=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90 ################# 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 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (19.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/01 job id: 42275805 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/02 job id: 42275809 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/03 job id: 42275810 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/04 job id: 42275812 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/05 job id: 42275816 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/06 job id: 42275818 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/07 job id: 42275820 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/08 job id: 42275823 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/09 job id: 42275827 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/10 job id: 42275831 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/11 job id: 42275836 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/12 job id: 42275839 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/13 job id: 42275842 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/14 job id: 42275845 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/15 job id: 42275847 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/16 job id: 42275850 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/17 job id: 42275853 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/18 job id: 42275855 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/19 job id: 42275856 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/20 job id: 42275872 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/21 job id: 42275875 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/22 job id: 42275876 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/23 job id: 42275877 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/24 job id: 42275878 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/25 job id: 42275879 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/26 job id: 42275881 --- 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/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 0.022 0.035 135.246 135.247 farming_run 1 2.0 134.668 134.670 135.217 135.221 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.472471E+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 0.009 0.027 116.428 116.429 qs_energies 1 2.0 0.000 0.000 116.160 116.161 mp2_main 1 3.0 0.000 0.000 114.020 114.020 mp2_gpw_main 1 4.0 0.019 0.026 113.045 113.046 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.712 94.256 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.591 56.136 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.887 47.391 get_2c_integrals 1 6.0 0.000 0.000 37.448 37.950 integrate_v_rspace 273 8.0 0.433 0.447 25.205 30.440 pw_transfer 6555 10.6 0.372 0.394 27.527 27.976 grid_integrate_task_list 273 9.0 21.008 26.738 21.008 26.738 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.168 26.644 fft_wrap_pw1pw2_100 2178 12.4 1.226 1.426 23.727 24.207 compute_2c_integrals 1 7.0 0.002 0.002 19.932 19.933 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.062 19.607 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.456 19.059 19.603 rpa_ri_compute_en 1 5.0 0.000 0.000 19.222 19.518 cp_fm_cholesky_decompose 12 8.2 17.535 18.093 17.535 18.093 cholesky_decomp 1 7.0 0.000 0.000 16.355 16.878 fft3d_s 5443 13.4 16.223 16.675 16.245 16.697 ao_to_mo_and_store_B_mult_1 272 7.0 10.870 15.598 10.870 15.598 calculate_wavefunction 272 8.0 5.441 5.596 12.587 13.213 rpa_num_int 1 6.0 0.000 0.000 10.691 10.691 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.577 10.689 calc_mat_Q 8 8.0 0.000 0.000 9.371 9.514 contract_S_to_Q 8 9.0 0.000 0.000 8.791 8.933 calc_potential_gpw 544 9.5 0.005 0.006 8.348 8.832 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.320 8.642 parallel_gemm_fm 14 9.1 0.000 0.000 8.370 8.431 parallel_gemm_fm_cosma 14 10.1 8.370 8.431 8.370 8.431 potential_pw2rs 545 10.0 0.108 0.110 7.726 8.388 create_integ_mat 1 6.0 0.006 0.008 7.894 7.894 collocate_single_gaussian 272 10.0 0.040 0.043 7.531 7.838 array2fm 1 7.0 0.000 0.000 6.812 7.361 pw_scatter_s 2720 13.7 4.445 4.609 4.445 4.609 pw_gather_s 2722 13.2 3.902 4.258 3.902 4.258 array2fm_buffer_send 1 8.0 3.020 3.179 3.020 3.179 pw_poisson_solve 545 10.5 1.138 1.185 2.206 2.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.045146, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 0.028 0.039 396.635 396.637 farming_run 1 2.0 395.764 395.776 396.596 396.599 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.218789E+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 0.032 209.222 209.223 qs_energies 1 2.0 0.000 0.000 208.921 208.930 scf_env_do_scf 1 3.0 0.000 0.000 105.864 105.864 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.947 104.954 rebuild_ks_matrix 4 6.0 0.000 0.000 104.946 104.953 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 104.946 104.953 hfx_ks_matrix 4 8.0 0.001 0.001 104.567 104.571 integrate_four_center 4 9.0 0.145 0.464 104.567 104.570 mp2_main 1 3.0 0.000 0.000 102.759 102.767 mp2_gpw_main 1 4.0 0.036 0.068 101.875 101.883 integrate_four_center_main 4 10.0 0.129 0.564 96.510 99.158 integrate_four_center_bin 263 11.0 96.381 98.949 96.381 98.949 init_scf_loop 1 4.0 0.000 0.000 91.617 91.617 mp2_ri_gpw_compute_in 1 5.0 0.066 0.085 75.026 76.066 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.026 54.582 55.619 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.162 42.189 47.357 integrate_v_rspace 95 8.0 0.399 0.576 28.557 33.582 pw_transfer 2240 10.6 0.145 0.167 29.929 30.361 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.936 29.381 ao_to_mo_and_store_B_mult_1 91 7.0 10.704 29.127 10.704 29.127 grid_integrate_task_list 95 9.0 23.842 29.077 23.842 29.077 mp2_ri_gpw_compute_en 1 5.0 0.055 0.067 26.678 28.456 fft_wrap_pw1pw2_100 730 12.4 1.282 1.433 26.650 27.142 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.847 1.925 24.996 25.005 get_2c_integrals 1 6.0 0.001 0.009 20.347 20.381 compute_2c_integrals 1 7.0 0.003 0.004 19.323 19.333 compute_2c_integrals_loop_lm 1 8.0 0.002 0.007 18.925 19.208 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.938 18.923 19.208 fft3d_s 1823 13.4 18.395 18.767 18.408 18.780 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.245 14.245 calculate_wavefunction 91 8.0 2.021 2.057 9.746 9.988 potential_pw2rs 186 10.0 0.034 0.035 8.618 9.203 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.573 8.710 9.067 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.281 8.542 local_gemm 172 8.0 8.157 8.527 8.157 8.527 mp2_ri_gpw_compute_en_comm 22 7.0 0.493 0.513 8.047 8.425 calc_potential_gpw 182 9.5 0.002 0.002 7.976 8.180 collocate_single_gaussian 91 10.0 0.020 0.071 7.896 8.067 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.475 6.346 6.475 mp_sendrecv_dm3 2068 8.0 6.092 6.432 6.092 6.432 pw_gather_s 912 13.2 4.908 5.511 4.908 5.511 mp_sync 38 10.4 2.714 4.783 2.714 4.783 pw_scatter_s 910 13.7 3.984 4.236 3.984 4.236 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.873991, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 452.030464E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.038 53.565 53.567 qs_mol_dyn_low 1 2.0 0.003 0.003 53.330 53.337 qs_forces 11 3.9 0.003 0.006 53.262 53.265 qs_energies 11 4.9 0.001 0.003 51.810 51.828 scf_env_do_scf 11 5.9 0.000 0.001 45.720 45.720 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.587 43.588 dbcsr_multiply_generic 2286 12.5 0.092 0.095 34.702 35.279 qs_scf_new_mos 108 7.5 0.000 0.000 33.294 33.590 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.294 33.589 ot_scf_mini 108 9.5 0.002 0.002 31.596 31.780 multiply_cannon 2286 13.5 0.195 0.205 26.554 28.435 multiply_cannon_loop 2286 14.5 1.473 1.567 25.702 27.548 velocity_verlet 10 3.0 0.001 0.001 26.029 26.030 ot_mini 108 10.5 0.001 0.001 20.341 20.597 qs_ot_get_derivative 108 11.5 0.001 0.001 17.310 17.550 mp_waitall_1 267858 16.1 9.293 16.051 9.293 16.051 multiply_cannon_metrocomm3 54864 15.5 0.067 0.071 6.087 13.769 multiply_cannon_multrec 54864 15.5 4.289 6.659 7.611 11.124 rebuild_ks_matrix 119 8.3 0.000 0.000 7.951 8.104 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.951 8.104 multiply_cannon_sync_h2d 54864 15.5 5.944 7.870 5.944 7.870 mp_sum_l 7207 12.9 5.759 7.674 5.759 7.674 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.041 7.175 qs_ot_get_p 119 10.4 0.001 0.001 6.492 6.844 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.019 6.154 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.579 6.054 init_scf_run 11 5.9 0.000 0.001 4.812 4.812 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.812 4.812 dbcsr_mm_accdrv_process 76910 16.1 1.126 1.849 3.245 4.637 sum_up_and_integrate 119 10.3 0.012 0.015 4.582 4.588 integrate_v_rspace 119 11.3 0.002 0.002 4.570 4.577 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.290 4.392 calculate_rho_elec 119 8.7 0.011 0.016 4.290 4.391 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 2.101 3.836 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.488 3.568 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.920 3.115 apply_single 119 13.6 0.000 0.000 2.919 3.115 calculate_dm_sparse 119 9.5 0.000 0.000 2.917 3.054 rs_pw_transfer 974 11.9 0.011 0.012 2.917 3.043 ot_diis_step 108 11.5 0.006 0.006 2.726 2.726 jit_kernel_multiply 13 15.8 2.060 2.711 2.060 2.711 calculate_first_density_matrix 1 7.0 0.001 0.002 2.623 2.631 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.622 2.623 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.499 2.579 density_rs2pw 119 9.7 0.003 0.004 2.338 2.457 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.413 2.415 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.389 2.390 cp_fm_redistribute_end 50 14.0 2.170 2.365 2.174 2.367 cp_fm_diag_elpa_base 50 14.0 0.191 2.308 0.192 2.320 acc_transpose_blocks 54864 15.5 0.213 0.241 1.780 2.311 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.123 2.174 make_m2s 4572 13.5 0.053 0.055 2.083 2.161 grid_integrate_task_list 119 12.3 2.023 2.135 2.023 2.135 wfi_extrapolate 11 7.9 0.001 0.001 2.107 2.107 mp_sum_d 4129 12.0 1.433 2.079 1.433 2.079 init_scf_loop 11 6.9 0.000 0.001 2.078 2.079 make_images 4572 14.5 0.132 0.138 2.001 2.078 potential_pw2rs 119 12.3 0.004 0.004 1.941 1.954 pw_transfer 1439 11.6 0.051 0.055 1.769 1.838 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.694 1.764 mp_alltoall_d11v 2130 13.8 1.484 1.707 1.484 1.707 fft3d_ps 1201 14.6 0.355 0.456 1.472 1.536 mp_waitany 12084 13.8 1.282 1.469 1.282 1.469 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.383 1.401 grid_collocate_task_list 119 9.7 1.290 1.349 1.290 1.349 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.270 1.342 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.793 1.229 acc_transpose_blocks_kernels 54864 16.5 0.240 0.382 0.856 1.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.567000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 489.361408E+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 1100735. 