=== 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: 34639cb15a96dbc4801a8ed01621e36a84bba0a6 ################# 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/01 job id: 42350843 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/02 job id: 42350844 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/03 job id: 42350845 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/04 job id: 42350846 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/05 job id: 42350847 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/06 job id: 42350848 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/07 job id: 42350850 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/08 job id: 42350851 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/09 job id: 42350852 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/10 job id: 42350853 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/11 job id: 42350854 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/12 job id: 42350855 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/13 job id: 42350856 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/14 job id: 42350857 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/15 job id: 42350858 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/16 job id: 42350859 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/17 job id: 42350860 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/18 job id: 42350862 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/19 job id: 42350863 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/20 job id: 42350864 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/21 job id: 42350866 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/22 job id: 42350867 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/23 job id: 42350868 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/24 job id: 42350869 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/25 job id: 42350871 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/26 job id: 42350872 --- 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 0.035 133.626 133.626 farming_run 1 2.0 133.202 133.204 133.595 133.599 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.461785E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 115.970 115.970 qs_energies 1 2.0 0.000 0.000 115.745 115.747 mp2_main 1 3.0 0.000 0.000 113.704 113.706 mp2_gpw_main 1 4.0 0.019 0.025 112.793 112.795 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.985 94.359 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.589 55.964 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.171 41.896 47.199 get_2c_integrals 1 6.0 0.000 0.000 37.394 38.224 integrate_v_rspace 273 8.0 0.434 0.447 25.254 30.383 pw_transfer 6555 10.6 0.372 0.397 27.603 28.244 fft_wrap_pw1pw2 5465 11.4 0.043 0.046 26.186 26.809 grid_integrate_task_list 273 9.0 21.055 26.678 21.055 26.678 fft_wrap_pw1pw2_100 2178 12.4 1.282 1.425 23.748 24.362 compute_2c_integrals 1 7.0 0.002 0.002 19.743 19.744 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.187 19.449 mp2_eri_2c_integrate_gpw 1 9.0 2.345 2.441 19.185 19.448 rpa_ri_compute_en 1 5.0 0.000 0.000 18.699 18.979 cp_fm_cholesky_decompose 12 8.2 17.641 18.508 17.641 18.508 cholesky_decomp 1 7.0 0.000 0.000 16.501 17.368 fft3d_s 5443 13.4 16.267 16.759 16.290 16.780 ao_to_mo_and_store_B_mult_1 272 7.0 10.868 15.587 10.868 15.587 calculate_wavefunction 272 8.0 5.423 5.573 12.560 13.158 rpa_num_int 1 6.0 0.000 0.000 10.596 10.596 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.528 10.541 calc_mat_Q 8 8.0 0.000 0.000 9.335 9.429 contract_S_to_Q 8 9.0 0.000 0.000 8.758 8.850 calc_potential_gpw 544 9.5 0.005 0.006 8.345 8.733 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.401 8.584 potential_pw2rs 545 10.0 0.108 0.109 7.801 8.443 parallel_gemm_fm 14 9.1 0.000 0.000 8.335 8.422 parallel_gemm_fm_cosma 14 10.1 8.334 8.422 8.334 8.422 collocate_single_gaussian 272 10.0 0.040 0.042 7.620 7.797 create_integ_mat 1 6.0 0.022 0.027 7.634 7.634 array2fm 1 7.0 0.000 0.000 6.589 7.086 pw_scatter_s 2720 13.7 4.385 4.550 4.385 4.550 pw_gather_s 2722 13.2 3.875 4.064 3.875 4.064 array2fm_buffer_send 1 8.0 2.936 3.172 2.936 3.172 pw_poisson_solve 545 10.5 1.134 1.188 2.261 2.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.794941, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2734.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 0.030 0.038 396.402 396.403 farming_run 1 2.0 395.638 395.658 396.358 396.363 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222316E+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 0.048 209.208 209.209 qs_energies 1 2.0 0.000 0.001 208.988 208.994 scf_env_do_scf 1 3.0 0.000 0.000 106.345 106.345 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.426 105.436 rebuild_ks_matrix 4 6.0 0.000 0.000 105.424 105.434 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.069 105.424 105.434 hfx_ks_matrix 4 8.0 0.001 0.001 105.043 105.046 integrate_four_center 4 9.0 0.144 0.458 105.043 105.046 mp2_main 1 3.0 0.000 0.002 102.350 102.356 mp2_gpw_main 1 4.0 0.033 0.047 101.416 101.427 integrate_four_center_main 4 10.0 0.097 0.514 96.543 99.128 integrate_four_center_bin 264 11.0 96.446 99.054 96.446 99.054 init_scf_loop 1 4.0 0.000 0.000 92.146 92.146 mp2_ri_gpw_compute_in 1 5.0 0.065 0.073 74.827 75.861 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.540 55.575 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.158 42.227 47.351 integrate_v_rspace 95 8.0 0.398 0.562 28.602 33.591 pw_transfer 2240 10.6 0.147 0.164 29.872 30.295 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.888 29.300 grid_integrate_task_list 95 9.0 23.886 29.084 23.886 29.084 mp2_ri_gpw_compute_en 1 5.0 0.056 0.066 26.441 28.246 ao_to_mo_and_store_B_mult_1 91 7.0 10.627 27.159 10.627 27.159 fft_wrap_pw1pw2_100 730 12.4 1.265 1.422 26.623 27.055 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.893 24.761 24.771 get_2c_integrals 1 6.0 0.000 0.002 20.198 20.221 compute_2c_integrals 1 7.0 0.003 0.004 19.180 19.183 compute_2c_integrals_loop_lm 1 8.0 0.002 0.005 18.826 19.048 mp2_eri_2c_integrate_gpw 1 9.0 1.748 1.937 18.825 19.047 fft3d_s 1823 13.4 18.418 18.850 18.432 18.864 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.196 14.196 calculate_wavefunction 91 8.0 2.013 2.051 9.740 9.989 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.593 8.737 9.174 potential_pw2rs 186 10.0 0.034 0.036 8.603 9.137 local_gemm 172 8.0 8.180 8.616 8.180 8.616 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.002 8.208 8.517 mp2_ri_gpw_compute_en_comm 22 7.0 0.514 0.543 7.789 8.161 calc_potential_gpw 182 9.5 0.002 0.002 7.919 8.128 collocate_single_gaussian 91 10.0 0.017 0.021 7.869 8.085 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.410 6.349 6.410 mp_sendrecv_dm3 2068 8.0 5.785 6.132 5.785 6.132 pw_gather_s 912 13.2 4.901 5.377 4.901 5.377 mp_sync 38 10.4 3.140 5.366 3.140 5.366 pw_scatter_s 910 13.7 3.936 4.220 3.936 4.220 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.412105, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.227648E+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 0.018 0.036 55.007 55.009 qs_mol_dyn_low 1 2.0 0.003 0.004 54.748 54.755 qs_forces 11 3.9 0.002 0.002 54.686 54.687 qs_energies 11 4.9 0.002 0.006 53.231 53.244 scf_env_do_scf 11 5.9 0.000 0.002 47.062 47.063 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.929 44.929 dbcsr_multiply_generic 2286 12.5 0.094 0.098 36.042 36.509 qs_scf_new_mos 108 7.5 0.000 0.001 34.784 35.080 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.783 35.080 ot_scf_mini 108 9.5 0.002 0.002 33.002 33.187 multiply_cannon 2286 13.5 0.179 0.185 27.277 28.962 multiply_cannon_loop 2286 14.5 1.485 1.588 26.513 28.198 velocity_verlet 10 3.0 0.001 0.002 26.070 26.070 ot_mini 108 10.5 0.001 0.001 21.319 21.575 qs_ot_get_derivative 108 11.5 0.001 0.001 18.198 18.413 mp_waitall_1 267858 16.1 10.115 15.989 10.115 15.989 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 6.934 14.003 multiply_cannon_multrec 54864 15.5 4.154 6.418 7.623 11.008 mp_sum_l 7207 12.9 6.657 8.181 6.657 8.181 rebuild_ks_matrix 119 8.3 0.000 0.000 7.909 8.065 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.909 8.065 multiply_cannon_sync_h2d 54864 15.5 5.735 7.572 5.735 7.572 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.990 7.131 qs_ot_get_p 119 10.4 0.001 0.001 6.719 7.026 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.673 6.769 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.821 6.279 init_scf_run 11 5.9 0.000 0.001 4.930 4.930 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.929 4.930 dbcsr_mm_accdrv_process 76910 16.1 1.156 1.862 3.393 4.774 sum_up_and_integrate 119 10.3 0.012 0.015 4.509 4.514 integrate_v_rspace 119 11.3 0.002 0.002 4.496 4.503 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.134 4.290 calculate_rho_elec 119 8.7 0.011 0.016 4.133 4.289 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.695 3.764 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 2.268 3.704 calculate_dm_sparse 119 9.5 0.000 0.000 3.109 3.234 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.938 3.149 apply_single 119 13.6 0.000 0.000 2.938 3.148 rs_pw_transfer 974 11.9 0.012 0.013 2.780 2.868 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.763 2.840 jit_kernel_multiply 13 15.8 2.177 2.826 2.177 2.826 ot_diis_step 108 11.5 0.006 0.006 2.780 2.781 calculate_first_density_matrix 1 7.0 0.000 0.000 2.779 2.780 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.702 2.702 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.451 2.453 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.433 2.434 cp_fm_redistribute_end 50 14.0 2.207 2.407 2.211 2.409 cp_fm_diag_elpa_base 50 14.0 0.195 2.362 0.196 2.372 density_rs2pw 119 9.7 0.004 0.004 2.206 2.293 acc_transpose_blocks 54864 15.5 0.219 0.245 1.762 2.268 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.085 2.134 grid_integrate_task_list 119 12.3 2.019 2.130 2.019 2.130 init_scf_loop 11 6.9 0.000 0.000 2.114 2.114 wfi_extrapolate 11 7.9 0.001 0.001 2.086 2.086 mp_sum_d 4129 12.0 1.484 2.038 1.484 2.038 potential_pw2rs 119 12.3 0.004 0.004 1.893 1.908 make_m2s 4572 13.5 0.053 0.056 1.796 1.842 pw_transfer 1439 11.6 0.051 0.055 1.731 1.804 make_images 4572 14.5 0.133 0.141 1.714 1.759 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.656 1.732 mp_alltoall_d11v 2130 13.8 1.358 1.625 1.358 1.625 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.494 1.514 fft3d_ps 1201 14.6 0.359 0.465 1.435 1.504 grid_collocate_task_list 119 9.7 1.291 1.357 1.291 1.357 fft_wrap_pw1pw2_140 487 13.2 0.079 0.094 1.248 1.328 mp_waitany 12084 13.8 1.224 1.314 1.224 1.314 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.847 1.230 acc_transpose_blocks_kernels 54864 16.5 0.248 0.386 0.815 1.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.009000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.968768E+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 1761540. 