=== 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: 9400c5a650e170a3f6cad4de0da2f9ce09dfe117 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/01 job id: 42165126 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/02 job id: 42165127 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/03 job id: 42165131 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/04 job id: 42165133 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/05 job id: 42165135 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/06 job id: 42165138 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/07 job id: 42165142 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/08 job id: 42165146 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/09 job id: 42165148 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/10 job id: 42165149 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/11 job id: 42165150 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/12 job id: 42165151 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/13 job id: 42165153 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/14 job id: 42165154 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/15 job id: 42165156 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/16 job id: 42165158 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/17 job id: 42165160 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/18 job id: 42165164 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/19 job id: 42165166 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/20 job id: 42165168 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/21 job id: 42165170 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/22 job id: 42165172 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/23 job id: 42165175 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/24 job id: 42165178 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/25 job id: 42165179 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/26 job id: 42165180 --- 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 133.056 133.057 farming_run 1 2.0 132.532 132.533 133.022 133.025 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463046E+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 0.014 0.027 115.822 115.822 qs_energies 1 2.0 0.000 0.000 115.582 115.584 mp2_main 1 3.0 0.000 0.000 113.267 113.268 mp2_gpw_main 1 4.0 0.020 0.026 112.139 112.142 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.183 93.410 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.363 55.590 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.174 41.650 46.715 get_2c_integrals 1 6.0 0.000 0.001 37.258 37.648 integrate_v_rspace 273 8.0 0.435 0.449 25.042 29.755 pw_transfer 6555 10.6 0.377 0.388 27.323 27.709 fft_wrap_pw1pw2 5465 11.4 0.046 0.049 25.985 26.357 grid_integrate_task_list 273 9.0 20.871 26.012 20.871 26.012 fft_wrap_pw1pw2_100 2178 12.4 1.175 1.242 23.524 23.900 compute_2c_integrals 1 7.0 0.003 0.004 19.293 19.294 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.889 19.008 mp2_eri_2c_integrate_gpw 1 9.0 2.415 2.451 18.886 19.004 rpa_ri_compute_en 1 5.0 0.001 0.001 18.843 18.988 cp_fm_cholesky_decompose 12 8.2 18.082 18.549 18.082 18.549 cholesky_decomp 1 7.0 0.000 0.000 16.812 17.216 fft3d_s 5443 13.4 16.117 16.329 16.139 16.352 ao_to_mo_and_store_B_mult_1 272 7.0 10.873 15.589 10.873 15.589 calculate_wavefunction 272 8.0 5.416 5.498 12.516 13.163 rpa_num_int 1 6.0 0.000 0.001 10.715 10.724 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.619 10.676 calc_mat_Q 8 8.0 0.000 0.000 9.333 9.432 contract_S_to_Q 8 9.0 0.000 0.000 8.753 8.854 calc_potential_gpw 544 9.5 0.005 0.006 8.259 8.639 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.201 8.446 parallel_gemm_fm 14 9.1 0.000 0.000 8.331 8.419 parallel_gemm_fm_cosma 14 10.1 8.331 8.419 8.331 8.419 potential_pw2rs 545 10.0 0.107 0.109 7.665 8.346 create_integ_mat 1 6.0 0.015 0.027 7.806 7.815 collocate_single_gaussian 272 10.0 0.040 0.042 7.455 7.651 array2fm 1 7.0 0.000 0.000 6.772 7.278 pw_scatter_s 2720 13.7 4.449 4.595 4.449 4.595 pw_gather_s 2722 13.2 3.854 4.180 3.854 4.180 array2fm_buffer_send 1 8.0 2.940 3.153 2.940 3.153 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.141552, 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/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.036 394.908 394.910 farming_run 1 2.0 394.157 394.165 394.867 394.871 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.225474E+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 17540648. 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 0.051 209.336 209.337 qs_energies 1 2.0 0.000 0.000 209.079 209.115 scf_env_do_scf 1 3.0 0.000 0.000 106.343 106.343 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.456 105.465 rebuild_ks_matrix 4 6.0 0.000 0.000 105.455 105.463 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.067 105.455 105.463 hfx_ks_matrix 4 8.0 0.001 0.001 105.078 105.081 integrate_four_center 4 9.0 0.144 0.479 105.077 105.080 mp2_main 1 3.0 0.003 0.023 102.442 102.477 mp2_gpw_main 1 4.0 0.041 0.108 101.594 101.608 integrate_four_center_main 4 10.0 0.107 0.542 96.348 99.125 integrate_four_center_bin 269 11.0 96.240 99.109 96.240 99.109 init_scf_loop 1 4.0 0.000 0.000 91.850 91.851 mp2_ri_gpw_compute_in 1 5.0 0.070 0.123 74.901 75.941 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.018 54.537 55.578 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.267 47.387 integrate_v_rspace 95 8.0 0.397 0.564 28.620 33.535 pw_transfer 2240 10.6 0.146 0.172 29.950 30.348 fft_wrap_pw1pw2 1868 11.4 0.019 0.022 28.952 29.358 grid_integrate_task_list 95 9.0 23.906 29.020 23.906 29.020 mp2_ri_gpw_compute_en 1 5.0 0.055 0.069 26.472 28.169 ao_to_mo_and_store_B_mult_1 91 7.0 10.584 27.565 10.584 27.565 fft_wrap_pw1pw2_100 730 12.4 1.278 1.421 26.681 27.099 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.850 1.910 24.770 24.780 get_2c_integrals 1 6.0 0.002 0.022 20.257 20.298 compute_2c_integrals 1 7.0 0.005 0.023 19.232 19.238 compute_2c_integrals_loop_lm 1 8.0 0.005 0.025 18.912 19.094 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.859 18.907 19.092 fft3d_s 1823 13.4 18.443 18.738 18.456 18.752 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.490 14.490 calculate_wavefunction 91 8.0 2.021 2.055 9.755 10.025 potential_pw2rs 186 10.0 0.033 0.035 8.615 9.292 mp2_ri_gpw_compute_en_expansio 172 7.0 0.561 0.608 8.740 9.270 local_gemm 172 8.0 8.179 8.702 8.179 8.702 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.262 8.578 calc_potential_gpw 182 9.5 0.002 0.002 7.965 8.223 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.520 7.786 8.206 collocate_single_gaussian 91 10.0 0.017 0.033 7.894 8.118 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.423 6.348 6.423 mp_sendrecv_dm3 2068 8.0 5.819 6.214 5.819 6.214 mp_sync 38 10.4 3.396 5.830 3.396 5.830 pw_gather_s 912 13.2 4.915 5.533 4.915 5.533 pw_scatter_s 910 13.7 3.949 4.299 3.949 4.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.595591, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 450.166784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 49.254 49.254 qs_mol_dyn_low 1 2.0 0.003 0.004 49.066 49.074 qs_forces 11 3.9 0.002 0.002 48.999 49.001 qs_energies 11 4.9 0.001 0.002 47.509 47.525 scf_env_do_scf 11 5.9 0.000 0.001 41.677 41.678 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 39.649 39.649 dbcsr_multiply_generic 2286 12.5 0.094 0.106 31.228 31.635 qs_scf_new_mos 108 7.5 0.000 0.001 29.609 29.908 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.608 29.908 ot_scf_mini 108 9.5 0.002 0.002 28.016 28.205 multiply_cannon 2286 13.5 0.185 0.192 24.881 26.421 multiply_cannon_loop 2286 14.5 1.495 1.583 24.159 25.740 velocity_verlet 10 3.0 0.002 0.002 24.351 24.351 ot_mini 108 10.5 0.001 0.001 17.755 18.013 qs_ot_get_derivative 108 11.5 0.001 0.001 14.860 15.043 mp_waitall_1 267858 16.1 7.764 13.659 7.764 13.659 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 5.347 11.705 multiply_cannon_multrec 54864 15.5 4.259 6.694 7.520 10.902 rebuild_ks_matrix 119 8.3 0.000 0.000 7.905 8.044 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.905 8.043 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.959 7.089 multiply_cannon_sync_h2d 54864 15.5 5.926 6.718 5.926 6.718 qs_ot_get_p 119 10.4 0.001 0.001 5.857 6.140 mp_sum_l 7207 12.9 4.289 5.782 4.289 5.782 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.838 5.271 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.127 5.244 init_scf_run 11 5.9 0.000 0.001 4.624 4.624 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.623 4.624 sum_up_and_integrate 119 10.3 0.012 0.015 4.535 4.542 integrate_v_rspace 119 11.3 0.002 0.002 4.522 4.531 dbcsr_mm_accdrv_process 76910 16.1 1.180 1.837 3.183 4.437 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.153 4.247 calculate_rho_elec 119 8.7 0.011 0.016 4.153 4.247 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.300 3.335 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.866 3.050 apply_single 119 13.6 0.000 0.000 2.866 3.050 calculate_dm_sparse 119 9.5 0.000 0.000 2.801 2.941 rs_pw_transfer 974 11.9 0.012 0.013 2.853 2.937 ot_diis_step 108 11.5 0.006 0.006 2.646 2.646 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.563 2.564 multiply_cannon_metrocomm1 54864 15.5 0.052 0.059 1.443 2.544 jit_kernel_multiply 13 15.8 1.940 2.514 1.940 2.514 calculate_first_density_matrix 1 7.0 0.000 0.001 2.423 2.428 density_rs2pw 119 9.7 0.004 0.004 2.266 2.374 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.357 2.357 cp_fm_redistribute_end 50 14.0 2.141 2.335 2.145 2.337 cp_fm_diag_elpa_base 50 14.0 0.189 2.291 0.190 2.304 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.179 2.242 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.221 2.223 acc_transpose_blocks 54864 15.5 0.225 0.241 1.741 2.164 wfi_extrapolate 11 7.9 0.001 0.001 2.139 2.139 grid_integrate_task_list 119 12.3 2.030 2.114 2.030 2.114 init_scf_loop 11 6.9 0.000 0.000 2.010 2.011 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.930 1.980 potential_pw2rs 119 12.3 0.004 0.004 1.894 1.906 make_m2s 4572 13.5 0.054 0.055 1.746 1.805 pw_transfer 1439 11.6 0.051 0.057 1.723 1.800 mp_sum_d 4129 12.0 1.209 1.776 1.209 1.776 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.648 1.727 make_images 4572 14.5 0.134 0.140 1.664 1.723 fft3d_ps 1201 14.6 0.357 0.464 1.426 1.501 mp_alltoall_d11v 2130 13.8 1.312 1.473 1.312 1.473 mp_waitany 12084 13.8 1.272 1.424 1.272 1.424 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.369 1.385 grid_collocate_task_list 119 9.7 1.287 1.362 1.287 1.362 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.277 1.354 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.667 1.066 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.990 1.007 acc_transpose_blocks_kernels 54864 16.5 0.255 0.387 0.743 1.001 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=49.254000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 486.031360E+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 604573. 