10.426 10.528 prepare_preconditioner 11 7.9 0.000 0.000 10.379 10.395 make_preconditioner 11 8.9 0.000 0.000 10.379 10.395 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.566 10.047 dbcsr_mm_accdrv_process 52304 16.0 6.730 8.516 8.261 9.237 mp_waitall_1 126806 16.4 4.930 7.042 4.930 7.042 sum_up_and_integrate 110 10.3 0.067 0.070 6.359 6.371 integrate_v_rspace 110 11.3 0.003 0.003 6.292 6.302 qs_ot_get_p 110 10.4 0.001 0.001 5.948 6.098 make_m2s 4110 13.4 0.059 0.061 5.576 5.868 make_images 4110 14.4 0.580 0.701 5.436 5.722 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.586 5.604 calculate_rho_elec 110 8.6 0.077 0.081 5.586 5.603 init_scf_run 11 5.9 0.000 0.001 5.590 5.590 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.589 5.590 cp_fm_upper_to_full 70 13.8 3.333 4.735 3.333 4.735 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.072 4.198 apply_single 110 13.6 0.000 0.000 4.072 4.198 ot_diis_step 99 11.5 0.011 0.011 4.183 4.183 qs_ot_p2m_diag 48 11.0 0.054 0.063 4.090 4.105 dbcsr_complete_redistribute 325 12.2 0.414 0.458 2.763 3.926 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.680 3.753 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.597 3.597 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.433 3.482 grid_integrate_task_list 110 12.3 3.283 3.405 3.283 3.405 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.186 3.335 multiply_cannon_sync_h2d 24660 15.4 3.158 3.295 3.158 3.295 pw_transfer 1331 11.6 0.064 0.073 3.173 3.196 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.110 3.163 make_images_data 4110 15.4 0.046 0.050 2.774 3.104 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.067 3.093 wfi_extrapolate 11 7.9 0.001 0.001 3.007 3.008 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.365 3.005 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.003 3.005 calculate_dm_sparse 110 9.5 0.001 0.001 2.946 2.980 cp_fm_cholesky_invert 11 10.9 2.965 2.975 2.965 2.975 cp_fm_redistribute_end 48 14.0 1.497 2.966 1.499 2.968 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.805 2.941 cp_fm_diag_elpa_base 48 14.0 1.383 2.821 1.466 2.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.905 2.907 density_rs2pw 110 9.6 0.004 0.004 2.761 2.900 mp_alltoall_i22 605 13.7 1.687 2.860 1.687 2.860 rs_pw_transfer 902 11.9 0.010 0.011 2.524 2.658 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.564 2.589 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.520 2.564 calculate_first_density_matrix 1 7.0 0.000 0.001 2.492 2.498 grid_collocate_task_list 110 9.6 2.221 2.389 2.221 2.389 fft3d_ps 1111 14.6 1.062 1.091 2.369 2.388 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.224 2.224 potential_pw2rs 110 12.3 0.013 0.013 2.164 2.174 mp_alltoall_d11v 2046 13.8 1.837 2.042 1.837 2.042 jit_kernel_multiply 11 15.6 1.199 1.982 1.199 1.982 cp_fm_cholesky_decompose 22 10.9 1.791 1.838 1.791 1.838 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.770 1.796 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.610 1.622 mp_allgather_i34 2055 14.4 0.522 1.608 0.522 1.608 mp_sum_l 6514 12.8 1.047 1.598 1.047 1.598 acc_transpose_blocks 24660 15.4 0.106 0.109 1.524 1.575 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.556 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.845 1.465 mp_irecv_dv 62702 16.1 0.747 1.390 0.747 1.390 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.788000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.727273, yerr=7.033568 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 818.511872E+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 0.057 0.061 56.891 56.891 qs_mol_dyn_low 1 2.0 0.003 0.003 56.575 56.585 qs_forces 11 3.9 0.014 0.014 56.507 56.509 qs_energies 11 4.9 0.001 0.002 52.889 52.893 scf_env_do_scf 11 5.9 0.000 0.001 44.306 44.306 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.837 35.838 velocity_verlet 10 3.0 0.002 0.002 31.592 31.595 dbcsr_multiply_generic 2055 12.4 0.104 0.105 23.249 23.404 qs_scf_new_mos 99 7.5 0.001 0.001 21.634 21.670 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.634 21.669 ot_scf_mini 99 9.5 0.002 0.002 20.354 20.376 multiply_cannon 2055 13.4 0.246 0.256 17.504 18.681 multiply_cannon_loop 2055 14.4 0.322 0.335 16.076 16.374 rebuild_ks_matrix 110 8.3 0.000 0.000 11.496 11.508 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.496 11.507 ot_mini 99 10.5 0.001 0.001 10.776 10.792 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.258 10.268 init_scf_loop 11 6.9 0.000 0.000 8.421 8.422 multiply_cannon_multrec 8220 15.4 3.218 4.476 7.495 8.339 mp_waitall_1 106626 16.5 6.377 7.981 6.377 7.981 qs_ot_get_derivative 99 11.5 0.001 0.001 6.911 6.930 prepare_preconditioner 11 7.9 0.000 0.000 6.768 6.771 make_preconditioner 11 8.9 0.000 0.000 6.768 6.771 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.339 6.415 sum_up_and_integrate 110 10.3 0.080 0.082 6.223 6.235 integrate_v_rspace 110 11.3 0.003 0.003 6.143 6.154 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.668 5.685 calculate_rho_elec 110 8.6 0.115 0.116 5.667 5.684 qs_ot_get_p 110 10.4 0.001 0.001 5.579 5.599 init_scf_run 11 5.9 0.000 0.001 5.454 5.454 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.454 5.454 dbcsr_mm_accdrv_process 17442 15.9 2.925 4.053 4.145 5.100 make_m2s 4110 13.4 0.039 0.040 4.367 4.621 make_images 4110 14.4 0.639 0.695 4.238 4.489 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.033 4.423 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.115 4.121 ot_diis_step 99 11.5 0.012 0.012 3.834 3.834 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.764 3.765 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.720 3.743 apply_single 110 13.6 0.000 0.000 3.720 3.743 grid_integrate_task_list 110 12.3 3.362 3.469 3.362 3.469 cp_fm_cholesky_invert 11 10.9 3.450 3.455 3.450 3.455 pw_transfer 1331 11.6 0.064 0.069 3.265 3.276 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.158 3.173 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.168 3.169 cp_fm_redistribute_end 48 14.0 0.814 3.131 0.820 3.132 cp_fm_diag_elpa_base 48 14.0 2.123 2.910 2.302 3.076 multiply_cannon_sync_h2d 8220 15.4 2.922 2.980 2.922 2.980 make_images_data 4110 15.4 0.038 0.043 2.457 2.848 density_rs2pw 110 9.6 0.004 0.004 2.721 2.805 wfi_extrapolate 11 7.9 0.001 0.001 2.727 2.727 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.338 2.715 fft_wrap_pw1pw2_140 451 13.1 0.211 0.214 2.658 2.679 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.676 2.677 calculate_first_density_matrix 1 7.0 0.001 0.001 2.638 2.639 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.601 2.602 calculate_dm_sparse 110 9.5 0.001 0.001 2.508 2.546 grid_collocate_task_list 