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 0.096 0.233 39.684 39.686 qs_mol_dyn_low 1 2.0 0.004 0.005 39.072 39.079 qs_forces 11 3.9 0.002 0.003 38.962 38.963 qs_energies 11 4.9 0.012 0.037 37.260 37.264 scf_env_do_scf 11 5.9 0.001 0.001 31.598 31.599 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.966 28.968 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.557 21.925 qs_scf_new_mos 108 7.5 0.001 0.001 19.933 20.187 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.932 20.186 ot_scf_mini 108 9.5 0.003 0.003 19.027 19.209 velocity_verlet 10 3.0 0.001 0.002 18.489 18.490 multiply_cannon 2286 13.5 0.207 0.218 16.422 18.038 multiply_cannon_loop 2286 14.5 0.905 0.975 15.215 16.679 ot_mini 108 10.5 0.001 0.001 11.877 12.117 mp_waitall_1 217478 16.2 6.209 11.321 6.209 11.321 qs_ot_get_derivative 108 11.5 0.001 0.001 9.402 9.584 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.155 9.447 multiply_cannon_multrec 27432 15.5 1.962 4.489 5.845 8.601 rebuild_ks_matrix 119 8.3 0.000 0.000 7.195 7.332 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.194 7.331 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.350 6.474 dbcsr_mm_accdrv_process 47894 16.0 3.004 5.078 3.815 5.648 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.623 4.467 qs_ot_get_p 119 10.4 0.001 0.001 4.216 4.449 init_scf_run 11 5.9 0.000 0.001 4.243 4.243 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.242 4.243 sum_up_and_integrate 119 10.3 0.024 0.026 4.235 4.242 integrate_v_rspace 119 11.3 0.002 0.003 4.210 4.219 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.060 4.136 apply_single 119 13.6 0.000 0.000 3.060 4.136 mp_sum_l 7207 12.9 2.094 4.060 2.094 4.060 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.660 3.699 calculate_rho_elec 119 8.7 0.021 0.024 3.660 3.699 make_m2s 4572 13.5 0.052 0.054 2.601 2.879 rs_pw_transfer 974 11.9 0.010 0.011 2.737 2.838 multiply_cannon_sync_h2d 27432 15.5 2.185 2.836 2.185 2.836 make_images 4572 14.5 0.199 0.236 2.513 2.790 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.712 2.731 calculate_first_density_matrix 1 7.0 0.001 0.003 2.711 2.713 init_scf_loop 11 6.9 0.001 0.004 2.606 2.614 ot_diis_step 108 11.5 0.011 0.011 2.423 2.424 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.277 2.278 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.182 2.277 calculate_dm_sparse 119 9.5 0.000 0.000 2.104 2.182 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.147 2.149 density_rs2pw 119 9.7 0.004 0.004 2.026 2.116 potential_pw2rs 119 12.3 0.006 0.007 1.978 1.987 grid_integrate_task_list 119 12.3 1.832 1.928 1.832 1.928 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.923 1.923 cp_fm_redistribute_end 50 14.0 1.586 1.898 1.590 1.899 cp_fm_diag_elpa_base 50 14.0 0.301 1.840 0.308 1.870 pw_transfer 1439 11.6 0.063 0.066 1.772 1.806 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.734 1.777 jit_kernel_multiply 9 16.2 0.759 1.732 0.759 1.732 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.683 1.717 prepare_preconditioner 11 7.9 0.000 0.000 1.668 1.693 make_preconditioner 11 8.9 0.000 0.002 1.668 1.693 make_images_data 4572 15.5 0.045 0.052 1.191 1.693 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.565 1.621 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.036 1.572 acc_transpose_blocks 27432 15.5 0.108 0.112 1.244 1.557 wfi_extrapolate 11 7.9 0.001 0.001 1.471 1.471 mp_allgather_i34 2286 14.5 0.651 1.431 0.651 1.431 fft3d_ps 1201 14.6 0.495 0.548 1.390 1.422 mp_alltoall_d11v 2130 13.8 1.241 1.416 1.241 1.416 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.308 1.316 grid_collocate_task_list 119 9.7 1.242 1.305 1.242 1.305 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.265 1.299 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.210 1.258 qs_energies_init_hamiltonians 11 5.9 0.070 0.081 1.102 1.119 mp_sum_d 4129 12.0 0.613 1.072 0.613 1.072 acc_transpose_blocks_kernels 27432 16.5 0.183 0.271 0.713 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.903 0.916 dbcsr_complete_redistribute 329 12.2 0.222 0.290 0.781 0.846 make_images_sizes 4572 15.5 0.005 0.005 0.592 0.842 mp_alltoall_i44 4572 16.5 0.587 0.837 0.587 0.837 rs_pw_transfer_PW2RS_50 119 14.3 0.586 0.604 0.775 0.817 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.686000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.272727, yerr=1.285649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 523.943936E+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 0.042 0.105 32.688 32.690 qs_mol_dyn_low 1 2.0 0.003 0.003 32.227 32.236 qs_forces 11 3.9 0.002 0.003 32.165 32.167 qs_energies 11 4.9 0.002 0.007 30.597 30.600 scf_env_do_scf 11 5.9 0.001 0.002 25.542 25.542 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.924 22.925 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.434 16.501 velocity_verlet 10 3.0 0.002 0.004 15.344 15.346 qs_scf_new_mos 108 7.5 0.001 0.001 14.813 14.829 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.812 14.828 ot_scf_mini 108 9.5 0.002 0.003 14.081 14.094 multiply_cannon 2286 13.5 0.194 0.200 13.041 13.764 multiply_cannon_loop 2286 14.5 0.641 0.667 12.220 12.981 ot_mini 108 10.5 0.001 0.001 8.772 8.782 qs_ot_get_derivative 108 11.5 0.001 0.001 7.280 7.292 multiply_cannon_multrec 18288 15.5 1.946 2.886 6.764 7.087 rebuild_ks_matrix 119 8.3 0.000 0.000 6.415 6.429 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.414 6.429 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.675 5.687 dbcsr_mm_accdrv_process 38222 16.0 4.036 5.307 4.736 5.543 mp_waitall_1 169478 16.3 3.059 4.271 3.059 4.271 sum_up_and_integrate 119 10.3 0.030 0.031 4.029 4.034 integrate_v_rspace 119 11.3 0.002 0.003 3.998 4.008 init_scf_run 11 5.9 0.000 0.001 3.788 3.788 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.787 3.788 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.329 3.334 calculate_rho_elec 119 8.7 0.030 0.031 3.328 3.333 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.720 3.333 qs_ot_get_p 119 10.4 0.001 0.001 3.114 3.131 init_scf_loop 11 6.9 0.000 0.002 2.596 2.597 calculate_first_density_matrix 1 7.0 0.000 0.002 2.562 2.565 rs_pw_transfer 974 11.9 0.009 0.010 2.383 2.472 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.431 2.415 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.022 2.355 apply_single 119 13.6 0.000 0.000 2.022 2.355 make_m2s 4572 13.5 0.044 0.045 1.909 2.080 density_rs2pw 119 9.7 0.004 0.004 1.912 2.031 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.009 2.015 make_images 4572 14.5 0.189 0.201 1.824 1.995 grid_integrate_task_list 119 12.3 1.799 1.914 1.799 1.914 calculate_dm_sparse 119 9.5 0.000 0.000 1.822 1.832 potential_pw2rs 119 12.3 0.007 0.008 1.802 1.817 pw_transfer 1439 11.6 0.063 0.066 1.785 1.798 prepare_preconditioner 11 7.9 0.000 0.000 1.791 1.793 make_preconditioner 11 8.9 0.000 0.000 1.791 1.793 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.741 1.742 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.645 1.727 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.695 1.708 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.701 1.703 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.605 1.614 multiply_cannon_sync_h2d 18288 15.5 1.392 1.565 1.392 1.565 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.552 1.559 jit_kernel_multiply 8 16.2 0.648 1.516 0.648 1.516 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.487 1.488 ot_diis_step 108 11.5 0.011 0.011 1.473 1.473 cp_fm_redistribute_end 50 14.0 1.103 1.465 1.104 1.466 cp_fm_diag_elpa_base 50 14.0 0.345 1.400 0.359 1.439 mp_sum_l 7207 12.9 1.121 1.431 1.121 1.431 fft3d_ps 1201 14.6 0.507 0.526 1.382 1.395 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.313 1.327 grid_collocate_task_list 119 9.7 1.204 1.282 1.204 1.282 acc_transpose_blocks 18288 15.5 0.075 0.077 1.242 1.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.186 1.190 wfi_extrapolate 11 7.9 0.001 0.001 1.177 1.177 make_images_data 4572 15.5 0.045 0.049 0.852 1.071 qs_energies_init_hamiltonians 11 5.9 0.009 0.071 1.028 1.030 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.719 0.960 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.889 0.913 mp_alltoall_d11v 2130 13.8 0.727 0.819 0.727 0.819 acc_transpose_blocks_kernels 18288 16.5 0.212 0.219 0.802 0.817 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.811 0.813 mp_alltoall_z22v 1201 16.6 0.711 0.775 0.711 0.775 cp_fm_cholesky_invert 11 10.9 0.764 0.768 0.764 0.768 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.729 parallel_gemm_fm 81 9.0 0.000 0.000 0.663 0.666 parallel_gemm_fm_cosma 81 10.0 0.663 0.666 0.663 0.666 mp_allgather_i34 2286 14.5 0.285 0.663 0.285 0.663 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.690000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.545455, yerr=2.462949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 547.319808E+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 0.022 0.060 35.774 35.775 qs_mol_dyn_low 1 2.0 0.003 0.004 35.221 35.228 qs_forces 11 3.9 0.012 0.031 34.926 34.927 qs_energies 11 4.9 0.003 0.008 33.211 33.225 scf_env_do_scf 11 5.9 0.001 0.002 27.944 27.946 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.503 24.503 dbcsr_multiply_generic 2286 12.5 0.101 0.106 18.352 18.444 velocity_verlet 10 3.0 0.002 0.002 17.851 17.863 qs_scf_new_mos 108 7.5 0.001 0.001 16.260 16.310 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.260 16.310 ot_scf_mini 108 9.5 0.003 0.004 15.324 15.375 multiply_cannon 2286 13.5 0.227 0.257 14.611 15.063 multiply_cannon_loop 2286 14.5 0.937 0.966 13.642 14.066 ot_mini 108 10.5 0.001 0.001 9.475 9.543 multiply_cannon_multrec 27432 15.5 2.354 3.007 8.633 8.978 qs_ot_get_derivative 108 11.5 0.001 0.001 7.668 7.720 dbcsr_mm_accdrv_process 47916 15.9 5.221 7.305 6.185 7.397 rebuild_ks_matrix 119 8.3 0.000 0.000 6.509 6.563 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.508 6.563 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.790 5.837 sum_up_and_integrate 119 10.3 0.035 0.038 3.785 3.792 init_scf_run 11 5.9 0.000 0.001 3.766 3.766 scf_env_initial_rho_setup 11 6.9 0.001 0.005 3.766 3.766 integrate_v_rspace 119 11.3 0.002 0.003 3.750 3.759 init_scf_loop 11 6.9 0.001 0.004 3.419 3.420 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.369 3.404 calculate_rho_elec 119 8.7 0.040 0.046 3.369 3.403 qs_ot_get_p 119 10.4 0.001 0.001 3.220 3.299 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.690 3.176 mp_waitall_1 145218 16.4 2.319 2.890 2.319 2.890 prepare_preconditioner 11 7.9 0.000 0.000 2.569 2.577 make_preconditioner 11 8.9 0.000 0.002 2.569 2.576 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.122 2.567 apply_single 119 13.6 0.000 0.000 2.121 2.566 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.169 2.497 calculate_first_density_matrix 1 7.0 0.001 0.003 2.391 2.395 make_m2s 4572 13.5 0.054 0.056 2.263 2.389 make_images 4572 14.5 0.273 0.334 2.155 2.280 rs_pw_transfer 974 11.9 0.009 0.009 2.104 2.226 calculate_dm_sparse 119 9.5 0.000 0.000 2.056 2.115 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.964 1.991 density_rs2pw 119 9.7 0.003 0.004 1.849 1.960 qs_ot_p2m_diag 50 11.0 0.015 0.022 1.916 1.924 grid_integrate_task_list 119 12.3 1.821 1.904 1.821 1.904 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.898 1.899 pw_transfer 1439 11.6 0.063 0.067 1.818 1.846 ot_diis_step 108 11.5 0.012 0.012 1.767 1.767 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.728 1.760 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.665 1.679 mp_sum_l 7207 12.9 1.054 1.667 1.054 1.667 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.603 1.603 potential_pw2rs 119 12.3 0.009 0.009 1.568 1.572 acc_transpose_blocks 27432 15.5 0.113 0.116 1.483 1.502 jit_kernel_multiply 8 15.7 0.903 1.501 0.903 1.501 fft3d_ps 1201 14.6 0.536 0.588 1.409 1.433 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.390 1.424 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.857 1.421 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.354 1.355 cp_fm_redistribute_end 50 14.0 0.891 1.328 0.893 1.328 wfi_extrapolate 11 7.9 0.001 0.001 1.323 1.323 cp_fm_diag_elpa_base 50 14.0 0.413 1.261 0.433 1.302 grid_collocate_task_list 119 9.7 1.221 1.292 1.221 1.292 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.241 1.255 qs_energies_init_hamiltonians 11 5.9 0.044 0.066 1.229 1.241 dbcsr_complete_redistribute 329 12.2 0.126 0.181 0.850 1.140 cp_fm_upper_to_full 72 13.5 0.817 1.130 0.817 1.130 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.105 1.126 multiply_cannon_sync_h2d 27432 15.5 0.987 1.067 0.987 1.067 make_images_data 4572 15.5 0.045 0.049 0.890 1.027 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.758 0.948 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.639 0.920 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.874 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.855 0.866 mp_alltoall_d11v 2130 13.8 0.766 0.845 0.766 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.815 mp_alltoall_z22v 1201 16.6 0.754 0.780 0.754 0.780 mp_alltoall_i22 627 13.8 0.465 0.764 0.465 0.764 cp_fm_cholesky_invert 11 10.9 0.755 0.759 0.755 0.759 qs_env_update_s_mstruct 11 6.9 0.018 0.024 0.663 0.