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 0.023 0.038 40.006 40.008 qs_mol_dyn_low 1 2.0 0.003 0.004 39.721 39.729 qs_forces 11 3.9 0.002 0.005 39.661 39.662 qs_energies 11 4.9 0.017 0.019 37.901 37.905 scf_env_do_scf 11 5.9 0.001 0.001 32.428 32.429 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.798 29.799 dbcsr_multiply_generic 2286 12.5 0.101 0.107 22.088 22.489 qs_scf_new_mos 108 7.5 0.001 0.001 20.573 20.819 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.573 20.818 ot_scf_mini 108 9.5 0.003 0.003 19.658 19.833 velocity_verlet 10 3.0 0.001 0.003 19.351 19.352 multiply_cannon 2286 13.5 0.208 0.218 16.675 18.351 multiply_cannon_loop 2286 14.5 0.912 0.984 15.487 17.024 ot_mini 108 10.5 0.001 0.003 12.384 12.624 mp_waitall_1 217478 16.2 6.438 11.662 6.438 11.662 qs_ot_get_derivative 108 11.5 0.001 0.002 9.858 10.036 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.309 9.653 multiply_cannon_multrec 27432 15.5 1.967 4.505 6.014 8.842 rebuild_ks_matrix 119 8.3 0.000 0.000 7.390 7.533 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 7.390 7.533 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.486 6.618 dbcsr_mm_accdrv_process 47894 16.0 3.005 5.169 3.978 5.869 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.916 4.770 qs_ot_get_p 119 10.4 0.001 0.001 4.339 4.566 sum_up_and_integrate 119 10.3 0.024 0.028 4.373 4.378 integrate_v_rspace 119 11.3 0.002 0.003 4.348 4.357 mp_sum_l 7207 12.9 2.242 4.208 2.242 4.208 init_scf_run 11 5.9 0.000 0.001 4.198 4.198 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.198 4.198 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.079 4.169 apply_single 119 13.6 0.000 0.000 3.079 4.169 qs_rho_update_rho_low 119 7.7 0.001 0.004 3.733 3.768 calculate_rho_elec 119 8.7 0.021 0.024 3.732 3.767 make_m2s 4572 13.5 0.052 0.054 2.726 2.968 make_images 4572 14.5 0.202 0.239 2.637 2.878 rs_pw_transfer 974 11.9 0.010 0.011 2.745 2.860 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.785 2.805 multiply_cannon_sync_h2d 27432 15.5 2.144 2.738 2.144 2.738 init_scf_loop 11 6.9 0.001 0.003 2.608 2.610 calculate_first_density_matrix 1 7.0 0.001 0.002 2.549 2.552 ot_diis_step 108 11.5 0.011 0.014 2.467 2.468 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.343 2.344 calculate_dm_sparse 119 9.5 0.000 0.000 2.191 2.262 jit_kernel_multiply 11 16.1 0.920 2.255 0.920 2.255 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.156 2.248 density_rs2pw 119 9.7 0.004 0.004 2.085 2.173 potential_pw2rs 119 12.3 0.006 0.007 2.101 2.118 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.058 2.061 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.981 1.982 pw_transfer 1439 11.6 0.063 0.066 1.925 1.968 cp_fm_redistribute_end 50 14.0 1.630 1.951 1.634 1.953 grid_integrate_task_list 119 12.3 1.824 1.937 1.824 1.937 cp_fm_diag_elpa_base 50 14.0 0.309 1.884 0.317 1.914 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.835 1.880 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.808 1.852 prepare_preconditioner 11 7.9 0.000 0.000 1.662 1.690 make_preconditioner 11 8.9 0.000 0.001 1.662 1.690 make_images_data 4572 15.5 0.045 0.051 1.222 1.666 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.560 1.618 wfi_extrapolate 11 7.9 0.001 0.002 1.591 1.592 fft3d_ps 1201 14.6 0.498 0.549 1.543 1.582 acc_transpose_blocks 27432 15.5 0.108 0.113 1.222 1.544 hybrid_alltoall_any 4725 16.4 0.050 0.110 1.075 1.518 mp_allgather_i34 2286 14.5 0.632 1.424 0.632 1.424 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.367 1.410 mp_alltoall_d11v 2130 13.8 1.337 1.407 1.337 1.407 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.387 1.399 grid_collocate_task_list 119 9.7 1.242 1.303 1.242 1.303 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.206 1.253 mp_sum_d 4129 12.0 0.650 1.094 0.650 1.094 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.965 0.981 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.969 0.971 make_images_sizes 4572 15.5 0.005 0.005 0.680 0.946 mp_alltoall_i44 4572 16.5 0.675 0.942 0.675 0.942 acc_transpose_blocks_kernels 27432 16.5 0.184 0.274 0.680 0.912 mp_alltoall_z22v 1201 16.6 0.791 0.866 0.791 0.866 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.604 0.803 0.854 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.008000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.909091, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 520.069120E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.032 32.836 32.839 qs_mol_dyn_low 1 2.0 0.003 0.003 32.525 32.533 qs_forces 11 3.9 0.002 0.003 32.468 32.469 qs_energies 11 4.9 0.015 0.016 30.883 30.885 scf_env_do_scf 11 5.9 0.001 0.001 25.876 25.876 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.154 23.155 dbcsr_multiply_generic 2286 12.5 0.094 0.099 16.909 17.021 velocity_verlet 10 3.0 0.001 0.001 15.737 15.738 qs_scf_new_mos 108 7.5 0.001 0.001 15.068 15.095 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.068 15.094 ot_scf_mini 108 9.5 0.002 0.003 14.321 14.335 multiply_cannon 2286 13.5 0.195 0.203 13.398 14.233 multiply_cannon_loop 2286 14.5 0.644 0.677 12.550 13.427 ot_mini 108 10.5 0.001 0.001 8.969 8.984 qs_ot_get_derivative 108 11.5 0.001 0.001 7.430 7.443 multiply_cannon_multrec 18288 15.5 1.915 2.911 6.970 7.293 rebuild_ks_matrix 119 8.3 0.000 0.000 6.419 6.442 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.419 6.442 dbcsr_mm_accdrv_process 38222 16.0 4.301 5.558 4.970 5.942 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.672 5.693 mp_waitall_1 169478 16.3 3.105 4.193 3.105 4.193 sum_up_and_integrate 119 10.3 0.030 0.031 4.013 4.018 integrate_v_rspace 119 11.3 0.002 0.003 3.983 3.992 init_scf_run 11 5.9 0.000 0.001 3.747 3.747 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.747 3.747 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.827 3.448 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.304 3.318 calculate_rho_elec 119 8.7 0.030 0.031 3.304 3.317 qs_ot_get_p 119 10.4 0.001 0.001 3.132 3.162 init_scf_loop 11 6.9 0.001 0.002 2.700 2.706 calculate_first_density_matrix 1 7.0 0.001 0.002 2.533 2.534 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.492 2.489 rs_pw_transfer 974 11.9 0.009 0.010 2.323 2.428 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.051 2.365 apply_single 119 13.6 0.000 0.000 2.051 2.364 make_m2s 4572 13.5 0.044 0.045 1.898 2.024 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.008 2.016 calculate_dm_sparse 119 9.5 0.000 0.000 1.976 1.995 density_rs2pw 119 9.7 0.004 0.004 1.870 1.976 make_images 4572 14.5 0.192 0.203 1.812 1.938 grid_integrate_task_list 119 12.3 1.811 1.915 1.811 1.915 prepare_preconditioner 11 7.9 0.000 0.000 1.872 1.874 make_preconditioner 11 8.9 0.000 0.001 1.872 1.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.835 1.837 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.727 1.805 pw_transfer 1439 11.6 0.063 0.066 1.782 1.799 potential_pw2rs 119 12.3 0.007 0.008 1.779 1.788 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.739 1.739 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.692 1.710 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.643 1.649 mp_sum_l 7207 12.9 1.231 1.634 1.231 1.634 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.600 1.608 multiply_cannon_sync_h2d 18288 15.5 1.352 1.529 1.352 1.529 ot_diis_step 108 11.5 0.011 0.011 1.510 1.510 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.501 1.502 cp_fm_redistribute_end 50 14.0 1.115 1.481 1.117 1.481 cp_fm_diag_elpa_base 50 14.0 0.349 1.414 0.363 1.454 jit_kernel_multiply 7 16.1 0.618 1.445 0.618 1.445 fft3d_ps 1201 14.6 0.506 0.522 1.380 1.398 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.324 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.326 1.332 acc_transpose_blocks 18288 15.5 0.076 0.078 1.276 1.300 grid_collocate_task_list 119 9.7 1.204 1.270 1.204 1.270 wfi_extrapolate 11 7.9 0.001 0.001 1.166 1.166 make_images_data 4572 15.5 0.045 0.049 0.850 1.012 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.969 0.970 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.931 0.954 hybrid_alltoall_any 4725 16.4 0.054 0.113 0.731 0.916 mp_alltoall_d11v 2130 13.8 0.756 0.878 0.756 0.878 acc_transpose_blocks_kernels 18288 16.5 0.211 0.221 0.826 0.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.824 cp_fm_cholesky_invert 11 10.9 0.808 0.812 0.808 0.812 mp_alltoall_z22v 1201 16.6 0.713 0.790 0.713 0.790 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.729 mp_allgather_i34 2286 14.5 0.308 0.687 0.308 0.687 mp_waitany 9880 13.7 0.554 0.666 0.554 0.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.839000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.090909, yerr=0.995859 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 550.191104E+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 0.018 0.030 35.184 35.186 qs_mol_dyn_low 1 2.0 0.003 0.004 34.925 34.933 qs_forces 11 3.9 0.002 0.003 34.865 34.867 qs_energies 11 4.9 0.016 0.017 33.163 33.169 scf_env_do_scf 11 5.9 0.001 0.001 27.867 27.868 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.332 24.333 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.498 18.616 velocity_verlet 10 3.0 0.002 0.002 17.978 17.980 qs_scf_new_mos 108 7.5 0.001 0.001 16.179 16.230 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.179 16.230 ot_scf_mini 108 9.5 0.002 0.003 15.222 15.272 multiply_cannon 2286 13.5 0.236 0.278 14.805 15.162 multiply_cannon_loop 2286 14.5 0.942 0.973 13.825 14.167 ot_mini 108 10.5 0.001 0.001 9.391 9.449 multiply_cannon_multrec 27432 15.5 2.386 3.104 8.939 9.275 dbcsr_mm_accdrv_process 47916 15.9 5.592 7.331 6.458 7.732 qs_ot_get_derivative 108 11.5 0.001 0.001 7.558 7.606 rebuild_ks_matrix 119 8.3 0.000 0.000 6.485 6.540 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.484 6.539 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.763 5.811 init_scf_run 11 5.9 0.000 0.001 3.917 3.917 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.916 3.917 sum_up_and_integrate 119 10.3 0.035 0.038 3.745 3.754 integrate_v_rspace 119 11.3 0.002 0.003 3.710 3.721 init_scf_loop 11 6.9 0.001 0.004 3.510 3.512 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.310 3.341 calculate_rho_elec 119 8.7 0.040 0.046 3.310 3.341 qs_ot_get_p 119 10.4 0.001 0.002 3.218 3.288 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.867 3.260 prepare_preconditioner 11 7.9 0.000 0.000 2.637 2.645 make_preconditioner 11 8.9 0.000 0.001 2.637 2.645 mp_waitall_1 145218 16.4 2.120 2.629 2.120 2.629 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.232 2.565 calculate_first_density_matrix 1 7.0 0.001 0.002 2.532 2.533 make_m2s 4572 13.5 0.054 0.056 2.263 2.350 calculate_dm_sparse 119 9.5 0.000 0.000 2.200 2.264 make_images 4572 14.5 0.274 0.336 2.156 2.242 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.046 2.175 apply_single 119 13.6 0.000 0.000 2.046 2.175 rs_pw_transfer 974 11.9 0.009 0.009 1.999 2.114 jit_kernel_multiply 10 16.3 0.805 2.080 0.805 2.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.949 1.950 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.913 1.938 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.911 1.921 grid_integrate_task_list 119 12.3 1.822 1.919 1.822 1.919 density_rs2pw 119 9.7 0.004 0.004 1.784 1.891 pw_transfer 1439 11.6 0.063 0.066 1.768 1.796 ot_diis_step 108 11.5 0.012 0.013 1.793 1.794 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.764 1.776 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.678 1.711 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.598 1.599 potential_pw2rs 119 12.3 0.008 0.010 1.529 1.535 acc_transpose_blocks 27432 15.5 0.112 0.115 1.495 1.524 fft_wrap_pw1pw2_140 487 13.2 0.084 0.091 1.364 1.397 fft3d_ps 1201 14.6 0.534 0.588 1.360 1.386 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.363 1.374 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.366 1.367 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.762 1.348 cp_fm_redistribute_end 50 14.0 0.901 1.343 0.902 1.344 wfi_extrapolate 11 7.9 0.001 0.001 1.337 1.337 cp_fm_diag_elpa_base 50 14.0 0.420 1.279 0.439 1.320 grid_collocate_task_list 119 9.7 1.219 1.293 1.219 1.293 mp_sum_l 7207 12.9 1.006 1.268 1.006 1.268 cp_fm_upper_to_full 72 13.5 0.819 1.137 0.819 1.137 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.114 1.128 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.090 1.091 dbcsr_complete_redistribute 329 12.2 0.125 0.154 0.790 1.067 multiply_cannon_sync_h2d 27432 15.5 0.979 1.030 0.979 1.030 make_images_data 4572 15.5 0.045 0.048 0.894 0.990 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.772 0.912 mp_alltoall_d11v 2130 13.8 0.797 0.909 0.797 0.909 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.858 0.875 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.792 0.870 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.586 0.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.814 0.819 cp_fm_cholesky_invert 11 10.9 0.797 0.801 0.797 0.801 mp_alltoall_z22v 1201 16.6 0.708 0.733 0.708 0.733 mp_alltoall_i22 627 13.8 0.442 0.732 0.442 0.