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 0.015 0.033 42.111 42.112 qs_mol_dyn_low 1 2.0 0.003 0.003 41.831 41.839 qs_forces 11 3.9 0.002 0.003 41.713 41.714 qs_energies 11 4.9 0.004 0.012 39.894 39.898 scf_env_do_scf 11 5.9 0.001 0.001 34.270 34.270 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 31.313 31.315 dbcsr_multiply_generic 2286 12.5 0.099 0.102 22.908 23.354 qs_scf_new_mos 108 7.5 0.001 0.001 21.520 21.790 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.519 21.789 ot_scf_mini 108 9.5 0.003 0.004 20.571 20.756 velocity_verlet 10 3.0 0.001 0.002 20.133 20.135 multiply_cannon 2286 13.5 0.211 0.219 17.111 18.768 multiply_cannon_loop 2286 14.5 0.906 0.979 15.805 17.569 ot_mini 108 10.5 0.001 0.001 12.540 12.786 mp_waitall_1 217478 16.2 7.185 12.365 7.185 12.365 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.714 10.181 qs_ot_get_derivative 108 11.5 0.001 0.001 9.975 10.162 multiply_cannon_multrec 27432 15.5 1.938 4.306 5.854 8.561 rebuild_ks_matrix 119 8.3 0.000 0.000 7.829 8.006 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.828 8.005 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.903 7.068 dbcsr_mm_accdrv_process 47894 16.0 3.109 5.436 3.847 5.703 qs_ot_get_p 119 10.4 0.001 0.002 4.826 5.080 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.793 4.648 sum_up_and_integrate 119 10.3 0.024 0.027 4.590 4.598 integrate_v_rspace 119 11.3 0.002 0.002 4.566 4.577 mp_sum_l 7207 12.9 2.451 4.439 2.451 4.439 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.115 4.213 apply_single 119 13.6 0.000 0.000 3.115 4.212 init_scf_run 11 5.9 0.000 0.001 4.169 4.170 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.169 4.169 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.936 3.967 calculate_rho_elec 119 8.7 0.021 0.024 3.935 3.966 make_m2s 4572 13.5 0.052 0.054 2.902 3.172 rs_pw_transfer 974 11.9 0.010 0.011 3.060 3.158 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.084 3.108 make_images 4572 14.5 0.200 0.237 2.814 3.082 init_scf_loop 11 6.9 0.000 0.002 2.905 2.910 multiply_cannon_sync_h2d 27432 15.5 2.160 2.757 2.160 2.757 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.575 2.576 calculate_first_density_matrix 1 7.0 0.001 0.004 2.515 2.517 ot_diis_step 108 11.5 0.011 0.011 2.502 2.503 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.366 2.467 density_rs2pw 119 9.7 0.004 0.004 2.280 2.382 potential_pw2rs 119 12.3 0.006 0.007 2.239 2.250 calculate_dm_sparse 119 9.5 0.000 0.000 2.091 2.180 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.137 2.137 cp_fm_redistribute_end 50 14.0 1.761 2.103 1.765 2.105 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.096 2.099 pw_transfer 1439 11.6 0.063 0.066 2.010 2.047 cp_fm_diag_elpa_base 50 14.0 0.330 2.014 0.338 2.047 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.922 1.961 grid_integrate_task_list 119 12.3 1.834 1.936 1.834 1.936 prepare_preconditioner 11 7.9 0.000 0.000 1.870 1.896 make_preconditioner 11 8.9 0.000 0.000 1.870 1.896 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.833 1.880 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.764 1.822 make_images_data 4572 15.5 0.045 0.051 1.272 1.788 jit_kernel_multiply 9 16.0 0.685 1.751 0.685 1.751 fft3d_ps 1201 14.6 0.496 0.555 1.629 1.661 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.102 1.626 wfi_extrapolate 11 7.9 0.001 0.001 1.599 1.599 acc_transpose_blocks 27432 15.5 0.107 0.111 1.256 1.586 mp_allgather_i34 2286 14.5 0.747 1.570 0.747 1.570 mp_alltoall_d11v 2130 13.8 1.420 1.518 1.420 1.518 qs_ot_get_orbitals 108 10.5 0.000 0.001 1.408 1.457 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.393 1.430 grid_collocate_task_list 119 9.7 1.240 1.330 1.240 1.330 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.273 1.287 mp_sum_d 4129 12.0 0.762 1.221 0.762 1.221 make_images_sizes 4572 15.5 0.005 0.005 0.821 1.102 mp_alltoall_i44 4572 16.5 0.817 1.097 0.817 1.097 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.043 1.045 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.993 1.009 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.729 0.987 cp_fm_cholesky_invert 11 10.9 0.931 0.936 0.931 0.936 mp_alltoall_z22v 1201 16.6 0.854 0.930 0.854 0.930 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.606 0.865 0.913 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.493 0.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.112000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.727273, yerr=2.048987 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.337728E+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 0.059 0.092 35.049 35.050 qs_mol_dyn_low 1 2.0 0.003 0.003 34.613 34.620 qs_forces 11 3.9 0.003 0.005 34.535 34.540 qs_energies 11 4.9 0.002 0.005 32.929 32.935 scf_env_do_scf 11 5.9 0.001 0.001 27.944 27.944 scf_env_do_scf_inner_loop 108 6.5 0.017 0.124 24.975 24.975 dbcsr_multiply_generic 2286 12.5 0.092 0.095 17.520 17.691 velocity_verlet 10 3.0 0.001 0.002 16.678 16.680 qs_scf_new_mos 108 7.5 0.001 0.001 16.199 16.214 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.198 16.213 ot_scf_mini 108 9.5 0.002 0.003 15.390 15.413 multiply_cannon 2286 13.5 0.195 0.204 13.437 14.349 multiply_cannon_loop 2286 14.5 0.641 0.664 12.520 13.431 ot_mini 108 10.5 0.001 0.001 9.485 9.507 qs_ot_get_derivative 108 11.5 0.001 0.001 7.854 7.871 multiply_cannon_multrec 18288 15.5 1.932 3.045 6.761 7.184 rebuild_ks_matrix 119 8.3 0.000 0.000 6.722 6.749 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.721 6.749 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.971 5.995 dbcsr_mm_accdrv_process 38222 16.0 4.082 5.408 4.746 5.584 mp_waitall_1 169478 16.3 3.800 4.888 3.800 4.888 sum_up_and_integrate 119 10.3 0.030 0.031 4.281 4.285 integrate_v_rspace 119 11.3 0.002 0.003 4.251 4.259 init_scf_run 11 5.9 0.000 0.001 3.723 3.723 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.723 3.723 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.652 3.659 calculate_rho_elec 119 8.7 0.030 0.031 3.651 3.658 qs_ot_get_p 119 10.4 0.001 0.001 3.576 3.608 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.912 3.560 init_scf_loop 11 6.9 0.000 0.002 2.946 2.947 rs_pw_transfer 974 11.9 0.009 0.009 2.756 2.870 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.642 2.621 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.126 2.459 apply_single 119 13.6 0.000 0.000 2.126 2.459 calculate_first_density_matrix 1 7.0 0.001 0.003 2.370 2.371 make_m2s 4572 13.5 0.044 0.045 2.152 2.340 density_rs2pw 119 9.7 0.004 0.004 2.220 2.319 make_images 4572 14.5 0.191 0.205 2.067 2.255 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.216 2.225 prepare_preconditioner 11 7.9 0.000 0.000 2.035 2.039 make_preconditioner 11 8.9 0.002 0.012 2.035 2.039 potential_pw2rs 119 12.3 0.007 0.008 2.025 2.039 pw_transfer 1439 11.6 0.063 0.066 1.955 1.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.884 1.964 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.930 1.930 calculate_dm_sparse 119 9.5 0.000 0.000 1.902 1.920 mp_sum_l 7207 12.9 1.558 1.916 1.558 1.916 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.865 1.888 grid_integrate_task_list 119 12.3 1.793 1.874 1.793 1.874 jit_kernel_multiply 10 16.2 0.613 1.854 0.613 1.854 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.792 1.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.801 1.802 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.653 1.655 cp_fm_redistribute_end 50 14.0 1.227 1.625 1.228 1.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.598 1.606 ot_diis_step 108 11.5 0.011 0.011 1.595 1.595 cp_fm_diag_elpa_base 50 14.0 0.380 1.541 0.394 1.581 fft3d_ps 1201 14.6 0.508 0.527 1.553 1.576 multiply_cannon_sync_h2d 18288 15.5 1.311 1.482 1.311 1.482 fft_wrap_pw1pw2_140 487 13.2 0.085 0.090 1.366 1.383 acc_transpose_blocks 18288 15.5 0.074 0.076 1.311 1.339 wfi_extrapolate 11 7.9 0.001 0.001 1.300 1.300 grid_collocate_task_list 119 9.7 1.203 1.267 1.203 1.267 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.190 1.194 make_images_data 4572 15.5 0.045 0.048 0.952 1.157 hybrid_alltoall_any 4725 16.4 0.054 0.112 0.812 1.041 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.009 1.023 mp_alltoall_d11v 2130 13.8 0.846 1.005 0.846 1.005 qs_energies_init_hamiltonians 11 5.9 0.010 0.025 0.997 1.003 mp_alltoall_z22v 1201 16.6 0.851 0.928 0.851 0.928 cp_fm_cholesky_invert 11 10.9 0.916 0.921 0.916 0.921 acc_transpose_blocks_kernels 18288 16.5 0.210 0.216 0.873 0.895 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.828 0.831 mp_allgather_i34 2286 14.5 0.378 0.801 0.378 0.801 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.659 0.739 mp_waitany 9880 13.7 0.575 0.701 0.575 0.701 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.050000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=1.966664 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 562.028544E+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 0.019 0.032 36.416 36.417 qs_mol_dyn_low 1 2.0 0.003 0.003 36.165 36.172 qs_forces 11 3.9 0.002 0.004 36.055 36.056 qs_energies 11 4.9 0.005 0.011 34.287 34.294 scf_env_do_scf 11 5.9 0.001 0.001 28.986 28.987 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 25.352 25.355 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.509 18.627 velocity_verlet 10 3.0 0.001 0.001 18.337 18.341 qs_scf_new_mos 108 7.5 0.001 0.001 16.824 16.887 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.824 16.887 ot_scf_mini 108 9.5 0.003 0.003 15.872 15.928 multiply_cannon 2286 13.5 0.229 0.261 14.529 15.047 multiply_cannon_loop 2286 14.5 0.937 0.970 13.519 13.911 ot_mini 108 10.5 0.001 0.001 9.526 9.597 multiply_cannon_multrec 27432 15.5 2.358 3.052 8.653 9.030 qs_ot_get_derivative 108 11.5 0.001 0.001 7.707 7.762 dbcsr_mm_accdrv_process 47916 15.9 5.437 7.012 6.197 7.443 rebuild_ks_matrix 119 8.3 0.000 0.000 6.721 6.768 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.721 6.768 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.961 6.003 sum_up_and_integrate 119 10.3 0.035 0.038 3.946 3.956 integrate_v_rspace 119 11.3 0.002 0.003 3.912 3.922 init_scf_run 11 5.9 0.000 0.001 3.854 3.855 scf_env_initial_rho_setup 11 6.9 0.002 0.006 3.854 3.855 qs_ot_get_p 119 10.4 0.001 0.002 3.713 3.785 init_scf_loop 11 6.9 0.001 0.005 3.608 3.612 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.531 3.567 calculate_rho_elec 119 8.7 0.040 0.046 3.531 3.566 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.769 3.259 mp_waitall_1 145218 16.4 2.328 2.934 2.328 2.934 prepare_preconditioner 11 7.9 0.000 0.000 2.697 2.706 make_preconditioner 11 8.9 0.001 0.002 2.697 2.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.281 2.619 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.131 2.589 apply_single 119 13.6 0.000 0.000 2.131 2.588 make_m2s 4572 13.5 0.054 0.056 2.426 2.534 make_images 4572 14.5 0.273 0.332 2.319 2.426 calculate_first_density_matrix 1 7.0 0.003 0.012 2.394 2.398 rs_pw_transfer 974 11.9 0.009 0.009 2.228 2.335 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.240 2.253 calculate_dm_sparse 119 9.5 0.000 0.000 2.045 2.108 density_rs2pw 119 9.7 0.003 0.004 1.987 2.103 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.073 2.099 pw_transfer 1439 11.6 0.062 0.066 1.955 1.988 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.934 1.936 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.912 1.915 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.866 1.903 grid_integrate_task_list 119 12.3 1.823 1.878 1.823 1.878 jit_kernel_multiply 10 16.1 0.700 1.824 0.700 1.824 ot_diis_step 108 11.5 0.012 0.012 1.777 1.777 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.733 1.746 potential_pw2rs 119 12.3 0.008 0.010 1.690 1.699 mp_sum_l 7207 12.9 1.140 1.666 1.140 1.666 acc_transpose_blocks 27432 15.5 0.111 0.113 1.538 1.575 fft3d_ps 1201 14.6 0.534 0.587 1.547 1.575 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.558 1.559 cp_fm_redistribute_end 50 14.0 1.022 1.521 1.024 1.522 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.460 1.498 cp_fm_diag_elpa_base 50 14.0 0.474 1.450 0.495 1.492 wfi_extrapolate 11 7.9 0.001 0.001 1.397 1.397 grid_collocate_task_list 119 9.7 1.221 1.304 1.221 1.304 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.226 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.680 1.196 dbcsr_complete_redistribute 329 12.2 0.121 0.149 0.887 1.168 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 1.164 1.166 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.131 1.155 cp_fm_upper_to_full 72 13.5 0.821 1.142 0.821 1.142 make_images_data 4572 15.5 0.045 0.048 0.960 1.099 multiply_cannon_sync_h2d 27432 15.5 0.987 1.070 0.987 1.070 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.828 1.030 mp_alltoall_d11v 2130 13.8 0.908 1.019 0.908 1.019 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.915 0.936 mp_alltoall_z22v 1201 16.6 0.895 0.920 0.895 0.920 copy_fm_to_dbcsr 176 11.2 0.001 0.004 0.628 0.900 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.857 0.861 cp_fm_cholesky_invert 11 10.9 0.804 0.807 0.804 0.807 mp_alltoall_i22 627 13.8 0.465 0.759 0.465 0.