=== 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: 2940f7cc1a499533767160797f0663fb9d581809 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (19.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/01 job id: 42251121 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/02 job id: 42251122 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/03 job id: 42251123 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/04 job id: 42251131 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/05 job id: 42251132 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/06 job id: 42251133 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/07 job id: 42251134 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/08 job id: 42251136 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/09 job id: 42251138 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/10 job id: 42251139 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/11 job id: 42251140 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/12 job id: 42251142 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/13 job id: 42251143 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/14 job id: 42251144 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/15 job id: 42251145 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/16 job id: 42251146 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/17 job id: 42251147 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/18 job id: 42251150 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/19 job id: 42251151 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/20 job id: 42251152 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/21 job id: 42251153 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/22 job id: 42251154 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/23 job id: 42251155 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/24 job id: 42251156 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/25 job id: 42251157 --- 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/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/26 job id: 42251158 --- 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/2940f7cc1a499533767160797f0663fb9d581809_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 0.033 132.589 132.590 farming_run 1 2.0 132.059 132.060 132.562 132.565 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.475236E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.024 114.612 114.612 qs_energies 1 2.0 0.000 0.000 114.343 114.344 mp2_main 1 3.0 0.000 0.000 112.463 112.465 mp2_gpw_main 1 4.0 0.019 0.025 111.609 111.611 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.811 93.216 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.006 55.527 55.932 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.829 47.181 get_2c_integrals 1 6.0 0.000 0.000 36.518 37.112 integrate_v_rspace 273 8.0 0.435 0.448 25.174 30.354 pw_transfer 6555 10.6 0.376 0.386 27.328 27.803 grid_integrate_task_list 273 9.0 20.984 26.669 20.984 26.669 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.012 26.470 fft_wrap_pw1pw2_100 2178 12.4 1.158 1.238 23.563 24.039 compute_2c_integrals 1 7.0 0.002 0.002 19.193 19.194 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.843 18.923 mp2_eri_2c_integrate_gpw 1 9.0 2.423 2.468 18.841 18.919 rpa_ri_compute_en 1 5.0 0.000 0.000 18.688 18.889 cp_fm_cholesky_decompose 12 8.2 17.272 17.893 17.272 17.893 cholesky_decomp 1 7.0 0.000 0.000 16.165 16.781 fft3d_s 5443 13.4 16.155 16.623 16.177 16.644 ao_to_mo_and_store_B_mult_1 272 7.0 10.867 15.592 10.867 15.592 calculate_wavefunction 272 8.0 5.422 5.572 12.563 13.165 rpa_num_int 1 6.0 0.000 0.000 10.541 10.541 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.470 10.516 calc_mat_Q 8 8.0 0.000 0.000 9.315 9.435 contract_S_to_Q 8 9.0 0.000 0.000 8.737 8.855 calc_potential_gpw 544 9.5 0.005 0.005 8.291 8.606 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.183 8.432 parallel_gemm_fm 14 9.1 0.000 0.000 8.324 8.420 parallel_gemm_fm_cosma 14 10.1 8.324 8.420 8.324 8.420 potential_pw2rs 545 10.0 0.108 0.111 7.634 8.319 create_integ_mat 1 6.0 0.022 0.026 7.647 7.647 collocate_single_gaussian 272 10.0 0.041 0.043 7.417 7.601 array2fm 1 7.0 0.000 0.000 6.691 7.107 pw_scatter_s 2720 13.7 4.454 4.592 4.454 4.592 pw_gather_s 2722 13.2 3.872 4.221 3.872 4.221 array2fm_buffer_send 1 8.0 2.960 3.111 2.960 3.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.607683, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 0.032 0.041 397.784 397.785 farming_run 1 2.0 396.792 396.802 397.731 397.735 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223422E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.044 209.768 209.769 qs_energies 1 2.0 0.000 0.000 209.420 209.425 scf_env_do_scf 1 3.0 0.000 0.000 106.230 106.230 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.334 105.342 rebuild_ks_matrix 4 6.0 0.000 0.000 105.332 105.340 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.065 105.332 105.340 hfx_ks_matrix 4 8.0 0.001 0.001 104.959 104.962 integrate_four_center 4 9.0 0.143 0.457 104.958 104.961 mp2_main 1 3.0 0.000 0.002 102.907 102.913 mp2_gpw_main 1 4.0 0.032 0.044 101.932 101.939 integrate_four_center_main 4 10.0 0.132 0.600 96.470 99.036 integrate_four_center_bin 263 11.0 96.339 99.005 96.339 99.005 init_scf_loop 1 4.0 0.000 0.000 92.027 92.027 mp2_ri_gpw_compute_in 1 5.0 0.066 0.088 75.054 76.119 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.013 54.571 55.633 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.159 42.141 47.220 integrate_v_rspace 95 8.0 0.396 0.566 28.489 33.399 pw_transfer 2240 10.6 0.143 0.162 29.956 30.321 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.940 29.372 grid_integrate_task_list 95 9.0 23.784 28.874 23.784 28.874 mp2_ri_gpw_compute_en 1 5.0 0.056 0.064 26.726 28.455 ao_to_mo_and_store_B_mult_1 91 7.0 10.738 28.218 10.738 28.218 fft_wrap_pw1pw2_100 730 12.4 1.312 1.464 26.681 27.092 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.887 25.020 25.028 get_2c_integrals 1 6.0 0.000 0.003 20.373 20.413 compute_2c_integrals 1 7.0 0.003 0.005 19.349 19.360 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.997 19.202 mp2_eri_2c_integrate_gpw 1 9.0 1.749 1.885 18.995 19.200 fft3d_s 1823 13.4 18.427 18.655 18.440 18.668 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.201 14.201 calculate_wavefunction 91 8.0 2.028 2.063 9.758 9.994 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.587 8.759 9.394 potential_pw2rs 186 10.0 0.034 0.036 8.657 9.163 local_gemm 172 8.0 8.202 8.807 8.202 8.807 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.303 8.605 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.518 8.033 8.438 calc_potential_gpw 182 9.5 0.002 0.002 7.952 8.165 collocate_single_gaussian 91 10.0 0.017 0.024 7.942 8.144 mp_sendrecv_dm3 2068 8.0 6.068 6.464 6.068 6.464 mp2_ri_gpw_compute_en_ener 172 7.0 6.347 6.430 6.347 6.430 mp_sync 38 10.4 3.151 5.578 3.151 5.578 pw_gather_s 912 13.2 4.905 5.401 4.905 5.401 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.929236, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1506.