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.775000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=519.636364, yerr=3.867645 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 603.234304E+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 0.052 0.066 28.053 28.053 qs_mol_dyn_low 1 2.0 0.003 0.003 27.587 27.594 qs_forces 11 3.9 0.018 0.028 27.527 27.528 qs_energies 11 4.9 0.002 0.004 25.827 25.847 scf_env_do_scf 11 5.9 0.001 0.005 20.799 20.799 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.315 18.315 velocity_verlet 10 3.0 0.002 0.002 14.014 14.016 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.276 12.355 qs_scf_new_mos 108 7.5 0.001 0.001 10.843 10.866 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.843 10.865 multiply_cannon 2286 13.5 0.231 0.240 9.730 10.215 ot_scf_mini 108 9.5 0.002 0.002 10.171 10.197 multiply_cannon_loop 2286 14.5 0.332 0.344 8.820 9.036 multiply_cannon_multrec 9144 15.5 1.614 1.843 5.888 6.107 ot_mini 108 10.5 0.001 0.001 5.759 5.790 rebuild_ks_matrix 119 8.3 0.000 0.000 5.756 5.778 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.755 5.778 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.133 5.154 qs_ot_get_derivative 108 11.5 0.001 0.001 4.472 4.499 dbcsr_mm_accdrv_process 12550 15.8 3.194 4.210 4.174 4.250 sum_up_and_integrate 119 10.3 0.037 0.041 3.525 3.531 init_scf_run 11 5.9 0.000 0.001 3.507 3.507 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.507 3.507 integrate_v_rspace 119 11.3 0.003 0.004 3.488 3.494 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.159 3.185 calculate_rho_elec 119 8.7 0.059 0.061 3.158 3.184 qs_ot_get_p 119 10.4 0.001 0.001 2.637 2.674 init_scf_loop 11 6.9 0.000 0.000 2.458 2.458 calculate_first_density_matrix 1 7.0 0.000 0.000 2.398 2.403 mp_waitall_1 121218 16.5 1.636 2.063 1.636 2.063 grid_integrate_task_list 119 12.3 1.847 1.906 1.847 1.906 make_m2s 4572 13.5 0.034 0.035 1.744 1.904 jit_kernel_multiply 10 15.8 0.942 1.861 0.942 1.861 calculate_dm_sparse 119 9.5 0.000 0.000 1.808 1.831 make_images 4572 14.5 0.267 0.300 1.655 1.813 prepare_preconditioner 11 7.9 0.000 0.000 1.727 1.733 make_preconditioner 11 8.9 0.000 0.000 1.727 1.733 rs_pw_transfer 974 11.9 0.008 0.008 1.631 1.732 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.714 1.717 density_rs2pw 119 9.7 0.004 0.004 1.590 1.709 pw_transfer 1439 11.6 0.063 0.065 1.694 1.701 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.617 1.644 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.604 1.611 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.574 1.588 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.550 1.552 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.519 1.519 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.392 1.402 grid_collocate_task_list 119 9.7 1.271 1.384 1.271 1.384 qs_energies_init_hamiltonians 11 5.9 0.056 0.064 1.306 1.323 potential_pw2rs 119 12.3 0.010 0.011 1.314 1.316 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.305 1.313 fft3d_ps 1201 14.6 0.540 0.553 1.271 1.278 ot_diis_step 108 11.5 0.013 0.013 1.274 1.274 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.246 1.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.232 1.241 cp_fm_redistribute_end 50 14.0 0.619 1.226 0.620 1.226 cp_fm_diag_elpa_base 50 14.0 0.563 1.155 0.605 1.212 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.147 1.169 apply_single 119 13.6 0.000 0.000 1.146 1.169 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.140 1.152 wfi_extrapolate 11 7.9 0.001 0.001 1.060 1.060 hybrid_alltoall_any 4725 16.4 0.062 0.177 0.774 1.032 make_images_data 4572 15.5 0.039 0.042 0.813 1.021 acc_transpose_blocks 9144 15.5 0.039 0.040 0.979 0.988 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.923 cp_fm_cholesky_invert 11 10.9 0.870 0.873 0.870 0.873 mp_alltoall_d11v 2130 13.8 0.767 0.839 0.767 0.839 multiply_cannon_sync_h2d 9144 15.5 0.705 0.785 0.705 0.785 qs_env_update_s_mstruct 11 6.9 0.023 0.024 0.697 0.766 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.743 0.746 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.737 0.745 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.722 0.729 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.357 0.657 mp_allgather_i34 2286 14.5 0.242 0.648 0.242 0.648 mp_alltoall_z22v 1201 16.6 0.605 0.637 0.605 0.637 jit_kernel_transpose 5 15.6 0.620 0.628 0.620 0.628 qs_create_task_list 11 7.9 0.001 0.001 0.543 0.568 generate_qs_task_list 11 8.9 0.189 0.212 0.543 0.568 mp_waitany 5200 13.7 0.458 0.566 0.458 0.566 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.053000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.272727, yerr=5.690357 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 738.455552E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 40.652 40.652 qs_mol_dyn_low 1 2.0 0.003 0.003 40.439 40.445 qs_forces 11 3.9 0.002 0.002 40.366 40.367 qs_energies 11 4.9 0.001 0.001 38.426 38.429 scf_env_do_scf 11 5.9 0.001 0.002 32.785 32.785 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.924 24.925 velocity_verlet 10 3.0 0.002 0.002 22.844 22.850 dbcsr_multiply_generic 2286 12.5 0.101 0.102 17.488 17.557 qs_scf_new_mos 108 7.5 0.001 0.001 16.018 16.113 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.017 16.113 ot_scf_mini 108 9.5 0.002 0.002 14.940 15.037 multiply_cannon 2286 13.5 0.307 0.315 13.599 14.398 multiply_cannon_loop 2286 14.5 0.343 0.350 12.359 13.207 ot_mini 108 10.5 0.001 0.001 8.991 9.101 multiply_cannon_multrec 9144 15.5 3.503 5.091 8.680 8.878 init_scf_loop 11 6.9 0.000 0.000 7.833 7.835 qs_ot_get_derivative 108 11.5 0.001 0.001 6.979 7.073 rebuild_ks_matrix 119 8.3 0.000 0.000 6.831 6.966 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.830 6.966 prepare_preconditioner 11 7.9 0.000 0.000 6.921 6.934 make_preconditioner 11 8.9 0.000 0.000 6.921 6.934 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.468 6.814 dbcsr_mm_accdrv_process 12550 15.8 4.033 5.631 5.056 6.359 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.177 6.300 cp_fm_upper_to_full 72 14.2 3.272 4.674 3.272 4.674 sum_up_and_integrate 119 10.3 0.064 0.066 3.701 3.708 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.688 3.692 calculate_rho_elec 119 8.7 0.118 0.121 3.687 3.692 integrate_v_rspace 119 11.3 0.003 0.003 3.637 3.644 init_scf_run 11 5.9 0.000 0.001 3.614 3.614 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.614 3.614 mp_waitall_1 97218 16.6 2.465 3.497 2.465 3.497 qs_ot_get_p 119 10.4 0.001 0.001 3.234 3.367 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.420 2.866 dbcsr_complete_redistribute 329 12.2 0.287 0.294 2.003 2.800 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.684 2.481 make_m2s 4572 13.5 0.038 0.038 2.298 2.451 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.196 2.435 apply_single 119 13.6 0.000 0.000 2.196 2.435 make_images 4572 14.5 0.354 0.385 2.178 2.332 mp_alltoall_i22 627 13.8 1.485 2.280 1.485 2.280 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.448 2.238 calculate_first_density_matrix 1 7.0 0.000 0.000 2.201 2.201 calculate_dm_sparse 119 9.5 0.000 0.000 2.174 2.190 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.244 2.160 pw_transfer 1439 11.6 0.066 0.067 2.068 2.070 grid_integrate_task_list 119 12.3 2.017 2.032 2.017 2.032 ot_diis_step 108 11.5 0.014 0.015 1.986 1.986 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.973 1.975 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.958 1.959 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.908 density_rs2pw 119 9.7 0.003 0.003 1.795 1.817 jit_kernel_multiply 10 15.4 0.996 1.777 0.996 1.777 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.761 1.762 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.702 1.751 mp_sum_l 7207 12.9 1.050 1.708 1.050 1.708 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.683 1.683 fft_wrap_pw1pw2_140 487 13.2 0.087 0.088 1.638 1.641 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.590 1.612 fft3d_ps 1201 14.6 0.567 0.577 1.607 1.609 grid_collocate_task_list 119 9.7 1.454 1.472 1.454 1.472 cp_fm_cholesky_invert 11 10.9 1.436 1.439 1.436 1.439 rs_pw_transfer 974 11.9 0.009 0.009 1.392 1.421 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.378 1.379 cp_fm_diag_elpa_base 50 14.0 1.228 1.285 1.377 1.377 wfi_extrapolate 11 7.9 0.001 0.001 1.357 1.357 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.123 1.346 make_images_data 4572 15.5 0.043 0.046 1.062 1.278 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.249 potential_pw2rs 119 12.3 0.014 0.015 1.231 1.234 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.136 1.153 mp_alltoall_d11v 2130 13.8 1.136 1.148 1.136 1.148 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.101 1.121 multiply_cannon_sync_h2d 9144 15.5 1.040 1.045 1.040 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.034 qs_create_task_list 11 7.9 0.004 0.024 0.962 0.974 generate_qs_task_list 11 8.9 0.372 0.392 0.958 0.973 mp_alltoall_z22v 1201 16.6 0.904 0.922 0.904 0.922 acc_transpose_blocks 9144 15.5 0.039 0.039 0.896 0.907 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.865 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.652000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=692.545455, yerr=18.102554 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.105216E+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 1803893. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 81.918 81.919 qs_mol_dyn_low 1 2.0 0.003 0.004 81.687 81.696 qs_forces 11 3.9 0.002 0.002 81.620 81.621 qs_energies 11 4.9 0.002 0.003 78.805 78.823 scf_env_do_scf 11 5.9 0.000 0.001 70.031 70.034 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.485 64.485 dbcsr_multiply_generic 2055 12.4 0.106 0.108 51.896 52.217 qs_scf_new_mos 99 7.5 0.000 0.001 47.559 47.713 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.558 47.712 ot_scf_mini 99 9.5 0.002 0.002 45.143 45.279 multiply_cannon 2055 13.4 0.175 0.180 42.508 43.702 multiply_cannon_loop 2055 14.4 1.555 1.595 41.540 42.724 velocity_verlet 10 3.0 0.001 0.002 41.280 41.280 ot_mini 99 10.5 0.001 0.001 27.277 27.378 qs_ot_get_derivative 99 11.5 0.001 0.001 20.435 20.540 multiply_cannon_multrec 49320 15.4 12.226 12.967 17.219 17.995 rebuild_ks_matrix 110 8.3 0.000 0.001 14.443 14.566 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.442 14.566 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.702 12.807 mp_waitall_1 241148 16.1 11.608 12.703 11.608 12.703 multiply_cannon_sync_h2d 49320 15.4 10.099 10.790 10.099 10.790 qs_ot_get_p 110 10.4 0.001 0.001 9.292 9.405 multiply_cannon_metrocomm3 49320 15.4 0.077 0.081 6.874 8.193 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.176 7.719 apply_single 110 13.6 0.000 0.001 7.176 7.718 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.579 7.087 sum_up_and_integrate 110 10.3 0.037 0.043 7.006 7.020 integrate_v_rspace 110 11.3 0.003 0.003 6.969 6.992 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.919 6.976 init_scf_run 11 5.9 0.000 0.001 6.758 6.759 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.758 6.759 ot_diis_step 99 11.5 0.006 0.006 6.573 6.573 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.163 6.296 calculate_rho_elec 110 8.6 0.020 0.024 6.163 6.296 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.085 6.109 mp_sum_l 6514 12.8 4.969 5.844 4.969 5.844 init_scf_loop 11 6.9 0.000 0.000 5.519 5.519 dbcsr_mm_accdrv_process 87628 16.1 2.047 2.168 4.872 5.177 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.044 5.044 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.556 4.557 cp_fm_redistribute_end 48 14.0 3.967 4.530 3.970 4.530 cp_fm_diag_elpa_base 48 14.0 0.553 4.432 0.556 4.457 make_m2s 4110 13.4 0.060 0.066 3.954 4.043 wfi_extrapolate 11 7.9 0.001 0.001 4.038 4.038 make_images 4110 14.4 0.178 0.193 3.859 3.952 calculate_dm_sparse 110 9.5 0.000 0.001 3.804 3.942 rs_pw_transfer 902 11.9 0.012 0.014 3.751 3.921 multiply_cannon_metrocomm1 49320 15.4 0.061 0.065 2.571 3.704 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.581 3.585 density_rs2pw 110 9.6 0.004 0.004 3.277 3.473 prepare_preconditioner 11 7.9 0.000 0.000 3.375 3.400 make_preconditioner 11 8.9 0.000 0.000 3.375 3.400 grid_integrate_task_list 110 12.3 3.243 3.394 3.243 3.394 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.280 3.342 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.227 3.288 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.144 3.199 pw_transfer 1331 11.6 0.053 0.065 2.867 2.930 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.780 2.846 calculate_first_density_matrix 1 7.0 0.000 0.000 2.632 2.641 potential_pw2rs 110 12.3 0.006 0.007 2.599 2.623 jit_kernel_multiply 13 15.9 2.547 2.572 2.547 2.572 mp_alltoall_d11v 2046 13.8 2.108 2.507 2.108 2.507 fft_wrap_pw1pw2_140 451 13.1 0.168 0.188 2.329 2.395 fft3d_ps 1111 14.6 0.748 0.831 2.319 2.372 acc_transpose_blocks 49320 15.4 0.209 0.222 2.140 2.227 grid_collocate_task_list 110 9.6 2.088 2.206 2.088 2.206 mp_sum_d 3883 11.9 1.437 1.999 1.437 1.999 mp_waitany 14300 13.8 1.807 1.985 1.807 1.985 make_images_data 4110 15.4 0.043 0.046 1.802 1.933 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.880 1.907 hybrid_alltoall_any 4261 16.3 0.081 0.472 1.563 1.840 cp_fm_cholesky_invert 11 10.9 1.804 1.807 1.804 1.807 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.667 1.708 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.919000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.909091, yerr=1.676281 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.306112E+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 2358931. 