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.186000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.181818, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 611.799040E+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 0.016 0.032 27.876 27.877 qs_mol_dyn_low 1 2.0 0.003 0.004 27.700 27.707 qs_forces 11 3.9 0.002 0.003 27.643 27.644 qs_energies 11 4.9 0.001 0.001 25.953 25.956 scf_env_do_scf 11 5.9 0.001 0.009 21.349 21.349 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.756 18.757 velocity_verlet 10 3.0 0.002 0.004 14.273 14.276 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.205 12.270 qs_scf_new_mos 108 7.5 0.001 0.001 11.135 11.164 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.134 11.164 ot_scf_mini 108 9.5 0.002 0.002 10.454 10.484 multiply_cannon 2286 13.5 0.230 0.239 9.536 10.022 multiply_cannon_loop 2286 14.5 0.334 0.344 8.588 8.739 multiply_cannon_multrec 9144 15.5 1.577 1.820 5.730 6.009 rebuild_ks_matrix 119 8.3 0.000 0.000 5.851 5.872 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.851 5.872 ot_mini 108 10.5 0.001 0.001 5.778 5.812 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.229 5.248 qs_ot_get_derivative 108 11.5 0.001 0.001 4.421 4.451 dbcsr_mm_accdrv_process 12550 15.8 3.075 4.091 4.051 4.130 sum_up_and_integrate 119 10.3 0.037 0.041 3.584 3.590 integrate_v_rspace 119 11.3 0.003 0.003 3.547 3.553 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.240 3.245 calculate_rho_elec 119 8.7 0.059 0.061 3.240 3.245 init_scf_run 11 5.9 0.000 0.001 3.179 3.180 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.179 3.179 qs_ot_get_p 119 10.4 0.001 0.001 2.857 2.901 init_scf_loop 11 6.9 0.000 0.000 2.563 2.563 mp_waitall_1 121218 16.5 1.782 2.245 1.782 2.245 calculate_first_density_matrix 1 7.0 0.000 0.000 2.032 2.033 make_m2s 4572 13.5 0.034 0.036 1.840 1.966 grid_integrate_task_list 119 12.3 1.853 1.905 1.853 1.905 make_images 4572 14.5 0.269 0.304 1.751 1.876 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.850 1.853 prepare_preconditioner 11 7.9 0.000 0.000 1.817 1.821 make_preconditioner 11 8.9 0.000 0.000 1.817 1.821 rs_pw_transfer 974 11.9 0.008 0.008 1.686 1.793 density_rs2pw 119 9.7 0.004 0.004 1.652 1.755 pw_transfer 1439 11.6 0.063 0.065 1.734 1.743 jit_kernel_multiply 10 15.7 0.938 1.738 0.938 1.738 calculate_dm_sparse 119 9.5 0.000 0.000 1.709 1.728 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.697 1.722 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.643 1.653 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.646 1.647 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.505 1.519 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.511 1.512 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.361 1.362 potential_pw2rs 119 12.3 0.010 0.011 1.352 1.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.343 1.350 ot_diis_step 108 11.5 0.013 0.013 1.341 1.341 grid_collocate_task_list 119 9.7 1.273 1.341 1.273 1.341 cp_fm_redistribute_end 50 14.0 0.676 1.339 0.677 1.339 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.328 1.337 cp_fm_diag_elpa_base 50 14.0 0.617 1.266 0.661 1.324 fft3d_ps 1201 14.6 0.539 0.551 1.312 1.319 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.233 1.248 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.209 1.233 apply_single 119 13.6 0.000 0.000 1.209 1.233 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.208 1.209 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.125 1.130 wfi_extrapolate 11 7.9 0.001 0.001 1.101 1.101 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.835 1.053 make_images_data 4572 15.5 0.038 0.042 0.873 1.040 cp_fm_cholesky_invert 11 10.9 0.937 0.940 0.937 0.940 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.921 mp_alltoall_d11v 2130 13.8 0.830 0.897 0.830 0.897 acc_transpose_blocks 9144 15.5 0.039 0.040 0.886 0.893 multiply_cannon_sync_h2d 9144 15.5 0.709 0.778 0.709 0.778 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.342 0.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.745 0.748 mp_allgather_i34 2286 14.5 0.279 0.743 0.279 0.743 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.734 0.743 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.663 0.715 mp_alltoall_z22v 1201 16.6 0.646 0.680 0.646 0.680 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.640 0.643 mp_waitany 5200 13.7 0.461 0.571 0.461 0.571 qs_create_task_list 11 7.9 0.001 0.001 0.543 0.569 generate_qs_task_list 11 8.9 0.190 0.212 0.543 0.568 dbcsr_complete_redistribute 329 12.2 0.166 0.176 0.537 0.565 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.877000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=578.636364, yerr=5.139380 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 730.800128E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.039 40.985 40.987 qs_mol_dyn_low 1 2.0 0.003 0.003 40.772 40.779 qs_forces 11 3.9 0.002 0.002 40.712 40.713 qs_energies 11 4.9 0.001 0.001 38.774 38.777 scf_env_do_scf 11 5.9 0.001 0.002 33.231 33.231 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.375 25.376 velocity_verlet 10 3.0 0.002 0.002 22.902 22.908 dbcsr_multiply_generic 2286 12.5 0.101 0.102 17.611 17.732 qs_scf_new_mos 108 7.5 0.001 0.001 16.419 16.519 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.418 16.518 ot_scf_mini 108 9.5 0.002 0.002 15.331 15.431 multiply_cannon 2286 13.5 0.300 0.306 13.628 14.473 multiply_cannon_loop 2286 14.5 0.345 0.354 12.359 13.185 ot_mini 108 10.5 0.001 0.001 9.105 9.220 multiply_cannon_multrec 9144 15.5 3.384 4.797 8.535 8.657 init_scf_loop 11 6.9 0.000 0.000 7.828 7.829 qs_ot_get_derivative 108 11.5 0.001 0.001 7.047 7.144 rebuild_ks_matrix 119 8.3 0.000 0.000 6.852 6.994 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.852 6.994 prepare_preconditioner 11 7.9 0.000 0.000 6.908 6.921 make_preconditioner 11 8.9 0.000 0.000 6.908 6.921 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.409 6.800 dbcsr_mm_accdrv_process 12550 15.8 4.082 5.614 5.026 6.342 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.205 6.334 cp_fm_upper_to_full 72 14.2 3.271 4.715 3.271 4.715 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.739 3.742 calculate_rho_elec 119 8.7 0.118 0.121 3.738 3.742 sum_up_and_integrate 119 10.3 0.064 0.065 3.670 3.678 integrate_v_rspace 119 11.3 0.003 0.003 3.606 3.616 qs_ot_get_p 119 10.4 0.001 0.001 3.455 3.600 mp_waitall_1 97218 16.6 2.722 3.549 2.722 3.549 init_scf_run 11 5.9 0.000 0.001 3.544 3.544 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.544 3.544 dbcsr_complete_redistribute 329 12.2 0.300 0.303 2.064 2.940 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.495 2.887 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.750 2.624 make_m2s 4572 13.5 0.038 0.038 2.371 2.534 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.172 2.475 apply_single 119 13.6 0.000 0.000 2.172 2.474 make_images 4572 14.5 0.353 0.381 2.251 2.414 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.493 2.360 mp_alltoall_i22 627 13.8 1.483 2.340 1.483 2.340 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.400 2.184 calculate_dm_sparse 119 9.5 0.000 0.000 2.147 2.159 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.112 2.114 calculate_first_density_matrix 1 7.0 0.000 0.000 2.086 2.087 pw_transfer 1439 11.6 0.066 0.068 2.081 2.084 grid_integrate_task_list 119 12.3 2.007 2.032 2.007 2.032 ot_diis_step 108 11.5 0.014 0.014 2.029 2.029 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.986 1.987 density_rs2pw 119 9.7 0.003 0.004 1.835 1.854 mp_sum_l 7207 12.9 1.071 1.854 1.071 1.854 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.832 1.832 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.821 1.821 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.764 1.813 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.736 1.737 jit_kernel_multiply 10 15.5 0.918 1.735 0.918 1.735 fft_wrap_pw1pw2_140 487 13.2 0.088 0.089 1.652 1.656 fft3d_ps 1201 14.6 0.569 0.583 1.618 1.623 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.542 1.583 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.507 1.508 cp_fm_diag_elpa_base 50 14.0 1.357 1.411 1.506 1.506 grid_collocate_task_list 119 9.7 1.449 1.467 1.449 1.467 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.207 1.459 rs_pw_transfer 974 11.9 0.009 0.009 1.404 1.422 cp_fm_cholesky_invert 11 10.9 1.412 1.415 1.412 1.415 wfi_extrapolate 11 7.9 0.001 0.001 1.399 1.400 make_images_data 4572 15.5 0.042 0.045 1.151 1.371 potential_pw2rs 119 12.3 0.014 0.015 1.212 1.216 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.210 mp_alltoall_d11v 2130 13.8 1.170 1.202 1.170 1.202 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.146 1.164 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.075 1.086 multiply_cannon_sync_h2d 9144 15.5 1.034 1.037 1.034 1.037 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.988 1.032 qs_create_task_list 11 7.9 0.001 0.001 0.938 0.951 generate_qs_task_list 11 8.9 0.372 0.391 0.938 0.950 mp_alltoall_z22v 1201 16.6 0.912 0.930 0.912 0.930 acc_transpose_blocks 9144 15.5 0.039 0.039 0.890 0.897 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.861 0.874 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.987000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=687.454545, yerr=14.234169 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 502.218752E+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 803800. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 82.211 82.212 qs_mol_dyn_low 1 2.0 0.003 0.004 81.979 81.988 qs_forces 11 3.9 0.002 0.002 81.911 81.912 qs_energies 11 4.9 0.001 0.002 79.097 79.116 scf_env_do_scf 11 5.9 0.000 0.001 70.312 70.315 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.735 64.736 dbcsr_multiply_generic 2055 12.4 0.105 0.107 51.966 52.169 qs_scf_new_mos 99 7.5 0.000 0.001 47.841 47.949 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.840 47.948 ot_scf_mini 99 9.5 0.002 0.002 45.480 45.557 multiply_cannon 2055 13.4 0.181 0.187 42.438 43.381 multiply_cannon_loop 2055 14.4 1.561 1.597 41.424 42.367 velocity_verlet 10 3.0 0.001 0.001 42.109 42.110 ot_mini 99 10.5 0.001 0.001 27.662 27.742 qs_ot_get_derivative 99 11.5 0.001 0.001 20.826 20.923 multiply_cannon_multrec 49320 15.4 12.132 12.889 17.169 17.957 rebuild_ks_matrix 110 8.3 0.000 0.001 14.381 14.479 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.381 14.479 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.647 12.736 mp_waitall_1 241148 16.1 11.640 12.666 11.640 12.666 multiply_cannon_sync_h2d 49320 15.4 9.934 10.470 9.934 10.470 qs_ot_get_p 110 10.4 0.001 0.001 9.324 9.458 multiply_cannon_metrocomm3 49320 15.4 0.076 0.081 6.878 7.887 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.221 7.774 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.171 7.712 apply_single 110 13.6 0.000 0.000 7.170 7.712 sum_up_and_integrate 110 10.3 0.037 0.044 7.009 7.024 integrate_v_rspace 110 11.3 0.003 0.003 6.972 6.996 init_scf_run 11 5.9 0.000 0.001 6.759 6.759 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.759 6.759 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.536 6.576 ot_diis_step 99 11.5 0.006 0.006 6.575 6.575 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.219 6.381 calculate_rho_elec 110 8.6 0.020 0.024 6.218 6.380 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.065 6.107 mp_sum_l 6514 12.8 5.177 5.981 5.177 5.981 init_scf_loop 11 6.9 0.000 0.000 5.545 5.546 dbcsr_mm_accdrv_process 87628 16.1 2.091 2.193 4.915 5.214 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.054 5.054 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.570 4.571 cp_fm_redistribute_end 48 14.0 3.983 4.546 3.987 4.548 cp_fm_diag_elpa_base 48 14.0 0.554 4.446 0.558 4.475 wfi_extrapolate 11 7.9 0.001 0.001 4.030 4.030 make_m2s 4110 13.4 0.061 0.064 3.893 4.019 make_images 4110 14.4 0.180 0.194 3.798 3.925 rs_pw_transfer 902 11.9 0.012 0.013 3.747 3.896 calculate_dm_sparse 110 9.5 0.000 0.001 3.727 3.820 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.631 3.557 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.549 3.553 density_rs2pw 110 9.6 0.004 0.004 3.313 3.487 grid_integrate_task_list 110 12.3 3.248 3.446 3.248 3.446 prepare_preconditioner 11 7.9 0.000 0.000 3.397 3.414 make_preconditioner 11 8.9 0.000 0.000 3.397 3.414 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.257 3.308 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.191 3.244 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.173 3.219 pw_transfer 1331 11.6 0.053 0.066 2.904 2.977 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.818 2.894 calculate_first_density_matrix 1 7.0 0.000 0.000 2.644 2.652 potential_pw2rs 110 12.3 0.006 0.007 2.602 2.636 mp_alltoall_d11v 2046 13.8 2.126 2.586 2.126 2.586 jit_kernel_multiply 13 15.9 2.537 2.551 2.537 2.551 fft_wrap_pw1pw2_140 451 13.1 0.168 0.189 2.354 2.430 fft3d_ps 1111 14.6 0.748 0.833 2.356 2.422 acc_transpose_blocks 49320 15.4 0.213 0.227 2.180 2.252 grid_collocate_task_list 110 9.6 2.087 2.235 2.087 2.235 mp_waitany 14300 13.8 1.779 2.061 1.779 2.061 mp_sum_d 3883 11.9 1.467 2.045 1.467 2.045 make_images_data 4110 15.4 0.043 0.046 1.745 1.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.874 1.888 cp_fm_cholesky_invert 11 10.9 1.867 1.870 1.867 1.870 hybrid_alltoall_any 4261 16.3 0.081 0.477 1.524 1.777 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.646 1.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.212000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=2.790184 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.218368E+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 1788305. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.033 67.898 67.900 qs_mol_dyn_low 1 2.0 0.003 0.005 67.657 67.667 qs_forces 11 3.9 0.003 0.009 67.577 67.578 qs_energies 11 4.9 0.001 0.001 64.280 64.284 scf_env_do_scf 11 5.9 0.000 0.001 55.868 55.872 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.158 48.158 dbcsr_multiply_generic 2055 12.4 0.113 0.116 37.516 37.717 velocity_verlet 10 3.0 0.003 0.016 35.972 35.973 qs_scf_new_mos 99 7.5 0.001 0.001 32.428 32.596 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.428 32.595 multiply_cannon 2055 13.4 0.222 0.247 30.940 32.139 ot_scf_mini 99 9.5 0.003 0.004 30.766 30.935 multiply_cannon_loop 2055 14.4 0.937 0.968 29.707 30.800 ot_mini 99 10.5 0.001 0.002 18.134 18.309 multiply_cannon_multrec 24660 15.4 7.610 9.613 13.808 15.596 rebuild_ks_matrix 110 8.3 0.000 0.001 13.579 13.704 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 13.578 13.703 qs_ot_get_derivative 99 11.5 0.001 0.002 12.301 12.466 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.976 12.086 mp_waitall_1 186928 16.3 7.777 10.394 7.777 10.394 multiply_cannon_sync_h2d 24660 15.4 7.080 8.204 7.080 8.204 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.165 7.912 init_scf_loop 11 6.9 0.000 0.000 7.673 7.673 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.541 7.229 apply_single 110 13.6 0.000 0.001 6.541 7.229 sum_up_and_integrate 110 10.3 0.053 0.060 6.450 6.467 integrate_v_rspace 110 11.3 0.002 0.003 6.397 6.409 dbcsr_mm_accdrv_process 52282 16.1 4.603 5.623 6.038 6.320 qs_ot_get_p 110 10.4 0.001 0.001 6.019 6.225 init_scf_run 11 5.9 0.000 0.001 6.038 6.038 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.037 6.038 ot_diis_step 99 11.5 0.010 0.014 5.785 5.785 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.678 5.687 calculate_rho_elec 110 8.6 0.039 0.047 5.677 5.687 prepare_preconditioner 11 7.9 0.000 0.000 5.656 5.676 make_preconditioner 11 8.9 0.000 0.000 5.656 5.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.224 5.386 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.631 5.327 make_m2s 4110 13.4 0.057 0.059 4.221 4.635 make_images 4110 14.4 0.402 0.455 4.113 4.523 qs_ot_p2m_diag 48 11.0 0.029 0.046 4.177 4.197 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.749 3.750 wfi_extrapolate 11 7.9 0.001 0.003 3.508 3.508 pw_transfer 1331 11.6 0.065 0.069 3.267 3.402 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.161 3.296 grid_integrate_task_list 110 12.3 3.149 3.275 3.149 3.275 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.153 3.246 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.239 3.244 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.215 3.216 cp_fm_redistribute_end 48 14.0 2.397 3.189 2.399 3.189 cp_fm_diag_elpa_base 48 14.0 0.755 3.068 0.787 3.153 density_rs2pw 110 9.6 0.004 0.004 2.984 3.095 rs_pw_transfer 902 11.9 0.012 0.014 2.863 2.995 calculate_dm_sparse 110 9.5 0.001 0.001 2.959 2.992 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.896 2.941 fft_wrap_pw1pw2_140 451 13.1 0.200 0.219 2.662 2.793 make_images_data 4110 15.4 0.047 0.051 2.373 2.790 hybrid_alltoall_any 4261 16.3 0.101 0.442 2.071 2.700 fft3d_ps 1111 14.6 1.073 1.265 2.519 2.642 cp_fm_cholesky_invert 11 10.9 2.627 2.634 2.627 2.634 calculate_first_density_matrix 1 7.0 0.000 0.001 2.444 2.447 mp_sum_l 6514 12.8 1.732 2.447 1.732 2.447 potential_pw2rs 110 12.3 0.008 0.009 2.335 2.353 grid_collocate_task_list 110 9.6 2.120 2.295 2.120 2.295 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.949 1.968 mp_alltoall_d11v 2046 13.8 1.759 1.907 1.759 1.907 jit_kernel_multiply 10 16.3 1.078 1.895 1.078 1.895 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.769 1.880 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.864 1.866 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.768 1.781 mp_irecv_dv 57340 16.2 0.645 1.665 0.645 1.665 mp_allgather_i34 2055 14.4 0.561 1.592 0.561 1.592 acc_transpose_blocks 24660 15.4 0.108 0.111 1.531 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.474 dbcsr_complete_redistribute 325 12.2 0.240 0.298 1.157 1.421 cp_fm_cholesky_decompose 22 10.9 1.384 1.393 1.384 1.393 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=67.900000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=7.708533 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 655.396864E+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 0.030 0.079 60.313 60.314 qs_mol_dyn_low 1 2.0 0.003 0.003 59.914 59.923 qs_forces 11 3.9 0.002 0.004 59.848 59.851 qs_energies 11 4.9 0.002 0.005 56.653 56.659 scf_env_do_scf 11 5.9 0.001 0.001 48.737 48.738 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 39.859 39.860 velocity_verlet 10 3.0 0.001 0.002 32.920 32.922 dbcsr_multiply_generic 2055 12.4 0.107 0.111 28.618 28.944 qs_scf_new_mos 99 7.5 0.001 0.001 25.306 25.421 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.305 25.420 ot_scf_mini 99 9.5 0.002 0.003 24.080 24.208 multiply_cannon 2055 13.4 0.212 0.222 22.177 23.411 multiply_cannon_loop 2055 14.4 0.623 0.635 20.940 22.367 ot_mini 99 10.5 0.001 0.001 13.737 13.870 rebuild_ks_matrix 110 8.3 0.000 0.000 12.244 12.415 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.244 12.414 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.796 10.950 multiply_cannon_multrec 16440 15.4 3.806 4.880 9.563 10.569 mp_waitall_1 146766 16.3 7.367 10.364 7.367 10.364 qs_ot_get_derivative 99 11.5 0.001 0.001 9.196 9.328 init_scf_loop 11 6.9 0.000 0.002 8.840 8.844 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.507 7.448 prepare_preconditioner 11 7.9 0.000 0.000 7.050 7.069 make_preconditioner 11 8.9 0.000 0.001 7.050 7.069 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.384 6.723 sum_up_and_integrate 110 10.3 0.060 0.061 6.480 6.498 integrate_v_rspace 110 11.3 0.002 0.003 6.420 6.439 dbcsr_mm_accdrv_process 34862 16.1 4.793 5.270 5.612 5.774 qs_ot_get_p 110 10.4 0.001 0.001 5.509 5.659 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.551 5.564 calculate_rho_elec 110 8.6 0.058 0.058 5.551 5.563 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.085 5.559 apply_single 110 13.6 0.000 0.000 5.085 5.559 init_scf_run 11 5.9 0.000 0.001 5.489 5.489 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.489 5.489 make_m2s 4110 13.4 0.050 0.051 4.233 4.604 ot_diis_step 99 11.5 0.010 0.011 4.507 4.507 make_images 4110 14.4 0.396 0.513 4.118 4.489 multiply_cannon_sync_h2d 16440 15.4 3.637 4.341 3.637 4.341 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.827 3.834 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.140 3.779 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.485 3.486 grid_integrate_task_list 110 12.3 3.217 3.389 3.217 3.389 pw_transfer 1331 11.6 0.064 0.070 3.221 3.234 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.115 3.129 density_rs2pw 110 9.6 0.004 0.004 2.830 3.006 wfi_extrapolate 11 7.9 0.001 0.001 2.994 2.994 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.958 2.960 cp_fm_redistribute_end 48 14.0 1.842 2.934 1.844 2.935 rs_pw_transfer 902 11.9 0.010 0.011 2.717 2.917 make_images_data 4110 15.4 0.044 0.048 2.430 2.915 cp_fm_diag_elpa_base 48 14.0 1.029 2.799 1.087 2.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.821 2.823 cp_fm_cholesky_invert 11 10.9 2.813 2.820 2.813 2.820 hybrid_alltoall_any 4261 16.3 0.104 0.370 2.147 2.812 fft_wrap_pw1pw2_140 451 13.1 0.210 0.217 2.635 2.648 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.913 2.547 calculate_dm_sparse 110 9.5 0.001 0.001 2.497 2.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.499 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.381 2.462 fft3d_ps 1111 14.6 1.064 1.075 2.430 2.443 mp_irecv_dv 48980 15.7 0.843 2.423 0.843 2.423 calculate_first_density_matrix 1 7.0 0.001 0.003 2.409 2.410 grid_collocate_task_list 110 9.6 2.180 2.361 2.180 2.361 potential_pw2rs 110 12.3 0.010 0.010 2.278 2.293 mp_sum_l 6514 12.8 1.620 2.290 1.620 2.290 mp_alltoall_d11v 2046 13.8 1.815 2.179 1.815 2.179 dbcsr_complete_redistribute 325 12.2 0.346 0.393 1.565 2.017 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.977 1.981 cp_fm_upper_to_full 70 13.6 1.368 1.825 1.368 1.825 cp_fm_cholesky_decompose 22 10.9 1.684 1.707 1.684 1.707 mp_allgather_i34 2055 14.4 0.559 1.681 0.559 1.681 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.649 1.667 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.358 1.488 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.031 1.476 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.459 1.469 mp_waitany 17072 13.8 1.160 1.342 1.160 1.342 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.286 1.300 acc_transpose_blocks 16440 15.4 0.074 0.076 1.220 1.260 rs_gather_matrices 110 12.3 0.138 0.150 0.872 1.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.314000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.272727, yerr=7.699952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 724.307968E+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 0.071 0.094 65.791 65.795 qs_mol_dyn_low 1 2.0 0.003 0.003 65.323 65.332 qs_forces 11 3.9 0.002 0.002 65.135 65.136 qs_energies 11 4.9 0.002 0.006 61.755 61.758 scf_env_do_scf 11 5.9 0.001 0.001 53.457 53.460 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.368 41.370 velocity_verlet 10 3.0 0.001 0.002 37.285 37.292 dbcsr_multiply_generic 2055 12.4 0.117 0.120 29.708 29.933 qs_scf_new_mos 99 7.5 0.001 0.001 26.891 27.000 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.890 26.999 ot_scf_mini 99 9.5 0.003 0.004 25.254 25.348 multiply_cannon 2055 13.4 0.245 0.261 22.524 23.764 multiply_cannon_loop 2055 14.4 0.887 0.914 21.113 21.714 ot_mini 99 10.5 0.001 0.001 14.465 14.575 multiply_cannon_multrec 24660 15.4 4.188 6.679 12.606 13.760 init_scf_loop 11 6.9 0.001 0.002 12.046 12.049 rebuild_ks_matrix 110 8.3 0.000 0.000 11.943 12.039 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.943 12.039 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.562 10.646 qs_ot_get_derivative 99 11.5 0.001 0.001 10.282 10.376 prepare_preconditioner 11 7.9 0.000 0.000 10.289 10.305 make_preconditioner 11 8.9 0.000 0.002 10.289 10.305 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.456 9.966 dbcsr_mm_accdrv_process 52304 16.0 6.686 8.430 8.271 9.164 mp_waitall_1 126806 16.4 4.672 6.585 4.672 6.585 sum_up_and_integrate 110 10.3 0.068 0.072 6.366 6.380 integrate_v_rspace 110 11.3 0.003 0.003 6.298 6.311 qs_ot_get_p 110 10.4 0.001 0.001 5.798 5.930 make_m2s 4110 13.4 0.059 0.061 5.505 5.739 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.696 5.709 calculate_rho_elec 110 8.6 0.077 0.081 5.696 5.709 make_images 4110 14.4 0.580 0.697 5.364 5.591 init_scf_run 11 5.9 0.000 0.001 5.529 5.530 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.529 5.529 cp_fm_upper_to_full 70 13.8 3.347 4.779 3.347 4.779 ot_diis_step 99 11.5 0.011 0.011 4.144 4.145 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.044 4.110 apply_single 110 13.6 0.000 0.000 4.043 4.109 dbcsr_complete_redistribute 325 12.2 0.423 0.488 2.831 4.010 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.950 3.964 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.459 3.460 grid_integrate_task_list 110 12.3 3.278 3.456 3.278 3.456 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.239 3.398 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.336 3.379 pw_transfer 1331 11.6 0.064 0.071 3.319 3.348 multiply_cannon_sync_h2d 24660 15.4 3.178 3.325 3.178 3.325 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.602 3.280 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.214 3.245 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.111 3.164 density_rs2pw 110 9.6 0.004 0.004 2.808 3.078 make_images_data 4110 15.4 0.046 0.050 2.741 3.023 hybrid_alltoall_any 4261 16.3 0.119 0.456 2.436 3.007 wfi_extrapolate 11 7.9 0.001 0.001 2.997 2.997 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.824 2.970 calculate_dm_sparse 110 9.5 0.001 0.001 2.928 2.959 mp_alltoall_i22 605 13.7 1.712 2.930 1.712 2.930 cp_fm_cholesky_invert 11 10.9 2.872 2.880 2.872 2.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.869 2.871 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.824 2.826 cp_fm_redistribute_end 48 14.0 1.408 2.791 1.410 2.792 cp_fm_diag_elpa_base 48 14.0 1.299 2.651 1.379 2.763 rs_pw_transfer 902 11.9 0.010 0.011 2.459 2.758 fft_wrap_pw1pw2_140 451 13.1 0.202 0.215 2.699 2.732 fft3d_ps 1111 14.6 1.066 1.100 2.517 2.535 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.442 2.474 calculate_first_density_matrix 1 7.0 0.000 0.000 2.440 2.443 grid_collocate_task_list 110 9.6 2.218 2.366 2.218 2.366 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.268 2.269 jit_kernel_multiply 12 15.9 1.251 2.192 1.251 2.192 mp_alltoall_d11v 2046 13.8 1.985 2.180 1.985 2.180 potential_pw2rs 110 12.3 0.013 0.013 2.143 2.156 cp_fm_cholesky_decompose 22 10.9 1.727 1.785 1.727 1.785 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.730 1.759 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.628 1.639 mp_sum_l 6514 12.8 1.010 1.618 1.010 1.618 mp_allgather_i34 2055 14.4 0.509 1.580 0.509 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.548 acc_transpose_blocks 24660 15.4 0.105 0.108 1.510 1.534 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.844 1.402 mp_irecv_dv 62702 16.1 0.747 1.332 0.747 1.332 mp_waitany 13376 13.8 1.070 1.329 1.070 1.