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.850240E+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 0.053 0.105 53.881 53.882 qs_mol_dyn_low 1 2.0 0.003 0.004 53.495 53.502 qs_forces 11 3.9 0.003 0.010 52.862 52.863 qs_energies 11 4.9 0.002 0.009 51.391 51.407 scf_env_do_scf 11 5.9 0.000 0.001 45.138 45.139 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.043 43.043 dbcsr_multiply_generic 2286 12.5 0.092 0.097 34.398 34.905 qs_scf_new_mos 108 7.5 0.000 0.000 32.897 33.190 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.897 33.189 ot_scf_mini 108 9.5 0.002 0.003 31.223 31.409 multiply_cannon 2286 13.5 0.194 0.205 26.727 28.409 multiply_cannon_loop 2286 14.5 1.499 1.570 25.977 27.688 velocity_verlet 10 3.0 0.001 0.001 26.484 26.490 ot_mini 108 10.5 0.001 0.001 20.071 20.309 qs_ot_get_derivative 108 11.5 0.001 0.001 17.059 17.247 mp_waitall_1 267858 16.1 9.204 15.964 9.204 15.964 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.346 13.817 multiply_cannon_multrec 54864 15.5 4.173 6.458 7.682 11.135 rebuild_ks_matrix 119 8.3 0.000 0.000 7.878 8.005 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.878 8.004 multiply_cannon_sync_h2d 54864 15.5 5.840 7.551 5.840 7.551 mp_sum_l 7207 12.9 5.620 7.393 5.620 7.393 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.956 7.072 qs_ot_get_p 119 10.4 0.001 0.001 6.448 6.748 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.035 6.451 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.690 5.781 init_scf_run 11 5.9 0.000 0.001 4.945 4.945 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.944 4.945 dbcsr_mm_accdrv_process 76910 16.1 1.195 1.860 3.432 4.799 sum_up_and_integrate 119 10.3 0.012 0.015 4.517 4.525 integrate_v_rspace 119 11.3 0.002 0.002 4.505 4.512 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.204 4.316 calculate_rho_elec 119 8.7 0.011 0.016 4.204 4.316 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.514 3.568 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.904 3.545 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.927 3.231 apply_single 119 13.6 0.000 0.000 2.927 3.231 calculate_dm_sparse 119 9.5 0.000 0.000 2.996 3.140 rs_pw_transfer 974 11.9 0.012 0.013 2.820 2.921 jit_kernel_multiply 13 15.8 2.174 2.877 2.174 2.877 calculate_first_density_matrix 1 7.0 0.001 0.003 2.743 2.749 acc_transpose_blocks 54864 15.5 0.223 0.244 2.016 2.735 ot_diis_step 108 11.5 0.006 0.006 2.725 2.726 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.675 2.676 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.463 2.531 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.425 2.426 cp_fm_redistribute_end 50 14.0 2.200 2.399 2.204 2.401 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.390 2.392 density_rs2pw 119 9.7 0.004 0.004 2.271 2.384 cp_fm_diag_elpa_base 50 14.0 0.195 2.350 0.195 2.363 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.101 2.150 wfi_extrapolate 11 7.9 0.001 0.001 2.140 2.140 grid_integrate_task_list 119 12.3 2.028 2.127 2.028 2.127 mp_sum_d 4129 12.0 1.407 2.117 1.407 2.117 init_scf_loop 11 6.9 0.001 0.004 2.075 2.076 potential_pw2rs 119 12.3 0.004 0.004 1.875 1.895 make_m2s 4572 13.5 0.053 0.056 1.784 1.825 pw_transfer 1439 11.6 0.051 0.055 1.733 1.808 make_images 4572 14.5 0.132 0.138 1.703 1.743 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.658 1.735 acc_transpose_blocks_kernels 54864 16.5 0.252 0.388 1.038 1.622 mp_alltoall_d11v 2130 13.8 1.399 1.564 1.399 1.564 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.491 1.523 fft3d_ps 1201 14.6 0.357 0.460 1.436 1.510 grid_collocate_task_list 119 9.7 1.290 1.348 1.290 1.348 fft_wrap_pw1pw2_140 487 13.2 0.079 0.093 1.269 1.345 mp_waitany 12084 13.8 1.235 1.344 1.235 1.344 jit_kernel_transpose 5 15.5 0.786 1.256 0.786 1.256 dbcsr_dot_sd 1205 11.9 0.047 0.059 0.784 1.190 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.882000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.473152E+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 606650. 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 0.018 0.060 39.123 39.124 qs_mol_dyn_low 1 2.0 0.004 0.010 38.895 38.903 qs_forces 11 3.9 0.011 0.017 38.826 38.828 qs_energies 11 4.9 0.001 0.002 37.123 37.127 scf_env_do_scf 11 5.9 0.001 0.002 32.050 32.051 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.493 29.494 dbcsr_multiply_generic 2286 12.5 0.099 0.103 21.550 21.921 qs_scf_new_mos 108 7.5 0.001 0.001 20.091 20.341 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.091 20.340 ot_scf_mini 108 9.5 0.002 0.003 19.172 19.351 velocity_verlet 10 3.0 0.001 0.002 18.533 18.534 multiply_cannon 2286 13.5 0.210 0.221 16.378 17.953 multiply_cannon_loop 2286 14.5 0.906 0.973 15.187 16.652 ot_mini 108 10.5 0.001 0.001 11.958 12.199 mp_waitall_1 217478 16.2 6.152 11.313 6.152 11.313 qs_ot_get_derivative 108 11.5 0.001 0.001 9.479 9.658 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.167 9.520 multiply_cannon_multrec 27432 15.5 1.971 4.276 5.853 8.648 rebuild_ks_matrix 119 8.3 0.000 0.000 7.136 7.278 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.136 7.278 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.302 6.432 dbcsr_mm_accdrv_process 47894 16.0 3.196 5.196 3.812 5.659 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.681 4.529 qs_ot_get_p 119 10.4 0.001 0.001 4.238 4.468 sum_up_and_integrate 119 10.3 0.024 0.026 4.175 4.180 integrate_v_rspace 119 11.3 0.002 0.002 4.151 4.158 mp_sum_l 7207 12.9 2.139 4.134 2.139 4.134 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.053 4.120 apply_single 119 13.6 0.000 0.000 3.053 4.120 init_scf_run 11 5.9 0.000 0.001 3.859 3.860 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.859 3.859 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.636 3.671 calculate_rho_elec 119 8.7 0.021 0.024 3.635 3.670 make_m2s 4572 13.5 0.052 0.053 2.622 2.877 multiply_cannon_sync_h2d 27432 15.5 2.197 2.803 2.197 2.803 make_images 4572 14.5 0.200 0.238 2.534 2.787 qs_ot_p2m_diag 50 11.0 0.008 0.013 2.713 2.732 rs_pw_transfer 974 11.9 0.010 0.011 2.605 2.698 init_scf_loop 11 6.9 0.000 0.001 2.528 2.529 ot_diis_step 108 11.5 0.011 0.014 2.421 2.422 calculate_first_density_matrix 1 7.0 0.001 0.001 2.316 2.317 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.189 2.282 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.269 2.270 calculate_dm_sparse 119 9.5 0.000 0.001 2.065 2.140 density_rs2pw 119 9.7 0.004 0.004 2.007 2.094 grid_integrate_task_list 119 12.3 1.839 1.937 1.839 1.937 potential_pw2rs 119 12.3 0.006 0.007 1.920 1.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.915 1.917 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.916 1.916 cp_fm_redistribute_end 50 14.0 1.580 1.891 1.584 1.893 cp_fm_diag_elpa_base 50 14.0 0.300 1.835 0.307 1.865 pw_transfer 1439 11.6 0.063 0.066 1.756 1.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.661 1.701 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.666 1.698 prepare_preconditioner 11 7.9 0.000 0.000 1.621 1.648 make_preconditioner 11 8.9 0.000 0.000 1.621 1.648 make_images_data 4572 15.5 0.045 0.052 1.181 1.633 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.519 1.576 jit_kernel_multiply 8 16.1 0.565 1.506 0.565 1.506 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.021 1.504 wfi_extrapolate 11 7.9 0.001 0.001 1.493 1.494 acc_transpose_blocks 27432 15.5 0.107 0.111 1.187 1.488 fft3d_ps 1201 14.6 0.495 0.549 1.374 1.403 mp_allgather_i34 2286 14.5 0.632 1.390 0.632 1.390 mp_alltoall_d11v 2130 13.8 1.228 1.386 1.228 1.386 grid_collocate_task_list 119 9.7 1.240 1.321 1.240 1.321 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.241 1.273 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.222 1.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.265 mp_sum_d 4129 12.0 0.599 1.041 0.599 1.041 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.895 0.909 make_images_sizes 4572 15.5 0.005 0.005 0.630 0.889 mp_alltoall_i44 4572 16.5 0.625 0.884 0.625 0.884 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.660 0.868 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.765 0.