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.417000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=533.000000, yerr=3.643175 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 599.359488E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 29.135 29.135 qs_mol_dyn_low 1 2.0 0.003 0.003 28.954 28.960 qs_forces 11 3.9 0.002 0.002 28.865 28.868 qs_energies 11 4.9 0.001 0.002 27.149 27.155 scf_env_do_scf 11 5.9 0.000 0.001 22.331 22.331 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.628 19.629 velocity_verlet 10 3.0 0.001 0.001 14.903 14.907 dbcsr_multiply_generic 2286 12.5 0.090 0.092 12.766 12.902 qs_scf_new_mos 108 7.5 0.001 0.001 11.812 11.845 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.812 11.845 ot_scf_mini 108 9.5 0.002 0.002 11.127 11.158 multiply_cannon 2286 13.5 0.230 0.242 9.956 10.377 multiply_cannon_loop 2286 14.5 0.331 0.342 9.029 9.292 ot_mini 108 10.5 0.001 0.001 6.203 6.242 multiply_cannon_multrec 9144 15.5 1.605 2.035 5.783 6.170 rebuild_ks_matrix 119 8.3 0.000 0.000 5.940 5.975 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.940 5.975 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.305 5.337 qs_ot_get_derivative 108 11.5 0.001 0.001 4.824 4.855 dbcsr_mm_accdrv_process 12550 15.8 3.151 4.103 4.075 4.157 sum_up_and_integrate 119 10.3 0.038 0.042 3.635 3.639 integrate_v_rspace 119 11.3 0.003 0.003 3.597 3.602 init_scf_run 11 5.9 0.000 0.001 3.370 3.371 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.370 3.370 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.363 3.370 calculate_rho_elec 119 8.7 0.060 0.061 3.362 3.369 qs_ot_get_p 119 10.4 0.001 0.001 3.071 3.120 init_scf_loop 11 6.9 0.000 0.000 2.678 2.679 mp_waitall_1 121218 16.5 2.216 2.678 2.216 2.678 calculate_first_density_matrix 1 7.0 0.001 0.001 2.158 2.159 make_m2s 4572 13.5 0.034 0.035 1.891 2.058 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.967 1.970 make_images 4572 14.5 0.267 0.302 1.802 1.967 grid_integrate_task_list 119 12.3 1.858 1.945 1.858 1.945 prepare_preconditioner 11 7.9 0.000 0.000 1.922 1.925 make_preconditioner 11 8.9 0.000 0.000 1.922 1.925 rs_pw_transfer 974 11.9 0.008 0.008 1.799 1.903 density_rs2pw 119 9.7 0.003 0.004 1.777 1.888 pw_transfer 1439 11.6 0.063 0.066 1.818 1.829 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.799 1.827 calculate_dm_sparse 119 9.5 0.000 0.000 1.741 1.762 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.760 1.761 jit_kernel_multiply 10 15.6 0.886 1.753 0.886 1.753 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.728 1.737 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.617 1.637 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.633 1.635 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.449 1.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.420 1.431 cp_fm_redistribute_end 50 14.0 0.727 1.419 0.729 1.421 potential_pw2rs 119 12.3 0.010 0.010 1.405 1.409 fft3d_ps 1201 14.6 0.541 0.553 1.393 1.400 fft_wrap_pw1pw2_140 487 13.2 0.081 0.084 1.385 1.397 cp_fm_diag_elpa_base 50 14.0 0.646 1.328 0.690 1.386 grid_collocate_task_list 119 9.7 1.273 1.381 1.273 1.381 ot_diis_step 108 11.5 0.012 0.013 1.363 1.364 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.224 1.254 apply_single 119 13.6 0.000 0.000 1.224 1.254 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.216 1.219 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.188 1.201 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.158 1.164 wfi_extrapolate 11 7.9 0.001 0.001 1.164 1.164 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.830 1.110 make_images_data 4572 15.5 0.039 0.042 0.876 1.101 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.474 1.042 cp_fm_cholesky_invert 11 10.9 1.004 1.008 1.004 1.008 acc_transpose_blocks 9144 15.5 0.039 0.039 0.941 0.953 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.923 mp_alltoall_d11v 2130 13.8 0.849 0.921 0.849 0.921 multiply_cannon_sync_h2d 9144 15.5 0.709 0.786 0.709 0.786 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.760 0.770 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.759 0.761 mp_alltoall_z22v 1201 16.6 0.725 0.757 0.725 0.757 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.672 0.718 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.701 0.709 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.511 0.700 mp_allgather_i34 2286 14.5 0.260 0.689 0.260 0.689 dbcsr_complete_redistribute 329 12.2 0.175 0.193 0.612 0.644 mp_sum_l 7207 12.9 0.500 0.639 0.500 0.639 jit_kernel_transpose 5 15.6 0.585 0.593 0.585 0.593 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.135000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.272727, yerr=5.046888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 753.610752E+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 0.015 0.034 41.690 41.690 qs_mol_dyn_low 1 2.0 0.003 0.003 41.475 41.481 qs_forces 11 3.9 0.002 0.002 41.345 41.346 qs_energies 11 4.9 0.001 0.001 39.390 39.393 scf_env_do_scf 11 5.9 0.000 0.001 33.301 33.301 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.227 25.229 velocity_verlet 10 3.0 0.001 0.001 23.456 23.471 dbcsr_multiply_generic 2286 12.5 0.100 0.106 18.014 18.127 qs_scf_new_mos 108 7.5 0.001 0.001 16.169 16.255 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.168 16.255 ot_scf_mini 108 9.5 0.002 0.002 15.072 15.166 multiply_cannon 2286 13.5 0.303 0.316 13.950 14.734 multiply_cannon_loop 2286 14.5 0.343 0.350 12.636 13.439 ot_mini 108 10.5 0.001 0.001 8.991 9.098 multiply_cannon_multrec 9144 15.5 3.450 5.088 8.687 8.788 init_scf_loop 11 6.9 0.000 0.000 8.044 8.046 prepare_preconditioner 11 7.9 0.000 0.000 7.119 7.133 make_preconditioner 11 8.9 0.000 0.000 7.119 7.132 rebuild_ks_matrix 119 8.3 0.000 0.000 6.937 7.068 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.937 7.068 qs_ot_get_derivative 108 11.5 0.001 0.001 6.957 7.053 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.591 7.008 dbcsr_mm_accdrv_process 12550 15.8 4.024 5.666 5.110 6.457 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.276 6.395 cp_fm_upper_to_full 72 14.2 3.277 4.751 3.277 4.751 init_scf_run 11 5.9 0.000 0.001 4.065 4.066 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.065 4.065 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.841 3.895 calculate_rho_elec 119 8.7 0.118 0.121 3.841 3.894 sum_up_and_integrate 119 10.3 0.064 0.065 3.738 3.745 integrate_v_rspace 119 11.3 0.003 0.003 3.674 3.683 mp_waitall_1 97218 16.6 2.779 3.650 2.779 3.650 qs_ot_get_p 119 10.4 0.001 0.001 3.288 3.417 dbcsr_complete_redistribute 329 12.2 0.295 0.298 2.108 2.991 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.533 2.949 make_m2s 4572 13.5 0.037 0.038 2.458 2.684 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.780 2.648 make_images 4572 14.5 0.350 0.379 2.338 2.564 calculate_first_density_matrix 1 7.0 0.000 0.001 2.504 2.560 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.185 2.470 apply_single 119 13.6 0.000 0.000 2.184 2.470 mp_alltoall_i22 627 13.8 1.559 2.450 1.559 2.450 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.523 2.385 calculate_dm_sparse 119 9.5 0.000 0.000 2.268 2.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.174 2.174 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.360 2.137 pw_transfer 1439 11.6 0.066 0.067 2.110 2.112 grid_integrate_task_list 119 12.3 2.024 2.049 2.024 2.049 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.015 2.017 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.004 2.006 ot_diis_step 108 11.5 0.014 0.014 2.002 2.004 density_rs2pw 119 9.7 0.003 0.003 1.867 1.876 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.824 1.846 jit_kernel_multiply 10 15.4 1.058 1.840 1.058 1.840 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.718 1.764 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.752 1.753 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.722 1.722 mp_sum_l 7207 12.9 1.068 1.711 1.068 1.711 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.671 1.676 fft3d_ps 1201 14.6 0.567 0.578 1.649 1.651 cp_fm_cholesky_invert 11 10.9 1.532 1.536 1.532 1.536 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.236 1.526 make_images_data 4572 15.5 0.042 0.045 1.194 1.482 grid_collocate_task_list 119 9.7 1.459 1.469 1.459 1.469 rs_pw_transfer 974 11.9 0.009 0.009 1.435 1.452 wfi_extrapolate 11 7.9 0.001 0.001 1.443 1.444 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.416 1.418 cp_fm_diag_elpa_base 50 14.0 1.264 1.326 1.415 1.417 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.320 1.370 mp_alltoall_d11v 2130 13.8 1.272 1.344 1.272 1.344 potential_pw2rs 119 12.3 0.014 0.015 1.241 1.244 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.156 1.172 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.092 1.107 multiply_cannon_sync_h2d 9144 15.5 1.043 1.049 1.043 1.049 acc_transpose_blocks 9144 15.5 0.038 0.039 1.039 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.028 mp_alltoall_z22v 1201 16.6 0.946 0.965 0.946 0.965 qs_create_task_list 11 7.9 0.001 0.002 0.950 0.961 generate_qs_task_list 11 8.9 0.372 0.393 0.950 0.960 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.873 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.690000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=700.636364, yerr=15.143392 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.321728E+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 402407. 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 0.014 0.029 86.159 86.160 qs_mol_dyn_low 1 2.0 0.003 0.003 85.901 85.911 qs_forces 11 3.9 0.002 0.002 85.829 85.830 qs_energies 11 4.9 0.001 0.001 82.894 82.907 scf_env_do_scf 11 5.9 0.000 0.001 73.469 73.472 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.438 67.438 dbcsr_multiply_generic 2055 12.4 0.104 0.107 53.736 53.951 qs_scf_new_mos 99 7.5 0.000 0.001 49.769 49.899 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.769 49.898 ot_scf_mini 99 9.5 0.002 0.002 47.314 47.469 multiply_cannon 2055 13.4 0.184 0.194 43.326 44.195 velocity_verlet 10 3.0 0.001 0.001 44.174 44.175 multiply_cannon_loop 2055 14.4 1.563 1.602 42.229 42.997 ot_mini 99 10.5 0.001 0.001 28.774 28.932 qs_ot_get_derivative 99 11.5 0.001 0.001 21.903 22.022 multiply_cannon_multrec 49320 15.4 11.997 12.786 17.068 17.896 rebuild_ks_matrix 110 8.3 0.000 0.000 14.903 15.061 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.903 15.061 mp_waitall_1 241148 16.1 13.027 14.157 13.027 14.157 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.107 13.251 multiply_cannon_sync_h2d 49320 15.4 9.816 10.479 9.816 10.479 qs_ot_get_p 110 10.4 0.001 0.001 9.840 10.002 multiply_cannon_metrocomm3 49320 15.4 0.075 0.079 7.263 8.339 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.610 8.123 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.140 7.648 apply_single 110 13.6 0.000 0.001 7.139 7.648 sum_up_and_integrate 110 10.3 0.036 0.043 7.430 7.443 integrate_v_rspace 110 11.3 0.002 0.003 7.393 7.415 init_scf_run 11 5.9 0.000 0.001 7.325 7.325 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.324 7.325 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.931 7.019 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.602 6.739 calculate_rho_elec 110 8.6 0.020 0.024 6.602 6.738 mp_sum_l 6514 12.8 5.760 6.688 5.760 6.688 ot_diis_step 99 11.5 0.006 0.006 6.558 6.559 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.277 6.303 init_scf_loop 11 6.9 0.000 0.000 5.999 5.999 dbcsr_mm_accdrv_process 87628 16.1 2.118 2.254 4.951 5.283 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.172 5.173 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.607 4.608 cp_fm_redistribute_end 48 14.0 4.008 4.573 4.011 4.574 cp_fm_diag_elpa_base 48 14.0 0.556 4.459 0.560 4.487 rs_pw_transfer 902 11.9 0.012 0.014 4.253 4.428 wfi_extrapolate 11 7.9 0.001 0.001 4.337 4.337 make_m2s 4110 13.4 0.061 0.065 4.200 4.314 make_images 4110 14.4 0.178 0.193 4.104 4.221 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 3.081 4.171 calculate_dm_sparse 110 9.5 0.000 0.001 3.852 3.947 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.824 3.827 prepare_preconditioner 11 7.9 0.000 0.000 3.782 3.800 make_preconditioner 11 8.9 0.000 0.000 3.782 3.800 density_rs2pw 110 9.6 0.004 0.004 3.649 3.788 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.553 3.605 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.433 3.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.441 3.485 grid_integrate_task_list 110 12.3 3.255 3.410 3.255 3.410 pw_transfer 1331 11.6 0.053 0.063 3.146 3.216 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.059 3.129 potential_pw2rs 110 12.3 0.006 0.007 2.942 2.978 calculate_first_density_matrix 1 7.0 0.000 0.001 2.899 2.908 fft3d_ps 1111 14.6 0.752 0.836 2.599 2.660 mp_alltoall_d11v 2046 13.8 2.296 2.645 2.296 2.645 fft_wrap_pw1pw2_140 451 13.1 0.167 0.186 2.522 2.593 jit_kernel_multiply 13 15.9 2.548 2.583 2.548 2.583 acc_transpose_blocks 49320 15.4 0.211 0.220 2.363 2.460 grid_collocate_task_list 110 9.6 2.087 2.190 2.087 2.190 mp_waitany 14300 13.8 1.868 2.176 1.868 2.176 cp_fm_cholesky_invert 11 10.9 2.124 2.129 2.124 2.129 mp_sum_d 3883 11.9 1.620 2.080 1.620 2.080 make_images_data 4110 15.4 0.043 0.046 1.872 2.033 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.953 hybrid_alltoall_any 4261 16.3 0.081 0.478 1.621 1.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.160000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.672171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 588.447744E+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 2962689. 