110 9.6 2.331 2.437 2.331 2.437 fft3d_ps 1111 14.6 1.112 1.127 2.414 2.428 rs_pw_transfer 902 11.9 0.010 0.010 2.323 2.404 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.208 2.220 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.031 2.038 potential_pw2rs 110 12.3 0.015 0.016 1.997 2.000 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.772 1.987 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.919 1.930 cp_fm_cholesky_decompose 22 10.9 1.872 1.884 1.872 1.884 mp_alltoall_d11v 2046 13.8 1.633 1.810 1.633 1.810 mp_allgather_i34 2055 14.4 0.611 1.728 0.611 1.728 dbcsr_complete_redistribute 325 12.2 0.574 0.601 1.537 1.643 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.519 1.632 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.573 1.578 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.429 1.440 qs_create_task_list 11 7.9 0.000 0.001 1.230 1.328 generate_qs_task_list 11 8.9 0.381 0.450 1.229 1.327 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.823 1.218 mp_waitany 9240 13.8 1.103 1.210 1.103 1.210 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.165 1.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.891000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=772.636364, yerr=8.003099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 1.360605E+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 0.019 0.041 86.352 86.352 qs_mol_dyn_low 1 2.0 0.003 0.003 86.037 86.046 qs_forces 11 3.9 0.002 0.002 85.969 85.969 qs_energies 11 4.9 0.001 0.001 81.913 81.914 scf_env_do_scf 11 5.9 0.000 0.001 71.729 71.729 velocity_verlet 10 3.0 0.002 0.002 55.518 55.524 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.637 42.638 init_scf_loop 11 6.9 0.000 0.000 29.017 29.019 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.865 28.914 prepare_preconditioner 11 7.9 0.000 0.000 27.053 27.059 make_preconditioner 11 8.9 0.000 0.000 27.053 27.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.109 26.535 qs_scf_new_mos 99 7.5 0.001 0.001 26.162 26.187 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.161 26.186 ot_scf_mini 99 9.5 0.002 0.002 24.388 24.409 multiply_cannon 2055 13.4 0.342 0.362 21.701 22.468 multiply_cannon_loop 2055 14.4 0.342 0.347 19.881 20.294 cp_fm_upper_to_full 70 14.2 13.049 18.722 13.049 18.722 ot_mini 99 10.5 0.001 0.001 13.562 13.578 rebuild_ks_matrix 110 8.3 0.000 0.001 13.207 13.239 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.206 13.239 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.991 12.021 dbcsr_complete_redistribute 325 12.2 1.023 1.047 7.517 10.927 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.498 9.901 multiply_cannon_multrec 8220 15.4 4.363 4.542 9.751 9.881 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.928 9.287 qs_ot_get_derivative 99 11.5 0.001 0.001 8.963 8.985 mp_alltoall_i22 605 13.7 5.560 8.977 5.560 8.977 mp_waitall_1 87304 16.6 8.010 8.839 8.010 8.839 sum_up_and_integrate 110 10.3 0.151 0.153 6.641 6.657 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.536 6.573 calculate_rho_elec 110 8.6 0.227 0.227 6.535 6.572 integrate_v_rspace 110 11.3 0.004 0.004 6.490 6.507 make_m2s 4110 13.4 0.043 0.043 5.463 5.997 init_scf_run 11 5.9 0.000 0.001 5.907 5.908 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.907 5.907 make_images 4110 14.4 0.881 0.926 5.276 5.810 qs_ot_get_p 110 10.4 0.001 0.001 5.604 5.624 dbcsr_mm_accdrv_process 11614 15.7 3.264 3.683 5.245 5.515 cp_fm_cholesky_invert 11 10.9 5.458 5.462 5.458 5.462 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.726 5.172 apply_single 110 13.6 0.000 0.000 4.726 5.171 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.724 5.153 ot_diis_step 99 11.5 0.015 0.016 4.576 4.576 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.054 4.061 multiply_cannon_sync_h2d 8220 15.4 3.944 3.954 3.944 3.954 hybrid_alltoall_any 4261 16.3 0.257 0.553 3.056 3.889 pw_transfer 1331 11.6 0.073 0.074 3.782 3.786 make_images_data 4110 15.4 0.041 0.045 3.095 3.781 grid_integrate_task_list 110 12.3 3.660 3.718 3.660 3.718 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.665 3.670 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.664 3.665 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.615 3.615 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.848 3.299 calculate_dm_sparse 110 9.5 0.001 0.001 3.270 3.291 wfi_extrapolate 11 7.9 0.001 0.001 3.242 3.242 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.173 3.181 density_rs2pw 110 9.6 0.004 0.004 3.006 3.020 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.009 3.010 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.004 3.009 cp_fm_diag_elpa_base 48 14.0 2.469 2.668 3.007 3.008 fft3d_ps 1111 14.6 1.267 1.294 2.874 2.880 grid_collocate_task_list 110 9.6 2.624 2.645 2.624 2.645 calculate_first_density_matrix 1 7.0 0.000 0.001 2.561 2.562 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.500 2.512 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.372 2.384 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.199 2.254 rs_pw_transfer 902 11.9 0.011 0.011 2.157 2.192 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.091 2.150 cp_fm_cholesky_decompose 22 10.9 2.096 2.125 2.096 2.125 mp_alltoall_d11v 2046 13.8 2.068 2.118 2.068 2.118 potential_pw2rs 110 12.3 0.022 0.022 2.074 2.077 jit_kernel_multiply 10 15.3 1.778 1.945 1.778 1.945 qs_create_task_list 11 7.9 0.001 0.001 1.901 1.945 generate_qs_task_list 11 8.9 0.739 0.795 1.900 1.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.821 1.825 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.796 1.811 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.758 1.789 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.352000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1210.090909, yerr=69.109208 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.081216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 3124303. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 204.307 204.308 qs_mol_dyn_low 1 2.0 0.003 0.003 203.948 203.962 qs_forces 11 3.9 0.003 0.004 203.844 203.846 qs_energies 11 4.9 0.001 0.002 198.351 198.364 scf_env_do_scf 11 5.9 0.001 0.001 181.569 181.573 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 160.649 160.651 dbcsr_multiply_generic 2507 12.6 0.185 0.200 124.574 125.819 velocity_verlet 10 3.0 0.001 0.002 122.939 122.940 qs_scf_new_mos 117 7.6 0.001 0.001 121.944 122.324 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.943 122.323 ot_scf_mini 117 9.6 0.003 0.003 115.334 115.672 multiply_cannon 2507 13.6 0.238 0.247 101.031 103.642 multiply_cannon_loop 2507 14.6 2.102 2.172 98.817 101.404 ot_mini 117 10.6 0.001 0.001 65.737 66.108 multiply_cannon_multrec 60168 15.6 33.169 36.230 41.731 44.509 qs_ot_get_derivative 117 11.6 0.001 0.001 40.665 41.001 rebuild_ks_matrix 128 8.3 0.001 0.001 33.268 33.694 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 33.267 33.693 mp_waitall_1 291448 16.2 28.847 32.120 28.847 32.120 multiply_cannon_sync_h2d 60168 15.6 27.271 30.700 27.271 30.700 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.939 30.327 qs_ot_get_p 128 10.4 0.001 0.001 27.628 27.994 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.356 25.425 apply_single 128 13.6 0.001 0.001 24.356 25.425 ot_diis_step 117 11.6 0.007 0.008 24.695 24.696 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.976 21.068 init_scf_loop 11 6.9 0.000 0.000 20.843 20.844 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.935 19.226 multiply_cannon_metrocomm3 60168 15.6 0.111 0.115 15.489 18.600 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.265 18.267 prepare_preconditioner 11 7.9 0.000 0.000 16.252 16.296 make_preconditioner 11 8.9 0.000 0.000 16.252 16.296 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.443 15.640 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.423 15.430 cp_fm_redistribute_end 83 14.4 12.177 15.349 12.192 15.353 cp_fm_diag_elpa_base 83 14.4 3.113 15.037 3.147 15.169 make_m2s 5014 13.6 0.104 0.111 13.987 14.420 make_images 5014 14.6 0.402 0.423 13.806 14.252 sum_up_and_integrate 128 10.3 0.090 0.107 13.954 13.970 integrate_v_rspace 128 11.3 0.003 0.004 13.864 13.883 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.622 12.723 calculate_rho_elec 128 8.7 0.045 0.063 12.621 12.722 init_scf_run 11 5.9 0.000 0.001 12.638 12.639 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.638 12.639 mp_sum_l 7870 13.0 8.286 9.842 8.286 9.842 cp_fm_cholesky_invert 11 10.9 9.264 9.272 9.264 9.272 wfi_extrapolate 11 7.9 0.001 0.001 9.114 9.114 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 6.309 9.108 calculate_dm_sparse 128 9.5 0.001 0.001 8.606 8.698 dbcsr_mm_accdrv_process 124484 16.2 3.302 3.448 8.123 8.622 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.001 8.161 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.862 8.006 make_images_data 5014 15.6 0.069 0.075 6.904 7.827 grid_integrate_task_list 128 12.3 7.057 7.504 7.057 7.504 hybrid_alltoall_any 5200 16.5 0.290 2.256 6.021 7.218 pw_transfer 1547 11.6 0.075 0.109 6.777 7.045 density_rs2pw 128 9.7 0.006 0.006 6.415 6.897 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.573 6.817 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.784 6.793 rs_pw_transfer 1046 11.9 0.017 0.019 5.650 6.173 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.939 6.071 fft_wrap_pw1pw2_140 523 13.2 0.445 0.507 5.679 5.860 mp_alltoall_d11v 2415 14.1 4.325 5.584 4.325 5.584 fft3d_ps 1291 14.7 2.093 2.569 5.366 5.553 grid_collocate_task_list 128 9.7 4.729 5.113 4.729 5.113 cp_fm_cholesky_decompose 22 10.9 4.748 4.763 4.748 4.763 mp_sum_d 4464 12.1 3.728 4.568 3.728 4.568 potential_pw2rs 128 12.3 0.009 0.011 4.436 4.456 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.308000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.727273, yerr=6.836944 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 829.505536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 5752009. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 189.668 189.669 qs_mol_dyn_low 1 2.0 0.003 0.003 189.240 189.253 qs_forces 11 3.9 0.012 0.027 189.157 189.159 qs_energies 11 4.9 0.001 0.002 182.475 182.493 scf_env_do_scf 11 5.9 0.001 0.001 166.172 166.182 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.836 132.839 velocity_verlet 10 3.0 0.001 0.002 119.477 119.479 dbcsr_multiply_generic 2507 12.6 0.183 0.188 97.151 98.322 qs_scf_new_mos 117 7.6 0.001 0.001 94.421 94.844 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.420 94.844 ot_scf_mini 117 9.6 0.004 0.004 89.631 90.103 multiply_cannon 2507 13.6 0.479 0.533 77.290 81.630 multiply_cannon_loop 2507 14.6 1.256 1.293 73.969 76.843 ot_mini 117 10.6 0.001 0.001 49.624 50.111 mp_waitall_1 226760 16.4 24.318 38.213 24.318 38.213 multiply_cannon_multrec 30084 15.6 22.118 26.405 31.769 36.527 init_scf_loop 11 6.9 0.000 0.000 33.238 33.239 rebuild_ks_matrix 128 8.3 0.001 0.001 32.630 33.236 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.629 33.236 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.418 29.965 prepare_preconditioner 11 7.9 0.000 0.000 28.896 28.960 make_preconditioner 11 8.9 0.000 0.000 28.896 28.960 qs_ot_get_derivative 117 11.6 0.001 0.002 27.830 28.300 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.588 28.108 multiply_cannon_metrocomm3 30084 15.6 0.095 0.101 15.421 27.690 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.874 22.957 apply_single 128 13.6 0.001 0.001 21.873 22.956 qs_ot_get_p 128 10.4 0.001 0.001 21.789 22.331 multiply_cannon_sync_h2d 30084 15.6 19.343 21.647 19.343 21.647 ot_diis_step 117 11.6 0.014 0.015 21.616 21.618 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.029 17.068 cp_fm_cholesky_invert 11 10.9 16.998 17.012 16.998 17.012 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.865 15.866 make_m2s 5014 13.6 0.089 0.094 14.029 15.626 make_images 5014 14.6 1.167 1.345 13.822 15.421 sum_up_and_integrate 128 10.3 0.114 0.131 14.337 14.365 integrate_v_rspace 128 11.3 0.003 0.003 14.224 14.255 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.807 12.837 calculate_rho_elec 128 8.7 0.088 0.105 12.806 12.837 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.714 12.721 cp_fm_redistribute_end 83 14.4 7.430 12.646 7.443 12.651 cp_fm_diag_elpa_base 83 14.4 4.971 12.187 5.191 12.539 init_scf_run 11 5.9 0.000 0.001 11.544 11.545 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.543 11.545 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.976 11.308 multiply_cannon_metrocomm4 27577 15.6 0.097 0.113 3.736 10.245 make_images_data 5014 15.6 0.067 0.074 8.362 10.161 mp_irecv_dv 69486 16.3 3.543 9.871 3.543 9.871 dbcsr_mm_accdrv_process 62242 16.2 4.576 5.388 9.115 9.717 hybrid_alltoall_any 5200 16.5 0.343 1.516 7.145 9.324 wfi_extrapolate 11 7.9 0.001 0.001 8.366 8.367 pw_transfer 1547 11.6 0.085 0.104 7.675 7.734 grid_integrate_task_list 128 12.3 7.189 7.603 7.189 7.603 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.452 7.510 cp_fm_cholesky_decompose 22 10.9 7.080 7.161 7.080 7.161 density_rs2pw 128 9.7 0.006 0.006 6.665 7.038 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.206 6.937 calculate_dm_sparse 128 9.5 0.001 0.001 6.439 6.577 fft_wrap_pw1pw2_140 523 13.2 0.470 0.518 6.493 6.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.195 6.205 rs_pw_transfer 1046 11.9 0.015 0.017 5.612 6.055 mp_sum_l 7870 13.0 4.055 6.026 4.055 6.026 fft3d_ps 1291 14.7 2.770 2.923 5.823 5.