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.795000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.090909, yerr=6.999410 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 841.568256E+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 0.019 0.034 55.583 55.584 qs_mol_dyn_low 1 2.0 0.003 0.003 55.296 55.305 qs_forces 11 3.9 0.002 0.003 55.231 55.232 qs_energies 11 4.9 0.003 0.003 51.629 51.634 scf_env_do_scf 11 5.9 0.000 0.001 43.171 43.171 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.178 35.178 velocity_verlet 10 3.0 0.002 0.002 31.131 31.134 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.111 23.258 qs_scf_new_mos 99 7.5 0.001 0.001 20.935 20.975 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.935 20.975 ot_scf_mini 99 9.5 0.002 0.002 19.684 19.720 multiply_cannon 2055 13.4 0.248 0.263 17.468 18.542 multiply_cannon_loop 2055 14.4 0.324 0.337 16.104 16.372 rebuild_ks_matrix 110 8.3 0.000 0.000 11.520 11.562 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.519 11.561 ot_mini 99 10.5 0.001 0.001 10.587 10.614 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.268 10.307 multiply_cannon_multrec 8220 15.4 3.183 4.493 7.436 8.589 mp_waitall_1 106626 16.5 6.421 8.225 6.421 8.225 init_scf_loop 11 6.9 0.000 0.000 7.946 7.947 qs_ot_get_derivative 99 11.5 0.001 0.001 6.705 6.740 prepare_preconditioner 11 7.9 0.000 0.000 6.303 6.308 make_preconditioner 11 8.9 0.000 0.000 6.303 6.308 sum_up_and_integrate 110 10.3 0.079 0.080 6.233 6.246 integrate_v_rspace 110 11.3 0.003 0.003 6.153 6.166 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.877 5.951 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.693 5.705 calculate_rho_elec 110 8.6 0.115 0.115 5.692 5.705 init_scf_run 11 5.9 0.000 0.001 5.325 5.325 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.325 5.325 qs_ot_get_p 110 10.4 0.001 0.001 5.119 5.150 dbcsr_mm_accdrv_process 17442 15.9 2.722 3.483 4.122 5.043 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.160 4.571 make_m2s 4110 13.4 0.038 0.040 4.309 4.570 make_images 4110 14.4 0.639 0.696 4.179 4.442 ot_diis_step 99 11.5 0.012 0.012 3.856 3.856 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.776 3.820 apply_single 110 13.6 0.000 0.000 3.776 3.819 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.714 3.718 grid_integrate_task_list 110 12.3 3.358 3.540 3.358 3.540 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.402 3.403 pw_transfer 1331 11.6 0.065 0.070 3.323 3.333 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.217 3.230 cp_fm_cholesky_invert 11 10.9 3.118 3.121 3.118 3.121 multiply_cannon_sync_h2d 8220 15.4 2.882 3.040 2.882 3.040 make_images_data 4110 15.4 0.038 0.044 2.451 2.882 density_rs2pw 110 9.6 0.004 0.004 2.730 2.846 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.828 2.829 cp_fm_redistribute_end 48 14.0 0.717 2.796 0.721 2.797 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.386 2.770 cp_fm_diag_elpa_base 48 14.0 1.889 2.609 2.068 2.762 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.736 2.749 wfi_extrapolate 11 7.9 0.001 0.001 2.741 2.741 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.660 2.660 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.593 2.594 calculate_dm_sparse 110 9.5 0.001 0.001 2.547 2.588 grid_collocate_task_list 110 9.6 2.327 2.518 2.327 2.518 calculate_first_density_matrix 1 7.0 0.000 0.000 2.493 2.494 fft3d_ps 1111 14.6 1.115 1.132 2.477 2.483 rs_pw_transfer 902 11.9 0.010 0.010 2.227 2.365 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.152 2.167 jit_kernel_multiply 11 15.8 1.085 2.002 1.085 2.002 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.768 1.981 potential_pw2rs 110 12.3 0.015 0.016 1.969 1.974 mp_alltoall_d11v 2046 13.8 1.746 1.965 1.746 1.965 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.945 1.960 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.864 1.882 cp_fm_cholesky_decompose 22 10.9 1.722 1.739 1.722 1.739 dbcsr_complete_redistribute 325 12.2 0.616 0.642 1.574 1.674 mp_allgather_i34 2055 14.4 0.544 1.667 0.544 1.667 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.503 1.619 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.586 1.589 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.504 1.514 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.478 1.337 qs_create_task_list 11 7.9 0.006 0.007 1.230 1.328 generate_qs_task_list 11 8.9 0.377 0.445 1.224 1.322 mp_irecv_dv 24056 15.7 0.455 1.296 0.455 1.296 mp_waitany 9240 13.8 1.071 1.225 1.071 1.225 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.190 1.217 mp_alltoall_z22v 1111 16.6 1.129 1.143 1.129 1.143 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.584000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=787.727273, yerr=14.722221 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 1.422967E+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 0.021 0.045 86.267 86.268 qs_mol_dyn_low 1 2.0 0.003 0.003 85.955 85.964 qs_forces 11 3.9 0.002 0.003 85.891 85.891 qs_energies 11 4.9 0.001 0.002 81.852 81.854 scf_env_do_scf 11 5.9 0.000 0.001 72.062 72.062 velocity_verlet 10 3.0 0.002 0.002 55.559 55.565 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.917 42.919 init_scf_loop 11 6.9 0.000 0.000 29.070 29.072 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.667 28.785 prepare_preconditioner 11 7.9 0.000 0.000 27.121 27.132 make_preconditioner 11 8.9 0.000 0.000 27.121 27.132 qs_scf_new_mos 99 7.5 0.001 0.001 26.509 26.586 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.508 26.585 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.160 26.582 ot_scf_mini 99 9.5 0.002 0.002 24.738 24.802 multiply_cannon 2055 13.4 0.352 0.374 21.540 22.283 multiply_cannon_loop 2055 14.4 0.344 0.347 19.707 20.007 cp_fm_upper_to_full 70 14.2 13.051 18.721 13.051 18.721 ot_mini 99 10.5 0.001 0.001 13.739 13.800 rebuild_ks_matrix 110 8.3 0.000 0.001 13.154 13.235 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.154 13.235 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.959 12.032 dbcsr_complete_redistribute 325 12.2 1.044 1.069 7.593 10.940 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.537 9.884 multiply_cannon_multrec 8220 15.4 4.349 4.522 9.549 9.637 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.945 9.270 qs_ot_get_derivative 99 11.5 0.001 0.001 9.067 9.133 mp_waitall_1 87304 16.6 7.876 8.922 7.876 8.922 mp_alltoall_i22 605 13.7 5.574 8.895 5.574 8.895 sum_up_and_integrate 110 10.3 0.150 0.152 6.626 6.641 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.481 6.517 calculate_rho_elec 110 8.6 0.227 0.227 6.480 6.516 integrate_v_rspace 110 11.3 0.004 0.004 6.476 6.491 qs_ot_get_p 110 10.4 0.001 0.001 5.791 5.860 make_m2s 4110 13.4 0.043 0.043 5.311 5.834 make_images 4110 14.4 0.883 0.932 5.124 5.646 init_scf_run 11 5.9 0.000 0.001 5.567 5.567 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.567 5.567 cp_fm_cholesky_invert 11 10.9 5.519 5.524 5.519 5.524 dbcsr_mm_accdrv_process 11614 15.7 3.374 3.683 5.056 5.298 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.758 5.265 apply_single 110 13.6 0.000 0.000 4.758 5.264 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.816 5.076 ot_diis_step 99 11.5 0.015 0.016 4.647 4.647 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.197 4.204 multiply_cannon_sync_h2d 8220 15.4 3.940 3.945 3.940 3.945 pw_transfer 1331 11.6 0.073 0.074 3.773 3.781 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.749 3.750 grid_integrate_task_list 110 12.3 3.659 3.713 3.659 3.713 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.658 3.666 hybrid_alltoall_any 4261 16.3 0.256 0.553 2.896 3.628 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.622 3.622 make_images_data 4110 15.4 0.041 0.044 2.897 3.540 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.885 3.347 wfi_extrapolate 11 7.9 0.001 0.001 3.233 3.233 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.168 3.181 calculate_dm_sparse 110 9.5 0.001 0.001 3.131 3.153 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.138 3.139 cp_fm_diag_elpa_base 48 14.0 2.592 2.790 3.137 3.137 density_rs2pw 110 9.6 0.004 0.004 2.972 2.995 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.868 2.874 fft3d_ps 1111 14.6 1.264 1.269 2.868 2.872 grid_collocate_task_list 110 9.6 2.627 2.656 2.627 2.656 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.436 2.470 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.342 2.368 calculate_first_density_matrix 1 7.0 0.000 0.000 2.230 2.231 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.162 2.223 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.170 rs_pw_transfer 902 11.9 0.011 0.011 2.125 2.169 mp_alltoall_d11v 2046 13.8 2.077 2.129 2.077 2.129 potential_pw2rs 110 12.3 0.021 0.022 2.066 2.073 cp_fm_cholesky_decompose 22 10.9 2.048 2.065 2.048 2.065 qs_create_task_list 11 7.9 0.001 0.001 1.893 1.937 generate_qs_task_list 11 8.9 0.737 0.790 1.892 1.936 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.813 1.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.799 1.808 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.268000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1249.545455, yerr=83.726138 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 630.710272E+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 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 1950215. 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 0.016 0.033 200.884 200.885 qs_mol_dyn_low 1 2.0 0.003 0.004 200.533 200.546 qs_forces 11 3.9 0.004 0.004 200.449 200.450 qs_energies 11 4.9 0.005 0.009 194.981 194.995 scf_env_do_scf 11 5.9 0.001 0.001 178.591 178.596 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 158.375 158.377 dbcsr_multiply_generic 2507 12.6 0.174 0.177 123.087 123.576 velocity_verlet 10 3.0 0.001 0.002 121.256 121.257 qs_scf_new_mos 117 7.6 0.001 0.001 119.822 120.039 qs_scf_loop_do_ot 117 8.6 0.001 0.001 119.821 120.038 ot_scf_mini 117 9.6 0.003 0.003 113.298 113.502 multiply_cannon 2507 13.6 0.237 0.244 100.333 101.922 multiply_cannon_loop 2507 14.6 2.120 2.197 98.180 99.646 ot_mini 117 10.6 0.001 0.004 65.052 65.245 multiply_cannon_multrec 60168 15.6 33.095 35.672 41.481 43.180 qs_ot_get_derivative 117 11.6 0.001 0.001 40.295 40.499 rebuild_ks_matrix 128 8.3 0.001 0.001 33.136 33.489 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.135 33.488 mp_waitall_1 291448 16.2 28.125 30.487 28.125 30.487 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.812 30.147 multiply_cannon_sync_h2d 60168 15.6 27.328 29.701 27.328 29.701 qs_ot_get_p 128 10.4 0.001 0.001 26.504 26.752 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.220 24.964 apply_single 128 13.6 0.001 0.001 24.219 24.964 ot_diis_step 117 11.6 0.008 0.008 24.369 24.370 init_scf_loop 11 6.9 0.000 0.000 20.140 20.141 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.882 20.003 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.875 19.011 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.394 17.395 multiply_cannon_metrocomm3 60168 15.6 0.112 0.117 15.346 17.177 prepare_preconditioner 11 7.9 0.000 0.000 15.584 15.630 make_preconditioner 11 8.9 0.000 0.000 15.584 15.