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 0.053 0.069 69.450 69.451 qs_mol_dyn_low 1 2.0 0.003 0.003 69.117 69.152 qs_forces 11 3.9 0.003 0.004 69.036 69.041 qs_energies 11 4.9 0.005 0.022 65.733 65.740 scf_env_do_scf 11 5.9 0.001 0.001 56.841 56.845 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.827 48.829 dbcsr_multiply_generic 2055 12.4 0.112 0.116 38.049 38.296 velocity_verlet 10 3.0 0.001 0.002 36.542 36.544 qs_scf_new_mos 99 7.5 0.001 0.001 32.888 33.036 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.887 33.036 multiply_cannon 2055 13.4 0.231 0.256 31.128 32.740 ot_scf_mini 99 9.5 0.003 0.004 31.218 31.351 multiply_cannon_loop 2055 14.4 0.929 0.950 29.712 30.975 ot_mini 99 10.5 0.001 0.001 18.379 18.522 multiply_cannon_multrec 24660 15.4 7.576 9.489 13.780 15.556 rebuild_ks_matrix 110 8.3 0.000 0.001 13.684 13.838 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.684 13.838 qs_ot_get_derivative 99 11.5 0.001 0.001 12.600 12.729 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.081 12.220 mp_waitall_1 186928 16.3 8.057 10.791 8.057 10.791 multiply_cannon_sync_h2d 24660 15.4 6.886 8.382 6.886 8.382 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.220 8.069 init_scf_loop 11 6.9 0.001 0.005 7.977 7.987 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.488 7.247 apply_single 110 13.6 0.000 0.001 6.488 7.247 sum_up_and_integrate 110 10.3 0.052 0.060 6.556 6.569 integrate_v_rspace 110 11.3 0.002 0.003 6.503 6.516 init_scf_run 11 5.9 0.000 0.001 6.424 6.426 scf_env_initial_rho_setup 11 6.9 0.002 0.013 6.423 6.426 qs_ot_get_p 110 10.4 0.001 0.002 6.238 6.395 dbcsr_mm_accdrv_process 52282 16.1 4.710 5.655 6.046 6.367 prepare_preconditioner 11 7.9 0.000 0.000 5.879 5.902 make_preconditioner 11 8.9 0.000 0.002 5.879 5.902 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.812 5.830 calculate_rho_elec 110 8.6 0.039 0.047 5.812 5.829 ot_diis_step 99 11.5 0.010 0.010 5.731 5.731 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.451 5.608 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.742 5.448 make_m2s 4110 13.4 0.057 0.060 4.350 4.823 make_images 4110 14.4 0.400 0.444 4.241 4.714 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.178 4.199 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.709 3.709 wfi_extrapolate 11 7.9 0.001 0.001 3.623 3.623 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.513 3.515 pw_transfer 1331 11.6 0.065 0.070 3.353 3.487 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.247 3.381 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.288 3.365 grid_integrate_task_list 110 12.3 3.159 3.291 3.159 3.291 density_rs2pw 110 9.6 0.004 0.004 3.102 3.271 rs_pw_transfer 902 11.9 0.012 0.014 3.005 3.185 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.121 3.123 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.044 3.112 cp_fm_redistribute_end 48 14.0 2.323 3.089 2.326 3.090 cp_fm_diag_elpa_base 48 14.0 0.730 2.960 0.761 3.050 calculate_dm_sparse 110 9.5 0.001 0.001 2.997 3.030 make_images_data 4110 15.4 0.047 0.051 2.436 2.923 hybrid_alltoall_any 4261 16.3 0.101 0.441 2.121 2.901 fft_wrap_pw1pw2_140 451 13.1 0.200 0.219 2.694 2.828 cp_fm_cholesky_invert 11 10.9 2.737 2.744 2.737 2.744 fft3d_ps 1111 14.6 1.070 1.264 2.604 2.725 calculate_first_density_matrix 1 7.0 0.001 0.003 2.712 2.717 mp_sum_l 6514 12.8 1.960 2.613 1.960 2.613 potential_pw2rs 110 12.3 0.008 0.009 2.416 2.430 grid_collocate_task_list 110 9.6 2.122 2.257 2.122 2.257 mp_alltoall_d11v 2046 13.8 1.811 2.120 1.811 2.120 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.979 1.997 jit_kernel_multiply 10 16.2 0.983 1.953 0.983 1.953 qs_energies_init_hamiltonians 11 5.9 0.001 0.011 1.942 1.947 mp_allgather_i34 2055 14.4 0.736 1.812 0.736 1.812 multiply_cannon_metrocomm4 22605 15.4 0.074 0.076 0.819 1.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.770 1.786 mp_irecv_dv 57340 16.2 0.696 1.690 0.696 1.690 acc_transpose_blocks 24660 15.4 0.106 0.111 1.630 1.652 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.560 1.573 dbcsr_complete_redistribute 325 12.2 0.278 0.343 1.231 1.507 cp_fm_cholesky_decompose 22 10.9 1.492 1.499 1.492 1.499 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.477 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.451000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.727273, yerr=7.699952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 662.876160E+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 0.061 0.177 64.415 64.416 qs_mol_dyn_low 1 2.0 0.003 0.003 63.916 63.935 qs_forces 11 3.9 0.003 0.014 63.815 63.817 qs_energies 11 4.9 0.004 0.024 60.602 60.606 scf_env_do_scf 11 5.9 0.001 0.006 52.326 52.326 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 42.063 42.068 velocity_verlet 10 3.0 0.001 0.002 35.207 35.210 dbcsr_multiply_generic 2055 12.4 0.108 0.113 29.845 30.174 qs_scf_new_mos 99 7.5 0.001 0.001 27.175 27.284 qs_scf_loop_do_ot 99 8.5 0.001 0.003 27.174 27.283 ot_scf_mini 99 9.5 0.003 0.005 25.903 26.025 multiply_cannon 2055 13.4 0.213 0.229 22.696 24.023 multiply_cannon_loop 2055 14.4 0.621 0.634 21.321 22.677 ot_mini 99 10.5 0.001 0.001 14.544 14.684 rebuild_ks_matrix 110 8.3 0.000 0.000 12.597 12.736 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.029 12.597 12.736 mp_waitall_1 146766 16.3 7.984 11.520 7.984 11.520 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.117 11.247 multiply_cannon_multrec 16440 15.4 3.911 5.103 9.706 10.781 init_scf_loop 11 6.9 0.001 0.006 10.217 10.229 qs_ot_get_derivative 99 11.5 0.001 0.001 9.928 10.058 prepare_preconditioner 11 7.9 0.000 0.000 8.258 8.276 make_preconditioner 11 8.9 0.000 0.003 8.258 8.276 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.610 7.951 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.590 7.910 sum_up_and_integrate 110 10.3 0.060 0.061 6.745 6.765 integrate_v_rspace 110 11.3 0.003 0.003 6.685 6.704 qs_ot_get_p 110 10.4 0.001 0.001 6.295 6.423 dbcsr_mm_accdrv_process 34862 16.1 4.754 5.368 5.651 5.785 init_scf_run 11 5.9 0.000 0.001 5.766 5.766 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.765 5.766 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.686 5.696 calculate_rho_elec 110 8.6 0.058 0.058 5.685 5.695 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.041 5.533 apply_single 110 13.6 0.000 0.000 5.041 5.533 make_m2s 4110 13.4 0.050 0.052 4.617 5.058 make_images 4110 14.4 0.393 0.512 4.501 4.942 ot_diis_step 99 11.5 0.010 0.011 4.571 4.571 multiply_cannon_sync_h2d 16440 15.4 3.646 4.335 3.646 4.335 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.324 4.329 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.358 4.014 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.950 3.951 cp_fm_cholesky_invert 11 10.9 3.561 3.568 3.561 3.568 grid_integrate_task_list 110 12.3 3.185 3.392 3.185 3.392 pw_transfer 1331 11.6 0.064 0.070 3.321 3.338 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.321 3.323 cp_fm_redistribute_end 48 14.0 2.059 3.278 2.061 3.278 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.216 3.235 cp_fm_diag_elpa_base 48 14.0 1.150 3.120 1.211 3.235 rs_pw_transfer 902 11.9 0.010 0.011 3.007 3.201 wfi_extrapolate 11 7.9 0.001 0.001 3.175 3.175 make_images_data 4110 15.4 0.043 0.048 2.613 3.149 density_rs2pw 110 9.6 0.004 0.004 2.966 3.136 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.313 2.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.956 2.958 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.640 2.705 fft_wrap_pw1pw2_140 451 13.1 0.210 0.212 2.641 2.659 calculate_dm_sparse 110 9.5 0.001 0.001 2.584 2.629 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.568 2.625 mp_sum_l 6514 12.8 1.946 2.570 1.946 2.570 fft3d_ps 1111 14.6 1.064 1.075 2.529 2.545 potential_pw2rs 110 12.3 0.010 0.010 2.509 2.519 calculate_first_density_matrix 1 7.0 0.001 0.004 2.495 2.499 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.876 2.427 mp_alltoall_d11v 2046 13.8 1.936 2.308 1.936 2.308 mp_irecv_dv 48980 15.7 0.807 2.302 0.807 2.302 grid_collocate_task_list 110 9.6 2.177 2.296 2.177 2.296 cp_fm_cholesky_decompose 22 10.9 2.215 2.228 2.215 2.228 dbcsr_complete_redistribute 325 12.2 0.317 0.370 1.575 2.016 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.993 1.995 cp_fm_upper_to_full 70 13.6 1.380 1.846 1.380 1.846 mp_allgather_i34 2055 14.4 0.695 1.820 0.695 1.820 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.686 1.704 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.526 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.509 1.524 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.018 1.460 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 1.370 1.442 mp_waitany 17072 13.8 1.224 1.441 1.224 1.441 acc_transpose_blocks 16440 15.4 0.072 0.075 1.283 1.296 rs_gather_matrices 110 12.3 0.137 0.151 0.941 1.295 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=64.416000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=627.545455, yerr=9.902834 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 740.663296E+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 0.045 0.109 68.438 68.439 qs_mol_dyn_low 1 2.0 0.003 0.003 67.938 67.947 qs_forces 11 3.9 0.002 0.005 67.869 67.870 qs_energies 11 4.9 0.002 0.005 64.447 64.451 scf_env_do_scf 11 5.9 0.001 0.001 55.270 55.272 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 42.700 42.701 velocity_verlet 10 3.0 0.001 0.002 38.618 38.620 dbcsr_multiply_generic 2055 12.4 0.116 0.122 30.905 31.157 qs_scf_new_mos 99 7.5 0.001 0.001 27.983 28.094 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.983 28.093 ot_scf_mini 99 9.5 0.003 0.004 26.321 26.435 multiply_cannon 2055 13.4 0.245 0.266 23.250 24.443 multiply_cannon_loop 2055 14.4 0.885 0.911 21.721 22.329 ot_mini 99 10.5 0.001 0.001 14.965 15.097 multiply_cannon_multrec 24660 15.4 4.207 7.003 13.102 14.310 init_scf_loop 11 6.9 0.001 0.002 12.519 12.522 rebuild_ks_matrix 110 8.3 0.000 0.000 12.123 12.230 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.123 12.229 qs_ot_get_derivative 99 11.5 0.001 0.001 10.729 10.846 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.709 10.803 prepare_preconditioner 11 7.9 0.000 0.000 10.692 10.707 make_preconditioner 11 8.9 0.000 0.002 10.692 10.707 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.821 10.348 dbcsr_mm_accdrv_process 52304 16.0 6.972 8.811 8.747 9.747 mp_waitall_1 126806 16.4 5.006 7.089 5.006 7.089 sum_up_and_integrate 110 10.3 0.068 0.071 6.527 6.545 integrate_v_rspace 110 11.3 0.003 0.003 6.459 6.481 qs_ot_get_p 110 10.4 0.001 0.001 6.236 6.376 init_scf_run 11 5.9 0.000 0.001 6.297 6.298 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.297 6.298 make_m2s 4110 13.4 0.059 0.062 5.751 6.092 make_images 4110 14.4 0.577 0.689 5.610 5.950 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.887 5.895 calculate_rho_elec 110 8.6 0.077 0.081 5.886 5.895 cp_fm_upper_to_full 70 13.8 3.338 4.781 3.338 4.781 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.227 4.242 ot_diis_step 99 11.5 0.011 0.011 4.192 4.193 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.081 4.187 apply_single 110 13.6 0.000 0.000 4.080 4.187 dbcsr_complete_redistribute 325 12.2 0.415 0.455 2.923 4.090 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.722 3.723 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.634 3.691 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.302 3.462 grid_integrate_task_list 110 12.3 3.273 3.454 3.273 3.454 pw_transfer 1331 11.6 0.064 0.073 3.417 3.447 multiply_cannon_sync_h2d 24660 15.4 3.182 3.401 3.182 3.401 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.494 3.364 calculate_dm_sparse 110 9.5 0.001 0.001 3.320 3.349 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.312 3.345 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.208 3.271 make_images_data 4110 15.4 0.046 0.050 2.856 3.261 wfi_extrapolate 11 7.9 0.001 0.001 3.184 3.184 cp_fm_cholesky_invert 11 10.9 3.152 3.164 3.152 3.164 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.477 3.160 density_rs2pw 110 9.6 0.004 0.004 3.003 3.096 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.081 3.083 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.072 3.074 cp_fm_redistribute_end 48 14.0 1.527 3.027 1.528 3.028 calculate_first_density_matrix 1 7.0 0.001 0.003 3.003 3.006 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.862 3.005 cp_fm_diag_elpa_base 48 14.0 1.415 2.886 1.496 2.995 mp_alltoall_i22 605 13.7 1.762 2.975 1.762 2.975 rs_pw_transfer 902 11.9 0.010 0.011 2.747 2.873 jit_kernel_multiply 12 15.7 1.444 2.739 1.444 2.739 fft_wrap_pw1pw2_140 451 13.1 0.201 0.211 2.699 2.