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 0.019 0.037 68.288 68.290 qs_mol_dyn_low 1 2.0 0.003 0.004 67.982 67.991 qs_forces 11 3.9 0.002 0.004 67.915 67.915 qs_energies 11 4.9 0.001 0.003 64.603 64.606 scf_env_do_scf 11 5.9 0.000 0.001 56.035 56.038 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.391 48.391 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.695 37.880 velocity_verlet 10 3.0 0.001 0.002 35.868 35.869 qs_scf_new_mos 99 7.5 0.001 0.001 32.611 32.756 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.610 32.756 multiply_cannon 2055 13.4 0.220 0.241 31.100 32.441 ot_scf_mini 99 9.5 0.003 0.003 30.955 31.083 multiply_cannon_loop 2055 14.4 0.933 0.952 29.799 30.798 ot_mini 99 10.5 0.001 0.001 18.179 18.318 multiply_cannon_multrec 24660 15.4 7.633 9.253 13.931 15.650 rebuild_ks_matrix 110 8.3 0.000 0.000 13.579 13.693 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.578 13.692 qs_ot_get_derivative 99 11.5 0.001 0.001 12.377 12.507 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.972 12.069 mp_waitall_1 186928 16.3 7.774 10.293 7.774 10.293 multiply_cannon_sync_h2d 24660 15.4 7.037 8.090 7.037 8.090 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.115 8.058 init_scf_loop 11 6.9 0.000 0.000 7.609 7.609 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.515 7.256 apply_single 110 13.6 0.000 0.001 6.515 7.255 sum_up_and_integrate 110 10.3 0.052 0.060 6.501 6.514 integrate_v_rspace 110 11.3 0.002 0.003 6.448 6.463 dbcsr_mm_accdrv_process 52282 16.1 4.745 5.523 6.137 6.439 qs_ot_get_p 110 10.4 0.001 0.001 6.149 6.304 init_scf_run 11 5.9 0.000 0.001 6.177 6.177 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.176 6.177 ot_diis_step 99 11.5 0.010 0.010 5.756 5.756 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.736 5.744 calculate_rho_elec 110 8.6 0.039 0.047 5.735 5.743 prepare_preconditioner 11 7.9 0.000 0.000 5.594 5.612 make_preconditioner 11 8.9 0.000 0.000 5.594 5.612 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.155 5.314 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.626 5.312 make_m2s 4110 13.4 0.056 0.059 4.186 4.678 make_images 4110 14.4 0.397 0.440 4.078 4.567 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.270 4.290 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.839 3.846 wfi_extrapolate 11 7.9 0.001 0.001 3.503 3.503 pw_transfer 1331 11.6 0.065 0.069 3.285 3.423 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.179 3.315 grid_integrate_task_list 110 12.3 3.161 3.290 3.161 3.290 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.270 3.272 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.185 3.253 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.203 3.205 density_rs2pw 110 9.6 0.004 0.004 3.028 3.189 cp_fm_redistribute_end 48 14.0 2.388 3.176 2.390 3.177 cp_fm_diag_elpa_base 48 14.0 0.753 3.054 0.784 3.140 rs_pw_transfer 902 11.9 0.012 0.013 2.920 3.087 calculate_dm_sparse 110 9.5 0.001 0.001 3.048 3.074 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.926 2.959 make_images_data 4110 15.4 0.047 0.051 2.341 2.842 fft_wrap_pw1pw2_140 451 13.1 0.199 0.218 2.665 2.803 hybrid_alltoall_any 4261 16.3 0.102 0.437 2.036 2.771 fft3d_ps 1111 14.6 1.074 1.269 2.537 2.658 cp_fm_cholesky_invert 11 10.9 2.614 2.621 2.614 2.621 calculate_first_density_matrix 1 7.0 0.000 0.002 2.588 2.591 mp_sum_l 6514 12.8 1.783 2.384 1.783 2.384 potential_pw2rs 110 12.3 0.008 0.009 2.356 2.370 grid_collocate_task_list 110 9.6 2.123 2.262 2.123 2.262 jit_kernel_multiply 10 16.3 1.041 2.022 1.041 2.022 mp_alltoall_d11v 2046 13.8 1.768 2.015 1.768 2.015 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.946 1.966 qs_energies_init_hamiltonians 11 5.9 0.007 0.021 1.882 1.883 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.770 1.784 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.769 1.652 mp_allgather_i34 2055 14.4 0.631 1.645 0.631 1.645 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.626 1.635 acc_transpose_blocks 24660 15.4 0.108 0.111 1.564 1.595 mp_irecv_dv 57340 16.2 0.645 1.535 0.645 1.535 dbcsr_complete_redistribute 325 12.2 0.240 0.289 1.238 1.507 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.374 1.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.290000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.454545, yerr=6.998229 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 661.823488E+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 0.023 0.041 59.684 59.686 qs_mol_dyn_low 1 2.0 0.004 0.005 59.286 59.295 qs_forces 11 3.9 0.003 0.005 59.219 59.219 qs_energies 11 4.9 0.002 0.005 56.035 56.039 scf_env_do_scf 11 5.9 0.001 0.001 48.201 48.202 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.503 39.504 velocity_verlet 10 3.0 0.002 0.002 32.421 32.423 dbcsr_multiply_generic 2055 12.4 0.108 0.112 28.413 28.653 qs_scf_new_mos 99 7.5 0.001 0.001 25.175 25.255 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.174 25.254 ot_scf_mini 99 9.5 0.003 0.003 23.946 24.043 multiply_cannon 2055 13.4 0.211 0.219 22.069 23.219 multiply_cannon_loop 2055 14.4 0.621 0.636 20.803 21.930 ot_mini 99 10.5 0.001 0.001 13.627 13.732 rebuild_ks_matrix 110 8.3 0.000 0.000 12.075 12.204 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.075 12.204 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.649 10.764 multiply_cannon_multrec 16440 15.4 3.968 5.048 9.732 10.685 mp_waitall_1 146766 16.3 7.065 10.422 7.065 10.422 qs_ot_get_derivative 99 11.5 0.001 0.001 9.125 9.224 init_scf_loop 11 6.9 0.001 0.002 8.662 8.666 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.175 7.282 prepare_preconditioner 11 7.9 0.000 0.000 6.893 6.913 make_preconditioner 11 8.9 0.000 0.001 6.893 6.913 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.195 6.569 sum_up_and_integrate 110 10.3 0.060 0.061 6.356 6.374 integrate_v_rspace 110 11.3 0.003 0.003 6.296 6.314 dbcsr_mm_accdrv_process 34862 16.1 4.722 5.291 5.618 5.814 qs_ot_get_p 110 10.4 0.001 0.002 5.518 5.648 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.445 5.455 calculate_rho_elec 110 8.6 0.058 0.058 5.444 5.454 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.011 5.449 apply_single 110 13.6 0.000 0.000 5.010 5.449 init_scf_run 11 5.9 0.000 0.001 5.419 5.420 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.419 5.420 make_m2s 4110 13.4 0.050 0.051 4.223 4.650 make_images 4110 14.4 0.389 0.507 4.108 4.536 ot_diis_step 99 11.5 0.011 0.011 4.473 4.473 multiply_cannon_sync_h2d 16440 15.4 3.681 4.337 3.681 4.337 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.869 3.873 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.118 3.728 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.520 3.520 grid_integrate_task_list 110 12.3 3.178 3.354 3.178 3.354 pw_transfer 1331 11.6 0.064 0.070 3.051 3.060 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.964 2.965 make_images_data 4110 15.4 0.044 0.048 2.445 2.959 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.945 2.955 wfi_extrapolate 11 7.9 0.001 0.001 2.950 2.950 cp_fm_redistribute_end 48 14.0 1.843 2.936 1.845 2.937 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.165 2.930 density_rs2pw 110 9.6 0.004 0.004 2.743 2.917 cp_fm_diag_elpa_base 48 14.0 1.030 2.798 1.088 2.904 rs_pw_transfer 902 11.9 0.010 0.012 2.703 2.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.773 2.774 cp_fm_cholesky_invert 11 10.9 2.727 2.733 2.727 2.733 calculate_dm_sparse 110 9.5 0.001 0.001 2.501 2.529 fft_wrap_pw1pw2_140 451 13.1 0.210 0.212 2.496 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.479 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.374 2.423 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.859 2.416 calculate_first_density_matrix 1 7.0 0.001 0.002 2.385 2.387 grid_collocate_task_list 110 9.6 2.179 2.344 2.179 2.344 mp_irecv_dv 48980 15.7 0.789 2.291 0.789 2.291 fft3d_ps 1111 14.6 1.061 1.075 2.259 2.272 potential_pw2rs 110 12.3 0.011 0.011 2.220 2.233 mp_sum_l 6514 12.8 1.511 2.096 1.511 2.096 mp_alltoall_d11v 2046 13.8 1.761 2.075 1.761 2.075 dbcsr_complete_redistribute 325 12.2 0.324 0.360 1.505 1.969 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.966 1.967 cp_fm_upper_to_full 70 13.6 1.374 1.835 1.374 1.835 mp_allgather_i34 2055 14.4 0.590 1.696 0.590 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.626 1.641 cp_fm_cholesky_decompose 22 10.9 1.620 1.639 1.620 1.639 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.361 1.501 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.023 1.474 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.454 1.465 mp_waitany 17072 13.8 1.162 1.313 1.162 1.313 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.277 1.284 acc_transpose_blocks 16440 15.4 0.073 0.075 1.225 1.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.686000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.000000, yerr=10.126473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 723.382272E+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 0.020 0.037 65.093 65.094 qs_mol_dyn_low 1 2.0 0.003 0.004 64.770 64.779 qs_forces 11 3.9 0.002 0.004 64.600 64.601 qs_energies 11 4.9 0.003 0.017 61.220 61.224 scf_env_do_scf 11 5.9 0.001 0.001 52.809 52.812 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.866 40.867 velocity_verlet 10 3.0 0.001 0.002 37.099 37.105 dbcsr_multiply_generic 2055 12.4 0.116 0.118 29.789 30.042 qs_scf_new_mos 99 7.5 0.001 0.001 26.357 26.451 qs_scf_loop_do_ot 99 8.5 0.001 0.002 26.356 26.451 ot_scf_mini 99 9.5 0.003 0.003 24.735 24.825 multiply_cannon 2055 13.4 0.245 0.311 22.688 24.069 multiply_cannon_loop 2055 14.4 0.884 0.904 21.253 21.882 ot_mini 99 10.5 0.001 0.001 14.278 14.392 multiply_cannon_multrec 24660 15.4 4.207 6.677 12.793 13.991 rebuild_ks_matrix 110 8.3 0.000 0.000 12.039 12.153 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 12.038 12.152 init_scf_loop 11 6.9 0.001 0.004 11.901 11.902 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.647 10.749 qs_ot_get_derivative 99 11.5 0.001 0.001 10.134 10.228 prepare_preconditioner 11 7.9 0.000 0.000 10.128 10.147 make_preconditioner 11 8.9 0.000 0.001 10.128 10.147 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.297 9.812 dbcsr_mm_accdrv_process 52304 16.0 6.875 8.287 8.441 9.412 mp_waitall_1 126806 16.4 4.706 6.750 4.706 6.750 sum_up_and_integrate 110 10.3 0.068 0.071 6.379 6.393 integrate_v_rspace 110 11.3 0.003 0.003 6.311 6.325 make_m2s 4110 13.4 0.059 0.062 5.523 5.881 make_images 4110 14.4 0.577 0.695 5.382 5.738 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.681 5.690 calculate_rho_elec 110 8.6 0.077 0.081 5.680 5.689 init_scf_run 11 5.9 0.000 0.001 5.637 5.638 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.637 5.638 qs_ot_get_p 110 10.4 0.001 0.001 5.408 5.554 cp_fm_upper_to_full 70 13.8 3.376 4.813 3.376 4.813 ot_diis_step 99 11.5 0.011 0.011 4.108 4.108 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.012 4.082 apply_single 110 13.6 0.000 0.000 4.012 4.081 dbcsr_complete_redistribute 325 12.2 0.420 0.466 2.737 3.913 qs_ot_p2m_diag 48 11.0 0.054 0.062 3.604 3.618 grid_integrate_task_list 110 12.3 3.271 3.397 3.271 3.397 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.195 3.359 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.314 3.356 pw_transfer 1331 11.6 0.064 0.072 3.309 3.331 multiply_cannon_sync_h2d 24660 15.4 3.161 3.313 3.161 3.313 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.544 3.238 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.204 3.229 make_images_data 4110 15.4 0.046 0.051 2.834 3.224 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.342 3.204 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.131 3.131 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.003 3.059 wfi_extrapolate 11 7.9 0.001 0.001 2.999 2.999 calculate_dm_sparse 110 9.5 0.001 0.001 2.961 2.992 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.823 2.972 density_rs2pw 110 9.6 0.004 0.004 2.823 2.966 mp_alltoall_i22 605 13.7 1.703 2.910 1.703 2.910 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.885 2.887 cp_fm_cholesky_invert 11 10.9 2.768 2.776 2.768 2.776 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.701 2.729 rs_pw_transfer 902 11.9 0.010 0.011 2.477 2.670 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.585 2.587 cp_fm_redistribute_end 48 14.0 1.289 2.559 1.291 2.560 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.511 2.551 calculate_first_density_matrix 1 7.0 0.000 0.004 2.541 2.543 cp_fm_diag_elpa_base 48 14.0 1.186 2.424 1.266 2.536 fft3d_ps 1111 14.6 1.065 1.099 2.506 2.525 grid_collocate_task_list 110 9.6 2.222 2.407 2.222 2.407 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.256 