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.124000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=1.149919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 525.844480E+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 0.012 0.028 32.018 32.018 qs_mol_dyn_low 1 2.0 0.003 0.003 31.796 31.804 qs_forces 11 3.9 0.002 0.002 31.727 31.727 qs_energies 11 4.9 0.001 0.001 30.183 30.186 scf_env_do_scf 11 5.9 0.000 0.001 25.549 25.549 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.007 23.007 dbcsr_multiply_generic 2286 12.5 0.090 0.092 16.534 16.611 velocity_verlet 10 3.0 0.001 0.001 15.296 15.297 qs_scf_new_mos 108 7.5 0.001 0.001 14.907 14.922 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.907 14.921 ot_scf_mini 108 9.5 0.002 0.002 14.146 14.158 multiply_cannon 2286 13.5 0.206 0.215 13.150 13.896 multiply_cannon_loop 2286 14.5 0.641 0.664 12.308 13.075 ot_mini 108 10.5 0.001 0.001 8.794 8.811 qs_ot_get_derivative 108 11.5 0.001 0.001 7.290 7.303 multiply_cannon_multrec 18288 15.5 1.924 2.889 6.767 7.062 rebuild_ks_matrix 119 8.3 0.000 0.000 6.399 6.430 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.398 6.430 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.670 5.700 dbcsr_mm_accdrv_process 38222 16.0 4.191 5.464 4.758 5.586 mp_waitall_1 169478 16.3 3.086 4.307 3.086 4.307 sum_up_and_integrate 119 10.3 0.030 0.031 4.001 4.006 integrate_v_rspace 119 11.3 0.002 0.003 3.971 3.980 init_scf_run 11 5.9 0.000 0.001 3.477 3.477 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.477 3.477 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.730 3.331 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.312 3.322 calculate_rho_elec 119 8.7 0.030 0.031 3.311 3.322 qs_ot_get_p 119 10.4 0.001 0.001 3.152 3.175 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.435 2.527 init_scf_loop 11 6.9 0.000 0.000 2.525 2.525 rs_pw_transfer 974 11.9 0.009 0.009 2.324 2.380 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.012 2.340 apply_single 119 13.6 0.000 0.000 2.012 2.339 calculate_first_density_matrix 1 7.0 0.001 0.001 2.258 2.260 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.061 2.066 make_m2s 4572 13.5 0.044 0.045 1.874 2.023 density_rs2pw 119 9.7 0.004 0.004 1.906 1.959 grid_integrate_task_list 119 12.3 1.811 1.942 1.811 1.942 make_images 4572 14.5 0.191 0.205 1.789 1.936 calculate_dm_sparse 119 9.5 0.000 0.000 1.850 1.862 pw_transfer 1439 11.6 0.062 0.066 1.786 1.798 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.790 1.791 potential_pw2rs 119 12.3 0.007 0.008 1.759 1.768 prepare_preconditioner 11 7.9 0.000 0.000 1.747 1.749 make_preconditioner 11 8.9 0.000 0.000 1.747 1.749 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.717 1.719 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.697 1.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.598 1.688 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.593 1.601 jit_kernel_multiply 8 16.2 0.516 1.566 0.516 1.566 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.557 1.562 multiply_cannon_sync_h2d 18288 15.5 1.408 1.557 1.408 1.557 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.525 1.529 cp_fm_redistribute_end 50 14.0 1.136 1.505 1.139 1.507 mp_sum_l 7207 12.9 1.130 1.502 1.130 1.502 ot_diis_step 108 11.5 0.011 0.011 1.482 1.483 cp_fm_diag_elpa_base 50 14.0 0.350 1.421 0.364 1.459 fft3d_ps 1201 14.6 0.506 0.523 1.384 1.395 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.305 1.320 grid_collocate_task_list 119 9.7 1.202 1.293 1.202 1.293 acc_transpose_blocks 18288 15.5 0.074 0.076 1.208 1.227 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.183 1.190 wfi_extrapolate 11 7.9 0.001 0.001 1.173 1.174 make_images_data 4572 15.5 0.044 0.048 0.823 1.013 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.905 0.938 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.935 hybrid_alltoall_any 4725 16.4 0.054 0.111 0.698 0.914 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.797 0.800 mp_alltoall_z22v 1201 16.6 0.723 0.787 0.723 0.787 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 0.771 0.785 mp_alltoall_d11v 2130 13.8 0.700 0.781 0.700 0.781 cp_fm_cholesky_invert 11 10.9 0.716 0.720 0.716 0.720 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.652 0.719 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.454 0.677 mp_allgather_i34 2286 14.5 0.292 0.656 0.292 0.656 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.018000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.363636, yerr=3.624982 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 557.768704E+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 0.054 0.175 34.712 34.715 qs_mol_dyn_low 1 2.0 0.003 0.003 34.360 34.367 qs_forces 11 3.9 0.002 0.003 34.305 34.306 qs_energies 11 4.9 0.002 0.006 32.610 32.616 scf_env_do_scf 11 5.9 0.001 0.001 27.656 27.658 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.234 24.235 dbcsr_multiply_generic 2286 12.5 0.098 0.100 18.297 18.427 velocity_verlet 10 3.0 0.001 0.002 17.475 17.476 qs_scf_new_mos 108 7.5 0.001 0.001 16.248 16.309 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.247 16.308 multiply_cannon 2286 13.5 0.236 0.270 14.440 15.595 ot_scf_mini 108 9.5 0.002 0.003 15.313 15.373 multiply_cannon_loop 2286 14.5 0.937 0.970 13.507 14.548 ot_mini 108 10.5 0.001 0.001 9.556 9.627 multiply_cannon_multrec 27432 15.5 2.339 3.001 8.601 8.948 qs_ot_get_derivative 108 11.5 0.001 0.001 7.750 7.809 dbcsr_mm_accdrv_process 47916 15.9 5.283 7.113 6.165 7.367 rebuild_ks_matrix 119 8.3 0.000 0.000 6.390 6.437 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.390 6.437 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.676 5.719 sum_up_and_integrate 119 10.3 0.035 0.039 3.678 3.687 integrate_v_rspace 119 11.3 0.002 0.003 3.643 3.653 init_scf_run 11 5.9 0.000 0.001 3.611 3.612 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.611 3.611 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.833 3.563 init_scf_loop 11 6.9 0.001 0.004 3.403 3.404 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.203 3.241 calculate_rho_elec 119 8.7 0.040 0.046 3.202 3.240 qs_ot_get_p 119 10.4 0.001 0.001 3.153 3.232 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.371 3.011 apply_single 119 13.6 0.000 0.000 2.371 3.011 mp_waitall_1 145218 16.4 1.987 2.808 1.987 2.808 prepare_preconditioner 11 7.9 0.000 0.000 2.556 2.565 make_preconditioner 11 8.9 0.000 0.002 2.556 2.565 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.158 2.490 calculate_first_density_matrix 1 7.0 0.001 0.002 2.266 2.268 make_m2s 4572 13.5 0.054 0.055 2.156 2.253 make_images 4572 14.5 0.272 0.333 2.049 2.145 calculate_dm_sparse 119 9.5 0.000 0.000 2.018 2.078 mp_sum_l 7207 12.9 1.348 2.066 1.348 2.066 rs_pw_transfer 974 11.9 0.009 0.009 1.885 1.995 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.894 1.923 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.907 1.917 grid_integrate_task_list 119 12.3 1.835 1.895 1.835 1.895 acc_transpose_blocks 27432 15.5 0.111 0.114 1.492 1.