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.330 5.488 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.358 5.420 grid_collocate_task_list 128 9.7 4.909 5.348 4.909 5.348 potential_pw2rs 128 12.3 0.015 0.016 4.819 4.839 mp_allgather_i34 2507 14.6 1.893 4.707 1.893 4.707 mp_alltoall_d11v 2415 14.1 4.002 4.478 4.002 4.478 dbcsr_complete_redistribute 395 12.7 0.766 0.863 3.108 3.986 mp_sum_d 4464 12.1 2.578 3.925 2.578 3.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.669000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.545455, yerr=2.934548 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 937.041920E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.035 177.048 177.049 qs_mol_dyn_low 1 2.0 0.003 0.003 176.560 176.575 qs_forces 11 3.9 0.003 0.003 176.455 176.464 qs_energies 11 4.9 0.001 0.002 169.928 169.938 scf_env_do_scf 11 5.9 0.001 0.001 154.218 154.219 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.696 118.697 velocity_verlet 10 3.0 0.001 0.002 113.393 113.395 dbcsr_multiply_generic 2507 12.6 0.182 0.187 81.825 82.753 qs_scf_new_mos 117 7.6 0.001 0.001 82.079 82.446 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.079 82.446 ot_scf_mini 117 9.6 0.003 0.004 77.909 78.273 multiply_cannon 2507 13.6 0.502 0.525 61.855 66.374 multiply_cannon_loop 2507 14.6 0.861 0.890 58.784 61.249 ot_mini 117 10.6 0.001 0.001 42.595 42.965 init_scf_loop 11 6.9 0.000 0.000 35.418 35.419 mp_waitall_1 178456 16.5 25.658 35.395 25.658 35.395 prepare_preconditioner 11 7.9 0.000 0.000 31.403 31.461 make_preconditioner 11 8.9 0.000 0.000 31.403 31.461 rebuild_ks_matrix 128 8.3 0.001 0.001 30.140 30.581 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.140 30.580 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.053 30.478 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.218 27.620 multiply_cannon_multrec 20056 15.6 13.469 16.109 22.332 25.036 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.362 24.707 qs_ot_get_derivative 117 11.6 0.001 0.002 22.761 23.123 qs_ot_get_p 128 10.4 0.001 0.001 20.687 21.183 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.911 20.937 apply_single 128 13.6 0.001 0.001 19.910 20.936 ot_diis_step 117 11.6 0.018 0.019 19.730 19.730 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.432 16.440 make_m2s 5014 13.6 0.080 0.087 14.932 15.841 multiply_cannon_sync_h2d 20056 15.6 14.131 15.830 14.131 15.830 make_images 5014 14.6 1.190 1.286 14.700 15.605 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.406 15.407 cp_fm_cholesky_invert 11 10.9 14.753 14.763 14.753 14.763 sum_up_and_integrate 128 10.3 0.131 0.143 14.208 14.237 integrate_v_rspace 128 11.3 0.003 0.004 14.077 14.100 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.038 13.065 calculate_rho_elec 128 8.7 0.132 0.146 13.038 13.064 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.234 12.236 cp_fm_redistribute_end 83 14.4 4.644 12.169 4.660 12.171 cp_fm_diag_elpa_base 83 14.4 7.073 11.497 7.487 12.040 init_scf_run 11 5.9 0.000 0.001 10.681 10.681 scf_env_initial_rho_setup 11 6.9 0.002 0.003 10.680 10.681 make_images_data 5014 15.6 0.061 0.068 9.262 10.627 hybrid_alltoall_any 5200 16.5 0.431 1.974 7.981 9.759 multiply_cannon_metrocomm4 17549 15.6 0.063 0.073 3.454 9.406 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.923 9.189 mp_irecv_dv 50230 16.2 3.331 9.154 3.331 9.154 dbcsr_mm_accdrv_process 41502 16.2 4.567 5.399 8.320 8.432 pw_transfer 1547 11.6 0.084 0.104 7.808 7.914 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.585 7.695 grid_integrate_task_list 128 12.3 7.298 7.685 7.298 7.685 cp_fm_cholesky_decompose 22 10.9 7.554 7.583 7.554 7.583 cp_fm_upper_to_full 105 14.5 5.705 7.463 5.705 7.463 wfi_extrapolate 11 7.9 0.001 0.001 7.434 7.435 density_rs2pw 128 9.7 0.006 0.006 6.521 6.943 fft_wrap_pw1pw2_140 523 13.2 0.479 0.537 6.668 6.791 dbcsr_complete_redistribute 395 12.7 1.163 1.191 4.602 6.369 calculate_dm_sparse 128 9.5 0.001 0.001 5.918 6.029 fft3d_ps 1291 14.7 2.694 2.903 5.848 5.914 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.587 5.592 rs_pw_transfer 1046 11.9 0.014 0.015 5.131 5.566 grid_collocate_task_list 128 9.7 5.074 5.512 5.074 5.512 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.525 5.209 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.407 5.174 mp_alltoall_d11v 2415 14.1 4.319 4.936 4.319 4.936 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.675 4.805 mp_allgather_i34 2507 14.6 1.588 4.619 1.588 4.619 potential_pw2rs 128 12.3 0.020 0.022 4.559 4.574 mp_sum_l 7870 13.0 3.219 4.490 3.219 4.490 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.039 4.059 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.330 4.029 mp_alltoall_i22 716 14.1 1.947 3.812 1.947 3.812 qs_energies_init_hamiltonians 11 5.9 0.017 0.029 3.751 3.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.049000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.363636, yerr=10.455731 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.149190E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57191. MP_Allreduce 11251 1068. MP_Sync 168 MP_Alltoall 1700 12496376. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.110 187.571 187.574 qs_mol_dyn_low 1 2.0 0.003 0.003 187.041 187.073 qs_forces 11 3.9 0.003 0.003 186.926 186.932 qs_energies 11 4.9 0.002 0.005 179.844 179.855 scf_env_do_scf 11 5.9 0.001 0.001 163.231 163.243 velocity_verlet 10 3.0 0.002 0.002 123.939 123.942 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.677 115.679 qs_scf_new_mos 116 7.6 0.001 0.001 80.673 80.950 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.672 80.949 dbcsr_multiply_generic 2485 12.5 0.188 0.193 78.737 79.510 ot_scf_mini 116 9.6 0.003 0.004 76.249 76.543 multiply_cannon 2485 13.5 0.551 0.587 54.321 57.272 multiply_cannon_loop 2485 14.5 1.167 1.200 50.503 52.577 init_scf_loop 11 6.9 0.001 0.002 47.430 47.435 prepare_preconditioner 11 7.9 0.000 0.000 43.324 43.348 make_preconditioner 11 8.9 0.000 0.002 43.324 43.348 ot_mini 116 10.6 0.001 0.001 42.121 42.398 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.936 42.000 multiply_cannon_multrec 29820 15.5 14.025 19.415 25.835 30.631 rebuild_ks_matrix 127 8.3 0.001 0.001 28.719 29.039 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 28.718 29.038 mp_waitall_1 152434 16.5 17.325 26.502 17.325 26.502 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.866 26.148 qs_ot_get_derivative 116 11.6 0.001 0.002 22.710 23.001 make_m2s 4970 13.5 0.095 0.100 19.961 20.967 qs_ot_get_p 127 10.4 0.001 0.002 20.518 20.804 make_images 4970 14.5 1.953 2.250 19.658 