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.643 2.678 fft3d_ps 1111 14.6 1.066 1.097 2.614 2.636 grid_collocate_task_list 110 9.6 2.223 2.397 2.223 2.397 potential_pw2rs 110 12.3 0.013 0.013 2.328 2.350 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.312 2.313 mp_alltoall_d11v 2046 13.8 2.018 2.257 2.018 2.257 cp_fm_cholesky_decompose 22 10.9 1.907 1.945 1.907 1.945 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.913 1.924 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.809 1.844 mp_sum_l 6514 12.8 1.249 1.818 1.249 1.818 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.604 1.698 mp_allgather_i34 2055 14.4 0.626 1.688 0.626 1.688 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.664 1.675 acc_transpose_blocks 24660 15.4 0.105 0.106 1.640 1.671 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.833 1.569 mp_irecv_dv 62702 16.1 0.736 1.488 0.736 1.488 mp_alltoall_z22v 1111 16.6 1.332 1.396 1.332 1.396 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.439000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.454545, yerr=14.993111 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 844.722176E+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 0.015 0.062 56.938 56.939 qs_mol_dyn_low 1 2.0 0.003 0.003 56.685 56.694 qs_forces 11 3.9 0.002 0.003 56.584 56.585 qs_energies 11 4.9 0.001 0.001 52.916 52.922 scf_env_do_scf 11 5.9 0.000 0.001 44.527 44.527 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 36.347 36.348 velocity_verlet 10 3.0 0.001 0.001 32.245 32.249 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.570 23.738 qs_scf_new_mos 99 7.5 0.001 0.001 21.665 21.706 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.665 21.706 ot_scf_mini 99 9.5 0.002 0.002 20.385 20.405 multiply_cannon 2055 13.4 0.250 0.259 17.641 19.190 multiply_cannon_loop 2055 14.4 0.322 0.335 16.232 16.646 rebuild_ks_matrix 110 8.3 0.000 0.000 11.698 11.734 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.697 11.734 ot_mini 99 10.5 0.001 0.001 11.074 11.089 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.393 10.426 multiply_cannon_multrec 8220 15.4 3.237 4.632 7.465 8.441 mp_waitall_1 106626 16.5 6.732 8.400 6.732 8.400 init_scf_loop 11 6.9 0.000 0.000 8.130 8.132 qs_ot_get_derivative 99 11.5 0.001 0.001 7.242 7.260 prepare_preconditioner 11 7.9 0.000 0.000 6.436 6.440 make_preconditioner 11 8.9 0.000 0.000 6.436 6.440 sum_up_and_integrate 110 10.3 0.079 0.081 6.367 6.381 integrate_v_rspace 110 11.3 0.003 0.003 6.288 6.300 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.110 6.126 calculate_rho_elec 110 8.6 0.115 0.116 6.109 6.125 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.009 6.086 qs_ot_get_p 110 10.4 0.001 0.001 5.338 5.358 init_scf_run 11 5.9 0.000 0.001 5.224 5.224 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.224 5.224 dbcsr_mm_accdrv_process 17442 15.9 2.714 4.650 4.098 4.991 make_m2s 4110 13.4 0.038 0.040 4.503 4.750 make_images 4110 14.4 0.636 0.689 4.373 4.622 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.976 4.302 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.929 3.933 ot_diis_step 99 11.5 0.012 0.012 3.802 3.803 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.698 3.720 apply_single 110 13.6 0.000 0.000 3.698 3.719 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.618 3.619 grid_integrate_task_list 110 12.3 3.369 3.490 3.369 3.490 pw_transfer 1331 11.6 0.064 0.070 3.455 3.478 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.348 3.375 density_rs2pw 110 9.6 0.004 0.004 3.107 3.259 cp_fm_cholesky_invert 11 10.9 3.130 3.134 3.130 3.134 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.071 3.073 multiply_cannon_sync_h2d 8220 15.4 2.914 3.072 2.914 3.072 cp_fm_redistribute_end 48 14.0 0.790 3.045 0.796 3.046 cp_fm_diag_elpa_base 48 14.0 2.062 2.830 2.242 2.994 make_images_data 4110 15.4 0.038 0.043 2.559 2.984 hybrid_alltoall_any 4261 16.3 0.200 0.868 2.450 2.916 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.812 2.845 wfi_extrapolate 11 7.9 0.001 0.001 2.799 2.800 rs_pw_transfer 902 11.9 0.010 0.010 2.558 2.729 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.662 2.663 fft3d_ps 1111 14.6 1.120 1.161 2.606 2.637 calculate_dm_sparse 110 9.5 0.001 0.001 2.562 2.604 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.576 2.577 grid_collocate_task_list 110 9.6 2.326 2.413 2.326 2.413 calculate_first_density_matrix 1 7.0 0.000 0.001 2.314 2.316 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.247 2.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.189 2.204 mp_alltoall_d11v 2046 13.8 1.898 2.069 1.898 2.069 potential_pw2rs 110 12.3 0.015 0.016 2.046 2.053 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 1.992 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.953 1.963 cp_fm_cholesky_decompose 22 10.9 1.845 1.855 1.845 1.855 jit_kernel_multiply 10 15.8 1.071 1.748 1.071 1.748 mp_allgather_i34 2055 14.4 0.592 1.734 0.592 1.734 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.642 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.507 1.629 dbcsr_complete_redistribute 325 12.2 0.560 0.576 1.494 1.607 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.493 1.507 mp_waitany 9240 13.8 1.251 1.440 1.251 1.440 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.046 1.403 qs_create_task_list 11 7.9 0.000 0.001 1.233 1.332 generate_qs_task_list 11 8.9 0.380 0.447 1.232 1.332 mp_alltoall_z22v 1111 16.6 1.255 1.273 1.255 1.273 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.168 1.181 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.939000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=791.545455, yerr=16.626177 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 1.356698E+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 0.044 0.069 89.848 89.849 qs_mol_dyn_low 1 2.0 0.003 0.003 89.268 89.276 qs_forces 11 3.9 0.004 0.007 89.183 89.188 qs_energies 11 4.9 0.001 0.001 84.804 84.810 scf_env_do_scf 11 5.9 0.000 0.001 74.221 74.221 velocity_verlet 10 3.0 0.001 0.001 57.351 57.358 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 44.343 44.345 init_scf_loop 11 6.9 0.000 0.000 29.793 29.796 dbcsr_multiply_generic 2055 12.4 0.119 0.121 29.552 29.692 prepare_preconditioner 11 7.9 0.000 0.000 27.784 27.791 make_preconditioner 11 8.9 0.000 0.000 27.784 27.791 qs_scf_new_mos 99 7.5 0.001 0.001 27.692 27.742 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.691 27.742 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.744 27.261 ot_scf_mini 99 9.5 0.002 0.002 25.888 25.923 multiply_cannon 2055 13.4 0.349 0.371 21.927 22.685 multiply_cannon_loop 2055 14.4 0.341 0.343 20.041 20.417 cp_fm_upper_to_full 70 14.2 13.032 18.812 13.032 18.812 ot_mini 99 10.5 0.001 0.001 14.195 14.233 rebuild_ks_matrix 110 8.3 0.000 0.000 13.406 13.491 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.406 13.490 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.966 12.015 dbcsr_complete_redistribute 325 12.2 1.029 1.058 7.790 11.183 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.704 10.095 multiply_cannon_multrec 8220 15.4 4.416 4.590 9.836 9.945 qs_ot_get_derivative 99 11.5 0.001 0.001 9.720 9.752 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.024 9.391 mp_waitall_1 87304 16.6 8.179 9.222 8.179 9.222 mp_alltoall_i22 605 13.7 5.684 9.107 5.684 9.107 sum_up_and_integrate 110 10.3 0.151 0.152 6.914 6.926 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.750 6.784 calculate_rho_elec 110 8.6 0.227 0.227 6.750 6.784 integrate_v_rspace 110 11.3 0.004 0.004 6.763 6.775 qs_ot_get_p 110 10.4 0.001 0.001 6.538 6.597 init_scf_run 11 5.9 0.000 0.001 6.279 6.279 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.278 6.278 make_m2s 4110 13.4 0.042 0.043 5.593 6.134 make_images 4110 14.4 0.877 0.921 5.405 5.945 cp_fm_cholesky_invert 11 10.9 5.912 5.917 5.912 5.917 dbcsr_mm_accdrv_process 11614 15.7 3.350 3.679 5.275 5.494 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.583 5.060 apply_single 110 13.6 0.000 0.000 4.583 5.060 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.746 5.053 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.820 4.830 ot_diis_step 99 11.5 0.015 0.015 4.453 4.454 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.282 4.282 multiply_cannon_sync_h2d 8220 15.4 3.949 3.957 3.949 3.957 pw_transfer 1331 11.6 0.073 0.074 3.830 3.843 hybrid_alltoall_any 4261 16.3 0.256 0.554 3.039 3.790 make_images_data 4110 15.4 0.041 0.044 3.049 3.780 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.714 3.726 grid_integrate_task_list 110 12.3 3.667 3.715 3.667 3.715 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.660 3.666 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.642 3.643 cp_fm_diag_elpa_base 48 14.0 3.073 3.285 3.639 3.641 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.187 3.624 wfi_extrapolate 11 7.9 0.001 0.001 3.555 3.555 calculate_dm_sparse 110 9.5 0.001 0.001 3.301 3.324 fft_wrap_pw1pw2_140 451 13.1 0.216 0.223 3.185 3.198 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.171 3.175 density_rs2pw 110 9.6 0.004 0.004 3.103 3.124 fft3d_ps 1111 14.6 1.266 1.289 2.918 2.935 grid_collocate_task_list 110 9.6 2.630 2.656 2.630 2.656 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.628 2.648 calculate_first_density_matrix 1 7.0 0.000 0.001 2.616 2.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.551 2.577 mp_alltoall_d11v 2046 13.8 2.446 2.493 2.446 2.493 rs_pw_transfer 902 11.9 0.010 0.011 2.396 2.439 cp_fm_cholesky_decompose 22 10.9 2.306 2.333 2.306 2.333 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.197 2.243 potential_pw2rs 110 12.3 0.021 0.022 2.230 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.113 2.180 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.050 2.086 qs_create_task_list 11 7.9 0.010 0.010 1.912 1.957 generate_qs_task_list 11 8.9 0.740 0.795 1.902 1.948 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.891 1.934 jit_kernel_multiply 10 15.3 1.721 1.925 1.721 1.925 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.827 1.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.849000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1217.636364, yerr=66.916872 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 627.003392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 56701. MP_Allreduce 11263 785. MP_Sync 168 MP_Alltoall 2210 2436849. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 207.567 207.568 qs_mol_dyn_low 1 2.0 0.003 0.003 207.158 207.172 qs_forces 11 3.9 0.004 0.005 207.064 207.067 qs_energies 11 4.9 0.002 0.007 201.472 201.487 scf_env_do_scf 11 5.9 0.001 0.002 184.237 184.241 scf_env_do_scf_inner_loop 116 6.6 0.002 0.007 162.378 162.380 dbcsr_multiply_generic 2485 12.5 0.173 0.177 124.718 125.811 velocity_verlet 10 3.0 0.001 0.001 124.400 124.402 qs_scf_new_mos 116 7.6 0.001 0.001 122.877 123.182 qs_scf_loop_do_ot 116 8.6 0.001 0.001 122.876 123.182 ot_scf_mini 116 9.6 0.003 0.003 116.306 116.600 multiply_cannon 2485 13.5 0.237 0.243 100.589 103.507 multiply_cannon_loop 2485 14.5 2.091 2.171 98.188 100.796 ot_mini 116 10.6 0.001 0.001 66.277 66.613 multiply_cannon_multrec 59640 15.5 32.772 34.855 41.167 43.344 qs_ot_get_derivative 116 11.6 0.001 0.001 41.471 41.790 rebuild_ks_matrix 127 8.3 0.001 0.001 33.784 34.322 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.018 33.783 34.322 mp_waitall_1 288940 16.2 29.632 33.158 29.632 33.158 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.380 30.874 multiply_cannon_sync_h2d 59640 15.5 27.087 30.195 27.087 30.195 qs_ot_get_p 127 10.4 0.001 0.001 28.144 28.403 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.124 25.083 apply_single 127 13.6 0.001 0.001 24.124 25.083 ot_diis_step 116 11.6 0.007 0.008 24.452 24.454 init_scf_loop 11 6.9 0.000 0.001 21.778 21.780 qs_ot_p2m_diag 82 11.4 0.076 0.090 21.089 21.220 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.264 19.525 cp_dbcsr_syevd 82 12.4 0.004 0.005 18.521 18.522 multiply_cannon_metrocomm3 59640 15.5 0.109 0.114 15.380 18.063 prepare_preconditioner 11 7.9 0.000 0.000 17.055 17.104 make_preconditioner 11 8.9 0.000 0.000 17.055 17.104 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.263 16.441 cp_fm_diag_elpa 82 13.4 0.000 0.001 15.505 15.512 cp_fm_redistribute_end 82 14.4 12.234 15.419 12.248 15.424 cp_fm_diag_elpa_base 82 14.4 3.128 15.118 3.162 15.223 make_m2s 4970 13.5 0.102 0.109 14.489 14.961 make_images 4970 14.5 0.400 0.422 14.311 14.793 sum_up_and_integrate 127 10.3 0.089 0.108 14.427 14.442 integrate_v_rspace 127 11.3 0.003 0.004 14.338 14.357 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.099 13.176 calculate_rho_elec 127 8.7 0.044 0.063 13.098 13.175 init_scf_run 11 5.9 0.000 0.001 12.786 12.786 scf_env_initial_rho_setup 11 6.9 0.002 0.006 12.785 12.786 cp_fm_cholesky_invert 11 10.9 9.768 9.776 9.768 9.776 mp_sum_l 7804 13.0 8.394 9.740 8.394 9.740 wfi_extrapolate 11 7.9 0.001 0.001 9.370 9.370 multiply_cannon_metrocomm1 59640 15.5 0.084 0.089 6.534 8.962 dbcsr_mm_accdrv_process 123452 16.2 3.312 3.514 7.959 8.575 calculate_dm_sparse 127 9.5 0.001 0.001 8.410 8.532 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.094 8.217 make_images_data 4970 15.5 0.067 0.074 7.143 8.133 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.811 7.921 pw_transfer 1535 11.6 0.074 0.105 7.339 7.629 density_rs2pw 127 9.7 0.006 0.007 6.862 7.506 grid_integrate_task_list 127 12.3 7.045 7.472 7.045 7.472 hybrid_alltoall_any 5155 16.4 0.287 2.245 6.199 7.459 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 7.137 7.404 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.945 6.954 rs_pw_transfer 1038 11.9 0.017 0.019 6.140 6.689 fft3d_ps 1281 14.7 2.121 2.588 5.940 6.148 fft_wrap_pw1pw2_140 519 13.2 0.438 0.504 5.886 6.094 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.959 6.063 mp_alltoall_d11v 2401 14.1 4.583 5.689 4.583 5.689 cp_fm_cholesky_decompose 22 10.9 5.064 5.080 5.064 5.080 grid_collocate_task_list 127 9.7 4.686 5.046 4.686 5.046 potential_pw2rs 127 12.3 0.009 0.011 4.884 4.931 mp_sum_d 4448 12.1 3.808 4.630 3.808 4.630 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.568000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.454545, yerr=6.184491 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 827.187200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828432560 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56635. MP_Allreduce 11357 942. MP_Sync 172 MP_Alltoall 1983 5047701. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 84 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.060 194.887 194.888 qs_mol_dyn_low 1 2.0 0.004 0.009 194.310 194.323 qs_forces 11 3.9 0.003 0.005 194.148 194.149 qs_energies 11 4.9 0.003 0.018 187.229 187.243 scf_env_do_scf 11 5.9 0.001 0.002 170.710 170.720 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.738 136.742 velocity_verlet 10 3.0 0.001 0.001 122.790 122.798 dbcsr_multiply_generic 2529 12.6 0.185 0.194 99.566 100.756 qs_scf_new_mos 118 7.6 0.001 0.001 97.254 97.699 qs_scf_loop_do_ot 118 8.6 0.001 0.002 97.253 97.698 ot_scf_mini 118 9.6 0.004 0.006 92.190 92.707 multiply_cannon 2529 13.6 0.481 0.533 78.780 83.265 multiply_cannon_loop 2529 14.6 1.271 1.305 75.277 77.963 ot_mini 118 10.6 0.001 0.001 51.029 51.623 mp_waitall_1 228724 16.4 25.827 40.749 25.827 40.749 multiply_cannon_multrec 30348 15.6 21.998 25.953 31.733 36.124 init_scf_loop 11 6.9 0.001 0.002 33.882 33.887 rebuild_ks_matrix 129 8.3 0.001 0.001 33.295 33.815 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 33.295 33.814 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.923 30.408 multiply_cannon_metrocomm3 30348 15.6 0.096 0.100 16.504 29.697 prepare_preconditioner 11 7.9 0.000 0.000 29.535 29.603 make_preconditioner 11 8.9 0.000 0.002 29.535 29.602 qs_ot_get_derivative 118 11.6 0.002 0.002 28.754 29.265 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.245 28.769 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.186 23.286 apply_single 129 13.6 0.001 0.001 22.185 23.286 qs_ot_get_p 129 10.4 0.001 0.002 22.537 23.272 ot_diis_step 118 11.6 0.014 0.015 22.086 22.089 multiply_cannon_sync_h2d 30348 15.6 19.364 21.895 19.364 21.895 qs_ot_p2m_diag 84 11.4 0.189 0.218 17.364 17.399 cp_fm_cholesky_invert 11 10.9 17.263 17.276 17.263 17.276 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.209 16.210 make_m2s 5058 13.6 0.091 0.097 14.497 16.134 make_images 5058 14.6 1.170 1.361 14.285 15.916 sum_up_and_integrate 129 10.3 0.117 0.133 14.611 14.646 integrate_v_rspace 129 11.3 0.003 0.004 14.493 14.530 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.297 13.369 calculate_rho_elec 129 8.7 0.088 0.106 13.297 13.368 cp_fm_diag_elpa 84 13.4 0.000 0.001 13.039 13.048 cp_fm_redistribute_end 84 14.4 7.621 12.973 7.635 12.976 cp_fm_diag_elpa_base 84 14.4 5.099 12.465 5.322 12.857 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 11.437 11.805 init_scf_run 11 5.9 0.000 0.001 11.663 11.664 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.662 11.664 make_images_data 5058 15.6 0.066 0.073 8.664 10.592 multiply_cannon_metrocomm4 27819 15.6 0.097 0.111 3.774 10.527 mp_irecv_dv 70084 16.3 3.580 10.150 3.580 10.150 dbcsr_mm_accdrv_process 62758 16.2 4.662 5.416 9.199 9.809 hybrid_alltoall_any 5245 16.5 0.344 1.523 7.455 9.698 wfi_extrapolate 11 7.9 0.001 0.001 8.415 8.415 pw_transfer 1559 11.6 0.085 0.100 8.253 8.330 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.028 8.095 grid_integrate_task_list 129 12.3 7.255 7.532 7.255 7.532 cp_fm_cholesky_decompose 22 10.9 7.400 7.475 7.400 7.475 density_rs2pw 129 9.7 0.006 0.007 6.982 7.446 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.341 7.113 calculate_dm_sparse 129 9.5 0.001 0.001 6.715 6.919 fft_wrap_pw1pw2_140 527 13.2 0.473 0.519 6.836 6.900 mp_sum_l 7936 13.1 4.512 6.445 4.512 6.445 fft3d_ps 1301 14.7 2.796 2.945 6.383 6.432 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.271 6.282 rs_pw_transfer 1054 12.0 0.014 0.017 5.669 6.192 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.390 5.572 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.463 5.538 grid_collocate_task_list 129 9.7 4.950 5.349 4.950 5.349 mp_allgather_i34 2529 14.6 2.061 5.057 2.061 5.057 potential_pw2rs 129 12.3 0.015 0.016 4.989 5.010 mp_alltoall_d11v 2429 14.1 4.182 4.545 4.182 4.545 mp_sum_d 4503 12.2 2.905 4.251 2.905 4.251 dbcsr_complete_redistribute 397 12.7 0.772 0.853 3.132 3.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.888000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.090909, yerr=2.314168 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 934.883328E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56629. MP_Allreduce 11356 981. MP_Sync 172 MP_Alltoall 1724 9394234. MP_ISendRecv 15996 75008. MP_Wait 29984 MP_comm_split 84 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.035 181.903 181.904 qs_mol_dyn_low 1 2.0 0.003 0.003 181.503 181.517 qs_forces 11 3.9 0.004 0.006 181.379 181.387 qs_energies 11 4.9 0.003 0.004 174.837 174.848 scf_env_do_scf 11 5.9 0.001 0.001 158.633 158.636 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 122.675 122.676 velocity_verlet 10 3.0 0.001 0.001 116.203 116.206 qs_scf_new_mos 118 7.6 0.001 0.001 84.845 85.162 qs_scf_loop_do_ot 118 8.6 0.001 0.001 84.844 85.161 dbcsr_multiply_generic 2529 12.6 0.183 0.188 83.780 85.029 ot_scf_mini 118 9.6 0.003 0.004 80.565 80.976 multiply_cannon 2529 13.6 0.502 0.530 63.054 68.107 multiply_cannon_loop 2529 14.6 0.868 0.896 59.644 62.463 ot_mini 118 10.6 0.001 0.001 43.521 43.943 init_scf_loop 11 6.9 0.000 0.000 35.853 35.854 mp_waitall_1 180004 16.5 26.009 34.965 26.009 34.965 prepare_preconditioner 11 7.9 0.000 0.000 31.791 31.835 make_preconditioner 11 8.9 0.000 0.000 31.791 31.835 rebuild_ks_matrix 129 8.3 0.001 0.001 30.979 31.450 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.979 31.450 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.452 30.796 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.055 28.486 multiply_cannon_multrec 20232 15.6 13.652 16.768 22.451 25.566 multiply_cannon_metrocomm3 20232 15.6 0.058 0.061 15.094 24.510 qs_ot_get_derivative 118 11.6 0.001 0.002 23.505 23.924 qs_ot_get_p 129 10.4 0.001 0.001 22.159 22.645 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.017 21.065 apply_single 129 13.6 0.001 0.001 20.017 21.065 ot_diis_step 118 11.6 0.018 0.018 19.904 19.907 qs_ot_p2m_diag 84 11.4 0.268 0.275 17.588 17.602 cp_dbcsr_syevd 84 12.4 0.005 0.005 16.539 16.541 multiply_cannon_sync_h2d 20232 15.6 14.475 16.230 14.475 16.230 make_m2s 5058 13.6 0.082 0.089 15.239 16.227 make_images 5058 14.6 1.195 1.289 15.004 15.990 cp_fm_cholesky_invert 11 10.9 14.999 15.009 14.999 15.009 sum_up_and_integrate 129 10.3 0.135 0.145 14.713 14.737 integrate_v_rspace 129 11.3 0.003 0.004 14.578 14.605 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.454 13.487 calculate_rho_elec 129 8.7 0.133 0.148 13.453 13.486 cp_fm_diag_elpa 84 13.4 0.000 0.001 13.190 13.193 cp_fm_redistribute_end 84 14.4 4.977 13.105 4.992 13.107 cp_fm_diag_elpa_base 84 14.4 7.678 12.484 8.093 12.994 init_scf_run 11 5.9 0.000 0.001 11.127 11.127 scf_env_initial_rho_setup 11 6.9 0.003 0.005 11.126 11.127 make_images_data 5058 15.6 0.060 0.068 9.398 10.857 hybrid_alltoall_any 5245 16.5 0.432 1.968 8.164 10.106 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.420 9.727 multiply_cannon_metrocomm4 17703 15.6 0.063 0.073 3.386 9.190 mp_irecv_dv 50659 16.2 3.262 8.944 3.262 8.944 dbcsr_mm_accdrv_process 41846 16.2 4.428 5.216 8.249 8.395 pw_transfer 1559 11.6 0.085 0.104 8.219 8.318 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 7.994 8.100 grid_integrate_task_list 129 12.3 7.396 7.845 7.396 7.845 cp_fm_cholesky_decompose 22 10.9 7.772 7.800 7.772 7.800 wfi_extrapolate 11 7.9 0.001 0.001 7.476 7.476 cp_fm_upper_to_full 106 14.5 5.631 7.341 5.631 7.341 density_rs2pw 129 9.7 0.006 0.006 6.804 7.169 fft_wrap_pw1pw2_140 527 13.2 0.480 0.538 6.864 6.977 dbcsr_complete_redistribute 397 12.7 1.194 1.289 4.665 6.424 fft3d_ps 1301 14.7 2.712 2.928 6.247 6.300 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.122 6.129 calculate_dm_sparse 129 9.5 0.001 0.001 5.902 6.020 rs_pw_transfer 1054 12.0 0.014 0.014 5.365 5.767 grid_collocate_task_list 129 9.7 5.102 5.541 5.102 5.541 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.605 5.260 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.099 5.249 mp_sum_l 7936 13.1 3.647 5.203 3.647 5.203 mp_allgather_i34 2529 14.6 1.919 5.176 1.919 5.176 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.398 5.153 potential_pw2rs 129 12.3 0.021 0.023 4.911 4.932 mp_alltoall_d11v 2429 14.1 4.412 4.771 4.412 4.771 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.096 4.133 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.319 4.028 mp_alltoall_i22 720 14.1 1.960 3.869 1.960 3.869 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.733 3.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=181.904000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.545455, yerr=10.030532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.163878E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57274. MP_Allreduce 11236 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.190 188.723 188.724 qs_mol_dyn_low 1 2.0 0.003 0.003 188.050 188.063 qs_forces 11 3.9 0.003 0.004 187.935 187.940 qs_energies 11 4.9 0.002 0.007 180.858 180.871 scf_env_do_scf 11 5.9 0.001 0.001 163.944 163.954 velocity_verlet 10 3.0 0.001 0.002 124.415 124.418 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.727 116.728 qs_scf_new_mos 116 7.6 0.001 0.001 81.119 81.410 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.118 81.410 dbcsr_multiply_generic 2485 12.5 0.189 0.198 79.701 80.390 ot_scf_mini 116 9.6 0.003 0.004 76.627 76.906 multiply_cannon 2485 13.5 0.552 0.589 54.696 58.183 multiply_cannon_loop 2485 14.5 1.169 1.198 50.934 52.574 init_scf_loop 11 6.9 0.001 0.004 47.090 47.091 prepare_preconditioner 11 7.9 0.000 0.000 42.996 43.026 make_preconditioner 11 8.9 0.000 0.003 42.996 43.026 ot_mini 116 10.6 0.001 0.001 42.442 42.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.580 41.634 multiply_cannon_multrec 29820 15.5 13.829 19.033 25.732 30.353 rebuild_ks_matrix 127 8.3 0.001 0.001 29.139 29.401 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 29.138 29.400 mp_waitall_1 152434 16.5 18.099 29.348 18.099 29.348 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.318 26.563 qs_ot_get_derivative 116 11.6 0.001 0.002 22.807 23.102 make_m2s 4970 13.5 0.094 0.097 20.491 21.641 make_images 4970 14.5 1.944 2.260 20.189 21.337 qs_ot_get_p 127 10.4 0.001 0.001 20.331 20.653 ot_diis_step 116 11.6 0.017 0.018 19.509 19.511 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.967 19.508 apply_single 127 13.6 0.001 0.001 18.967 19.507 cp_fm_cholesky_invert 11 10.9 16.493 16.503 16.493 16.503 cp_fm_upper_to_full 104 14.7 11.129 16.481 11.129 16.481 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.115 16.168 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.597 15.911 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.797 14.801 sum_up_and_integrate 127 10.3 0.140 0.152 14.230 14.256 integrate_v_rspace 127 11.3 0.004 0.004 14.090 14.121 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.140 13.169 calculate_rho_elec 127 8.7 0.173 0.189 13.140 13.168 make_images_data 4970 15.5 0.062 0.066 11.066 13.070 dbcsr_complete_redistribute 393 12.7 1.515 1.635 9.119 12.880 multiply_cannon_sync_h2d 29820 15.5 11.658 12.462 11.658 12.462 hybrid_alltoall_any 5155 16.4 0.519 2.201 9.994 12.236 dbcsr_mm_accdrv_process 61748 16.2 7.323 8.379 11.480 11.997 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.664 11.667 cp_fm_redistribute_end 82 14.4 2.008 11.588 2.023 11.593 cp_fm_diag_elpa_base 82 14.4 8.956 10.977 9.537 11.459 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.682 11.448 init_scf_run 11 5.9 0.000 0.001 10.999 11.000 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.999 11.000 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.393 10.073 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.468 9.681 mp_alltoall_i22 712 14.1 5.645 9.372 5.645 9.372 pw_transfer 1535 11.6 0.083 0.096 7.928 8.019 cp_fm_cholesky_decompose 22 10.9 7.846 7.963 7.846 7.963 grid_integrate_task_list 127 12.3 7.463 7.930 7.463 7.930 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.706 7.801 wfi_extrapolate 11 7.9 0.001 0.001 7.653 7.653 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.811 7.111 fft_wrap_pw1pw2_140 519 13.2 0.475 0.484 6.763 6.885 mp_irecv_dv 75445 16.2 2.666 6.839 2.666 6.839 density_rs2pw 127 9.7 0.005 0.006 6.304 6.564 calculate_dm_sparse 127 9.5 0.001 0.001 6.170 6.277 fft3d_ps 1281 14.7 2.771 2.848 5.954 6.018 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.721 5.758 mp_alltoall_d11v 2401 14.1 4.976 5.617 4.976 5.617 grid_collocate_task_list 127 9.7 5.168 5.431 5.168 5.431 rs_pw_transfer 1038 11.9 0.013 0.014 4.744 5.005 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.604 4.690 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.381 4.460 potential_pw2rs 127 12.3 0.022 0.023 4.416 4.427 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.426 4.427 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.220 4.260 mp_allgather_i34 2485 14.5 1.876 3.781 1.876 3.