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.809 14.966 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.535 14.541 cp_fm_redistribute_end 83 14.4 11.476 14.463 11.487 14.465 cp_fm_diag_elpa_base 83 14.4 2.935 14.179 2.966 14.294 make_m2s 5014 13.6 0.104 0.111 13.722 14.232 make_images 5014 14.6 0.407 0.425 13.542 14.062 sum_up_and_integrate 128 10.3 0.090 0.109 13.916 13.935 integrate_v_rspace 128 11.3 0.003 0.004 13.826 13.848 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.560 12.665 calculate_rho_elec 128 8.7 0.045 0.063 12.559 12.664 init_scf_run 11 5.9 0.000 0.001 12.261 12.262 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.261 12.262 mp_sum_l 7870 13.0 7.766 9.087 7.766 9.087 wfi_extrapolate 11 7.9 0.001 0.001 9.044 9.044 cp_fm_cholesky_invert 11 10.9 8.801 8.808 8.801 8.808 dbcsr_mm_accdrv_process 124484 16.2 3.325 3.498 7.947 8.531 calculate_dm_sparse 128 9.5 0.001 0.001 8.335 8.407 multiply_cannon_metrocomm1 60168 15.6 0.088 0.094 5.938 7.840 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.717 7.800 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.666 7.798 make_images_data 5014 15.6 0.069 0.076 6.723 7.676 grid_integrate_task_list 128 12.3 7.122 7.508 7.122 7.508 hybrid_alltoall_any 5200 16.5 0.289 2.244 5.899 7.247 pw_transfer 1547 11.6 0.074 0.107 6.753 7.034 density_rs2pw 128 9.7 0.006 0.007 6.368 6.831 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 6.549 6.805 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.598 6.607 rs_pw_transfer 1046 11.9 0.017 0.020 5.566 6.066 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.765 5.911 fft_wrap_pw1pw2_140 523 13.2 0.445 0.509 5.645 5.819 fft3d_ps 1291 14.7 2.089 2.565 5.342 5.540 mp_alltoall_d11v 2415 14.1 4.169 5.480 4.169 5.480 grid_collocate_task_list 128 9.7 4.726 5.171 4.726 5.171 cp_fm_cholesky_decompose 22 10.9 4.611 4.624 4.611 4.624 potential_pw2rs 128 12.3 0.010 0.011 4.399 4.425 mp_sum_d 4473 12.2 3.518 4.215 3.518 4.215 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=200.885000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.090909, yerr=8.490150 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 826.290176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 5473037. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.127 190.731 190.753 qs_mol_dyn_low 1 2.0 0.003 0.005 190.221 190.240 qs_forces 11 3.9 0.004 0.007 190.133 190.136 qs_energies 11 4.9 0.010 0.059 183.453 183.472 scf_env_do_scf 11 5.9 0.001 0.002 167.105 167.119 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.837 133.842 velocity_verlet 10 3.0 0.002 0.004 119.942 119.945 dbcsr_multiply_generic 2507 12.6 0.184 0.189 97.046 98.352 qs_scf_new_mos 117 7.6 0.001 0.001 95.336 96.009 qs_scf_loop_do_ot 117 8.6 0.001 0.002 95.335 96.009 ot_scf_mini 117 9.6 0.004 0.005 90.517 91.249 multiply_cannon 2507 13.6 0.477 0.533 76.898 80.928 multiply_cannon_loop 2507 14.6 1.263 1.301 73.890 76.655 ot_mini 117 10.6 0.001 0.004 50.048 50.699 mp_waitall_1 226760 16.4 25.034 37.840 25.034 37.840 multiply_cannon_multrec 30084 15.6 21.868 26.484 31.508 36.366 init_scf_loop 11 6.9 0.001 0.005 33.176 33.180 rebuild_ks_matrix 128 8.3 0.001 0.001 32.520 33.084 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.039 32.520 33.083 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.317 29.834 prepare_preconditioner 11 7.9 0.000 0.000 28.796 28.843 make_preconditioner 11 8.9 0.000 0.001 28.796 28.843 qs_ot_get_derivative 117 11.6 0.001 0.003 28.005 28.726 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.788 28.473 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.477 28.008 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.071 23.312 apply_single 128 13.6 0.001 0.001 22.071 23.311 qs_ot_get_p 128 10.4 0.001 0.001 22.405 23.123 ot_diis_step 117 11.6 0.015 0.018 21.868 21.871 multiply_cannon_sync_h2d 30084 15.6 19.088 21.064 19.088 21.064 qs_ot_p2m_diag 83 11.4 0.187 0.216 17.571 17.614 cp_fm_cholesky_invert 11 10.9 16.777 16.790 16.777 16.790 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.403 16.404 make_m2s 5014 13.6 0.088 0.094 14.321 15.613 make_images 5014 14.6 1.172 1.358 14.114 15.407 sum_up_and_integrate 128 10.3 0.116 0.133 14.342 14.371 integrate_v_rspace 128 11.3 0.003 0.003 14.226 14.261 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.206 13.215 cp_fm_redistribute_end 83 14.4 7.725 13.138 7.737 13.141 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.039 13.070 calculate_rho_elec 128 8.7 0.088 0.105 13.038 13.069 cp_fm_diag_elpa_base 83 14.4 5.166 12.614 5.383 13.032 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.096 11.640 init_scf_run 11 5.9 0.000 0.001 11.479 11.481 scf_env_initial_rho_setup 11 6.9 0.001 0.006 11.479 11.481 multiply_cannon_metrocomm4 27577 15.6 0.097 0.112 3.792 10.864 mp_irecv_dv 69486 16.3 3.597 10.474 3.597 10.474 make_images_data 5014 15.6 0.066 0.073 8.722 10.297 dbcsr_mm_accdrv_process 62242 16.2 4.561 5.355 9.105 9.729 hybrid_alltoall_any 5200 16.5 0.342 1.509 7.574 9.551 wfi_extrapolate 11 7.9 0.001 0.003 8.352 8.353 pw_transfer 1547 11.6 0.084 0.095 7.770 7.838 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.547 7.614 grid_integrate_task_list 128 12.3 7.182 7.563 7.182 7.563 density_rs2pw 128 9.7 0.006 0.006 6.736 7.378 cp_fm_cholesky_decompose 22 10.9 7.047 7.120 7.047 7.120 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.218 6.997 fft_wrap_pw1pw2_140 523 13.2 0.468 0.511 6.594 6.678 calculate_dm_sparse 128 9.5 0.001 0.001 6.479 6.625 rs_pw_transfer 1046 11.9 0.014 0.016 5.611 6.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.087 6.097 fft3d_ps 1291 14.7 2.774 2.925 5.928 5.975 mp_sum_l 7870 13.0 4.090 5.911 4.090 5.911 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.340 5.417 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.212 5.361 grid_collocate_task_list 128 9.7 4.906 5.230 4.906 5.230 mp_alltoall_d11v 2415 14.1 4.471 5.015 4.471 5.015 potential_pw2rs 128 12.3 0.015 0.016 4.795 4.826 mp_allgather_i34 2507 14.6 1.575 4.627 1.575 4.627 dbcsr_complete_redistribute 395 12.7 0.772 0.864 3.232 4.086 mp_sum_d 4464 12.1 2.681 4.069 2.681 4.069 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.753000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.090909, yerr=1.831767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 943.734784E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.066 0.141 178.819 178.821 qs_mol_dyn_low 1 2.0 0.003 0.004 178.268 178.282 qs_forces 11 3.9 0.003 0.004 178.162 178.168 qs_energies 11 4.9 0.003 0.007 171.649 171.662 scf_env_do_scf 11 5.9 0.001 0.001 156.144 156.148 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 120.655 120.656 velocity_verlet 10 3.0 0.001 0.002 114.527 114.529 qs_scf_new_mos 117 7.6 0.001 0.001 83.897 84.263 qs_scf_loop_do_ot 117 8.6 0.001 0.002 83.896 84.262 dbcsr_multiply_generic 2507 12.6 0.182 0.190 81.992 83.214 ot_scf_mini 117 9.6 0.004 0.004 79.763 80.200 multiply_cannon 2507 13.6 0.499 0.521 61.676 65.869 multiply_cannon_loop 2507 14.6 0.869 0.896 58.467 61.491 ot_mini 117 10.6 0.001 0.001 42.924 43.361 mp_waitall_1 178456 16.5 25.616 36.125 25.616 36.125 init_scf_loop 11 6.9 0.001 0.002 35.384 35.385 prepare_preconditioner 11 7.9 0.000 0.000 31.358 31.420 make_preconditioner 11 8.9 0.000 0.002 31.358 31.420 rebuild_ks_matrix 128 8.3 0.001 0.001 30.258 30.705 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.257 30.705 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.994 30.342 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.321 27.740 multiply_cannon_multrec 20056 15.6 13.462 16.917 22.143 25.535 multiply_cannon_metrocomm3 20056 15.6 0.059 0.063 15.121 25.406 qs_ot_get_derivative 117 11.6 0.001 0.002 23.166 23.612 qs_ot_get_p 128 10.4 0.001 0.001 22.108 22.661 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.831 20.788 apply_single 128 13.6 0.001 0.001 19.831 20.787 ot_diis_step 117 11.6 0.018 0.018 19.647 19.647 qs_ot_p2m_diag 83 11.4 0.265 0.272 17.699 17.713 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.687 16.688 multiply_cannon_sync_h2d 20056 15.6 14.221 16.302 14.221 16.302 make_m2s 5014 13.6 0.080 0.087 15.041 16.031 make_images 5014 14.6 1.199 1.264 14.810 15.787 cp_fm_cholesky_invert 11 10.9 14.759 14.768 14.759 14.768 sum_up_and_integrate 128 10.3 0.134 0.145 14.298 14.324 integrate_v_rspace 128 11.3 0.003 0.004 14.164 14.190 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.463 13.464 cp_fm_redistribute_end 83 14.4 5.093 13.378 5.111 13.381 cp_fm_diag_elpa_base 83 14.4 7.839 12.778 8.247 13.262 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.130 13.181 calculate_rho_elec 128 8.7 0.132 0.146 13.130 13.180 make_images_data 5014 15.6 0.061 0.067 9.320 10.734 init_scf_run 11 5.9 0.000 0.001 10.438 10.438 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.438 10.438 hybrid_alltoall_any 5200 16.5 0.431 1.982 8.029 9.840 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.125 9.452 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.456 8.980 mp_irecv_dv 50230 16.2 3.334 8.736 3.334 8.736 dbcsr_mm_accdrv_process 41502 16.2 4.422 5.189 8.136 8.225 pw_transfer 1547 11.6 0.084 0.104 7.826 7.925 grid_integrate_task_list 128 12.3 7.329 7.717 7.329 7.717 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.603 7.709 cp_fm_cholesky_decompose 22 10.9 7.500 7.573 7.500 7.573 cp_fm_upper_to_full 105 14.5 5.679 7.449 5.679 7.449 wfi_extrapolate 11 7.9 0.001 0.001 7.449 7.449 density_rs2pw 128 9.7 0.006 0.006 6.547 6.933 fft_wrap_pw1pw2_140 523 13.2 0.478 0.525 6.645 6.761 dbcsr_complete_redistribute 395 12.7 1.182 1.217 4.648 6.440 fft3d_ps 1291 14.7 2.697 2.903 5.871 5.934 calculate_dm_sparse 128 9.5 0.001 0.001 5.724 5.838 rs_pw_transfer 1046 11.9 0.014 0.014 5.249 5.693 grid_collocate_task_list 128 9.7 5.081 5.613 5.081 5.613 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.487 5.494 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.627 5.337 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.435 5.221 mp_alltoall_d11v 2415 14.1 4.385 4.958 4.385 4.958 mp_allgather_i34 2507 14.6 1.725 4.890 1.725 4.890 mp_sum_l 7870 13.0 3.474 4.820 3.474 4.820 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.556 4.696 potential_pw2rs 128 12.3 0.020 0.023 4.644 4.680 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.037 4.072 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.344 4.061 mp_alltoall_i22 716 14.1 1.962 3.856 1.962 3.856 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.782 3.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.821000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=898.727273, yerr=7.186995 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/18/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 4.353791E+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 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 1.147953E+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 1.770320E+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 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4093 57084. MP_Allreduce 11271 1067. 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 0.072 0.121 187.547 187.606 qs_mol_dyn_low 1 2.0 0.003 0.004 186.933 186.946 qs_forces 11 3.9 0.003 0.003 186.825 186.831 qs_energies 11 4.9 0.022 0.044 179.794 179.811 scf_env_do_scf 11 5.9 0.002 0.023 162.885 162.897 velocity_verlet 10 3.0 0.002 0.002 124.202 124.205 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.187 116.188 qs_scf_new_mos 117 7.6 0.001 0.001 80.615 80.914 qs_scf_loop_do_ot 117 8.6 0.001 0.002 80.614 80.913 dbcsr_multiply_generic 2507 12.6 0.191 0.195 79.791 80.537 ot_scf_mini 117 9.6 0.003 0.004 76.094 76.416 multiply_cannon 2507 13.6 0.559 0.599 55.176 58.265 multiply_cannon_loop 2507 14.6 1.185 1.222 51.258 52.952 init_scf_loop 11 6.9 0.001 0.003 46.555 46.558 ot_mini 117 10.6 0.001 0.001 42.498 42.824 prepare_preconditioner 11 7.9 0.000 0.000 42.463 42.499 make_preconditioner 11 8.9 0.000 0.002 42.463 42.499 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.073 41.086 multiply_cannon_multrec 30084 15.6 14.009 19.541 26.127 31.109 rebuild_ks_matrix 128 8.3 0.001 0.001 29.048 29.302 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.048 29.302 mp_waitall_1 153770 16.5 17.750 27.533 17.750 27.533 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.214 26.446 qs_ot_get_derivative 117 11.6 0.001 0.002 22.743 23.065 make_m2s 5014 13.6 0.095 0.098 20.260 21.466 make_images 5014 14.6 1.981 2.308 19.954 21.159 qs_ot_get_p 128 10.4 0.001 0.002 19.806 20.193 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.192 19.761 apply_single 128 13.6 0.001 0.001 19.192 19.761 ot_diis_step 117 11.6 0.018 0.019 19.637 19.638 cp_fm_upper_to_full 105 14.7 11.136 16.473 11.136 16.473 cp_fm_cholesky_invert 11 10.9 16.236 16.245 16.236 16.245 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.652 15.704 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.670 15.638 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.361 14.363 sum_up_and_integrate 128 10.3 0.141 0.157 14.110 14.147 integrate_v_rspace 128 11.3 0.003 0.004 13.969 14.012 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.169 13.194 calculate_rho_elec 128 8.7 0.175 0.190 13.169 13.194 make_images_data 5014 15.6 0.063 0.067 10.887 12.933 dbcsr_complete_redistribute 395 12.7 1.520 1.649 9.049 12.862 multiply_cannon_sync_h2d 30084 15.6 11.642 12.576 11.642 12.576 dbcsr_mm_accdrv_process 62264 16.2 7.467 8.383 11.693 12.190 hybrid_alltoall_any 