20.667 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.791 19.357 apply_single 127 13.6 0.001 0.001 18.790 19.357 ot_diis_step 116 11.6 0.017 0.018 19.290 19.291 cp_fm_cholesky_invert 11 10.9 16.873 16.883 16.873 16.883 cp_fm_upper_to_full 104 14.7 11.193 16.554 11.193 16.554 qs_ot_p2m_diag 82 11.4 0.338 0.384 16.310 16.364 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.369 15.073 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.029 15.030 sum_up_and_integrate 127 10.3 0.140 0.151 14.006 14.041 integrate_v_rspace 127 11.3 0.003 0.004 13.866 13.906 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.937 12.964 calculate_rho_elec 127 8.7 0.174 0.189 12.936 12.963 dbcsr_complete_redistribute 393 12.7 1.515 1.649 9.028 12.843 multiply_cannon_sync_h2d 29820 15.5 11.628 12.683 11.628 12.683 make_images_data 4970 15.5 0.063 0.068 10.667 12.411 dbcsr_mm_accdrv_process 61748 16.2 7.212 8.047 11.389 11.956 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.930 11.932 cp_fm_redistribute_end 82 14.4 2.053 11.848 2.067 11.853 cp_fm_diag_elpa_base 82 14.4 9.165 11.238 9.751 11.723 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.645 11.433 hybrid_alltoall_any 5155 16.4 0.520 2.191 9.592 11.039 init_scf_run 11 5.9 0.000 0.001 10.686 10.688 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.686 10.687 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.366 10.078 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.360 9.573 mp_alltoall_i22 712 14.1 5.581 9.453 5.581 9.453 cp_fm_cholesky_decompose 22 10.9 7.825 7.923 7.825 7.923 grid_integrate_task_list 127 12.3 7.456 7.800 7.456 7.800 pw_transfer 1535 11.6 0.084 0.097 7.631 7.715 wfi_extrapolate 11 7.9 0.001 0.001 7.591 7.592 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.408 7.495 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.761 7.043 mp_irecv_dv 75445 16.2 2.618 6.786 2.618 6.786 density_rs2pw 127 9.7 0.005 0.006 6.145 6.691 fft_wrap_pw1pw2_140 519 13.2 0.478 0.487 6.496 6.589 calculate_dm_sparse 127 9.5 0.001 0.001 6.071 6.173 fft3d_ps 1281 14.7 2.778 2.878 5.651 5.731 grid_collocate_task_list 127 9.7 5.167 5.576 5.167 5.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.473 5.505 rs_pw_transfer 1038 11.9 0.013 0.014 4.673 5.197 mp_alltoall_d11v 2401 14.1 4.751 5.130 4.751 5.130 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.434 4.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.370 4.464 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.411 4.413 potential_pw2rs 127 12.3 0.022 0.022 4.317 4.339 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.165 4.218 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.574000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.909091, yerr=20.205555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.510797E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.050 171.089 171.090 qs_mol_dyn_low 1 2.0 0.003 0.003 170.570 170.607 qs_forces 11 3.9 0.003 0.004 170.467 170.470 qs_energies 11 4.9 0.004 0.005 163.141 163.151 scf_env_do_scf 11 5.9 0.001 0.001 146.103 146.116 velocity_verlet 10 3.0 0.002 0.003 112.109 112.112 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 110.404 110.405 dbcsr_multiply_generic 2507 12.6 0.179 0.185 74.751 75.358 qs_scf_new_mos 117 7.6 0.001 0.001 74.836 74.976 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.835 74.975 ot_scf_mini 117 9.6 0.003 0.004 70.386 70.547 multiply_cannon 2507 13.6 0.583 0.613 54.747 59.336 multiply_cannon_loop 2507 14.6 0.448 0.460 50.208 51.505 ot_mini 117 10.6 0.001 0.001 39.826 39.991 init_scf_loop 11 6.9 0.000 0.000 35.543 35.544 mp_waitall_1 129618 16.6 26.812 34.635 26.812 34.635 prepare_preconditioner 11 7.9 0.000 0.000 31.687 31.723 make_preconditioner 11 8.9 0.000 0.000 31.687 31.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.495 29.738 rebuild_ks_matrix 128 8.3 0.001 0.001 28.544 28.704 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.543 28.704 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.058 26.200 multiply_cannon_multrec 10028 15.6 10.379 15.078 17.745 21.888 multiply_cannon_metrocomm3 10028 15.6 0.022 0.024 12.857 20.524 qs_ot_get_derivative 117 11.6 0.002 0.002 19.843 20.004 ot_diis_step 117 11.6 0.020 0.020 19.912 19.912 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.432 19.805 apply_single 128 13.6 0.001 0.001 19.432 19.805 make_m2s 5014 13.6 0.065 0.071 16.337 18.616 make_images 5014 14.6 2.313 2.616 16.033 18.312 cp_fm_cholesky_invert 11 10.9 18.072 18.077 18.072 18.077 qs_ot_get_p 128 10.4 0.001 0.001 17.497 17.666 sum_up_and_integrate 128 10.3 0.180 0.191 13.959 14.004 integrate_v_rspace 128 11.3 0.004 0.004 13.779 13.831 qs_ot_p2m_diag 83 11.4 0.495 0.500 13.728 13.744 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.109 13.169 calculate_rho_elec 128 8.7 0.258 0.269 13.108 13.169 make_images_data 5014 15.6 0.052 0.060 10.005 12.617 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.588 12.589 hybrid_alltoall_any 5200 16.5 0.816 3.621 9.785 12.121 multiply_cannon_sync_h2d 10028 15.6 11.503 12.109 11.503 12.109 init_scf_run 11 5.9 0.000 0.001 10.267 10.267 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.267 10.267 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.564 9.571 cp_fm_diag_elpa_base 83 14.4 9.316 9.386 9.554 9.562 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.033 8.155 grid_integrate_task_list 128 12.3 7.745 8.030 7.745 8.030 cp_fm_cholesky_decompose 22 10.9 7.889 7.987 7.889 7.987 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 4.487 7.943 pw_transfer 1547 11.6 0.084 0.091 7.768 7.779 mp_allgather_i34 2507 14.6 2.736 7.700 2.736 7.700 dbcsr_mm_accdrv_process 20762 16.1 2.750 3.433 7.000 7.661 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.548 7.565 wfi_extrapolate 11 7.9 0.001 0.001 7.297 7.298 fft_wrap_pw1pw2_140 523 13.2 0.504 0.524 6.647 6.671 calculate_dm_sparse 128 9.5 0.001 0.001 6.135 6.232 density_rs2pw 128 9.7 0.005 0.006 5.912 6.123 fft3d_ps 1291 14.7 2.726 2.796 5.706 5.733 dbcsr_complete_redistribute 395 12.7 2.111 2.174 5.307 5.686 grid_collocate_task_list 128 9.7 5.486 5.656 5.486 5.656 multiply_cannon_metrocomm4 7521 15.6 0.023 0.027 1.873 5.372 mp_irecv_dv 28860 15.9 1.835 5.288 1.835 5.288 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.225 5.230 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.219 5.220 mp_alltoall_d11v 2415 14.1 4.854 5.218 4.854 5.218 rs_pw_transfer 1046 11.9 0.013 0.013 3.993 4.207 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.127 4.162 potential_pw2rs 128 12.3 0.027 0.028 4.017 4.033 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.559 3.869 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.462 