2.258 jit_kernel_multiply 11 15.6 1.237 2.148 1.237 2.148 potential_pw2rs 110 12.3 0.012 0.013 2.131 2.140 mp_alltoall_d11v 2046 13.8 1.851 2.125 1.851 2.125 cp_fm_cholesky_decompose 22 10.9 1.699 1.760 1.699 1.760 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.699 1.730 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.603 1.701 mp_allgather_i34 2055 14.4 0.535 1.690 0.535 1.690 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.639 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.576 1.593 acc_transpose_blocks 24660 15.4 0.106 0.107 1.544 1.570 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.861 1.551 mp_irecv_dv 62702 16.1 0.761 1.472 0.761 1.472 mp_sum_l 6514 12.8 0.892 1.464 0.892 1.464 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.210 1.312 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.094000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.636364, yerr=8.293182 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 827.154432E+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 0.016 0.032 54.171 54.171 qs_mol_dyn_low 1 2.0 0.003 0.003 53.936 53.945 qs_forces 11 3.9 0.002 0.002 53.871 53.871 qs_energies 11 4.9 0.001 0.002 50.258 50.263 scf_env_do_scf 11 5.9 0.000 0.001 42.142 42.142 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.413 34.413 velocity_verlet 10 3.0 0.002 0.002 30.467 30.470 dbcsr_multiply_generic 2055 12.4 0.104 0.105 22.620 22.755 qs_scf_new_mos 99 7.5 0.001 0.001 20.388 20.432 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.388 20.431 ot_scf_mini 99 9.5 0.002 0.002 19.153 19.174 multiply_cannon 2055 13.4 0.248 0.263 17.169 18.535 multiply_cannon_loop 2055 14.4 0.323 0.335 15.807 16.107 rebuild_ks_matrix 110 8.3 0.000 0.000 11.422 11.442 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.422 11.441 ot_mini 99 10.5 0.001 0.001 10.259 10.273 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.173 10.191 multiply_cannon_multrec 8220 15.4 3.200 4.490 7.409 8.278 mp_waitall_1 106626 16.5 6.075 7.746 6.075 7.746 init_scf_loop 11 6.9 0.000 0.000 7.682 7.683 qs_ot_get_derivative 99 11.5 0.001 0.001 6.469 6.488 sum_up_and_integrate 110 10.3 0.079 0.081 6.172 6.185 integrate_v_rspace 110 11.3 0.003 0.003 6.092 6.105 prepare_preconditioner 11 7.9 0.000 0.000 6.059 6.062 make_preconditioner 11 8.9 0.000 0.000 6.059 6.062 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.636 5.709 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.561 5.574 calculate_rho_elec 110 8.6 0.115 0.115 5.560 5.574 init_scf_run 11 5.9 0.000 0.001 5.008 5.008 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.008 5.008 qs_ot_get_p 110 10.4 0.001 0.001 4.974 4.997 dbcsr_mm_accdrv_process 17442 15.9 2.737 3.939 4.080 4.967 make_m2s 4110 13.4 0.038 0.040 4.218 4.509 make_images 4110 14.4 0.636 0.690 4.089 4.379 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.951 4.160 ot_diis_step 99 11.5 0.012 0.012 3.770 3.770 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.698 3.711 apply_single 110 13.6 0.000 0.000 3.698 3.711 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.640 3.643 grid_integrate_task_list 110 12.3 3.384 3.508 3.384 3.508 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.336 3.337 pw_transfer 1331 11.6 0.065 0.069 3.218 3.226 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.111 3.124 multiply_cannon_sync_h2d 8220 15.4 2.923 2.979 2.923 2.979 cp_fm_cholesky_invert 11 10.9 2.899 2.903 2.899 2.903 make_images_data 4110 15.4 0.038 0.044 2.414 2.836 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.815 2.815 hybrid_alltoall_any 4261 16.3 0.200 0.869 2.327 2.793 cp_fm_redistribute_end 48 14.0 0.716 2.786 0.721 2.787 density_rs2pw 110 9.6 0.004 0.004 2.624 2.771 cp_fm_diag_elpa_base 48 14.0 1.882 2.597 2.058 2.750 wfi_extrapolate 11 7.9 0.001 0.001 2.684 2.684 fft_wrap_pw1pw2_140 451 13.1 0.213 0.215 2.660 2.675 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.655 2.656 grid_collocate_task_list 110 9.6 2.329 2.514 2.329 2.514 calculate_dm_sparse 110 9.5 0.001 0.001 2.455 2.494 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.491 2.493 fft3d_ps 1111 14.6 1.118 1.150 2.369 2.381 rs_pw_transfer 902 11.9 0.010 0.010 2.163 2.319 calculate_first_density_matrix 1 7.0 0.000 0.000 2.239 2.240 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.124 2.137 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.775 1.997 potential_pw2rs 110 12.3 0.015 0.016 1.918 1.922 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.870 1.883 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.790 1.800 cp_fm_cholesky_decompose 22 10.9 1.720 1.729 1.720 1.729 mp_alltoall_d11v 2046 13.8 1.591 1.712 1.591 1.712 mp_allgather_i34 2055 14.4 0.547 1.648 0.547 1.648 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.497 1.619 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.583 1.586 jit_kernel_multiply 9 15.8 1.035 1.581 1.035 1.581 dbcsr_complete_redistribute 325 12.2 0.566 0.581 1.459 1.554 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.424 1.438 qs_create_task_list 11 7.9 0.000 0.001 1.221 1.321 generate_qs_task_list 11 8.9 0.378 0.446 1.220 1.321 mp_waitany 9240 13.8 1.057 1.207 1.057 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.130 1.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.171000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=776.454545, yerr=10.066720 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 1.371234E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 85.796 85.796 qs_mol_dyn_low 1 2.0 0.003 0.003 85.496 85.504 qs_forces 11 3.9 0.002 0.002 85.113 85.114 qs_energies 11 4.9 0.002 0.002 81.082 81.084 scf_env_do_scf 11 5.9 0.000 0.001 71.284 71.284 velocity_verlet 10 3.0 0.002 0.002 55.099 55.144 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.772 42.774 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.444 28.502 init_scf_loop 11 6.9 0.000 0.000 28.434 28.437 prepare_preconditioner 11 7.9 0.000 0.000 26.508 26.514 make_preconditioner 11 8.9 0.000 0.000 26.508 26.514 qs_scf_new_mos 99 7.5 0.001 0.001 26.423 26.468 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.422 26.467 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.723 25.981 ot_scf_mini 99 9.5 0.002 0.002 24.654 24.693 multiply_cannon 2055 13.4 0.351 0.383 21.417 22.375 multiply_cannon_loop 2055 14.4 0.342 0.346 19.597 20.024 cp_fm_upper_to_full 70 14.2 12.746 18.251 12.746 18.251 ot_mini 99 10.5 0.001 0.001 13.511 13.553 rebuild_ks_matrix 110 8.3 0.000 0.000 13.093 13.141 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.092 13.140 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.914 11.957 dbcsr_complete_redistribute 325 12.2 1.022 1.052 7.380 10.714 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.349 9.693 multiply_cannon_multrec 8220 15.4 4.360 4.542 9.559 9.651 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.770 9.077 qs_ot_get_derivative 99 11.5 0.001 0.001 8.975 9.013 mp_alltoall_i22 605 13.7 5.393 8.749 5.393 8.749 mp_waitall_1 87304 16.6 7.723 8.574 7.723 8.574 sum_up_and_integrate 110 10.3 0.151 0.153 6.612 6.624 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.447 6.481 calculate_rho_elec 110 8.6 0.227 0.227 6.446 6.480 integrate_v_rspace 110 11.3 0.004 0.004 6.461 6.472 qs_ot_get_p 110 10.4 0.001 0.001 5.930 5.983 make_m2s 4110 13.4 0.043 0.043 5.268 5.778 make_images 4110 14.4 0.881 0.936 5.081 5.590 init_scf_run 11 5.9 0.000 0.001 5.575 5.575 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.575 5.575 cp_fm_cholesky_invert 11 10.9 5.405 5.409 5.405 5.409 dbcsr_mm_accdrv_process 11614 15.7 3.287 3.633 5.055 5.310 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.695 5.163 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.644 5.083 apply_single 110 13.6 0.000 0.000 4.644 5.083 ot_diis_step 99 11.5 0.015 0.015 4.517 4.517 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.344 4.352 multiply_cannon_sync_h2d 8220 15.4 3.949 3.954 3.949 3.954 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.895 3.895 grid_integrate_task_list 110 12.3 3.667 3.726 3.667 3.726 pw_transfer 1331 11.6 0.074 0.074 3.687 3.695 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.620 3.621 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.571 3.578 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.865 3.574 make_images_data 4110 15.4 0.042 0.046 2.859 3.505 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.870 3.307 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.272 3.273 cp_fm_diag_elpa_base 48 14.0 2.737 2.936 3.271 3.271 wfi_extrapolate 11 7.9 0.001 0.001 3.236 3.236 calculate_dm_sparse 110 9.5 0.001 0.001 3.129 3.149 fft_wrap_pw1pw2_140 451 13.1 0.214 0.216 3.082 3.093 density_rs2pw 110 9.6 0.004 0.004 2.951 2.970 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.872 2.877 fft3d_ps 1111 14.6 1.265 1.274 2.780 2.785 grid_collocate_task_list 110 9.6 2.629 2.657 2.629 2.657 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.396 2.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.361 2.378 calculate_first_density_matrix 1 7.0 0.000 0.000 2.236 2.237 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.163 2.208 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.198 rs_pw_transfer 902 11.9 0.010 0.011 2.144 2.181 cp_fm_cholesky_decompose 22 10.9 2.067 2.082 2.067 2.082 mp_alltoall_d11v 2046 13.8 2.034 2.078 2.034 2.078 potential_pw2rs 110 12.3 0.021 0.021 2.052 2.055 qs_create_task_list 11 7.9 0.001 0.001 1.900 1.941 generate_qs_task_list 11 8.9 0.737 0.792 1.899 1.940 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.775 1.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.786 1.791 jit_kernel_multiply 10 15.3 1.568 1.732 1.568 1.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.796000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1213.363636, yerr=66.722318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.239808E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2309115. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.046 201.773 201.775 qs_mol_dyn_low 1 2.0 0.004 0.006 201.275 201.288 qs_forces 11 3.9 0.003 0.004 201.187 201.188 qs_energies 11 4.9 0.002 0.003 195.685 195.703 scf_env_do_scf 11 5.9 0.001 0.001 179.382 179.386 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.006 159.008 dbcsr_multiply_generic 2507 12.6 0.176 0.179 123.542 124.193 velocity_verlet 10 3.0 0.002 0.002 121.913 121.914 qs_scf_new_mos 117 7.6 0.001 0.001 120.237 120.570 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.236 120.569 ot_scf_mini 117 9.6 0.003 0.003 113.681 113.996 multiply_cannon 2507 13.6 0.238 0.246 100.539 102.358 multiply_cannon_loop 2507 14.6 2.066 2.144 98.387 100.106 ot_mini 117 10.6 0.001 0.001 65.448 65.780 multiply_cannon_multrec 60168 15.6 33.305 35.885 41.516 43.494 qs_ot_get_derivative 117 11.6 0.001 0.001 40.600 40.890 rebuild_ks_matrix 128 8.3 0.001 0.001 33.255 33.605 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 33.254 33.604 mp_waitall_1 291448 16.2 28.391 31.623 28.391 31.623 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.926 30.251 multiply_cannon_sync_h2d 60168 15.6 27.611 29.728 27.611 29.728 qs_ot_get_p 128 10.4 0.001 0.001 26.494 26.808 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.256 25.038 apply_single 128 13.6 0.001 0.001 24.256 25.038 ot_diis_step 117 11.6 0.007 0.008 24.477 24.478 init_scf_loop 11 6.9 0.000 0.000 20.297 20.298 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.824 19.885 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.664 18.851 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.377 17.378 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 15.470 17.290 prepare_preconditioner 11 7.9 0.000 0.000 15.668 15.702 make_preconditioner 11 8.9 0.000 0.000 15.668 15.702 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.919 15.100 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.505 14.511 cp_fm_redistribute_end 83 14.4 11.455 14.433 11.465 14.435 make_m2s 5014 13.6 0.103 0.112 13.877 14.347 cp_fm_diag_elpa_base 83 14.4 2.927 14.145 2.958 14.261 make_images 5014 14.6 0.403 0.422 13.697 14.177 sum_up_and_integrate 128 10.3 0.090 0.109 13.979 13.993 integrate_v_rspace 128 11.3 0.004 0.004 13.888 13.906 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.693 12.771 calculate_rho_elec 128 8.7 0.045 0.064 12.693 12.771 init_scf_run 11 5.9 0.000 0.001 12.204 12.204 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.204 12.204 mp_sum_l 7870 13.0 7.870 9.086 7.870 9.086 wfi_extrapolate 11 7.9 0.001 0.001 8.982 8.983 cp_fm_cholesky_invert 11 10.9 8.869 8.877 8.869 8.877 calculate_dm_sparse 128 9.5 0.001 0.001 8.361 8.460 dbcsr_mm_accdrv_process 124484 16.2 3.175 3.413 7.769 8.321 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.033 8.219 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.722 7.827 multiply_cannon_metrocomm1 60168 15.6 0.088 0.093 5.905 7.820 make_images_data 5014 15.6 0.069 0.079 6.797 7.726 grid_integrate_task_list 128 12.3 7.094 7.431 7.094 7.431 hybrid_alltoall_any 5200 16.5 0.289 2.252 5.962 7.154 pw_transfer 1547 11.6 0.075 0.107 6.846 7.123 density_rs2pw 128 9.7 0.006 0.007 6.482 7.024 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.642 6.886 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.667 6.674 rs_pw_transfer 1046 11.9 0.017 0.019 5.663 6.265 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.818 5.897 fft_wrap_pw1pw2_140 523 13.2 0.444 0.509 5.700 5.885 mp_alltoall_d11v 2415 14.1 4.230 5.740 4.230 5.740 fft3d_ps 1291 14.7 2.095 2.573 5.433 5.630 grid_collocate_task_list 128 9.7 4.726 5.089 4.726 5.089 cp_fm_cholesky_decompose 22 10.9 4.615 4.632 4.615 4.632 potential_pw2rs 128 12.3 0.010 0.011 4.451 4.481 mp_sum_d 4464 12.1 3.610 4.274 3.610 4.274 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=201.775000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.181818, yerr=6.766471 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 829.100032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 4953949. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 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 0.026 0.046 190.584 190.588 qs_mol_dyn_low 1 2.0 0.004 0.004 190.146 190.159 qs_forces 11 3.9 0.003 0.004 190.066 190.068 qs_energies 11 4.9 0.003 0.009 183.360 183.370 scf_env_do_scf 11 5.9 0.001 0.001 167.119 167.130 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 134.006 134.009 velocity_verlet 10 3.0 0.002 0.002 120.220 120.221 dbcsr_multiply_generic 2527 12.6 0.186 0.192 97.963 99.106 qs_scf_new_mos 118 7.6 0.001 0.001 95.247 95.657 qs_scf_loop_do_ot 118 8.6 0.001 0.002 95.246 95.656 ot_scf_mini 118 9.6 0.004 0.005 90.413 90.922 multiply_cannon 2527 13.6 0.479 0.528 78.040 81.718 multiply_cannon_loop 2527 14.6 1.272 1.310 74.700 77.114 ot_mini 118 10.6 0.001 0.001 49.870 50.340 mp_waitall_1 228564 16.4 24.595 38.338 24.595 38.338 multiply_cannon_multrec 30324 15.6 22.431 27.262 32.139 37.247 rebuild_ks_matrix 129 8.3 0.001 0.001 32.936 33.339 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.029 32.935 33.338 init_scf_loop 11 6.9 0.002 0.007 33.015 33.017 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.718 30.105 prepare_preconditioner 11 7.9 0.000 0.000 28.624 28.691 make_preconditioner 11 8.9 0.001 0.003 28.624 28.691 qs_ot_get_derivative 118 11.6 0.001 0.002 27.825 28.338 multiply_cannon_metrocomm3 30324 15.6 0.097 0.101 15.374 27.895 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.334 27.843 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.109 23.296 apply_single 129 13.6 0.001 0.001 22.109 23.295 qs_ot_get_p 129 10.4 0.001 0.002 22.068 22.643 multiply_cannon_sync_h2d 30324 15.6 19.743 22.100 19.743 22.100 ot_diis_step 118 11.6 0.014 0.015 21.871 21.873 qs_ot_p2m_diag 83 11.4 0.188 0.218 17.197 17.230 cp_fm_cholesky_invert 11 10.9 16.732 16.745 16.732 16.745 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.051 16.053 make_m2s 5054 13.6 0.089 0.094 14.272 15.880 make_images 5054 14.6 1.158 1.342 14.064 15.673 sum_up_and_integrate 129 10.3 0.117 0.133 14.367 14.394 integrate_v_rspace 129 11.3 0.003 0.004 14.249 14.282 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.007 13.014 cp_fm_redistribute_end 83 14.4 7.580 12.928 7.591 12.929 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.895 12.927 calculate_rho_elec 129 8.7 0.088 0.106 12.894 12.926 cp_fm_diag_elpa_base 83 14.4 5.103 12.458 5.321 12.810 init_scf_run 11 5.9 0.000 0.001 11.472 11.474 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.472 11.473 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.864 11.158 multiply_cannon_metrocomm4 27797 15.6 0.097 0.111 3.693 10.475 make_images_data 5054 15.6 0.068 0.074 8.606 10.426 mp_irecv_dv 70031 16.3 3.498 10.088 3.498 10.088 hybrid_alltoall_any 5240 16.5 0.345 1.517 7.337 9.951 dbcsr_mm_accdrv_process 62734 16.2 4.667 5.612 9.156 9.751 wfi_extrapolate 11 7.9 0.001 0.001 8.327 8.327 pw_transfer 1559 11.6 0.086 0.106 7.570 7.642 grid_integrate_task_list 129 12.3 7.247 7.571 7.247 7.571 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.344 7.413 density_rs2pw 129 9.7 0.006 0.007 6.715 7.132 cp_fm_cholesky_decompose 22 10.9 7.040 7.132 7.040 7.132 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.309 7.102 calculate_dm_sparse 129 9.5 0.001 0.001 6.504 6.664 fft_wrap_pw1pw2_140 527 13.2 0.472 0.514 6.354 6.432 rs_pw_transfer 1054 12.0 0.015 0.016 5.698 6.165 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.150 6.160 fft3d_ps 1301 14.7 2.792 2.944 5.701 5.746 mp_sum_l 7930 13.1 3.862 5.564 3.862 5.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.310 5.445 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.361 5.424 grid_collocate_task_list 129 9.7 4.950 5.331 4.950 5.331 mp_allgather_i34 2527 14.6 1.907 4.984 1.907 4.984 potential_pw2rs 129 12.3 0.015 0.018 4.786 4.809 mp_alltoall_d11v 2423 14.1 3.867 4.331 3.867 4.331 mp_sum_d 4493 12.2 2.644 4.016 2.644 4.016 dbcsr_complete_redistribute 395 12.7 0.773 0.843 2.982 3.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.588000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.363636, yerr=3.471251 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 929.599488E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.049 175.661 175.664 qs_mol_dyn_low 1 2.0 0.004 0.004 175.040 175.054 qs_forces 11 3.9 0.003 0.004 174.939 174.945 qs_energies 11 4.9 0.004 0.016 168.427 168.439 scf_env_do_scf 11 5.9 0.001 0.001 152.825 152.828 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.707 117.709 velocity_verlet 10 3.0 0.001 0.002 111.944 111.946 dbcsr_multiply_generic 2485 12.5 0.181 0.185 81.125 82.361 qs_scf_new_mos 116 7.6 0.001 0.001 81.346 81.687 qs_scf_loop_do_ot 116 8.6 0.001 0.002 81.345 81.686 ot_scf_mini 116 9.6 0.004 0.005 77.246 77.631 multiply_cannon 2485 13.5 0.495 0.520 61.216 66.355 multiply_cannon_loop 2485 14.5 0.856 0.883 58.217 61.187 ot_mini 116 10.6 0.001 0.001 42.328 42.717 init_scf_loop 11 6.9 0.001 0.005 35.004 35.006 mp_waitall_1 176908 16.5 25.309 34.996 25.309 34.996 prepare_preconditioner 11 7.9 0.000 0.000 30.965 31.008 make_preconditioner 11 8.9 0.000 0.003 30.965 31.008 rebuild_ks_matrix 127 8.3 0.001 0.001 29.963 30.387 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.023 29.962 30.386 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.536 29.999 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.021 27.412 multiply_cannon_multrec 19880 15.5 13.465 16.141 22.064 24.760 multiply_cannon_metrocomm3 19880 15.5 0.058 0.062 15.018 24.611 qs_ot_get_derivative 116 11.6 0.001 0.002 22.790 23.171 qs_ot_get_p 127 10.4 0.001 0.001 20.378 20.866 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.588 20.623 apply_single 127 13.6 0.001 0.001 19.588 20.623 ot_diis_step 116 11.6 0.017 0.019 19.432 19.432 qs_ot_p2m_diag 82 11.4 0.262 0.268 16.059 16.064 multiply_cannon_sync_h2d 19880 15.5 14.267 15.867 14.267 15.867 make_m2s 4970 13.5 0.080 0.086 14.882 15.801 make_images 4970 14.5 1.158 1.250 14.651 15.558 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.064 15.065 cp_fm_cholesky_invert 11 10.9 14.455 14.464 14.455 14.464 sum_up_and_integrate 127 10.3 0.133 0.144 14.098 14.125 integrate_v_rspace 127 11.3 0.004 0.005 13.964 13.995 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.996 13.028 calculate_rho_elec 127 8.7 0.131 0.147 12.995 13.028 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.946 11.947 cp_fm_redistribute_end 82 14.4 4.509 11.869 4.525 11.872 cp_fm_diag_elpa_base 82 14.4 6.925 11.295 7.324 11.765 make_images_data 4970 15.5 0.061 0.067 9.262 10.574 init_scf_run 11 5.9 0.001 0.002 10.483 10.484 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.483 10.483 hybrid_alltoall_any 5155 16.4 0.429 1.954 8.075 9.829 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.911 9.210 multiply_cannon_metrocomm4 17395 15.5 0.062 0.070 3.345 8.891 mp_irecv_dv 49801 16.2 3.223 8.634 3.223 8.634 dbcsr_mm_accdrv_process 41158 16.2 4.495 5.253 8.057 8.173 pw_transfer 1535 11.6 0.084 0.104 7.793 7.894 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 7.572 7.680 grid_integrate_task_list 127 12.3 7.251 7.642 7.251 7.642 wfi_extrapolate 11 7.9 0.001 0.001 7.407 7.407 cp_fm_cholesky_decompose 22 10.9 7.362 7.395 7.362 7.395 cp_fm_upper_to_full 104 14.5 5.671 7.301 5.671 7.301 density_rs2pw 127 9.7 0.006 0.006 6.476 6.943 fft_wrap_pw1pw2_140 519 13.2 0.474 0.522 6.630 6.745 dbcsr_complete_redistribute 393 12.7 1.214 1.266 4.763 6.511 fft3d_ps 1281 14.7 2.677 2.876 5.850 5.904 calculate_dm_sparse 127 9.5 0.001 0.001 5.698 5.815 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.525 5.530 rs_pw_transfer 1038 11.9 0.013 0.014 5.060 5.516 grid_collocate_task_list 127 9.7 5.032 5.493 5.032 5.493 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.565 5.317 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.579 5.278 mp_alltoall_d11v 2401 14.1 4.358 4.940 4.358 4.940 mp_sum_l 7804 13.0 3.230 4.922 3.230 4.922 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.550 4.682 mp_allgather_i34 2485 14.5 1.537 4.646 1.537 4.646 potential_pw2rs 127 12.3 0.020 0.022 4.488 4.503 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.409 4.132 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.954 3.983 mp_alltoall_i22 712 14.1 2.044 3.977 2.044 3.977 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.821 3.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.664000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.181818, yerr=8.757523 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.128927E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.042 188.535 188.536 qs_mol_dyn_low 1 2.0 0.003 0.004 188.035 188.048 qs_forces 11 3.9 0.008 0.009 187.883 187.888 qs_energies 11 4.9 0.006 0.020 180.831 180.843 scf_env_do_scf 11 5.9 0.001 0.002 164.092 164.107 velocity_verlet 10 3.0 0.002 0.002 124.695 124.700 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.660 116.661 qs_scf_new_mos 117 7.6 0.001 0.001 81.485 81.818 qs_scf_loop_do_ot 117 8.6 0.001 0.002 81.485 81.818 dbcsr_multiply_generic 2507 12.6 0.190 0.194 79.061 79.847 ot_scf_mini 117 9.6 0.003 0.004 77.030 77.328 multiply_cannon 2507 13.6 0.556 0.591 54.402 57.816 multiply_cannon_loop 2507 14.6 1.182 1.210 50.765 52.684 init_scf_loop 11 6.9 0.001 0.002 47.308 47.309 prepare_preconditioner 11 7.9 0.000 0.000 43.270 43.306 make_preconditioner 11 8.9 0.000 0.002 43.270 43.306 ot_mini 117 10.6 0.001 0.001 42.200 42.484 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.872 41.910 multiply_cannon_multrec 30084 15.6 14.236 19.702 26.238 31.155 rebuild_ks_matrix 128 8.3 0.001 0.001 28.807 29.108 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.044 28.806 29.108 mp_waitall_1 153770 16.5 17.178 27.309 17.178 27.309 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.004 26.282 qs_ot_get_derivative 117 11.6 0.001 0.002 22.697 22.998 qs_ot_get_p 128 10.4 0.001 0.001 21.060 21.387 make_m2s 5014 13.6 0.095 0.099 20.213 21.213 make_images 5014 14.6 1.963 2.274 19.907 20.905 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.847 19.406 apply_single 128 13.6 0.001 0.001 18.846 19.406 ot_diis_step 117 11.6 0.017 0.018 19.384 19.385 qs_ot_p2m_diag 83 11.4 0.342 0.389 16.927 16.979 cp_fm_cholesky_invert 11 10.9 16.830 16.840 16.830 16.840 cp_fm_upper_to_full 105 14.7 11.137 16.466 11.137 16.466 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.640 15.642 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.190 15.370 sum_up_and_integrate 128 10.3 0.141 0.155 13.950 13.978 integrate_v_rspace 128 11.3 0.004 0.004 13.809 13.843 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.976 13.015 calculate_rho_elec 128 8.7 0.175 0.190 12.975 13.015 dbcsr_complete_redistribute 395 12.7 1.512 1.637 8.974 12.791 multiply_cannon_sync_h2d 30084 15.6 11.643 12.704 11.643 12.704 make_images_data 5014 15.6 0.064 0.068 10.806 12.523 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.478 12.481 cp_fm_redistribute_end 83 14.4 2.145 12.394 2.161 12.400 cp_fm_diag_elpa_base 83 14.4 9.614 11.779 10.205 12.258 dbcsr_mm_accdrv_process 62264 16.2 7.350 8.489 11.576 12.108 hybrid_alltoall_any 5200 16.5 0.529 2.223 9.602 11.537 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.608 11.402 init_scf_run 11 5.9 0.000 0.001 10.784 10.785 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.784 