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.724000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.636364, yerr=24.606994 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.523696E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516877E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096701. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.058 177.538 177.539 qs_mol_dyn_low 1 2.0 0.003 0.003 177.119 177.131 qs_forces 11 3.9 0.003 0.003 177.022 177.026 qs_energies 11 4.9 0.001 0.002 169.720 169.729 scf_env_do_scf 11 5.9 0.001 0.001 152.495 152.508 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 116.099 116.100 velocity_verlet 10 3.0 0.001 0.001 116.038 116.042 qs_scf_new_mos 118 7.6 0.001 0.001 79.551 79.668 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.550 79.667 dbcsr_multiply_generic 2534 12.6 0.181 0.187 76.693 77.187 ot_scf_mini 118 9.6 0.004 0.004 74.950 75.050 multiply_cannon 2534 13.6 0.588 0.622 55.867 60.504 multiply_cannon_loop 2534 14.6 0.453 0.462 50.738 52.025 ot_mini 118 10.6 0.001 0.001 41.201 41.285 mp_waitall_1 130986 16.6 27.486 36.338 27.486 36.338 init_scf_loop 11 6.9 0.000 0.000 36.224 36.225 prepare_preconditioner 11 7.9 0.000 0.000 32.326 32.355 make_preconditioner 11 8.9 0.000 0.000 32.326 32.355 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.148 30.432 rebuild_ks_matrix 129 8.3 0.001 0.001 29.146 29.300 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.146 29.299 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.672 26.814 qs_ot_get_derivative 118 11.6 0.002 0.002 20.828 20.929 multiply_cannon_multrec 10136 15.6 10.473 14.628 17.825 20.783 qs_ot_get_p 129 10.4 0.001 0.001 20.468 20.594 multiply_cannon_metrocomm3 10136 15.6 0.022 0.024 12.757 20.403 ot_diis_step 118 11.6 0.020 0.020 20.288 20.289 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.759 20.133 apply_single 129 13.6 0.001 0.001 19.758 20.133 make_m2s 5068 13.6 0.067 0.071 16.745 19.347 make_images 5068 14.6 2.317 2.650 16.435 19.038 cp_fm_cholesky_invert 11 10.9 18.433 18.439 18.433 18.439 qs_ot_p2m_diag 84 11.4 0.501 0.507 16.289 16.308 cp_dbcsr_syevd 84 12.4 0.005 0.005 15.124 15.126 sum_up_and_integrate 129 10.3 0.181 0.191 14.339 14.394 integrate_v_rspace 129 11.3 0.004 0.004 14.158 14.221 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.533 13.577 calculate_rho_elec 129 8.7 0.260 0.271 13.533 13.576 make_images_data 5068 15.6 0.052 0.061 10.237 13.231 hybrid_alltoall_any 5255 16.5 0.827 3.670 9.889 12.913 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.995 11.999 cp_fm_diag_elpa_base 84 14.4 11.742 11.826 11.988 11.991 multiply_cannon_sync_h2d 10136 15.6 11.643 11.953 11.643 11.953 init_scf_run 11 5.9 0.000 0.001 10.404 10.405 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.404 10.404 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.524 8.602 cp_fm_cholesky_decompose 22 10.9 8.187 8.265 8.187 8.265 mp_allgather_i34 2534 14.6 3.314 8.246 3.314 8.246 grid_integrate_task_list 129 12.3 7.851 8.159 7.851 8.159 pw_transfer 1559 11.6 0.083 0.090 8.064 8.080 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.842 7.862 multiply_cannon_metrocomm1 10136 15.6 0.029 0.030 4.870 7.819 dbcsr_mm_accdrv_process 20954 16.1 2.678 3.496 6.985 7.675 wfi_extrapolate 11 7.9 0.001 0.001 7.433 7.433 fft_wrap_pw1pw2_140 527 13.2 0.506 0.527 6.848 6.873 density_rs2pw 129 9.7 0.005 0.006 6.220 6.587 calculate_dm_sparse 129 9.5 0.001 0.001 6.269 6.342 fft3d_ps 1301 14.7 2.753 2.826 5.990 6.030 grid_collocate_task_list 129 9.7 5.542 5.829 5.542 5.829 dbcsr_complete_redistribute 397 12.7 2.115 2.157 5.415 5.817 mp_alltoall_d11v 2429 14.1 5.030 5.409 5.030 5.409 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.277 5.298 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.243 5.244 rs_pw_transfer 1054 12.0 0.012 0.013 4.244 4.609 potential_pw2rs 129 12.3 0.027 0.028 4.213 4.230 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.126 4.170 multiply_cannon_metrocomm4 7602 15.6 0.023 0.026 1.868 4.074 mp_irecv_dv 29142 15.9 1.831 4.001 1.831 4.001 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.829 3.892 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.516 3.858 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.563 3.848 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.656 3.739 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.596 3.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.539000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1428.818182, yerr=55.884326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 3.080360E+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 0.027 0.073 291.071 291.072 qs_mol_dyn_low 1 2.0 0.003 0.003 290.512 290.535 qs_forces 11 3.9 0.003 0.003 290.415 290.421 qs_energies 11 4.9 0.002 0.002 281.805 281.819 scf_env_do_scf 11 5.9 0.001 0.001 260.025 260.035 velocity_verlet 10 3.0 0.001 0.001 211.232 211.241 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.855 133.857 init_scf_loop 11 6.9 0.000 0.000 125.910 125.912 prepare_preconditioner 11 7.9 0.000 0.000 121.221 121.253 make_preconditioner 11 8.9 0.000 0.000 121.221 121.253 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.020 118.375 qs_scf_new_mos 118 7.6 0.001 0.001 91.513 91.615 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.512 91.614 ot_scf_mini 118 9.6 0.004 0.004 86.690 86.721 dbcsr_multiply_generic 2535 12.6 0.214 0.222 83.395 83.971 cp_fm_upper_to_full 106 14.8 54.623 77.880 54.623 77.880 multiply_cannon 2535 13.6 0.704 0.752 58.846 59.552 multiply_cannon_loop 2535 14.6 0.476 0.481 55.195 56.480 ot_mini 118 10.6 0.001 0.001 44.541 44.568 dbcsr_complete_redistribute 397 12.7 4.006 4.042 29.941 43.196 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.545 39.732 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.155 37.092 mp_alltoall_i22 720 14.1 21.972 35.244 21.972 35.244 rebuild_ks_matrix 129 8.3 0.001 0.001 33.767 33.799 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.766 33.799 cp_fm_cholesky_invert 11 10.9 33.472 33.478 33.472 33.478 mp_waitall_1 106626 16.7 28.290 32.249 28.290 32.249 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.538 31.577 qs_ot_get_p 129 10.4 0.001 0.001 26.657 26.671 qs_ot_get_derivative 118 11.6 0.002 0.002 24.383 24.415 qs_ot_p2m_diag 84 11.4 0.889 0.897 22.495 22.522 make_m2s 5070 13.6 0.076 0.080 19.950 20.876 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.756 20.757 make_images 5070 14.6 3.788 3.914 19.469 20.397 ot_diis_step 118 11.6 0.022 0.023 20.132 20.132 multiply_cannon_metrocomm3 10140 15.6 0.023 0.023 18.799 19.932 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.325 19.515 apply_single 129 13.6 0.001 0.001 19.325 19.515 multiply_cannon_multrec 10140 15.6 10.543 12.294 18.032 18.146 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.474 17.475 cp_fm_diag_elpa_base 84 14.4 12.873 14.579 17.470 17.471 sum_up_and_integrate 129 10.3 0.323 0.325 15.764 15.860 multiply_cannon_sync_h2d 10140 15.6 15.780 15.802 15.780 15.802 integrate_v_rspace 129 11.3 0.004 0.004 15.441 15.536 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.072 15.094 calculate_rho_elec 129 8.7 0.487 0.488 15.071 15.093 hybrid_alltoall_any 5257 16.5 1.312 3.059 10.740 12.593 make_images_data 5070 15.6 0.060 0.066 10.562 12.346 init_scf_run 11 5.9 0.000 0.001 11.834 11.835 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.834 11.835 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.610 9.642 dbcsr_mm_accdrv_process 20958 16.1 3.850 5.895 7.251 9.181 cp_fm_cholesky_decompose 22 10.9 8.926 8.948 8.926 8.948 grid_integrate_task_list 129 12.3 8.616 8.793 8.616 8.793 wfi_extrapolate 11 7.9 0.001 0.001 8.727 8.727 pw_transfer 1559 11.6 0.091 0.093 8.403 8.411 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.168 8.177 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.948 7.949 fft_wrap_pw1pw2_140 527 13.2 0.543 0.544 7.236 7.253 mp_alltoall_d11v 2429 14.1 7.058 7.203 7.058 7.203 calculate_dm_sparse 129 9.5 0.001 0.001 6.665 6.756 grid_collocate_task_list 129 9.7 6.384 6.419 6.384 6.419 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.318 6.391 fft3d_ps 1301 14.7 2.786 2.796 6.197 6.207 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.065 6.109 density_rs2pw 129 9.7 0.005 0.005 6.032 6.061 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.072000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2715.727273, yerr=190.486025 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 1.261421E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.068 84.630 84.631 qs_energies 1 2.0 0.000 0.000 84.173 84.180 ls_scf 1 3.0 0.000 0.000 83.228 83.235 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.354 72.530 multiply_cannon 111 7.7 0.018 0.021 55.838 57.225 multiply_cannon_loop 111 8.7 0.212 0.227 52.374 53.929 ls_scf_main 1 4.0 0.000 0.000 52.145 52.145 density_matrix_trs4 2 5.0 0.002 0.003 46.920 47.006 ls_scf_init_scf 1 4.0 0.000 0.000 28.107 28.108 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.037 27.092 mp_waitall_1 11316 10.9 21.594 25.120 21.594 25.120 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.949 24.964 multiply_cannon_multrec 2664 9.7 8.228 9.056 15.618 17.412 multiply_cannon_sync_h2d 2664 9.7 14.155 15.722 14.155 15.722 make_m2s 222 7.7 0.008 0.011 12.998 13.569 make_images 222 8.7 0.099 0.110 12.976 13.548 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 8.969 11.986 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.342 8.531 make_images_data 222 9.7 0.004 0.005 7.602 8.082 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.642 7.003 7.961 hybrid_alltoall_any 227 10.6 0.216 1.842 6.504 7.774 dbcsr_mm_accdrv_process_sort 4760 11.4 6.288 7.198 6.288 7.198 calculate_norms 4752 9.8 5.596 6.155 5.596 6.155 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.831 4.955 mp_sum_l 807 5.4 2.986 4.239 2.986 4.239 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.021 3.886 mp_irecv_dv 6231 10.9 2.004 3.857 2.004 3.857 arnoldi_extremal 4 6.8 0.000 0.000 3.350 3.377 arnoldi_normal_ev 4 7.8 0.001 0.003 3.350 3.377 make_images_sizes 222 9.7 0.000 0.000 0.668 3.305 mp_alltoall_i44 222 10.7 0.668 3.305 0.668 3.305 build_subspace 16 8.4 0.009 0.012 3.258 3.260 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.229 3.229 ls_scf_post 1 4.0 0.000 0.000 2.976 2.983 ls_scf_store_result 1 5.0 0.000 0.000 2.790 2.839 dbcsr_special_finalize 555 9.7 0.005 0.006 2.375 2.738 dbcsr_merge_single_wm 555 10.7 0.463 0.599 2.367 2.730 make_images_pack 222 9.7 2.205 2.632 2.207 2.634 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.356 2.594 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.456 2.064 2.458 dbcsr_sort_data 658 11.4 2.163 2.458 2.163 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.143 2.204 buffer_matrices_ensure_size 222 8.7 1.768 2.071 1.768 2.071 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.705 1.706 rebuild_ks_matrix 3 7.3 0.000 0.000 1.696 1.697 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.696 1.697 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.631000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 2.109862E+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 0.027 0.053 91.407 91.408 qs_energies 1 2.0 0.000 0.000 90.950 90.954 ls_scf 1 3.0 0.000 0.000 89.641 89.645 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.445 75.772 multiply_cannon 111 7.7 0.029 0.039 53.251 56.874 ls_scf_main 1 4.0 0.000 0.000 55.062 55.073 multiply_cannon_loop 111 8.7 0.116 0.123 49.967 53.580 density_matrix_trs4 2 5.0 0.002 0.003 49.380 49.599 