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.825 1.826 jit_kernel_multiply 10 15.6 0.821 1.815 0.821 1.815 density_rs2pw 119 9.7 0.003 0.004 1.702 1.812 pw_transfer 1439 11.6 0.063 0.065 1.738 1.768 ot_diis_step 108 11.5 0.012 0.012 1.767 1.767 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.649 1.682 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.831 1.642 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.604 1.616 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.593 1.594 potential_pw2rs 119 12.3 0.008 0.010 1.472 1.477 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.350 1.384 fft3d_ps 1201 14.6 0.533 0.587 1.329 1.356 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.351 1.352 cp_fm_redistribute_end 50 14.0 0.891 1.327 0.892 1.328 cp_fm_diag_elpa_base 50 14.0 0.414 1.262 0.433 1.302 wfi_extrapolate 11 7.9 0.001 0.001 1.299 1.299 grid_collocate_task_list 119 9.7 1.220 1.290 1.220 1.290 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.870 1.249 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.200 1.210 cp_fm_upper_to_full 72 13.5 0.818 1.132 0.818 1.132 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.096 1.111 dbcsr_complete_redistribute 329 12.2 0.153 0.200 0.819 1.092 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.088 1.089 multiply_cannon_sync_h2d 27432 15.5 1.000 1.062 1.000 1.062 jit_kernel_transpose 5 15.6 0.602 0.974 0.602 0.974 make_images_data 4572 15.5 0.045 0.048 0.843 0.954 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.741 0.911 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.624 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.872 mp_alltoall_d11v 2130 13.8 0.724 0.825 0.724 0.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.808 cp_fm_cholesky_invert 11 10.9 0.747 0.751 0.747 0.751 mp_alltoall_i22 627 13.8 0.435 0.727 0.435 0.727 mp_alltoall_z22v 1201 16.6 0.678 0.706 0.678 0.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.715000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.363636, yerr=3.796475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 606.867456E+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 0.011 0.028 28.087 28.088 qs_mol_dyn_low 1 2.0 0.003 0.003 27.873 27.880 qs_forces 11 3.9 0.002 0.002 27.816 27.816 qs_energies 11 4.9 0.001 0.002 26.109 26.112 scf_env_do_scf 11 5.9 0.000 0.001 21.196 21.196 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.665 18.665 velocity_verlet 10 3.0 0.001 0.001 14.365 14.367 dbcsr_multiply_generic 2286 12.5 0.089 0.091 12.512 12.595 qs_scf_new_mos 108 7.5 0.001 0.001 11.000 11.020 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.999 11.020 multiply_cannon 2286 13.5 0.229 0.238 9.861 10.371 ot_scf_mini 108 9.5 0.002 0.002 10.330 10.354 multiply_cannon_loop 2286 14.5 0.331 0.343 8.945 9.127 multiply_cannon_multrec 9144 15.5 1.659 1.889 6.061 6.280 rebuild_ks_matrix 119 8.3 0.000 0.000 5.893 5.916 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.892 5.916 ot_mini 108 10.5 0.001 0.001 5.818 5.847 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.269 5.289 qs_ot_get_derivative 108 11.5 0.001 0.001 4.521 4.545 dbcsr_mm_accdrv_process 12550 15.8 3.353 3.921 4.299 4.374 sum_up_and_integrate 119 10.3 0.038 0.042 3.621 3.626 integrate_v_rspace 119 11.3 0.003 0.003 3.584 3.588 init_scf_run 11 5.9 0.000 0.001 3.489 3.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.489 3.489 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.239 3.245 calculate_rho_elec 119 8.7 0.059 0.061 3.239 3.245 qs_ot_get_p 119 10.4 0.001 0.001 2.703 2.743 init_scf_loop 11 6.9 0.000 0.000 2.509 2.510 calculate_first_density_matrix 1 7.0 0.001 0.001 2.330 2.331 mp_waitall_1 121218 16.5 1.729 2.195 1.729 2.195 make_m2s 4572 13.5 0.034 0.035 1.855 2.023 grid_integrate_task_list 119 12.3 1.851 1.932 1.851 1.932 make_images 4572 14.5 0.268 0.306 1.766 1.932 calculate_dm_sparse 119 9.5 0.000 0.000 1.788 1.808 pw_transfer 1439 11.6 0.063 0.065 1.789 1.802 rs_pw_transfer 974 11.9 0.008 0.008 1.713 1.802 prepare_preconditioner 11 7.9 0.000 0.000 1.766 1.770 make_preconditioner 11 8.9 0.000 0.000 1.766 1.770 density_rs2pw 119 9.7 0.003 0.004 1.671 1.759 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.726 1.728 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.699 1.709 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.693 1.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.654 1.683 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.592 1.610 jit_kernel_multiply 8 15.7 0.908 1.586 0.908 1.586 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.521 1.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.510 1.519 potential_pw2rs 119 12.3 0.010 0.011 1.399 1.402 fft3d_ps 1201 14.6 0.542 0.553 1.364 1.376 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.358 1.368 grid_collocate_task_list 119 9.7 1.271 1.335 1.271 1.335 ot_diis_step 108 11.5 0.012 0.013 1.282 1.283 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.251 1.253 cp_fm_redistribute_end 50 14.0 0.623 1.232 0.624 1.232 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.212 1.218 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.214 1.215 cp_fm_diag_elpa_base 50 14.0 0.563 1.158 0.606 1.213 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.157 1.207 apply_single 119 13.6 0.000 0.000 1.156 1.207 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.151 1.162 wfi_extrapolate 11 7.9 0.001 0.001 1.114 1.114 make_images_data 4572 15.5 0.038 0.042 0.861 1.080 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.795 1.060 acc_transpose_blocks 9144 15.5 0.039 0.040 0.951 0.959 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.920 cp_fm_cholesky_invert 11 10.9 0.888 0.892 0.888 0.892 mp_alltoall_d11v 2130 13.8 0.763 0.884 0.763 0.884 multiply_cannon_sync_h2d 9144 15.5 0.705 0.791 0.705 0.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.745 0.748 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.738 0.746 mp_alltoall_z22v 1201 16.6 0.696 0.723 0.696 0.723 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.669 0.717 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.711 0.713 mp_allgather_i34 2286 14.5 0.250 0.681 0.250 0.681 multiply_cannon_metrocomm3 9144 15.5 0.018 0.018 0.299 0.630 jit_kernel_transpose 5 15.6 0.594 0.596 0.594 0.596 dbcsr_complete_redistribute 329 12.2 0.167 0.179 0.552 0.580 mp_waitany 5200 13.7 0.479 0.574 0.479 0.574 qs_create_task_list 11 7.9 0.000 0.001 0.543 0.569 generate_qs_task_list 11 8.9 0.190 0.213 0.542 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.088000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.181818, yerr=3.485507 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 758.640640E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 40.938 40.939 qs_mol_dyn_low 1 2.0 0.003 0.003 40.688 40.696 qs_forces 11 3.9 0.002 0.002 40.628 40.630 qs_energies 11 4.9 0.001 0.001 38.682 38.686 scf_env_do_scf 11 5.9 0.001 0.001 33.052 33.052 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.055 25.056 velocity_verlet 10 3.0 0.001 0.001 23.052 23.057 dbcsr_multiply_generic 2286 12.5 0.097 0.097 17.769 17.941 qs_scf_new_mos 108 7.5 0.001 0.001 16.131 16.227 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.131 16.226 ot_scf_mini 108 9.5 0.002 0.002 15.058 15.154 multiply_cannon 2286 13.5 0.302 0.312 13.871 14.796 multiply_cannon_loop 2286 14.5 0.342 0.349 12.634 13.610 ot_mini 108 10.5 0.001 0.001 9.147 9.260 multiply_cannon_multrec 9144 15.5 3.374 4.850 8.712 9.054 init_scf_loop 11 6.9 0.000 0.000 7.969 7.970 qs_ot_get_derivative 108 11.5 0.001 0.001 7.158 7.256 prepare_preconditioner 11 7.9 0.000 0.000 7.044 7.057 make_preconditioner 11 8.9 0.000 0.000 7.044 7.057 rebuild_ks_matrix 119 8.3 0.000 0.000 6.862 6.999 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.861 6.998 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.563 6.937 dbcsr_mm_accdrv_process 12550 15.8 4.202 5.647 5.210 6.780 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.208 6.333 cp_fm_upper_to_full 72 14.2 3.308 4.738 3.308 4.738 mp_waitall_1 97218 16.6 2.744 3.931 2.744 3.931 sum_up_and_integrate 119 10.3 0.064 0.065 3.713 3.718 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.696 3.701 calculate_rho_elec 119 8.7 0.118 0.121 3.696 3.701 integrate_v_rspace 119 11.3 0.003 0.003 3.649 3.656 init_scf_run 11 5.9 0.000 0.001 3.608 3.608 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.608 3.608 qs_ot_get_p 119 10.4 0.001 0.001 3.207 3.340 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.837 3.312 dbcsr_complete_redistribute 329 12.2 0.290 0.296 2.024 2.894 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.218 2.875 apply_single 119 13.6 0.000 0.000 2.218 2.875 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.706 2.567 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 1.483 2.557 make_m2s 4572 13.5 0.037 0.038 2.292 2.437 mp_alltoall_i22 627 13.8 1.517 2.404 1.517 2.404 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.476 2.327 make_images 4572 14.5 0.353 0.384 2.172 2.318 calculate_first_density_matrix 1 7.0 0.001 0.001 2.205 2.206 calculate_dm_sparse 119 9.5 0.000 0.000 2.170 2.188 mp_sum_l 7207 12.9 1.088 2.177 1.088 2.177 grid_integrate_task_list 119 12.3 2.000 2.040 2.000 2.040 pw_transfer 1439 11.6 0.066 0.066 2.022 2.025 ot_diis_step 108 11.5 0.014 0.014 1.968 1.968 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.927 1.929 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.910 1.910 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.896 1.897 density_rs2pw 119 9.7 0.003 0.003 1.783 1.