3.777 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.741 3.773 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.586 3.691 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.467 3.506 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.090000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1414.363636, yerr=48.511054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 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 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.044508E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.106 291.924 291.925 qs_mol_dyn_low 1 2.0 0.003 0.003 291.329 291.358 qs_forces 11 3.9 0.003 0.003 291.244 291.246 qs_energies 11 4.9 0.002 0.002 282.627 282.636 scf_env_do_scf 11 5.9 0.001 0.001 260.475 260.487 velocity_verlet 10 3.0 0.002 0.002 211.591 211.599 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.033 135.036 init_scf_loop 11 6.9 0.000 0.000 125.183 125.188 prepare_preconditioner 11 7.9 0.000 0.000 120.484 120.526 make_preconditioner 11 8.9 0.000 0.000 120.484 120.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.309 117.620 qs_scf_new_mos 118 7.6 0.001 0.001 92.686 92.841 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.686 92.841 ot_scf_mini 118 9.6 0.004 0.004 87.796 87.887 dbcsr_multiply_generic 2535 12.6 0.214 0.221 85.008 85.782 cp_fm_upper_to_full 106 14.8 53.398 76.562 53.398 76.562 multiply_cannon 2535 13.6 0.710 0.763 59.828 60.592 multiply_cannon_loop 2535 14.6 0.477 0.485 55.909 57.940 ot_mini 118 10.6 0.001 0.001 45.755 45.870 dbcsr_complete_redistribute 397 12.7 4.001 4.068 29.953 42.855 copy_fm_to_dbcsr 210 11.7 0.001 0.001 26.492 39.467 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.130 36.963 mp_alltoall_i22 720 14.1 21.961 35.165 21.961 35.165 mp_waitall_1 106626 16.7 29.584 34.199 29.584 34.199 rebuild_ks_matrix 129 8.3 0.001 0.001 33.825 33.966 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.824 33.966 cp_fm_cholesky_invert 11 10.9 33.888 33.894 33.888 33.894 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.618 31.740 qs_ot_get_p 129 10.4 0.001 0.001 26.283 26.399 qs_ot_get_derivative 118 11.6 0.002 0.002 24.414 24.500 qs_ot_p2m_diag 84 11.4 0.889 0.895 22.167 22.197 make_m2s 5070 13.6 0.077 0.080 20.406 21.629 multiply_cannon_metrocomm3 10140 15.6 0.023 0.024 19.576 21.375 ot_diis_step 118 11.6 0.022 0.023 21.294 21.294 make_images 5070 14.6 3.801 3.947 19.926 21.151 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.369 20.754 apply_single 129 13.6 0.001 0.001 20.369 20.754 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.420 20.423 multiply_cannon_multrec 10140 15.6 10.525 12.298 17.984 18.068 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.189 17.191 cp_fm_diag_elpa_base 84 14.4 12.715 14.352 17.186 17.187 sum_up_and_integrate 129 10.3 0.323 0.326 15.739 15.829 multiply_cannon_sync_h2d 10140 15.6 15.777 15.785 15.777 15.785 integrate_v_rspace 129 11.3 0.004 0.004 15.416 15.507 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.988 15.007 calculate_rho_elec 129 8.7 0.487 0.488 14.987 15.006 hybrid_alltoall_any 5257 16.5 1.313 3.058 11.275 13.374 make_images_data 5070 15.6 0.061 0.065 11.110 13.120 init_scf_run 11 5.9 0.000 0.001 12.187 12.187 scf_env_initial_rho_setup 11 6.9 0.002 0.003 12.187 12.187 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.600 9.669 dbcsr_mm_accdrv_process 20958 16.1 3.800 5.652 7.220 9.062 cp_fm_cholesky_decompose 22 10.9 8.898 8.962 8.898 8.962 wfi_extrapolate 11 7.9 0.001 0.001 8.802 8.802 grid_integrate_task_list 129 12.3 8.614 8.776 8.614 8.776 pw_transfer 1559 11.6 0.091 0.092 8.362 8.374 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.126 8.139 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.948 7.949 fft_wrap_pw1pw2_140 527 13.2 0.546 0.551 7.193 7.206 mp_alltoall_d11v 2429 14.1 6.923 7.049 6.923 7.049 calculate_dm_sparse 129 9.5 0.001 0.001 6.734 6.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.400 6.468 grid_collocate_task_list 129 9.7 6.384 6.423 6.384 6.423 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.127 6.189 fft3d_ps 1301 14.7 2.785 2.807 6.150 6.161 density_rs2pw 129 9.7 0.005 0.005 6.011 6.043 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.925000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2685.818182, yerr=182.330279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 1.259995E+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 28164870. 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 0.018 0.034 85.212 85.213 qs_energies 1 2.0 0.000 0.000 84.758 84.765 ls_scf 1 3.0 0.000 0.000 83.848 83.855 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.788 72.994 multiply_cannon 111 7.7 0.018 0.021 56.006 57.415 multiply_cannon_loop 111 8.7 0.208 0.223 52.568 54.205 ls_scf_main 1 4.0 0.000 0.000 52.367 52.368 density_matrix_trs4 2 5.0 0.002 0.003 46.834 46.932 ls_scf_init_scf 1 4.0 0.000 0.000 28.403 28.404 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.346 27.394 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.142 25.165 mp_waitall_1 11316 10.9 22.532 25.139 22.532 25.139 multiply_cannon_multrec 2664 9.7 8.177 8.876 15.541 17.194 multiply_cannon_sync_h2d 2664 9.7 13.609 15.131 13.609 15.131 make_m2s 222 7.7 0.008 0.011 13.134 13.630 make_images 222 8.7 0.099 0.110 13.112 13.609 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.712 12.267 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.501 8.424 hybrid_alltoall_any 227 10.6 0.215 1.836 6.619 8.264 make_images_data 222 9.7 0.004 0.005 7.643 8.248 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.597 6.984 7.922 dbcsr_mm_accdrv_process_sort 4760 11.4 6.279 7.171 6.279 7.171 calculate_norms 4752 9.8 5.537 6.171 5.537 6.171 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.127 5.254 mp_sum_l 807 5.4 3.113 4.849 3.113 4.849 make_images_sizes 222 9.7 0.000 0.000 0.790 3.675 mp_alltoall_i44 222 10.7 0.789 3.675 0.789 3.675 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.073 3.520 mp_irecv_dv 6231 10.9 2.056 3.497 2.056 3.497 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.333 3.489 arnoldi_extremal 4 6.8 0.000 0.000 3.340 3.370 arnoldi_normal_ev 4 7.8 0.001 0.003 3.340 3.370 build_subspace 16 8.4 0.009 0.012 3.236 3.239 ls_scf_post 1 4.0 0.000 0.000 3.078 3.085 ls_scf_store_result 1 5.0 0.000 0.000 2.885 2.940 dbcsr_special_finalize 555 9.7 0.005 0.006 2.346 2.812 dbcsr_merge_single_wm 555 10.7 0.462 0.590 2.338 2.805 make_images_pack 222 9.7 2.209 2.634 2.211 2.636 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.341 2.568 dbcsr_sort_data 658 11.4 2.134 2.512 2.134 2.512 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.451 2.072 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.234 2.346 buffer_matrices_ensure_size 222 8.7 1.764 2.081 1.764 2.081 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.741 1.742 rebuild_ks_matrix 3 7.3 0.000 0.000 1.731 1.732 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.731 1.