10.785 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.376 10.031 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.356 9.581 mp_alltoall_i22 716 14.1 5.623 9.420 5.623 9.420 cp_fm_cholesky_decompose 22 10.9 7.858 7.964 7.858 7.964 grid_integrate_task_list 128 12.3 7.535 7.764 7.535 7.764 pw_transfer 1547 11.6 0.085 0.099 7.606 7.699 wfi_extrapolate 11 7.9 0.001 0.001 7.625 7.625 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.381 7.478 multiply_cannon_metrocomm4 25070 15.6 0.075 0.083 2.745 6.836 fft_wrap_pw1pw2_140 523 13.2 0.481 0.489 6.464 6.647 mp_irecv_dv 76098 16.2 2.600 6.570 2.600 6.570 density_rs2pw 128 9.7 0.005 0.006 6.138 6.500 calculate_dm_sparse 128 9.5 0.001 0.001 6.205 6.309 fft3d_ps 1291 14.7 2.799 2.890 5.607 5.680 grid_collocate_task_list 128 9.7 5.205 5.674 5.205 5.674 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.429 5.479 rs_pw_transfer 1046 11.9 0.013 0.014 4.699 4.998 mp_alltoall_d11v 2415 14.1 4.648 4.961 4.648 4.961 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.400 4.491 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 4.468 4.473 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.360 4.442 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.199 4.262 potential_pw2rs 128 12.3 0.023 0.023 4.225 4.241 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.536000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1069.181818, yerr=14.831283 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.509474E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.032 169.932 169.933 qs_mol_dyn_low 1 2.0 0.003 0.004 169.556 169.568 qs_forces 11 3.9 0.003 0.003 169.465 169.469 qs_energies 11 4.9 0.013 0.024 162.197 162.204 scf_env_do_scf 11 5.9 0.001 0.004 145.199 145.213 velocity_verlet 10 3.0 0.002 0.002 111.495 111.498 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.623 109.625 dbcsr_multiply_generic 2507 12.6 0.181 0.186 74.064 74.654 qs_scf_new_mos 117 7.6 0.001 0.001 74.162 74.274 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.162 74.273 ot_scf_mini 117 9.6 0.004 0.004 69.741 69.831 multiply_cannon 2507 13.6 0.585 0.619 54.302 59.005 multiply_cannon_loop 2507 14.6 0.450 0.459 49.705 50.791 ot_mini 117 10.6 0.001 0.001 39.571 39.650 init_scf_loop 11 6.9 0.000 0.000 35.422 35.424 mp_waitall_1 129618 16.6 26.115 33.547 26.115 33.547 prepare_preconditioner 11 7.9 0.000 0.000 31.584 31.614 make_preconditioner 11 8.9 0.000 0.000 31.584 31.614 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.369 29.656 rebuild_ks_matrix 128 8.3 0.001 0.001 28.433 28.538 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.432 28.538 qs_ks_update_qs_env 128 7.6 0.001 0.001 25.975 26.079 multiply_cannon_multrec 10028 15.6 10.400 14.078 17.706 20.688 ot_diis_step 117 11.6 0.020 0.020 19.904 19.904 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.433 19.835 apply_single 128 13.6 0.001 0.001 19.433 19.834 qs_ot_get_derivative 117 11.6 0.002 0.002 19.598 19.692 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.300 19.410 make_m2s 5014 13.6 0.065 0.072 16.192 18.357 make_images 5014 14.6 2.299 2.650 15.889 18.052 cp_fm_cholesky_invert 11 10.9 18.001 18.007 18.001 18.007 qs_ot_get_p 128 10.4 0.001 0.001 17.276 17.366 sum_up_and_integrate 128 10.3 0.180 0.190 13.969 14.013 integrate_v_rspace 128 11.3 0.004 0.004 13.788 13.842 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.588 13.605 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.072 13.107 calculate_rho_elec 128 8.7 0.258 0.269 13.071 13.107 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.458 12.459 make_images_data 5014 15.6 0.053 0.062 9.924 12.435 multiply_cannon_sync_h2d 10028 15.6 11.641 12.280 11.641 12.280 hybrid_alltoall_any 5200 16.5 0.837 3.780 9.749 12.178 init_scf_run 11 5.9 0.000 0.001 10.266 10.267 scf_env_initial_rho_setup 11 6.9 0.016 0.031 10.266 10.266 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.464 9.467 cp_fm_diag_elpa_base 83 14.4 9.219 9.300 9.460 9.463 grid_integrate_task_list 128 12.3 7.778 8.155 7.778 8.155 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.875 7.948 cp_fm_cholesky_decompose 22 10.9 7.856 7.945 7.856 7.945 pw_transfer 1547 11.6 0.083 0.089 7.696 7.711 dbcsr_mm_accdrv_process 20762 16.1 2.646 3.456 6.936 7.580 mp_allgather_i34 2507 14.6 2.802 7.502 2.802 7.502 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.476 7.493 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.450 7.420 wfi_extrapolate 11 7.9 0.001 0.001 7.266 7.266 fft_wrap_pw1pw2_140 523 13.2 0.501 0.519 6.588 6.612 calculate_dm_sparse 128 9.5 0.001 0.001 6.095 6.196 density_rs2pw 128 9.7 0.005 0.006 5.898 6.122 grid_collocate_task_list 128 9.7 5.494 5.717 5.494 5.717 dbcsr_complete_redistribute 395 12.7 2.114 2.250 5.313 5.708 fft3d_ps 1291 14.7 2.719 2.791 5.640 5.667 mp_alltoall_d11v 2415 14.1 4.770 5.460 4.770 5.460 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.180 5.197 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.192 5.192 rs_pw_transfer 1046 11.9 0.013 0.013 4.014 4.254 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.087 4.124 potential_pw2rs 128 12.3 0.026 0.028 4.003 4.014 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.846 4.003 mp_irecv_dv 28860 15.9 1.808 3.928 1.808 3.928 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.567 3.858 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.496 3.833 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.674 3.706 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.548 3.644 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.423 3.445 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.933000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1417.000000, yerr=51.615713 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_performance_tests/20/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 11.606413E+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 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.925457E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 59070. MP_Allreduce 11015 1514. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.041 286.265 286.266 qs_mol_dyn_low 1 2.0 0.003 0.004 285.750 285.762 qs_forces 11 3.9 0.003 0.003 285.656 285.658 qs_energies 11 4.9 0.002 0.002 277.070 277.080 scf_env_do_scf 11 5.9 0.001 0.002 255.411 255.430 velocity_verlet 10 3.0 0.002 0.002 206.848 206.856 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.376 131.378 init_scf_loop 11 6.9 0.000 0.000 123.775 123.779 prepare_preconditioner 11 7.9 0.000 0.000 119.015 119.045 make_preconditioner 11 8.9 0.000 0.000 119.015 119.045 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.168 116.179 qs_scf_new_mos 116 7.6 0.001 0.001 90.042 90.215 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.041 90.214 ot_scf_mini 116 9.6 0.004 0.004 85.360 85.503 dbcsr_multiply_generic 2485 12.5 0.212 0.218 82.020 82.598 cp_fm_upper_to_full 104 14.8 52.187 74.962 52.187 74.962 multiply_cannon 2485 13.5 0.703 0.759 57.973 58.734 multiply_cannon_loop 2485 14.5 0.470 0.476 54.447 55.825 ot_mini 116 10.6 0.001 0.001 43.742 43.933 dbcsr_complete_redistribute 393 12.7 3.987 4.033 29.478 42.173 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.118 38.856 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.802 36.471 mp_alltoall_i22 712 14.1 21.620 34.384 21.620 34.384 cp_fm_cholesky_invert 11 10.9 33.737 33.744 33.737 33.744 rebuild_ks_matrix 127 8.3 0.001 0.001 33.111 33.205 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.111 33.205 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.857 30.945 mp_waitall_1 104546 16.7 27.637 30.685 27.637 30.685 qs_ot_get_p 127 10.4 0.001 0.001 26.350 26.611 qs_ot_get_derivative 116 11.6 0.002 0.002 23.810 23.955 qs_ot_p2m_diag 82 11.4 0.868 0.876 22.262 22.291 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.550 20.552 make_m2s 4970 13.5 0.075 0.080 19.359 20.145 ot_diis_step 116 11.6 0.022 0.022 19.889 19.890 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.518 19.754 make_images 4970 14.5 3.756 3.999 18.886 19.675 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.088 19.349 apply_single 127 13.6 0.001 0.001 19.087 19.349 multiply_cannon_multrec 9940 15.5 10.443 12.166 17.841 17.950 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.369 17.372 cp_fm_diag_elpa_base 82 14.4 13.074 14.626 17.365 17.368 sum_up_and_integrate 127 10.3 0.318 0.321 15.536 15.624 multiply_cannon_sync_h2d 9940 15.5 15.535 15.572 15.535 15.572 integrate_v_rspace 127 11.3 0.004 0.005 15.217 15.307 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.685 14.719 calculate_rho_elec 127 8.7 0.479 0.480 14.684 14.719 hybrid_alltoall_any 5155 16.4 1.292 3.009 10.263 11.837 init_scf_run 11 5.9 0.000 0.001 11.755 11.756 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.755 11.756 make_images_data 4970 15.5 0.060 0.065 10.192 11.689 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.211 9.317 dbcsr_mm_accdrv_process 20590 16.0 3.874 5.737 7.164 9.013 cp_fm_cholesky_decompose 22 10.9 8.933 8.955 8.933 8.955 wfi_extrapolate 11 7.9 0.001 0.001 8.661 8.661 grid_integrate_task_list 127 12.3 8.490 8.644 8.490 8.644 pw_transfer 1535 11.6 0.090 0.090 8.073 8.080 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 7.922 7.923 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 7.841 7.847 mp_alltoall_d11v 2401 14.1 6.840 6.968 6.840 6.968 fft_wrap_pw1pw2_140 519 13.2 0.535 0.536 6.948 6.956 calculate_dm_sparse 127 9.5 0.001 0.001 6.561 6.668 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.295 6.396 grid_collocate_task_list 127 9.7 6.286 6.323 6.286 6.323 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.043 6.128 fft3d_ps 1281 14.7 2.731 2.735 5.897 5.913 density_rs2pw 127 9.7 0.005 0.005 5.869 5.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.266000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2615.545455, yerr=164.995517 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 1.261240E+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 36659982. 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 0.058 0.092 85.215 85.219 qs_energies 1 2.0 0.000 0.000 84.578 84.584 ls_scf 1 3.0 0.000 0.002 83.657 83.662 dbcsr_multiply_generic 111 6.7 0.016 0.026 72.758 72.896 multiply_cannon 111 7.7 0.018 0.020 56.066 57.151 multiply_cannon_loop 111 8.7 0.208 0.224 52.597 53.833 ls_scf_main 1 4.0 0.000 0.002 52.262 52.263 density_matrix_trs4 2 5.0 0.002 0.003 46.866 46.947 ls_scf_init_scf 1 4.0 0.001 0.003 28.367 28.371 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.290 27.349 mp_waitall_1 11316 10.9 22.439 25.407 22.439 25.407 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 25.164 25.190 multiply_cannon_multrec 2664 9.7 8.183 9.010 15.484 17.352 multiply_cannon_sync_h2d 2664 9.7 13.761 15.747 13.761 15.747 make_m2s 222 7.7 0.009 0.012 13.087 13.625 make_images 222 8.7 0.101 0.121 13.065 13.603 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.778 12.629 make_images_data 222 9.7 0.004 0.005 7.585 8.133 dbcsr_mm_accdrv_process 4760 10.4 0.522 0.645 6.919 7.905 hybrid_alltoall_any 227 10.6 0.216 1.827 6.488 7.752 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.401 7.731 dbcsr_mm_accdrv_process_sort 4760 11.4 6.196 7.079 6.196 7.079 calculate_norms 4752 9.8 5.514 6.173 5.514 6.173 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.025 5.157 mp_sum_l 807 5.4 3.070 4.628 3.070 4.628 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.288 3.582 arnoldi_extremal 4 6.8 0.001 0.003 3.350 3.378 arnoldi_normal_ev 4 7.8 0.002 0.008 3.349 3.378 make_images_sizes 222 9.7 0.000 0.000 0.702 3.310 mp_alltoall_i44 222 10.7 0.701 3.310 0.701 3.310 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.042 3.272 mp_irecv_dv 6231 10.9 2.024 3.243 2.024 3.243 build_subspace 16 8.4 0.009 0.014 3.218 3.223 ls_scf_post 1 4.0 0.000 0.003 3.027 3.034 ls_scf_store_result 1 5.0 0.000 0.000 2.848 2.885 dbcsr_special_finalize 555 9.7 0.005 0.006 2.445 2.826 dbcsr_merge_single_wm 555 10.7 0.476 0.613 2.437 2.818 make_images_pack 222 9.7 2.207 2.642 2.209 2.644 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.324 2.581 dbcsr_sort_data 658 11.4 2.218 2.542 2.218 2.542 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.468 2.063 2.470 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.173 2.241 buffer_matrices_ensure_size 222 8.7 1.780 2.167 1.780 2.167 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.716 1.720 rebuild_ks_matrix 3 7.3 0.000 0.000 1.707 1.711 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.008 1.707 1.711 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.219000, 