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.759 1.760 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.687 1.737 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.632 1.633 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.621 1.624 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.584 1.610 fft3d_ps 1201 14.6 0.567 0.578 1.561 1.564 cp_fm_cholesky_invert 11 10.9 1.472 1.475 1.472 1.475 grid_collocate_task_list 119 9.7 1.440 1.464 1.440 1.464 rs_pw_transfer 974 11.9 0.009 0.009 1.426 1.452 jit_kernel_multiply 6 15.6 0.981 1.436 0.981 1.436 wfi_extrapolate 11 7.9 0.001 0.001 1.342 1.342 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.336 1.337 cp_fm_diag_elpa_base 50 14.0 1.184 1.243 1.334 1.334 hybrid_alltoall_any 4725 16.4 0.087 0.150 1.112 1.328 make_images_data 4572 15.5 0.042 0.045 1.059 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.250 potential_pw2rs 119 12.3 0.014 0.015 1.232 1.235 mp_alltoall_d11v 2130 13.8 1.203 1.225 1.203 1.225 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.124 1.142 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.098 1.118 multiply_cannon_sync_h2d 9144 15.5 1.040 1.044 1.040 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.033 qs_create_task_list 11 7.9 0.001 0.001 0.950 0.962 generate_qs_task_list 11 8.9 0.374 0.394 0.949 0.962 acc_transpose_blocks 9144 15.5 0.038 0.038 0.890 0.901 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.865 0.877 mp_alltoall_z22v 1201 16.6 0.858 0.875 0.858 0.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.939000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=712.818182, yerr=17.605878 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 502.169600E+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 938415. 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 0.039 0.079 84.781 84.783 qs_mol_dyn_low 1 2.0 0.008 0.025 84.299 84.308 qs_forces 11 3.9 0.007 0.011 84.191 84.196 qs_energies 11 4.9 0.003 0.018 81.320 81.340 scf_env_do_scf 11 5.9 0.001 0.001 72.336 72.338 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.581 66.582 dbcsr_multiply_generic 2055 12.4 0.103 0.106 53.390 53.661 qs_scf_new_mos 99 7.5 0.000 0.001 49.421 49.541 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.421 49.540 ot_scf_mini 99 9.5 0.002 0.003 46.993 47.106 multiply_cannon 2055 13.4 0.182 0.188 43.499 44.433 multiply_cannon_loop 2055 14.4 1.564 1.603 42.504 43.449 velocity_verlet 10 3.0 0.001 0.001 42.841 42.842 ot_mini 99 10.5 0.001 0.001 28.825 28.920 qs_ot_get_derivative 99 11.5 0.001 0.001 21.950 22.061 multiply_cannon_multrec 49320 15.4 12.065 13.020 17.492 18.609 rebuild_ks_matrix 110 8.3 0.000 0.001 14.543 14.685 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.542 14.685 mp_waitall_1 241148 16.1 12.584 13.659 12.584 13.659 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.811 12.938 multiply_cannon_sync_h2d 49320 15.4 9.909 10.622 9.909 10.622 qs_ot_get_p 110 10.4 0.001 0.001 9.468 9.604 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.310 8.681 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.245 7.786 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.158 7.778 apply_single 110 13.6 0.000 0.001 7.158 7.777 sum_up_and_integrate 110 10.3 0.037 0.043 7.112 7.126 integrate_v_rspace 110 11.3 0.003 0.003 7.075 7.098 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.812 6.872 init_scf_run 11 5.9 0.000 0.001 6.817 6.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.817 6.817 ot_diis_step 99 11.5 0.006 0.006 6.562 6.563 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.338 6.480 calculate_rho_elec 110 8.6 0.020 0.024 6.338 6.479 mp_sum_l 6514 12.8 5.408 6.226 5.408 6.226 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.076 6.105 dbcsr_mm_accdrv_process 87628 16.1 2.120 2.210 5.306 5.926 init_scf_loop 11 6.9 0.001 0.005 5.723 5.724 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.054 5.054 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.548 4.550 cp_fm_redistribute_end 48 14.0 3.960 4.521 3.964 4.522 cp_fm_diag_elpa_base 48 14.0 0.551 4.417 0.554 4.443 rs_pw_transfer 902 11.9 0.012 0.014 3.916 4.189 make_m2s 4110 13.4 0.061 0.065 4.070 4.172 wfi_extrapolate 11 7.9 0.001 0.001 4.080 4.080 make_images 4110 14.4 0.178 0.192 3.975 4.080 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 3.020 3.918 calculate_dm_sparse 110 9.5 0.000 0.001 3.801 3.886 density_rs2pw 110 9.6 0.004 0.004 3.417 3.688 prepare_preconditioner 11 7.9 0.000 0.000 3.540 3.565 make_preconditioner 11 8.9 0.000 0.002 3.540 3.565 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.513 3.517 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.434 3.478 grid_integrate_task_list 110 12.3 3.241 3.394 3.241 3.394 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.311 3.353 jit_kernel_multiply 13 15.9 2.901 3.323 2.901 3.323 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.171 3.216 pw_transfer 1331 11.6 0.053 0.063 3.019 3.085 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.932 2.999 potential_pw2rs 110 12.3 0.006 0.007 2.701 2.745 calculate_first_density_matrix 1 7.0 0.001 0.002 2.645 2.655 mp_alltoall_d11v 2046 13.8 2.149 2.551 2.149 2.551 fft3d_ps 1111 14.6 0.749 0.831 2.469 2.528 fft_wrap_pw1pw2_140 451 13.1 0.168 0.190 2.420 2.485 acc_transpose_blocks 49320 15.4 0.211 0.222 2.140 2.205 grid_collocate_task_list 110 9.6 2.086 2.201 2.086 2.201 mp_waitany 14300 13.8 1.873 2.119 1.873 2.119 mp_sum_d 3883 11.9 1.564 2.075 1.564 2.075 make_images_data 4110 15.4 0.043 0.046 1.794 1.929 cp_fm_cholesky_invert 11 10.9 1.919 1.923 1.919 1.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.875 1.911 hybrid_alltoall_any 4261 16.3 0.081 0.478 1.560 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.783000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.314168 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 584.826880E+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 1941718. 