ls_scf_init_scf 1 4.0 0.000 0.000 31.009 31.010 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.877 29.959 mp_waitall_1 9246 10.9 21.072 29.910 21.072 29.910 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.464 27.482 multiply_cannon_multrec 1332 9.7 13.164 16.678 22.455 27.161 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.657 20.540 make_m2s 222 7.7 0.006 0.007 15.408 16.070 make_images 222 8.7 1.577 1.951 15.378 16.042 dbcsr_mm_accdrv_process 4041 10.4 0.280 0.459 8.890 10.513 dbcsr_mm_accdrv_process_sort 4041 11.4 8.483 10.053 8.483 10.053 make_images_data 222 9.7 0.004 0.004 8.823 9.682 hybrid_alltoall_any 227 10.6 0.522 2.469 8.182 9.372 mp_sum_l 807 5.4 5.499 9.093 5.499 9.093 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.220 7.740 mp_irecv_dv 3311 11.0 3.200 7.680 3.200 7.680 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.222 7.159 calculate_norms 2376 9.8 6.006 6.765 6.006 6.765 multiply_cannon_sync_h2d 1332 9.7 4.800 5.790 4.800 5.790 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.071 5.287 arnoldi_extremal 4 6.8 0.000 0.000 4.923 4.949 arnoldi_normal_ev 4 7.8 0.001 0.005 4.923 4.949 build_subspace 16 8.4 0.014 0.021 4.645 4.656 ls_scf_post 1 4.0 0.000 0.000 3.570 3.575 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.242 3.462 ls_scf_store_result 1 5.0 0.000 0.000 3.258 3.394 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.228 2.747 3.230 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.534 2.652 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.214 2.537 make_images_pack 222 9.7 2.024 2.437 2.027 2.439 mp_allgather_i34 111 8.7 0.986 2.374 0.986 2.374 dbcsr_sort_data 436 11.2 1.835 2.046 1.835 2.046 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.408000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 2.715083E+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 1.243 2.204 123.049 123.059 qs_energies 1 2.0 0.000 0.004 96.211 96.239 ls_scf 1 3.0 0.005 0.014 91.839 91.860 dbcsr_multiply_generic 111 6.7 0.020 0.056 76.015 76.280 ls_scf_main 1 4.0 0.000 0.002 56.990 57.000 multiply_cannon 111 7.7 0.060 0.142 52.585 56.387 multiply_cannon_loop 111 8.7 0.100 0.110 48.957 52.718 density_matrix_trs4 2 5.0 0.002 0.004 51.204 51.371 mp_waitall_1 7374 11.0 22.947 33.195 22.947 33.195 ls_scf_init_scf 1 4.0 0.012 0.032 31.301 31.306 ls_scf_init_matrix_S 1 5.0 0.003 0.017 29.772 29.864 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.010 27.131 27.152 qs_init_subsys 1 2.0 0.007 0.019 19.582 25.851 qs_env_setup 1 3.0 1.052 3.149 19.507 25.761 multiply_cannon_multrec 888 9.7 12.632 15.457 21.222 24.655 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.760 22.825 make_m2s 222 7.7 0.008 0.020 16.681 17.897 make_images 222 8.7 1.973 2.279 16.641 17.859 qs_env_rebuild_pw_env 2 3.5 0.000 0.000 15.970 15.998 pw_env_rebuild 1 5.0 3.864 6.771 15.970 15.998 mp_max_i 400 1.7 8.341 13.500 8.341 13.500 hybrid_alltoall_any 227 10.6 0.622 2.873 9.064 10.421 make_images_data 222 9.7 0.003 0.004 9.304 10.328 dbcsr_mm_accdrv_process 3754 10.4 0.260 0.443 8.129 9.384 dbcsr_mm_accdrv_process_sort 3754 11.4 7.749 8.941 7.749 8.941 mp_sum_l 807 5.4 5.231 8.539 5.231 8.539 multiply_cannon_sync_h2d 888 9.7 6.111 7.592 6.111 7.592 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.458 7.228 mp_irecv_dv 2335 11.1 2.443 7.185 2.443 7.185 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.043 6.798 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.506 6.711 distribute_molecules_2d 1 4.0 2.352 6.274 2.478 6.607 pw_grid_setup 4 6.0 0.000 0.000 6.407 6.407 pw_grid_setup_internal 4 7.0 0.404 3.157 6.407 6.407 arnoldi_extremal 4 6.8 0.000 0.000 5.114 5.138 arnoldi_normal_ev 4 7.8 0.004 0.009 5.114 5.138 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.865 5.066 build_subspace 16 8.4 0.014 0.021 4.753 4.760 calculate_norms 1584 9.8 4.265 4.606 4.265 4.606 qs_energies_init_hamiltonians 1 3.0 0.016 0.056 4.363 4.368 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.438 3.752 mp_allgather_i34 111 8.7 1.432 3.731 1.432 3.731 dbcsr_matrix_vector_mult_local 304 10.0 3.022 3.595 3.024 3.597 ls_scf_post 1 4.0 0.001 0.005 3.543 3.571 qs_env_update_s_mstruct 1 4.0 0.013 0.045 2.845 3.346 ls_scf_store_result 1 5.0 0.000 0.000 3.248 3.334 rs_pw_transfer 27 10.4 0.000 0.000 2.414 3.204 pw_grid_sort 4 8.0 2.588 3.133 2.643 3.189 mp_max_lv 4 8.0 2.752 3.146 2.752 3.146 mp_waitany 462 12.2 2.260 3.056 2.260 3.056 pw_grid_remap 4 8.0 0.006 0.007 0.559 3.020 rs_pw_transfer_RS2PW_150 5 9.2 0.042 0.066 2.231 3.015 mp_sum_im3 4 9.0 0.553 3.015 0.553 3.015 calculate_rho_core 1 5.0 0.210 0.507 2.319 2.885 ls_scf_dm_to_ks 2 5.0 0.001 0.017 2.725 2.830 rs_grid_create_descriptor 4 6.0 2.601 2.623 2.701 2.715 build_core_hamiltonian_matrix 1 4.0 0.020 0.077 1.263 2.695 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=123.059000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2186.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 3.366203E+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 0.041 0.070 97.184 97.185 qs_energies 1 2.0 0.000 0.000 96.494 96.508 ls_scf 1 3.0 0.006 0.023 94.852 94.863 dbcsr_multiply_generic 111 6.7 0.017 0.024 78.643 78.893 ls_scf_main 1 4.0 0.000 0.000 58.813 58.819 multiply_cannon 111 7.7 0.058 0.147 51.847 56.768 density_matrix_trs4 2 5.0 0.002 0.003 52.770 52.893 multiply_cannon_loop 111 8.7 0.114 0.125 46.758 49.853 ls_scf_init_scf 1 4.0 0.000 0.003 32.805 32.806 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.605 31.660 mp_waitall_1 6438 11.0 22.749 29.958 22.749 29.958 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.096 29.112 multiply_cannon_multrec 1332 9.7 14.281 17.672 22.206 25.640 make_m2s 222 7.7 0.007 0.008 21.227 22.684 make_images 222 8.7 3.149 3.622 21.176 22.636 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.205 17.461 make_images_data 222 9.7 0.004 0.005 11.839 13.514 hybrid_alltoall_any 227 10.6 0.799 3.839 11.125 12.968 dbcsr_mm_accdrv_process 3641 10.4 0.237 0.411 7.554 9.054 dbcsr_mm_accdrv_process_sort 3641 11.4 7.175 8.644 7.175 8.644 mp_sum_l 807 5.4 3.988 7.191 3.988 7.191 multiply_cannon_sync_h2d 1332 9.7 5.567 6.362 5.567 6.362 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.074 6.137 mp_irecv_dv 3229 10.9 2.051 6.054 2.051 6.054 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.083 5.640 arnoldi_extremal 4 6.8 0.000 0.000 5.264 5.280 arnoldi_normal_ev 4 7.8 0.001 0.005 5.264 5.280 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.534 5.222 build_subspace 16 8.4 0.015 0.022 4.912 4.922 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.591 4.767 mp_allgather_i34 111 8.7 2.216 4.716 2.216 4.716 calculate_norms 2376 9.8 4.197 4.534 4.197 4.534 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.607 3.887 dbcsr_matrix_vector_mult_local 304 10.0 3.183 3.685 3.185 3.686 dbcsr_sort_data 658 11.4 3.060 3.544 3.060 3.544 ls_scf_post 1 4.0 0.002 0.011 3.229 3.243 dbcsr_special_finalize 555 9.7 0.006 0.007 2.804 3.215 dbcsr_merge_single_wm 555 10.7 0.535 0.665 2.796 3.207 ls_scf_store_result 1 5.0 0.000 0.000 2.978 3.029 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.885 2.938 dbcsr_data_release 10477 10.7 1.589 2.479 1.589 2.479 dbcsr_finalize 304 7.8 0.049 0.061 1.805 2.071 dbcsr_merge_all 275 8.9 0.317 0.362 1.686 1.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.185000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 4.623262E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.055 99.058 99.058 qs_energies 1 2.0 0.000 0.000 98.239 98.245 ls_scf 1 3.0 0.000 0.000 96.326 96.331 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.957 78.149 ls_scf_main 1 4.0 0.000 0.000 61.877 61.877 multiply_cannon 111 7.7 0.095 0.184 55.225 60.745 density_matrix_trs4 2 5.0 0.002 0.003 54.870 54.976 multiply_cannon_loop 111 8.7 0.069 0.077 50.659 52.305 mp_waitall_1 5481 11.0 26.177 32.226 26.177 32.226 ls_scf_init_scf 1 4.0 0.000 0.000 30.851 30.857 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.605 29.653 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.435 27.450 multiply_cannon_multrec 444 9.7 13.986 16.931 21.075 22.973 make_m2s 222 7.7 0.004 0.005 17.631 20.139 make_images 222 8.7 3.722 4.428 17.570 20.080 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.007 15.968 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.142 14.435 make_images_data 222 9.7 0.003 0.004 9.825 12.289 hybrid_alltoall_any 227 10.6 0.791 3.770 9.623 12.109 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.343 6.794 7.923 dbcsr_mm_accdrv_process_sort 3003 11.4 6.478 7.578 6.478 7.578 multiply_cannon_sync_h2d 444 9.7 6.534 7.463 6.534 7.463 mp_allgather_i34 111 8.7 2.789 7.095 2.789 7.095 arnoldi_extremal 4 6.8 0.000 0.000 5.871 5.887 arnoldi_normal_ev 4 7.8 0.002 0.005 5.871 5.887 build_subspace 16 8.4 0.015 0.020 5.459 5.470 mp_sum_l 807 5.4 3.143 5.180 3.143 5.180 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.597 4.758 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.694 4.571 mp_irecv_dv 1241 11.2 1.676 4.543 1.676 4.543 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.174 4.365 dbcsr_matrix_vector_mult_local 304 10.0 3.686 4.154 3.688 4.156 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.215 3.854 calculate_norms 792 9.8 3.551 3.694 3.551 3.694 ls_scf_post 1 4.0 0.000 0.000 3.598 3.603 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.389 3.494 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.431 make_images_sizes 222 9.7 0.000 0.000 1.161 3.378 mp_alltoall_i44 222 10.7 1.161 3.377 1.161 3.377 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.319 dbcsr_merge_all 275 8.9 0.474 0.528 2.048 2.147 dbcsr_data_release 10123 10.8 1.334 2.022 1.334 2.022 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.995 1.996 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.058000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3672.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90_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 8.772583E+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 0.126 107.686 107.687 qs_energies 1 2.0 0.000 0.000 106.244 106.259 ls_scf 1 3.0 0.000 0.000 103.371 103.386 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.235 77.343 ls_scf_main 1 4.0 0.000 0.000 65.626 65.627 density_matrix_trs4 2 5.0 0.002 0.003 56.787 56.837 multiply_cannon 111 7.7 0.173 0.352 49.936 51.840 multiply_cannon_loop 111 8.7 0.067 0.069 46.528 47.430 ls_scf_init_scf 1 4.0 0.000 0.000 34.092 34.092 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.780 32.785 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.953 29.961 mp_waitall_1 4569 11.1 22.101 25.576 22.101 25.576 make_m2s 222 7.7 0.005 0.005 23.743 24.655 make_images 222 8.7 4.590 4.967 23.638 24.547 multiply_cannon_multrec 444 9.7 17.845 18.614 22.494 23.047 hybrid_alltoall_any 227 10.6 1.659 3.623 12.723 15.533 make_images_data 222 9.7 0.003 0.003 12.938 15.313 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.565 11.015 multiply_cannon_sync_h2d 444 9.7 8.846 8.887 8.846 8.887 arnoldi_extremal 4 6.8 0.000 0.000 7.608 7.622 arnoldi_normal_ev 4 7.8 0.007 0.013 7.608 7.622 build_subspace 16 8.4 0.026 0.036 7.031 7.040 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.510 5.702 dbcsr_matrix_vector_mult_local 304 10.0 5.009 5.295 5.011 5.298 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.270 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.798 4.896 dbcsr_mm_accdrv_process 1814 10.4 0.279 0.327 4.469 4.608 dbcsr_mm_accdrv_process_sort 1814 11.4 4.167 4.292 4.167 4.292 ls_scf_post 1 4.0 0.000 0.000 3.653 3.668 make_images_sizes 222 9.7 0.000 0.000 1.474 3.604 mp_alltoall_i44 222 10.7 1.474 3.603 1.474 3.603 ls_scf_store_result 1 5.0 0.000 0.000 3.405 3.413 calculate_norms 792 9.8 3.241 3.273 3.241 3.273 mp_allgather_i34 111 8.7 1.044 3.256 1.044 3.256 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.130 dbcsr_merge_all 275 8.9 0.893 0.919 2.870 2.910 dbcsr_complete_redistribute 5 7.6 1.438 1.473 2.772 2.900 dbcsr_data_release 12724 10.6 2.324 2.850 2.324 2.850 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.842 2.842 matrix_ls_to_qs 2 6.0 0.000 0.000 2.425 2.562 dbcsr_sort_data 325 11.1 2.442 2.500 2.442 2.500 dbcsr_new_transposed 4 7.5 0.242 0.251 2.426 2.444 dbcsr_frobenius_norm 74 6.6 2.057 2.130 2.199 2.246 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.204 dbcsr_add_anytype 103 7.2 0.859 0.891 2.133 2.204 mp_alltoall_d11v 48 9.2 2.133 2.191 2.133 2.191 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.189 2.191 dbcsr_redistribute 4 8.5 1.383 1.444 2.155 2.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.687000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: bdcfe8909c72340e1d8ec8d5ee5e3e5653de6b90 Summary: empty Status: OK