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.213000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 2.068955E+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 0.058 0.089 90.493 90.494 qs_energies 1 2.0 0.000 0.001 89.762 89.769 ls_scf 1 3.0 0.000 0.000 88.453 88.460 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.499 74.858 multiply_cannon 111 7.7 0.028 0.038 52.822 56.707 ls_scf_main 1 4.0 0.000 0.000 54.427 54.434 multiply_cannon_loop 111 8.7 0.117 0.125 49.524 52.822 density_matrix_trs4 2 5.0 0.002 0.003 48.827 49.046 ls_scf_init_scf 1 4.0 0.000 0.000 30.539 30.540 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.423 29.522 mp_waitall_1 9246 10.9 19.974 28.280 19.974 28.280 multiply_cannon_multrec 1332 9.7 13.218 16.662 22.506 27.053 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.027 27.052 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.110 19.776 make_m2s 222 7.7 0.006 0.008 14.785 15.390 make_images 222 8.7 1.582 1.985 14.755 15.359 dbcsr_mm_accdrv_process 4041 10.4 0.268 0.445 8.890 10.431 dbcsr_mm_accdrv_process_sort 4041 11.4 8.482 9.986 8.482 9.986 make_images_data 222 9.7 0.004 0.004 8.282 9.150 mp_sum_l 807 5.4 5.624 8.889 5.624 8.889 hybrid_alltoall_any 227 10.6 0.524 2.490 7.691 8.563 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.216 7.441 mp_irecv_dv 3311 11.0 3.196 7.380 3.196 7.380 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.302 6.969 calculate_norms 2376 9.8 6.039 6.734 6.039 6.734 multiply_cannon_sync_h2d 1332 9.7 4.864 6.119 4.864 6.119 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.005 5.198 arnoldi_extremal 4 6.8 0.000 0.000 4.790 4.822 arnoldi_normal_ev 4 7.8 0.001 0.005 4.790 4.822 build_subspace 16 8.4 0.014 0.021 4.522 4.525 ls_scf_post 1 4.0 0.000 0.000 3.487 3.494 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.205 3.430 ls_scf_store_result 1 5.0 0.000 0.000 3.165 3.309 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.228 2.746 3.230 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.212 2.861 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.484 2.589 make_images_pack 222 9.7 2.024 2.403 2.026 2.405 mp_allgather_i34 111 8.7 0.990 2.180 0.990 2.180 dbcsr_sort_data 436 11.2 1.796 2.034 1.796 2.034 dbcsr_data_new 4174 10.1 1.618 1.853 1.618 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.494000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1761.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 2.700464E+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 0.045 0.066 93.398 93.399 qs_energies 1 2.0 0.000 0.000 92.696 92.716 ls_scf 1 3.0 0.006 0.045 91.302 91.321 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.642 75.905 ls_scf_main 1 4.0 0.000 0.001 56.853 56.864 multiply_cannon 111 7.7 0.042 0.116 52.414 56.546 multiply_cannon_loop 111 8.7 0.100 0.106 48.741 52.572 density_matrix_trs4 2 5.0 0.002 0.003 50.999 51.180 mp_waitall_1 7374 11.0 22.753 32.529 22.753 32.529 ls_scf_init_scf 1 4.0 0.000 0.002 30.836 30.838 ls_scf_init_matrix_S 1 5.0 0.000 0.005 29.381 29.447 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.985 26.999 multiply_cannon_multrec 888 9.7 12.641 15.202 21.226 24.445 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.516 22.377 make_m2s 222 7.7 0.006 0.007 16.646 17.930 make_images 222 8.7 1.981 2.301 16.608 17.890 hybrid_alltoall_any 227 10.6 0.622 2.857 8.898 10.380 make_images_data 222 9.7 0.003 0.004 9.234 10.323 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.431 8.121 9.384 dbcsr_mm_accdrv_process_sort 3754 11.4 7.745 8.953 7.745 8.953 mp_sum_l 807 5.4 5.039 8.396 5.039 8.396 multiply_cannon_sync_h2d 888 9.7 6.022 7.381 6.022 7.381 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.467 6.996 mp_irecv_dv 2335 11.1 2.451 6.941 2.451 6.941 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.616 6.937 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.850 6.601 arnoldi_extremal 4 6.8 0.000 0.000 5.292 5.307 arnoldi_normal_ev 4 7.8 0.001 0.005 5.292 5.307 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.904 5.096 build_subspace 16 8.4 0.014 0.020 4.988 4.995 calculate_norms 1584 9.8 4.268 4.663 4.268 4.663 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.512 3.854 mp_allgather_i34 111 8.7 1.508 3.785 1.508 3.785 ls_scf_post 1 4.0 0.003 0.022 3.607 3.628 dbcsr_matrix_vector_mult_local 304 10.0 3.025 3.608 3.027 3.611 ls_scf_store_result 1 5.0 0.000 0.000 3.337 3.433 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.715 2.814 dbcsr_sort_data 325 11.1 1.860 2.141 1.860 2.141 make_images_pack 222 9.7 1.814 2.128 1.817 2.131 make_images_sizes 222 9.7 0.000 0.000 1.099 2.125 mp_alltoall_i44 222 10.7 1.099 2.125 1.099 2.125 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.051 2.053 rebuild_ks_matrix 3 7.3 0.000 0.000 2.033 2.035 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.029 2.033 2.035 dbcsr_data_release 9322 10.9 1.288 1.955 1.288 1.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.399000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2160.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 3.458724E+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 0.071 0.121 97.238 97.239 qs_energies 1 2.0 0.000 0.000 96.507 96.524 ls_scf 1 3.0 0.000 0.001 94.810 94.826 dbcsr_multiply_generic 111 6.7 0.018 0.027 78.324 78.609 ls_scf_main 1 4.0 0.001 0.021 58.605 58.606 multiply_cannon 111 7.7 0.050 0.096 51.459 56.322 density_matrix_trs4 2 5.0 0.002 0.003 52.596 52.729 multiply_cannon_loop 111 8.7 0.115 0.126 46.353 49.263 ls_scf_init_scf 1 4.0 0.000 0.002 32.948 32.950 ls_scf_init_matrix_S 1 5.0 0.001 0.005 31.582 31.654 mp_waitall_1 6438 11.0 22.598 29.951 22.598 29.951 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.093 29.112 multiply_cannon_multrec 1332 9.7 14.202 17.829 22.099 25.214 make_m2s 222 7.7 0.007 0.008 21.164 22.639 make_images 222 8.7 3.142 3.616 21.113 22.591 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.075 17.374 make_images_data 222 9.7 0.004 0.005 11.881 13.464 hybrid_alltoall_any 227 10.6 0.794 3.819 11.156 12.978 dbcsr_mm_accdrv_process 3641 10.4 0.205 0.420 7.539 9.059 dbcsr_mm_accdrv_process_sort 3641 11.4 7.161 8.638 7.161 8.638 mp_sum_l 807 5.4 4.137 7.301 4.137 7.301 multiply_cannon_sync_h2d 1332 9.7 5.538 6.250 5.538 6.250 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.062 5.930 mp_irecv_dv 3229 10.9 