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 0.013 0.031 69.328 69.338 qs_mol_dyn_low 1 2.0 0.003 0.003 68.988 68.997 qs_forces 11 3.9 0.002 0.003 68.919 68.922 qs_energies 11 4.9 0.001 0.002 65.555 65.562 scf_env_do_scf 11 5.9 0.001 0.001 57.036 57.039 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.230 49.230 dbcsr_multiply_generic 2055 12.4 0.112 0.115 38.154 38.322 velocity_verlet 10 3.0 0.001 0.001 36.326 36.327 qs_scf_new_mos 99 7.5 0.001 0.001 33.327 33.458 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.327 33.458 multiply_cannon 2055 13.4 0.222 0.246 31.435 32.598 ot_scf_mini 99 9.5 0.003 0.003 31.657 31.790 multiply_cannon_loop 2055 14.4 0.930 0.955 30.141 30.957 ot_mini 99 10.5 0.001 0.001 18.729 18.866 multiply_cannon_multrec 24660 15.4 7.594 9.428 13.908 15.452 rebuild_ks_matrix 110 8.3 0.000 0.001 13.663 13.776 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.663 13.775 qs_ot_get_derivative 99 11.5 0.001 0.001 12.936 13.069 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.023 12.126 mp_waitall_1 186928 16.3 8.331 10.675 8.331 10.675 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.546 8.449 multiply_cannon_sync_h2d 24660 15.4 7.006 8.174 7.006 8.174 init_scf_loop 11 6.9 0.000 0.001 7.770 7.770 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.496 7.185 apply_single 110 13.6 0.000 0.001 6.496 7.185 dbcsr_mm_accdrv_process 52282 16.1 4.690 5.644 6.155 6.724 sum_up_and_integrate 110 10.3 0.052 0.060 6.605 6.616 integrate_v_rspace 110 11.3 0.002 0.003 6.553 6.567 qs_ot_get_p 110 10.4 0.001 0.001 6.283 6.433 init_scf_run 11 5.9 0.000 0.001 6.146 6.146 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.145 6.146 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.171 5.900 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.843 5.853 calculate_rho_elec 110 8.6 0.039 0.047 5.843 5.852 prepare_preconditioner 11 7.9 0.000 0.000 5.755 5.775 make_preconditioner 11 8.9 0.000 0.000 5.755 5.775 ot_diis_step 99 11.5 0.010 0.010 5.735 5.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.327 5.478 make_m2s 4110 13.4 0.057 0.060 4.243 4.688 make_images 4110 14.4 0.399 0.444 4.135 4.576 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.337 4.359 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.890 3.890 wfi_extrapolate 11 7.9 0.001 0.001 3.528 3.528 pw_transfer 1331 11.6 0.065 0.070 3.348 3.484 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.241 3.377 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.331 3.333 grid_integrate_task_list 110 12.3 3.158 3.309 3.158 3.309 density_rs2pw 110 9.6 0.004 0.004 3.139 3.305 cp_fm_redistribute_end 48 14.0 2.480 3.300 2.483 3.300 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.297 3.298 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.221 3.288 cp_fm_diag_elpa_base 48 14.0 0.781 3.174 0.814 3.263 rs_pw_transfer 902 11.9 0.012 0.013 3.044 3.189 calculate_dm_sparse 110 9.5 0.001 0.001 2.972 3.005 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.941 2.987 fft_wrap_pw1pw2_140 451 13.1 0.199 0.217 2.686 2.821 make_images_data 4110 15.4 0.047 0.051 2.353 2.775 hybrid_alltoall_any 4261 16.3 0.102 0.447 2.081 2.736 fft3d_ps 1111 14.6 1.072 1.267 2.599 2.723 cp_fm_cholesky_invert 11 10.9 2.695 2.702 2.695 2.702 mp_sum_l 6514 12.8 1.918 2.571 1.918 2.571 calculate_first_density_matrix 1 7.0 0.001 0.001 2.530 2.535 potential_pw2rs 110 12.3 0.008 0.009 2.452 2.475 grid_collocate_task_list 110 9.6 2.120 2.257 2.120 2.257 jit_kernel_multiply 11 16.3 1.113 2.134 1.113 2.134 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.970 1.988 mp_alltoall_d11v 2046 13.8 1.824 1.986 1.824 1.986 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.852 1.856 multiply_cannon_metrocomm4 22605 15.4 0.074 0.076 0.787 1.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.805 1.817 mp_irecv_dv 57340 16.2 0.665 1.702 0.665 1.702 mp_allgather_i34 2055 14.4 0.622 1.648 0.622 1.648 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.550 acc_transpose_blocks 24660 15.4 0.106 0.110 1.502 1.523 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.482 dbcsr_complete_redistribute 325 12.2 0.246 0.307 1.174 1.447 mp_waitany 10164 13.8 1.209 1.427 1.209 1.427 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.338000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.000000, yerr=6.237715 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 657.977344E+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 0.050 0.083 60.123 60.124 qs_mol_dyn_low 1 2.0 0.003 0.003 59.598 59.608 qs_forces 11 3.9 0.009 0.056 59.533 59.534 qs_energies 11 4.9 0.005 0.029 56.304 56.308 scf_env_do_scf 11 5.9 0.001 0.001 48.160 48.161 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.540 39.541 velocity_verlet 10 3.0 0.001 0.002 32.424 32.425 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.491 28.735 qs_scf_new_mos 99 7.5 0.001 0.001 25.053 25.148 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.053 25.147 ot_scf_mini 99 9.5 0.002 0.003 23.820 23.919 multiply_cannon 2055 13.4 0.212 0.225 22.081 23.227 multiply_cannon_loop 2055 14.4 0.620 0.635 20.864 21.961 ot_mini 99 10.5 0.001 0.001 13.700 13.806 rebuild_ks_matrix 110 8.3 0.000 0.000 12.133 12.268 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.132 12.268 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.708 10.832 multiply_cannon_multrec 16440 15.4 3.917 4.898 9.713 10.643 mp_waitall_1 146766 16.3 7.109 9.853 7.109 9.853 qs_ot_get_derivative 99 11.5 0.001 0.001 9.233 9.340 init_scf_loop 11 6.9 0.001 0.006 8.583 8.584 prepare_preconditioner 11 7.9 0.000 0.000 6.823 6.837 make_preconditioner 11 8.9 0.000 0.003 6.823 6.837 multiply_cannon_metrocomm3 16440 15.4 0.042 0.043 4.214 6.801 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.179 6.535 sum_up_and_integrate 110 10.3 0.060 0.061 6.390 6.403 integrate_v_rspace 110 11.3 0.003 0.003 6.330 6.342 dbcsr_mm_accdrv_process 34862 16.1 4.580 5.320 5.648 5.743 init_scf_run 11 5.9 0.000 0.001 5.576 5.577 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.576 5.576 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.565 5.574 calculate_rho_elec 110 8.6 0.058 0.058 5.564 5.573 qs_ot_get_p 110 10.4 0.001 0.001 5.304 5.431 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.993 5.423 apply_single 110 13.6 0.000 0.000 4.993 5.423 make_m2s 4110 13.4 0.050 0.052 4.229 4.619 make_images 4110 14.4 0.393 0.513 4.113 4.503 ot_diis_step 99 11.5 0.010 0.011 4.429 4.429 multiply_cannon_sync_h2d 16440 15.4 3.735 4.322 3.735 4.322 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.108 3.763 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.625 3.629 grid_integrate_task_list 110 12.3 3.198 3.398 3.198 3.398 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.282 3.283 pw_transfer 1331 11.6 0.064 0.071 3.222 3.230 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.117 3.128 density_rs2pw 110 9.6 0.004 0.004 2.836 3.027 wfi_extrapolate 11 7.9 0.001 0.001 2.965 2.965 rs_pw_transfer 902 11.9 0.010 0.012 2.756 2.959 make_images_data 4110 15.4 0.043 0.048 2.450 2.931 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.152 2.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.753 2.754 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.748 2.750 cp_fm_redistribute_end 48 14.0 1.710 2.724 1.712 2.725 cp_fm_cholesky_invert 11 10.9 2.716 2.723 2.716 2.723 cp_fm_diag_elpa_base 48 14.0 0.951 2.585 1.008 2.692 fft_wrap_pw1pw2_140 451 13.1 0.209 0.212 2.656 2.672 calculate_dm_sparse 110 9.5 0.001 0.001 2.504 2.540 calculate_first_density_matrix 1 7.0 0.001 0.004 2.526 2.528 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.422 2.471 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.389 2.450 fft3d_ps 1111 14.6 1.065 1.099 2.431 2.448 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.876 2.393 grid_collocate_task_list 110 9.6 2.173 2.335 2.173 2.335 mp_irecv_dv 48980 15.7 0.805 2.272 0.805 2.272 potential_pw2rs 110 12.3 0.010 0.010 2.207 2.214 mp_alltoall_d11v 2046 13.8 1.774 2.118 1.774 2.118 mp_sum_l 6514 12.8 1.570 2.109 1.570 2.109 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.092 2.093 dbcsr_complete_redistribute 325 12.2 0.326 0.364 1.458 1.920 cp_fm_upper_to_full 70 13.6 1.375 1.831 1.375 1.831 cp_fm_cholesky_decompose 22 10.9 1.671 1.705 1.671 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.641 mp_allgather_i34 2055 14.4 0.540 1.598 0.540 1.598 jit_kernel_multiply 8 16.6 0.676 1.528 0.676 1.528 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.484 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.455 1.469 mp_waitany 17072 13.8 1.227 1.430 1.227 1.430 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.953 1.401 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.183 1.285 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.265 1.270 rs_gather_matrices 110 12.3 0.137 0.149 0.875 1.