5200 16.5 0.524 2.213 9.813 11.602 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.650 11.433 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.233 11.236 cp_fm_redistribute_end 83 14.4 1.931 11.161 1.946 11.166 cp_fm_diag_elpa_base 83 14.4 8.603 10.556 9.188 11.039 init_scf_run 11 5.9 0.000 0.001 10.917 10.919 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.917 10.919 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.367 10.008 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.401 9.633 mp_alltoall_i22 716 14.1 5.636 9.317 5.636 9.317 pw_transfer 1547 11.6 0.084 0.098 7.930 8.007 grid_integrate_task_list 128 12.3 7.515 7.973 7.515 7.973 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.706 7.788 cp_fm_cholesky_decompose 22 10.9 7.616 7.747 7.616 7.747 wfi_extrapolate 11 7.9 0.001 0.001 7.640 7.640 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.813 7.381 mp_irecv_dv 76098 16.2 2.669 7.117 2.669 7.117 fft_wrap_pw1pw2_140 523 13.2 0.484 0.490 6.815 6.903 density_rs2pw 128 9.7 0.006 0.006 6.261 6.531 calculate_dm_sparse 128 9.5 0.001 0.001 6.293 6.402 fft3d_ps 1291 14.7 2.800 2.858 5.934 6.008 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.569 5.610 grid_collocate_task_list 128 9.7 5.202 5.562 5.202 5.562 mp_alltoall_d11v 2415 14.1 4.901 5.283 4.901 5.283 rs_pw_transfer 1046 11.9 0.013 0.014 4.585 4.895 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.486 4.578 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.384 4.470 qs_energies_init_hamiltonians 11 5.9 0.013 0.021 4.429 4.430 potential_pw2rs 128 12.3 0.022 0.023 4.297 4.323 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.187 4.231 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.606000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1083.818182, yerr=22.085858 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 1.512735E+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 4010 58095. MP_Allreduce 11022 1174. 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 0.065 0.099 174.817 174.823 qs_mol_dyn_low 1 2.0 0.003 0.003 174.385 174.398 qs_forces 11 3.9 0.003 0.003 174.274 174.280 qs_energies 11 4.9 0.001 0.002 166.932 166.943 scf_env_do_scf 11 5.9 0.001 0.006 149.818 149.829 velocity_verlet 10 3.0 0.003 0.003 113.657 113.661 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.588 113.590 qs_scf_new_mos 117 7.6 0.001 0.001 77.623 77.756 qs_scf_loop_do_ot 117 8.6 0.001 0.001 77.622 77.755 dbcsr_multiply_generic 2507 12.6 0.181 0.185 75.518 76.025 ot_scf_mini 117 9.6 0.003 0.004 73.147 73.250 multiply_cannon 2507 13.6 0.598 0.650 55.181 59.569 multiply_cannon_loop 2507 14.6 0.449 0.461 50.452 51.531 ot_mini 117 10.6 0.001 0.001 40.279 40.407 init_scf_loop 11 6.9 0.000 0.000 36.065 36.067 mp_waitall_1 129618 16.6 27.157 34.708 27.157 34.708 prepare_preconditioner 11 7.9 0.000 0.000 32.153 32.177 make_preconditioner 11 8.9 0.000 0.000 32.153 32.177 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.931 30.187 rebuild_ks_matrix 128 8.3 0.001 0.001 28.832 28.922 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.831 28.922 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.344 26.423 multiply_cannon_multrec 10028 15.6 10.419 14.465 17.740 22.169 qs_ot_get_derivative 117 11.6 0.002 0.002 20.170 20.278 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.935 20.142 ot_diis_step 117 11.6 0.020 0.020 20.043 20.043 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.589 19.917 apply_single 128 13.6 0.001 0.001 19.589 19.916 qs_ot_get_p 128 10.4 0.001 0.001 19.712 19.841 make_m2s 5014 13.6 0.066 0.071 16.558 18.985 make_images 5014 14.6 2.314 2.679 16.254 18.672 cp_fm_cholesky_invert 11 10.9 18.425 18.431 18.425 18.431 qs_ot_p2m_diag 83 11.4 0.496 0.501 15.774 15.795 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.615 14.615 sum_up_and_integrate 128 10.3 0.180 0.190 14.143 14.197 integrate_v_rspace 128 11.3 0.003 0.004 13.962 14.013 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.253 13.302 calculate_rho_elec 128 8.7 0.258 0.269 13.252 13.301 make_images_data 5014 15.6 0.052 0.060 10.078 12.720 multiply_cannon_sync_h2d 10028 15.6 11.513 12.347 11.513 12.347 hybrid_alltoall_any 5200 16.5 0.834 3.798 9.848 12.020 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.486 11.495 cp_fm_diag_elpa_base 83 14.4 11.225 11.305 11.470 11.478 init_scf_run 11 5.9 0.000 0.001 10.351 10.352 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.351 10.351 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.206 8.302 grid_integrate_task_list 128 12.3 7.748 8.171 7.748 8.171 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.626 8.112 cp_fm_cholesky_decompose 22 10.9 7.978 8.087 7.978 8.087 pw_transfer 1547 11.6 0.083 0.089 7.861 7.896 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.640 7.670 mp_allgather_i34 2507 14.6 2.913 7.658 2.913 7.658 dbcsr_mm_accdrv_process 20762 16.1 2.666 3.476 6.953 7.615 wfi_extrapolate 11 7.9 0.001 0.001 7.312 7.312 fft_wrap_pw1pw2_140 523 13.2 0.503 0.525 6.722 6.755 density_rs2pw 128 9.7 0.005 0.006 6.014 6.356 calculate_dm_sparse 128 9.5 0.001 0.001 6.137 6.246 fft3d_ps 1291 14.7 2.728 2.803 5.802 5.849 grid_collocate_task_list 128 9.7 5.482 5.808 5.482 5.808 dbcsr_complete_redistribute 395 12.7 2.149 2.226 5.389 5.777 mp_alltoall_d11v 2415 14.1 4.942 5.725 4.942 5.725 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.320 5.332 multiply_cannon_metrocomm4 7521 15.6 0.023 0.027 1.868 5.240 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.204 5.206 mp_irecv_dv 28860 15.9 1.832 5.162 1.832 5.162 rs_pw_transfer 1046 11.9 0.012 0.013 4.125 4.487 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.226 4.251 potential_pw2rs 128 12.3 0.027 0.028 4.139 4.154 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.551 3.879 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.519 3.849 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.772 3.801 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.614 3.676 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.473 3.508 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.823000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1395.818182, yerr=45.644512 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 11.787405E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.158880E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.131 292.116 292.125 qs_mol_dyn_low 1 2.0 0.003 0.003 291.495 291.508 qs_forces 11 3.9 0.003 0.003 291.403 291.406 qs_energies 11 4.9 0.001 0.002 282.789 282.796 scf_env_do_scf 11 5.9 0.001 0.002 260.908 260.916 velocity_verlet 10 3.0 0.003 0.003 210.216 210.224 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.656 136.657 init_scf_loop 11 6.9 0.000 0.000 123.991 123.994 prepare_preconditioner 11 7.9 0.000 0.000 119.271 119.310 make_preconditioner 11 8.9 0.000 0.000 119.271 119.310 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.665 116.460 qs_scf_new_mos 118 7.6 0.001 0.001 94.266 94.398 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.265 94.397 ot_scf_mini 118 9.6 0.004 0.004 89.337 89.421 dbcsr_multiply_generic 2534 12.6 0.215 0.223 85.007 85.830 cp_fm_upper_to_full 106 14.8 52.881 75.532 52.881 75.532 multiply_cannon 2534 13.6 0.709 0.746 59.792 60.744 multiply_cannon_loop 2534 14.6 0.479 0.485 55.971 57.974 ot_mini 118 10.6 0.001 0.001 45.623 45.723 dbcsr_complete_redistribute 397 12.7 4.064 4.106 29.438 42.100 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.982 38.664 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.560 35.992 mp_alltoall_i22 720 14.1 21.369 34.193 21.369 34.193 mp_waitall_1 106586 16.7 29.358 34.101 29.358 34.101 rebuild_ks_matrix 129 8.3 0.001 0.001 33.839 33.972 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.838 33.971 cp_fm_cholesky_invert 11 10.9 33.660 33.667 33.660 33.667 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.601 31.725 qs_ot_get_p 129 10.4 0.001 0.001 27.890 27.983 qs_ot_get_derivative 118 11.6 0.001 0.002 24.582 24.663 qs_ot_p2m_diag 84 11.4 0.890 0.894 23.603 23.635 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.818 21.819 make_m2s 5068 13.6 0.077 0.081 20.302 21.316 multiply_cannon_metrocomm3 10136 15.6 0.023 0.024 19.502 21.303 ot_diis_step 118 11.6 0.022 0.023 20.988 20.989 make_images 5068 14.6 3.810 3.973 19.819 20.837 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.029 20.433 apply_single 129 13.6 0.001 0.001 20.029 20.432 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.503 18.505 cp_fm_diag_elpa_base 84 14.4 14.070 15.710 18.499 18.501 multiply_cannon_multrec 10136 15.6 10.621 12.443 18.075 18.159 sum_up_and_integrate 129 10.3 0.323 0.325 15.776 15.874 multiply_cannon_sync_h2d 10136 15.6 15.773 15.786 15.773 15.786 integrate_v_rspace 129 11.3 0.004 0.004 15.453 15.548 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.099 15.115 calculate_rho_elec 129 8.7 0.487 0.487 15.099 15.114 hybrid_alltoall_any 5255 16.5 1.309 3.055 11.068 12.947 make_images_data 5068 15.6 0.060 0.066 10.922 12.865 init_scf_run 11 5.9 0.000 0.001 11.917 11.917 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.917 11.917 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.745 9.809 cp_fm_cholesky_decompose 22 10.9 8.911 9.015 8.911 9.015 dbcsr_mm_accdrv_process 20954 16.1 3.810 5.598 7.214 9.009 wfi_extrapolate 11 7.9 0.001 0.001 8.808 8.808 grid_integrate_task_list 129 12.3 8.606 8.787 8.606 8.787 pw_transfer 1559 11.6 0.091 0.093 8.430 8.449 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.194 8.212 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.972 7.973 fft_wrap_pw1pw2_140 527 13.2 0.544 0.547 7.252 7.278 mp_alltoall_d11v 2429 14.1 7.046 7.185 7.046 7.185 calculate_dm_sparse 129 9.5 0.001 0.001 6.780 6.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.464 6.529 grid_collocate_task_list 129 9.7 6.389 6.427 6.389 6.427 fft3d_ps 1301 14.7 2.779 2.793 6.225 6.240 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.134 6.184 density_rs2pw 129 9.7 0.005 0.005 6.071 6.099 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.125000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2847.545455, yerr=164.496125 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 1.260114E+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 19390592. 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 0.025 0.034 84.930 84.931 qs_energies 1 2.0 0.000 0.000 84.487 84.509 ls_scf 1 3.0 0.000 0.000 83.559 83.581 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.568 72.708 multiply_cannon 111 7.7 0.017 0.020 55.778 56.870 multiply_cannon_loop 111 8.7 0.210 0.226 52.407 53.793 ls_scf_main 1 4.0 0.000 0.000 52.271 52.283 density_matrix_trs4 2 5.0 0.002 0.003 46.777 46.865 ls_scf_init_scf 1 4.0 0.000 0.000 28.220 28.222 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.161 27.204 mp_waitall_1 11316 10.9 22.495 25.294 22.495 25.294 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.918 24.936 multiply_cannon_multrec 2664 9.7 8.152 8.883 15.532 17.249 multiply_cannon_sync_h2d 2664 9.7 13.606 15.879 13.606 15.879 make_m2s 222 7.7 0.008 0.010 13.137 13.507 make_images 222 8.7 0.099 0.109 13.115 13.485 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.677 12.536 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.435 8.593 make_images_data 222 9.7 0.004 0.005 7.691 8.225 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.620 6.999 7.944 hybrid_alltoall_any 227 10.6 0.215 1.843 6.654 7.850 dbcsr_mm_accdrv_process_sort 4760 11.4 6.280 7.169 6.280 7.169 calculate_norms 4752 9.8 5.510 6.259 5.510 6.259 mp_sum_l 807 5.4 3.112 5.263 3.112 5.263 apply_matrix_preconditioner 6 5.3 0.001 0.003 5.094 5.215 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.296 4.027 make_images_sizes 222 9.7 0.000 0.000 0.686 3.601 mp_alltoall_i44 222 10.7 0.685 3.600 0.685 3.600 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.055 3.530 mp_irecv_dv 6231 10.9 2.037 3.503 2.037 3.503 arnoldi_extremal 4 6.8 0.000 0.000 3.248 3.273 arnoldi_normal_ev 4 7.8 0.001 0.004 3.248 3.273 build_subspace 16 8.4 0.009 0.012 3.154 3.157 ls_scf_post 1 4.0 0.000 0.000 3.067 3.090 ls_scf_store_result 1 5.0 0.000 0.000 2.873 2.920 dbcsr_special_finalize 555 9.7 0.005 0.006 2.406 2.763 dbcsr_merge_single_wm 555 10.7 0.462 0.572 2.398 2.755 make_images_pack 222 9.7 2.208 2.637 2.210 2.638 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.356 2.603 dbcsr_sort_data 658 11.4 2.193 2.517 2.193 2.517 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.469 2.073 2.471 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.226 2.310 buffer_matrices_ensure_size 222 8.7 1.761 2.091 1.761 2.091 