2.040 5.858 2.040 5.858 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.177 5.665 arnoldi_extremal 4 6.8 0.000 0.000 5.339 5.356 arnoldi_normal_ev 4 7.8 0.001 0.005 5.339 5.356 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.464 5.225 build_subspace 16 8.4 0.015 0.021 4.978 4.985 mp_allgather_i34 111 8.7 2.224 4.721 2.224 4.721 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.528 4.647 calculate_norms 2376 9.8 4.176 4.523 4.176 4.523 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.619 3.916 dbcsr_matrix_vector_mult_local 304 10.0 3.207 3.699 3.209 3.701 dbcsr_sort_data 658 11.4 3.073 3.459 3.073 3.459 ls_scf_post 1 4.0 0.004 0.018 3.257 3.274 dbcsr_special_finalize 555 9.7 0.006 0.007 2.820 3.203 dbcsr_merge_single_wm 555 10.7 0.536 0.662 2.811 3.194 ls_scf_store_result 1 5.0 0.000 0.000 2.999 3.051 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.904 2.961 dbcsr_data_release 10477 10.7 1.569 2.412 1.569 2.412 dbcsr_finalize 304 7.8 0.049 0.061 1.802 1.991 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.976 1.978 rebuild_ks_matrix 3 7.3 0.000 0.000 1.953 1.955 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.008 1.953 1.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.239000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2709.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 4.628382E+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 0.043 0.056 98.670 98.671 qs_energies 1 2.0 0.000 0.000 97.812 97.817 ls_scf 1 3.0 0.000 0.000 95.907 95.912 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.578 77.796 ls_scf_main 1 4.0 0.000 0.000 61.663 61.664 multiply_cannon 111 7.7 0.107 0.195 55.140 60.178 density_matrix_trs4 2 5.0 0.002 0.003 54.650 54.743 multiply_cannon_loop 111 8.7 0.070 0.076 50.568 52.169 mp_waitall_1 5481 11.0 26.201 32.668 26.201 32.668 ls_scf_init_scf 1 4.0 0.000 0.000 30.637 30.639 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.461 29.499 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.288 27.295 multiply_cannon_multrec 444 9.7 13.920 16.312 20.991 23.151 make_m2s 222 7.7 0.004 0.005 17.581 20.143 make_images 222 8.7 3.724 4.448 17.519 20.082 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.968 15.977 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.218 15.264 make_images_data 222 9.7 0.003 0.004 9.812 12.312 hybrid_alltoall_any 227 10.6 0.787 3.769 9.626 12.080 dbcsr_mm_accdrv_process 3003 10.4 0.177 0.351 6.783 7.903 multiply_cannon_sync_h2d 444 9.7 6.533 7.683 6.533 7.683 dbcsr_mm_accdrv_process_sort 3003 11.4 6.468 7.559 6.468 7.559 mp_allgather_i34 111 8.7 2.804 6.941 2.804 6.941 arnoldi_extremal 4 6.8 0.000 0.000 5.870 5.878 arnoldi_normal_ev 4 7.8 0.002 0.005 5.870 5.877 build_subspace 16 8.4 0.015 0.020 5.473 5.484 mp_sum_l 807 5.4 2.897 5.303 2.897 5.303 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.609 4.776 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.217 4.404 dbcsr_matrix_vector_mult_local 304 10.0 3.720 4.199 3.722 4.201 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.005 4.023 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.652 3.965 mp_irecv_dv 1241 11.2 1.632 3.940 1.632 3.940 calculate_norms 792 9.8 3.562 3.726 3.562 3.726 ls_scf_post 1 4.0 0.000 0.000 3.607 3.612 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.397 3.503 make_images_sizes 222 9.7 0.000 0.000 1.129 3.457 mp_alltoall_i44 222 10.7 1.129 3.457 1.129 3.457 ls_scf_store_result 1 5.0 0.000 0.000 3.369 3.433 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.332 dbcsr_merge_all 275 8.9 0.473 0.526 2.048 2.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.671000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3654.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/16f0f4032911ae89ed9404c17c23edc8a547187d_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 8.786391E+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 0.080 0.103 107.239 107.240 qs_energies 1 2.0 0.000 0.000 105.701 105.720 ls_scf 1 3.0 0.000 0.000 102.822 102.840 dbcsr_multiply_generic 111 6.7 0.023 0.026 76.796 76.925 ls_scf_main 1 4.0 0.000 0.000 65.297 65.298 density_matrix_trs4 2 5.0 0.002 0.003 56.447 56.508 multiply_cannon 111 7.7 0.147 0.253 49.607 51.205 multiply_cannon_loop 111 8.7 0.068 0.072 46.242 46.923 ls_scf_init_scf 1 4.0 0.000 0.000 33.852 33.853 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.456 32.470 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.633 29.643 mp_waitall_1 4569 11.1 21.834 25.441 21.834 25.441 make_m2s 222 7.7 0.005 0.005 23.668 24.550 make_images 222 8.7 4.588 4.973 23.562 24.443 multiply_cannon_multrec 444 9.7 17.834 18.512 22.461 23.087 hybrid_alltoall_any 227 10.6 1.662 3.648 12.660 15.500 make_images_data 222 9.7 0.003 0.003 12.914 15.237 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.349 10.843 multiply_cannon_sync_h2d 444 9.7 8.851 8.900 8.851 8.900 arnoldi_extremal 4 6.8 0.000 0.000 7.546 7.561 arnoldi_normal_ev 4 7.8 0.006 0.014 7.546 7.561 build_subspace 16 8.4 0.026 0.036 6.968 6.977 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.572 5.726 dbcsr_matrix_vector_mult_local 304 10.0 5.109 5.397 5.111 5.400 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.276 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.811 4.899 dbcsr_mm_accdrv_process 1814 10.4 0.230 0.317 4.454 4.590 dbcsr_mm_accdrv_process_sort 1814 11.4 4.156 4.295 4.156 4.295 ls_scf_post 1 4.0 0.000 0.000 3.672 3.690 make_images_sizes 222 9.7 0.000 0.000 1.422 3.570 mp_alltoall_i44 222 10.7 1.422 3.570 1.422 3.570 ls_scf_store_result 1 5.0 0.000 0.000 3.414 3.427 calculate_norms 792 9.8 3.236 3.272 3.236 3.272 dbcsr_finalize 304 7.8 0.082 0.090 3.075 3.124 mp_allgather_i34 111 8.7 1.006 3.043 1.006 3.043 dbcsr_complete_redistribute 5 7.6 1.438 1.481 2.766 2.911 dbcsr_merge_all 275 8.9 0.891 0.914 2.860 2.904 dbcsr_data_release 12724 10.6 2.330 2.860 2.330 2.860 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.849 2.849 matrix_ls_to_qs 2 6.0 0.000 0.000 2.414 2.579 dbcsr_sort_data 325 11.1 2.441 2.502 2.441 2.502 dbcsr_new_transposed 4 7.5 0.243 0.251 2.289 2.305 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.255 2.258 dbcsr_frobenius_norm 74 6.6 2.057 2.127 2.181 2.214 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.206 dbcsr_add_anytype 103 7.2 0.860 0.893 2.131 2.206 rebuild_ks_matrix 3 7.3 0.000 0.000 2.190 2.192 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.190 2.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.240000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6959.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 16f0f4032911ae89ed9404c17c23edc8a547187d Summary: empty Status: OK