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.124000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.636364, yerr=8.466756 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 725.876736E+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 0.012 0.032 64.659 64.660 qs_mol_dyn_low 1 2.0 0.003 0.003 64.390 64.399 qs_forces 11 3.9 0.002 0.002 64.324 64.325 qs_energies 11 4.9 0.001 0.001 60.959 60.964 scf_env_do_scf 11 5.9 0.000 0.001 52.740 52.743 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.789 40.790 velocity_verlet 10 3.0 0.001 0.001 37.055 37.057 dbcsr_multiply_generic 2055 12.4 0.114 0.116 29.529 29.732 qs_scf_new_mos 99 7.5 0.001 0.001 26.396 26.504 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.395 26.503 ot_scf_mini 99 9.5 0.003 0.003 24.759 24.868 multiply_cannon 2055 13.4 0.246 0.267 22.325 23.638 multiply_cannon_loop 2055 14.4 0.883 0.914 20.939 21.538 ot_mini 99 10.5 0.001 0.001 14.084 14.220 multiply_cannon_multrec 24660 15.4 4.186 6.569 12.629 13.914 rebuild_ks_matrix 110 8.3 0.000 0.000 11.944 12.040 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.943 12.040 init_scf_loop 11 6.9 0.000 0.000 11.909 11.909 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.576 10.662 prepare_preconditioner 11 7.9 0.000 0.000 10.192 10.208 make_preconditioner 11 8.9 0.000 0.000 10.192 10.208 qs_ot_get_derivative 99 11.5 0.001 0.001 9.936 10.052 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.382 9.860 dbcsr_mm_accdrv_process 52304 16.0 6.877 8.118 8.294 9.233 mp_waitall_1 126806 16.4 4.573 6.621 4.573 6.621 sum_up_and_integrate 110 10.3 0.068 0.072 6.318 6.330 integrate_v_rspace 110 11.3 0.003 0.003 6.250 6.262 make_m2s 4110 13.4 0.059 0.061 5.561 5.848 qs_ot_get_p 110 10.4 0.001 0.001 5.581 5.739 make_images 4110 14.4 0.576 0.698 5.422 5.705 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.577 5.588 calculate_rho_elec 110 8.6 0.077 0.081 5.576 5.587 init_scf_run 11 5.9 0.000 0.001 5.487 5.487 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.486 5.487 cp_fm_upper_to_full 70 13.8 3.354 4.751 3.354 4.751 ot_diis_step 99 11.5 0.011 0.012 4.109 4.110 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.001 4.080 apply_single 110 13.6 0.000 0.000 4.001 4.080 dbcsr_complete_redistribute 325 12.2 0.425 0.489 2.801 3.951 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.756 3.771 grid_integrate_task_list 110 12.3 3.272 3.398 3.272 3.398 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.397 3.380 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.240 3.378 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.293 3.343 multiply_cannon_sync_h2d 24660 15.4 3.160 3.323 3.160 3.323 pw_transfer 1331 11.6 0.064 0.071 3.262 3.285 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.279 3.279 make_images_data 4110 15.4 0.045 0.049 2.855 3.193 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.157 3.183 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.048 3.113 hybrid_alltoall_any 4261 16.3 0.119 0.459 2.364 3.113 calculate_dm_sparse 110 9.5 0.001 0.001 2.932 2.965 wfi_extrapolate 11 7.9 0.001 0.001 2.958 2.958 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.801 2.928 density_rs2pw 110 9.6 0.004 0.004 2.734 2.894 mp_alltoall_i22 605 13.7 1.667 2.849 1.667 2.849 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.824 2.826 cp_fm_cholesky_invert 11 10.9 2.813 2.821 2.813 2.821 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.756 2.758 cp_fm_redistribute_end 48 14.0 1.374 2.730 1.375 2.731 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.686 2.713 cp_fm_diag_elpa_base 48 14.0 1.272 2.597 1.353 2.711 rs_pw_transfer 902 11.9 0.010 0.011 2.367 2.530 fft3d_ps 1111 14.6 1.062 1.096 2.459 2.476 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.414 2.454 calculate_first_density_matrix 1 7.0 0.001 0.001 2.436 2.441 grid_collocate_task_list 110 9.6 2.221 2.363 2.221 2.363 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.223 2.223 potential_pw2rs 110 12.3 0.013 0.013 2.084 2.096 mp_alltoall_d11v 2046 13.8 1.851 2.009 1.851 2.009 jit_kernel_multiply 11 15.7 1.086 2.008 1.086 2.008 cp_fm_cholesky_decompose 22 10.9 1.726 1.762 1.726 1.762 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.725 1.759 mp_allgather_i34 2055 14.4 0.483 1.717 0.483 1.717 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.602 1.713 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.612 1.623 mp_sum_l 6514 12.8 0.956 1.605 0.956 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.572 acc_transpose_blocks 24660 15.4 0.104 0.107 1.491 1.514 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.859 1.497 mp_irecv_dv 62702 16.1 0.761 1.423 0.761 1.423 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.660000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.545455, yerr=7.240839 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 824.762368E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.055 53.971 53.971 qs_mol_dyn_low 1 2.0 0.003 0.003 53.652 53.660 qs_forces 11 3.9 0.002 0.002 53.590 53.591 qs_energies 11 4.9 0.001 0.002 50.001 50.007 scf_env_do_scf 11 5.9 0.000 0.001 41.549 41.550 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 33.948 33.948 velocity_verlet 10 3.0 0.001 0.001 30.241 30.244 dbcsr_multiply_generic 2055 12.4 0.103 0.105 22.805 22.905 qs_scf_new_mos 99 7.5 0.001 0.001 20.005 20.068 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.004 20.068 ot_scf_mini 99 9.5 0.002 0.002 18.777 18.829 multiply_cannon 2055 13.4 0.247 0.259 17.483 18.781 multiply_cannon_loop 2055 14.4 0.322 0.337 16.169 16.395 rebuild_ks_matrix 110 8.3 0.000 0.000 11.368 11.415 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.368 11.414 ot_mini 99 10.5 0.001 0.001 10.166 10.220 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.125 10.166 multiply_cannon_multrec 8220 15.4 3.230 4.485 7.741 8.826 init_scf_loop 11 6.9 0.000 0.000 7.555 7.556 mp_waitall_1 106626 16.5 5.943 7.520 5.943 7.520 qs_ot_get_derivative 99 11.5 0.001 0.001 6.360 6.414 sum_up_and_integrate 110 10.3 0.080 0.081 6.125 6.139 integrate_v_rspace 110 11.3 0.003 0.003 6.045 6.059 prepare_preconditioner 11 7.9 0.000 0.000 5.937 5.946 make_preconditioner 11 8.9 0.000 0.000 5.937 5.946 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.525 5.595 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.522 5.536 calculate_rho_elec 110 8.6 0.115 0.116 5.522 5.536 dbcsr_mm_accdrv_process 17442 15.9 2.870 4.186 4.381 5.364 init_scf_run 11 5.9 0.000 0.001 5.357 5.357 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.357 5.357 qs_ot_get_p 110 10.4 0.001 0.001 4.702 4.764 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.052 4.412 make_m2s 4110 13.4 0.038 0.040 4.093 4.344 make_images 4110 14.4 0.634 0.693 3.964 4.216 ot_diis_step 99 11.5 0.012 0.012 3.784 3.784 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.698 3.746 apply_single 110 13.6 0.000 0.000 3.698 3.746 grid_integrate_task_list 110 12.3 3.361 3.500 3.361 3.500 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.402 3.405 pw_transfer 1331 11.6 0.065 0.072 3.200 3.224 