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.739 1.743 rebuild_ks_matrix 3 7.3 0.000 0.000 1.730 1.734 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.729 1.733 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.931000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 2.100056E+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 0.033 0.046 90.983 90.984 qs_energies 1 2.0 0.000 0.000 90.487 90.492 ls_scf 1 3.0 0.000 0.001 89.173 89.178 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.234 75.599 multiply_cannon 111 7.7 0.028 0.036 53.125 56.961 ls_scf_main 1 4.0 0.000 0.000 54.838 54.845 multiply_cannon_loop 111 8.7 0.117 0.123 49.864 53.367 density_matrix_trs4 2 5.0 0.002 0.003 49.192 49.374 ls_scf_init_scf 1 4.0 0.000 0.002 30.815 30.816 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.659 29.784 mp_waitall_1 9246 10.9 21.139 29.258 21.139 29.258 multiply_cannon_multrec 1332 9.7 13.103 16.951 22.385 27.393 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.174 27.187 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.545 20.277 make_m2s 222 7.7 0.006 0.007 15.502 16.193 make_images 222 8.7 1.577 1.973 15.472 16.165 dbcsr_mm_accdrv_process 4041 10.4 0.251 0.447 8.886 10.453 dbcsr_mm_accdrv_process_sort 4041 11.4 8.474 10.006 8.474 10.006 make_images_data 222 9.7 0.004 0.004 8.950 9.768 hybrid_alltoall_any 227 10.6 0.520 2.481 8.345 9.405 mp_sum_l 807 5.4 5.304 7.905 5.304 7.905 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.199 7.702 mp_irecv_dv 3311 11.0 3.179 7.636 3.179 7.636 calculate_norms 2376 9.8 6.021 6.830 6.021 6.830 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.159 6.518 multiply_cannon_sync_h2d 1332 9.7 4.878 6.304 4.878 6.304 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.042 5.245 arnoldi_extremal 4 6.8 0.000 0.000 4.678 4.699 arnoldi_normal_ev 4 7.8 0.001 0.005 4.678 4.699 build_subspace 16 8.4 0.014 0.021 4.420 4.422 ls_scf_post 1 4.0 0.000 0.000 3.520 3.525 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.179 3.403 ls_scf_store_result 1 5.0 0.000 0.000 3.249 3.341 dbcsr_matrix_vector_mult_local 304 10.0 2.780 3.236 2.782 3.238 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.527 2.623 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.263 2.552 mp_allgather_i34 111 8.7 0.962 2.417 0.962 2.417 make_images_pack 222 9.7 2.022 2.362 2.024 2.365 dbcsr_sort_data 436 11.2 1.825 2.066 1.825 2.066 dbcsr_data_new 4174 10.1 1.611 1.842 1.611 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.984000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1790.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 2.734760E+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 0.085 0.181 93.622 93.623 qs_energies 1 2.0 0.000 0.000 92.791 92.812 ls_scf 1 3.0 0.000 0.001 91.382 91.420 dbcsr_multiply_generic 111 6.7 0.016 0.023 76.162 76.500 ls_scf_main 1 4.0 0.002 0.042 57.228 57.233 multiply_cannon 111 7.7 0.039 0.077 52.794 56.889 multiply_cannon_loop 111 8.7 0.100 0.105 49.245 53.077 density_matrix_trs4 2 5.0 0.002 0.003 51.297 51.482 mp_waitall_1 7374 11.0 23.752 33.711 23.752 33.711 ls_scf_init_scf 1 4.0 0.000 0.001 30.606 30.622 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.478 29.550 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.011 27.025 multiply_cannon_multrec 888 9.7 12.530 15.288 21.130 24.481 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.073 22.881 make_m2s 222 7.7 0.006 0.007 16.794 18.007 make_images 222 8.7 1.978 2.289 16.756 17.971 make_images_data 222 9.7 0.003 0.005 9.649 10.708 hybrid_alltoall_any 227 10.6 0.622 2.859 9.358 10.488 dbcsr_mm_accdrv_process 3754 10.4 0.258 0.414 8.113 9.320 dbcsr_mm_accdrv_process_sort 3754 11.4 7.743 8.906 7.743 8.906 mp_sum_l 807 5.4 5.024 8.603 5.024 8.603 multiply_cannon_sync_h2d 888 9.7 6.050 7.249 6.050 7.249 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.475 6.921 mp_irecv_dv 2335 11.1 2.459 6.876 2.459 6.876 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.641 6.766 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.796 6.672 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.950 5.136 arnoldi_extremal 4 6.8 0.000 0.000 5.092 5.110 arnoldi_normal_ev 4 7.8 0.001 0.005 5.091 5.110 build_subspace 16 8.4 0.014 0.021 4.793 4.800 calculate_norms 1584 9.8 4.263 4.576 4.263 4.576 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.486 3.786 mp_allgather_i34 111 8.7 1.396 3.731 1.396 3.731 dbcsr_matrix_vector_mult_local 304 10.0 3.066 3.610 3.068 3.612 ls_scf_post 1 4.0 0.003 0.024 3.548 3.572 ls_scf_store_result 1 5.0 0.000 0.000 3.294 3.371 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.762 2.884 make_images_pack 222 9.7 1.815 2.135 1.818 2.138 dbcsr_sort_data 325 11.1 1.899 2.092 1.899 2.092 dbcsr_data_release 9322 10.9 1.312 1.933 1.312 1.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.623000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2163.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 3.336872E+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 0.052 0.068 97.214 97.216 qs_energies 1 2.0 0.000 0.000 95.922 95.927 ls_scf 1 3.0 0.000 0.001 94.280 94.285 dbcsr_multiply_generic 111 6.7 0.017 0.023 78.276 78.543 ls_scf_main 1 4.0 0.000 0.001 58.567 58.568 multiply_cannon 111 7.7 0.059 0.159 51.708 55.861 density_matrix_trs4 2 5.0 0.002 0.003 52.538 52.641 multiply_cannon_loop 111 8.7 0.115 0.126 46.715 49.523 ls_scf_init_scf 1 4.0 0.000 0.002 32.450 32.452 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.309 31.379 mp_waitall_1 6438 11.0 22.819 29.678 22.819 29.678 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.836 28.849 multiply_cannon_multrec 1332 9.7 14.196 17.125 22.112 25.255 make_m2s 222 7.7 0.007 0.007 21.157 22.546 make_images 222 8.7 3.139 3.602 21.107 22.498 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.307 16.526 make_images_data 222 9.7 0.004 0.005 11.831 13.472 hybrid_alltoall_any 227 10.6 0.795 3.764 11.256 13.006 dbcsr_mm_accdrv_process 3641 10.4 0.236 0.412 7.548 9.058 dbcsr_mm_accdrv_process_sort 3641 11.4 7.172 8.640 7.172 8.640 mp_sum_l 807 5.4 3.850 6.974 3.850 6.974 multiply_cannon_sync_h2d 1332 9.7 5.611 6.296 5.611 6.296 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.073 5.954 mp_irecv_dv 3229 10.9 2.049 5.878 2.049 5.878 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.974 5.445 arnoldi_extremal 4 6.8 0.000 0.000 5.229 5.245 arnoldi_normal_ev 4 7.8 0.001 0.004 5.229 5.245 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.505 5.096 build_subspace 16 8.4 0.014 0.021 4.889 4.896 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.546 4.708 calculate_norms 2376 9.8 4.174 4.568 4.174 4.568 mp_allgather_i34 111 8.7 2.113 4.357 2.113 4.357 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.610 3.916 dbcsr_matrix_vector_mult_local 304 10.0 3.241 3.742 3.243 3.744 dbcsr_sort_data 658 11.4 3.097 3.425 3.097 3.425 ls_scf_post 1 4.0 0.001 0.031 3.263 3.269 dbcsr_special_finalize 555 9.7 0.006 0.007 2.855 3.251 dbcsr_merge_single_wm 555 10.7 0.538 0.666 2.846 3.243 ls_scf_store_result 1 5.0 0.000 0.000 2.996 3.063 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.904 2.956 dbcsr_data_release 10477 10.7 1.581 2.395 1.581 2.395 dbcsr_finalize 304 7.8 0.050 0.061 1.791 1.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.216000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2738.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 4.628275E+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 0.044 0.064 99.070 99.070 qs_energies 1 2.0 0.000 0.000 98.295 98.300 ls_scf 1 3.0 0.000 0.000 96.381 96.385 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.098 78.322 ls_scf_main 1 4.0 0.000 0.000 62.030 62.030 multiply_cannon 111 7.7 0.085 0.214 55.397 61.124 density_matrix_trs4 2 5.0 0.002 0.003 55.040 55.157 multiply_cannon_loop 111 8.7 0.070 0.076 50.775 52.748 mp_waitall_1 5481 11.0 26.250 32.321 26.250 32.321 ls_scf_init_scf 1 4.0 0.000 0.000 30.700 30.706 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.524 29.572 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.357 27.368 multiply_cannon_multrec 444 9.7 14.043 16.415 21.130 23.451 make_m2s 222 7.7 0.004 0.005 17.641 20.226 make_images 222 8.7 3.721 4.432 17.579 20.166 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.910 16.085 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.286 15.049 make_images_data 222 9.7 0.003 0.004 9.849 12.460 hybrid_alltoall_any 227 10.6 0.786 3.749 9.649 12.194 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.343 6.793 7.959 dbcsr_mm_accdrv_process_sort 3003 11.4 6.479 7.615 6.479 7.615 multiply_cannon_sync_h2d 444 9.7 6.550 7.551 6.550 7.551 mp_allgather_i34 111 8.7 2.855 6.999 2.855 6.999 arnoldi_extremal 4 6.8 0.000 0.000 5.857 5.870 arnoldi_normal_ev 4 7.8 0.002 0.005 5.857 5.870 mp_sum_l 807 5.4 3.115 5.597 3.115 5.597 build_subspace 16 8.4 0.015 0.019 5.463 5.471 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.627 4.811 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.715 4.543 mp_irecv_dv 1241 11.2 1.697 4.517 1.697 4.517 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.178 4.389 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.146 4.345 dbcsr_matrix_vector_mult_local 304 10.0 3.724 4.163 3.727 4.165 calculate_norms 792 9.8 3.539 3.676 3.539 3.676 ls_scf_post 1 4.0 0.000 0.000 3.651 3.656 make_images_sizes 222 9.7 0.000 0.000 1.157 3.483 mp_alltoall_i44 222 10.7 1.156 3.483 1.156 3.483 ls_scf_store_result 1 5.0 0.000 0.000 3.416 3.480 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.360 3.444 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.317 dbcsr_merge_all 275 8.9 0.474 0.534 2.045 2.153 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.070000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3640.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9400c5a650e170a3f6cad4de0da2f9ce09dfe117_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 8.675152E+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 0.093 0.112 107.037 107.038 qs_energies 1 2.0 0.000 0.000 105.602 105.618 ls_scf 1 3.0 0.000 0.000 102.714 102.729 dbcsr_multiply_generic 111 6.7 0.024 0.026 76.864 76.974 ls_scf_main 1 4.0 0.000 0.000 65.302 65.302 density_matrix_trs4 2 5.0 0.002 0.003 56.470 56.534 multiply_cannon 111 7.7 0.155 0.251 49.679 51.857 multiply_cannon_loop 111 8.7 0.068 0.073 46.301 46.972 ls_scf_init_scf 1 4.0 0.000 0.000 33.754 33.754 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.422 32.428 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.577 29.586 mp_waitall_1 4569 11.1 21.721 24.918 21.721 24.918 make_m2s 222 7.7 0.005 0.005 23.636 24.549 make_images 222 8.7 4.591 4.978 23.530 24.441 multiply_cannon_multrec 444 9.7 17.933 18.546 22.564 23.168 hybrid_alltoall_any 227 10.6 1.660 3.613 12.667 15.406 make_images_data 222 9.7 0.003 0.004 12.846 15.219 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.257 10.843 multiply_cannon_sync_h2d 444 9.7 8.853 8.907 8.853 8.907 arnoldi_extremal 4 6.8 0.000 0.000 7.426 7.436 arnoldi_normal_ev 4 7.8 0.003 0.009 7.426 7.436 build_subspace 16 8.4 0.026 0.037 6.868 6.881 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.518 5.667 dbcsr_matrix_vector_mult_local 304 10.0 5.093 5.378 5.095 5.380 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.034 5.281 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.784 4.876 dbcsr_mm_accdrv_process 1814 10.4 0.251 0.322 4.471 4.598 dbcsr_mm_accdrv_process_sort 1814 11.4 4.173 4.305 4.173 4.305 ls_scf_post 1 4.0 0.000 0.000 3.659 3.674 make_images_sizes 222 9.7 0.000 0.000 1.457 3.511 mp_alltoall_i44 222 10.7 1.457 3.511 1.457 3.511 ls_scf_store_result 1 5.0 0.000 0.000 3.410 3.419 calculate_norms 792 9.8 3.239 3.278 3.239 3.278 mp_allgather_i34 111 8.7 1.030 3.239 1.030 3.239 dbcsr_finalize 304 7.8 0.082 0.089 3.081 3.128 dbcsr_merge_all 275 8.9 0.891 0.915 2.866 2.908 dbcsr_complete_redistribute 5 7.6 1.443 1.477 2.781 2.876 dbcsr_data_release 12724 10.6 2.338 2.871 2.338 2.871 qs_energies_init_hamiltonians 1 3.0 0.007 0.010 2.858 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.428 2.529 dbcsr_sort_data 325 11.1 2.438 2.502 2.438 2.502 dbcsr_new_transposed 4 7.5 0.245 0.254 2.313 2.325 dbcsr_frobenius_norm 74 6.6 2.056 2.132 2.187 2.217 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.201 2.203 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.200 dbcsr_add_anytype 103 7.2 0.859 0.892 2.130 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.038000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6838.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9400c5a650e170a3f6cad4de0da2f9ce09dfe117 Summary: empty Status: OK