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/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 2.084790E+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 0.042 91.033 91.034 qs_energies 1 2.0 0.000 0.000 90.500 90.504 ls_scf 1 3.0 0.000 0.000 89.192 89.197 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.313 75.669 multiply_cannon 111 7.7 0.029 0.049 53.414 57.031 ls_scf_main 1 4.0 0.000 0.000 54.827 54.834 multiply_cannon_loop 111 8.7 0.116 0.123 50.134 53.292 density_matrix_trs4 2 5.0 0.002 0.003 49.130 49.359 ls_scf_init_scf 1 4.0 0.000 0.000 30.821 30.822 mp_waitall_1 9246 10.9 21.220 30.159 21.220 30.159 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.675 29.783 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.292 27.306 multiply_cannon_multrec 1332 9.7 13.211 16.820 22.414 27.149 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.867 20.956 make_m2s 222 7.7 0.006 0.008 15.320 16.032 make_images 222 8.7 1.574 1.970 15.290 16.003 dbcsr_mm_accdrv_process 4041 10.4 0.315 0.466 8.801 10.360 dbcsr_mm_accdrv_process_sort 4041 11.4 8.384 9.910 8.384 9.910 make_images_data 222 9.7 0.004 0.004 8.735 9.751 hybrid_alltoall_any 227 10.6 0.523 2.487 8.120 9.204 mp_sum_l 807 5.4 5.285 7.786 5.285 7.786 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.252 7.750 mp_irecv_dv 3311 11.0 3.232 7.700 3.232 7.700 calculate_norms 2376 9.8 5.976 6.754 5.976 6.754 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.065 6.111 multiply_cannon_sync_h2d 1332 9.7 4.785 5.737 4.785 5.737 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.116 5.324 arnoldi_extremal 4 6.8 0.000 0.000 4.668 4.699 arnoldi_normal_ev 4 7.8 0.001 0.005 4.668 4.699 build_subspace 16 8.4 0.014 0.021 4.408 4.411 ls_scf_post 1 4.0 0.000 0.000 3.545 3.548 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.162 3.389 ls_scf_store_result 1 5.0 0.000 0.000 3.262 3.369 dbcsr_matrix_vector_mult_local 304 10.0 2.771 3.259 2.773 3.261 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.244 2.860 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.518 2.593 make_images_pack 222 9.7 2.022 2.397 2.024 2.399 mp_allgather_i34 111 8.7 0.981 2.395 0.981 2.395 dbcsr_sort_data 436 11.2 1.803 2.077 1.803 2.077 dbcsr_data_new 4174 10.1 1.610 1.825 1.610 1.825 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.034000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1702.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 2.698449E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.070 93.284 93.286 qs_energies 1 2.0 0.000 0.000 92.615 92.625 ls_scf 1 3.0 0.000 0.001 91.194 91.198 dbcsr_multiply_generic 111 6.7 0.017 0.023 76.041 76.247 ls_scf_main 1 4.0 0.000 0.000 57.178 57.183 multiply_cannon 111 7.7 0.044 0.105 52.507 56.015 multiply_cannon_loop 111 8.7 0.100 0.118 48.869 53.251 density_matrix_trs4 2 5.0 0.002 0.003 51.340 51.541 mp_waitall_1 7374 11.0 22.967 33.221 22.967 33.221 ls_scf_init_scf 1 4.0 0.000 0.002 30.460 30.462 ls_scf_init_matrix_S 1 5.0 0.000 0.005 29.342 29.408 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.951 26.964 multiply_cannon_multrec 888 9.7 12.616 15.461 21.138 24.507 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.568 22.675 make_m2s 222 7.7 0.006 0.008 16.631 17.950 make_images 222 8.7 1.963 2.306 16.592 17.911 make_images_data 222 9.7 0.004 0.004 9.283 10.389 hybrid_alltoall_any 227 10.6 0.621 2.868 8.897 10.342 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.413 8.030 9.252 dbcsr_mm_accdrv_process_sort 3754 11.4 7.658 8.838 7.658 8.838 mp_sum_l 807 5.4 5.353 8.753 5.353 8.753 multiply_cannon_sync_h2d 888 9.7 6.101 7.377 6.101 7.377 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.737 7.161 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.443 7.057 mp_irecv_dv 2335 11.1 2.428 7.018 2.428 7.018 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.159 6.772 arnoldi_extremal 4 6.8 0.000 0.000 5.130 5.149 arnoldi_normal_ev 4 7.8 0.001 0.005 5.130 5.148 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.907 5.118 build_subspace 16 8.4 0.014 0.020 4.824 4.834 calculate_norms 1584 9.8 4.254 4.549 4.254 4.549 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.509 3.806 mp_allgather_i34 111 8.7 1.476 3.698 1.476 3.698 dbcsr_matrix_vector_mult_local 304 10.0 3.042 3.607 3.044 3.609 ls_scf_post 1 4.0 0.000 0.001 3.556 3.560 ls_scf_store_result 1 5.0 0.000 0.000 3.294 3.381 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.699 2.793 dbcsr_sort_data 325 11.1 1.870 2.165 1.870 2.165 make_images_sizes 222 9.7 0.000 0.000 1.046 2.141 mp_alltoall_i44 222 10.7 1.046 2.140 1.046 2.140 make_images_pack 222 9.7 1.811 2.115 1.813 2.118 dbcsr_data_release 9322 10.9 1.318 1.970 1.318 1.970 dbcsr_finalize 304 7.8 0.026 0.032 1.607 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.286000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2171.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 3.319738E+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 0.042 0.082 96.918 96.918 qs_energies 1 2.0 0.000 0.000 96.230 96.237 ls_scf 1 3.0 0.000 0.001 94.598 94.603 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.508 78.767 ls_scf_main 1 4.0 0.000 0.000 58.935 58.936 multiply_cannon 111 7.7 0.055 0.120 51.756 56.352 density_matrix_trs4 2 5.0 0.002 0.003 52.935 53.046 multiply_cannon_loop 111 8.7 0.114 0.127 46.706 50.102 ls_scf_init_scf 1 4.0 0.000 0.004 32.432 32.434 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.293 31.371 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.834 28.849 mp_waitall_1 6438 11.0 22.838 28.780 22.838 28.780 multiply_cannon_multrec 1332 9.7 14.162 17.315 21.978 24.830 make_m2s 222 7.7 0.007 0.008 20.899 22.338 make_images 222 8.7 3.126 3.583 20.849 22.289 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.402 17.364 make_images_data 222 9.7 0.004 0.004 11.649 13.241 hybrid_alltoall_any 227 10.6 0.797 3.785 10.936 12.745 dbcsr_mm_accdrv_process 3641 10.4 0.205 0.403 7.455 8.992 dbcsr_mm_accdrv_process_sort 3641 11.4 7.082 8.572 7.082 8.572 mp_sum_l 807 5.4 4.305 8.224 4.305 8.224 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.288 6.495 multiply_cannon_sync_h2d 1332 9.7 5.529 6.190 5.529 6.190 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.075 6.040 mp_irecv_dv 3229 10.9 2.052 5.972 2.052 5.972 arnoldi_extremal 4 6.8 0.000 0.000 5.285 5.301 arnoldi_normal_ev 4 7.8 0.001 0.005 5.285 5.301 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.614 5.281 build_subspace 16 8.4 0.014 0.021 4.948 4.955 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.550 4.762 calculate_norms 2376 9.8 4.171 4.516 4.171 4.516 mp_allgather_i34 111 8.7 2.172 4.474 2.172 4.474 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.588 3.896 dbcsr_matrix_vector_mult_local 304 10.0 3.189 3.698 3.191 3.700 dbcsr_sort_data 658 11.4 3.030 3.491 3.030 3.491 ls_scf_post 1 4.0 0.000 0.001 3.231 3.239 dbcsr_special_finalize 555 9.7 0.006 0.007 2.785 3.225 dbcsr_merge_single_wm 555 10.7 0.535 0.657 2.777 3.216 ls_scf_store_result 1 5.0 0.000 0.000 2.973 3.034 ls_scf_dm_to_ks 2 5.0 0.000 0.002 2.875 2.930 dbcsr_data_release 10477 10.7 1.584 2.419 1.584 2.419 dbcsr_finalize 304 7.8 0.049 0.061 1.786 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.918000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2702.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 4.620304E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.056 98.713 98.714 qs_energies 1 2.0 0.000 0.000 97.958 97.961 ls_scf 1 3.0 0.000 0.000 96.054 96.058 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.874 78.071 ls_scf_main 1 4.0 0.000 0.000 61.828 61.828 multiply_cannon 111 7.7 0.099 0.173 55.559 60.747 density_matrix_trs4 2 5.0 0.002 0.003 54.864 54.966 multiply_cannon_loop 111 8.7 0.069 0.079 51.118 52.700 mp_waitall_1 5481 11.0 26.716 31.246 26.716 31.246 ls_scf_init_scf 1 4.0 0.000 0.000 30.598 30.603 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.430 29.467 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.265 27.273 multiply_cannon_multrec 444 9.7 14.021 16.049 21.032 23.650 make_m2s 222 7.7 0.004 0.005 17.547 19.984 make_images 222 8.7 3.718 4.418 17.486 19.924 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.273 16.750 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.410 14.569 make_images_data 222 9.7 0.003 0.004 9.833 12.143 hybrid_alltoall_any 227 10.6 0.790 3.765 9.612 11.969 multiply_cannon_sync_h2d 444 9.7 6.604 8.154 6.604 8.154 dbcsr_mm_accdrv_process 3003 10.4 0.177 0.344 6.710 7.829 dbcsr_mm_accdrv_process_sort 3003 11.4 6.394 7.485 6.394 7.485 mp_allgather_i34 111 8.7 2.676 6.987 2.676 6.987 arnoldi_extremal 4 6.8 0.000 0.000 5.819 5.829 arnoldi_normal_ev 4 7.8 0.002 0.005 5.819 5.829 build_subspace 16 8.4 0.015 0.020 5.424 5.434 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.582 4.712 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.137 4.309 mp_sum_l 807 5.4 2.801 4.244 2.801 4.244 dbcsr_matrix_vector_mult_local 304 10.0 3.650 4.087 3.652 4.089 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.600 3.748 mp_irecv_dv 1241 11.2 1.579 3.722 1.579 3.722 calculate_norms 792 9.8 3.561 3.698 3.561 3.698 ls_scf_post 1 4.0 0.000 0.000 3.628 3.630 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.362 3.460 ls_scf_store_result 1 5.0 0.000 0.000 3.398 3.450 make_images_sizes 222 9.7 0.000 0.000 1.067 3.195 mp_alltoall_i44 222 10.7 1.067 3.195 1.067 3.195 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.918 3.150 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.331 dbcsr_merge_all 275 8.9 0.473 0.533 2.047 2.165 dbcsr_data_release 10123 10.8 1.346 1.987 1.346 1.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.714000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3608.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/34639cb15a96dbc4801a8ed01621e36a84bba0a6_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 8.720617E+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 0.101 107.282 107.283 qs_energies 1 2.0 0.000 0.000 105.821 105.842 ls_scf 1 3.0 0.000 0.000 102.933 102.953 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.315 77.434 ls_scf_main 1 4.0 0.000 0.000 65.558 65.559 density_matrix_trs4 2 5.0 0.002 0.003 56.751 56.818 multiply_cannon 111 7.7 0.096 0.195 49.875 51.823 multiply_cannon_loop 111 8.7 0.068 0.077 46.408 47.315 ls_scf_init_scf 1 4.0 0.000 0.000 33.718 33.718 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.384 32.405 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.562 29.572 mp_waitall_1 4569 11.1 22.165 26.163 22.165 26.163 make_m2s 222 7.7 0.005 0.005 23.854 24.841 make_images 222 8.7 4.575 4.952 23.748 24.733 multiply_cannon_multrec 444 9.7 17.902 18.517 22.481 23.054 make_images_data 222 9.7 0.003 0.003 13.096 15.563 hybrid_alltoall_any 227 10.6 1.657 3.634 12.906 15.558 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.480 10.926 multiply_cannon_sync_h2d 444 9.7 8.820 8.885 8.820 8.885 arnoldi_extremal 4 6.8 0.000 0.000 7.222 7.232 arnoldi_normal_ev 4 7.8 0.003 0.009 7.222 7.232 build_subspace 16 8.4 0.026 0.036 6.673 6.682 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.375 5.503 dbcsr_matrix_vector_mult_local 304 10.0 4.969 5.246 4.971 5.248 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.969 5.227 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.817 4.908 dbcsr_mm_accdrv_process 1814 10.4 0.278 0.313 4.415 4.538 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.242 4.114 4.242 ls_scf_post 1 4.0 0.000 0.000 3.657 3.677 mp_allgather_i34 111 8.7 1.120 3.549 1.120 3.549 make_images_sizes 222 9.7 0.000 0.000 1.450 3.501 mp_alltoall_i44 222 10.7 1.450 3.501 1.450 3.501 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.412 calculate_norms 792 9.8 3.268 3.316 3.268 3.316 dbcsr_finalize 304 7.8 0.082 0.089 3.075 3.118 dbcsr_complete_redistribute 5 7.6 1.442 1.487 2.773 2.916 dbcsr_merge_all 275 8.9 0.888 0.916 2.861 2.899 dbcsr_data_release 12724 10.6 2.322 2.873 2.322 2.873 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.858 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.429 2.581 dbcsr_sort_data 325 11.1 2.438 2.496 2.438 2.496 dbcsr_new_transposed 4 7.5 0.242 0.251 2.280 2.287 dbcsr_frobenius_norm 74 6.6 2.057 2.132 2.192 2.232 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.205 dbcsr_add_anytype 103 7.2 0.859 0.892 2.131 2.205 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.195 2.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.283000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6952.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 34639cb15a96dbc4801a8ed01621e36a84bba0a6 Summary: empty Status: OK