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.093 3.120 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.108 3.109 multiply_cannon_sync_h2d 8220 15.4 2.908 3.025 2.908 3.025 cp_fm_cholesky_invert 11 10.9 2.877 2.882 2.877 2.882 make_images_data 4110 15.4 0.038 0.043 2.343 2.746 density_rs2pw 110 9.6 0.004 0.004 2.580 2.697 fft_wrap_pw1pw2_140 451 13.1 0.212 0.217 2.661 2.692 wfi_extrapolate 11 7.9 0.001 0.001 2.683 2.683 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.649 2.650 hybrid_alltoall_any 4261 16.3 0.199 0.860 2.275 2.648 calculate_dm_sparse 110 9.5 0.001 0.001 2.597 2.645 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.633 2.635 calculate_first_density_matrix 1 7.0 0.001 0.001 2.590 2.591 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.585 2.585 cp_fm_redistribute_end 48 14.0 0.655 2.561 0.658 2.562 cp_fm_diag_elpa_base 48 14.0 1.721 2.373 1.896 2.533 grid_collocate_task_list 110 9.6 2.330 2.532 2.330 2.532 fft3d_ps 1111 14.6 1.120 1.160 2.350 2.376 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.269 2.285 rs_pw_transfer 902 11.9 0.010 0.010 2.095 2.239 jit_kernel_multiply 10 15.7 1.198 2.015 1.198 2.015 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.768 1.986 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.897 1.919 potential_pw2rs 110 12.3 0.015 0.016 1.877 1.879 mp_alltoall_d11v 2046 13.8 1.601 1.799 1.601 1.799 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.748 1.783 cp_fm_cholesky_decompose 22 10.9 1.650 1.660 1.650 1.660 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.493 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.572 1.585 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.573 1.578 mp_allgather_i34 2055 14.4 0.498 1.578 0.498 1.578 dbcsr_complete_redistribute 325 12.2 0.570 0.593 1.429 1.525 qs_create_task_list 11 7.9 0.001 0.001 1.220 1.320 generate_qs_task_list 11 8.9 0.379 0.447 1.220 1.320 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.478 1.228 mp_waitany 9240 13.8 1.041 1.191 1.041 1.191 mp_irecv_dv 24056 15.7 0.454 1.189 0.454 1.189 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.096 1.127 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=53.971000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=774.181818, yerr=10.911363 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 1.384268E+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 0.018 0.038 87.053 87.053 qs_mol_dyn_low 1 2.0 0.003 0.003 86.719 86.727 qs_forces 11 3.9 0.002 0.002 86.268 86.269 qs_energies 11 4.9 0.001 0.001 82.222 82.223 scf_env_do_scf 11 5.9 0.001 0.001 72.218 72.220 velocity_verlet 10 3.0 0.001 0.001 56.093 56.147 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.823 42.825 init_scf_loop 11 6.9 0.000 0.000 29.322 29.325 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.649 28.826 prepare_preconditioner 11 7.9 0.000 0.000 27.388 27.399 make_preconditioner 11 8.9 0.000 0.000 27.388 27.399 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.301 26.867 qs_scf_new_mos 99 7.5 0.001 0.001 26.568 26.635 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.567 26.635 ot_scf_mini 99 9.5 0.002 0.002 24.797 24.850 multiply_cannon 2055 13.4 0.345 0.371 21.465 22.161 multiply_cannon_loop 2055 14.4 0.342 0.345 19.630 20.053 cp_fm_upper_to_full 70 14.2 13.139 18.965 13.139 18.965 ot_mini 99 10.5 0.001 0.001 13.638 13.701 rebuild_ks_matrix 110 8.3 0.000 0.001 13.058 13.144 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.058 13.144 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.880 11.957 dbcsr_complete_redistribute 325 12.2 1.027 1.051 7.684 11.016 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.650 9.975 multiply_cannon_multrec 8220 15.4 4.341 4.514 9.701 9.811 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.072 9.354 qs_ot_get_derivative 99 11.5 0.001 0.001 9.066 9.119 mp_alltoall_i22 605 13.7 5.723 9.078 5.723 9.078 mp_waitall_1 87304 16.6 7.661 8.723 7.661 8.723 sum_up_and_integrate 110 10.3 0.150 0.152 6.587 6.603 integrate_v_rspace 110 11.3 0.004 0.004 6.437 6.453 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.407 6.445 calculate_rho_elec 110 8.6 0.227 0.227 6.407 6.445 qs_ot_get_p 110 10.4 0.001 0.001 5.971 6.049 make_m2s 4110 13.4 0.043 0.043 5.320 5.853 init_scf_run 11 5.9 0.000 0.001 5.771 5.771 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.771 5.771 make_images 4110 14.4 0.879 0.930 5.132 5.666 cp_fm_cholesky_invert 11 10.9 5.545 5.550 5.545 5.550 dbcsr_mm_accdrv_process 11614 15.7 3.331 3.840 5.216 5.485 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.609 5.150 apply_single 110 13.6 0.000 0.000 4.609 5.150 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.595 4.938 ot_diis_step 99 11.5 0.015 0.016 4.547 4.547 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.317 4.325 multiply_cannon_sync_h2d 8220 15.4 3.947 3.953 3.947 3.953 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.873 3.874 grid_integrate_task_list 110 12.3 3.662 3.718 3.662 3.718 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.928 3.686 pw_transfer 1331 11.6 0.073 0.073 3.656 3.661 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.636 3.637 make_images_data 4110 15.4 0.041 0.045 2.894 3.585 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.540 3.545 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.887 3.343 calculate_dm_sparse 110 9.5 0.001 0.001 3.275 3.304 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.264 3.265 cp_fm_diag_elpa_base 48 14.0 2.711 2.913 3.262 3.263 wfi_extrapolate 11 7.9 0.001 0.001 3.255 3.255 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.051 3.058 density_rs2pw 110 9.6 0.004 0.004 2.916 2.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.865 2.868 fft3d_ps 1111 14.6 1.265 1.275 2.748 2.754 grid_collocate_task_list 110 9.6 2.625 2.649 2.625 2.649 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.434 2.464 calculate_first_density_matrix 1 7.0 0.000 0.001 2.412 2.412 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.336 2.366 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.169 2.230 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.194 rs_pw_transfer 902 11.9 0.010 0.011 2.124 2.163 cp_fm_cholesky_decompose 22 10.9 2.088 2.112 2.088 2.112 mp_alltoall_d11v 2046 13.8 2.055 2.109 2.055 2.109 potential_pw2rs 110 12.3 0.021 0.022 2.029 2.032 qs_create_task_list 11 7.9 0.001 0.001 1.898 1.943 generate_qs_task_list 11 8.9 0.742 0.794 1.898 1.943 jit_kernel_multiply 10 15.4 1.683 1.870 1.683 1.870 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.807 1.864 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.800 1.815 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.781 1.791 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.053000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1236.636364, yerr=69.784438 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 626.708480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2238940. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 203.164 203.165 qs_mol_dyn_low 1 2.0 0.003 0.003 202.760 202.812 qs_forces 11 3.9 0.005 0.007 202.386 202.386 qs_energies 11 4.9 0.001 0.002 196.915 196.929 scf_env_do_scf 11 5.9 0.001 0.001 180.649 180.653 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 159.939 159.941 dbcsr_multiply_generic 2507 12.6 0.175 0.179 124.670 125.623 velocity_verlet 10 3.0 0.001 0.001 123.260 123.264 qs_scf_new_mos 117 7.6 0.001 0.001 121.287 121.574 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.286 121.574 ot_scf_mini 117 9.6 0.003 0.003 114.716 114.975 multiply_cannon 2507 13.6 0.239 0.249 101.079 104.193 multiply_cannon_loop 2507 14.6 2.067 2.148 98.831 101.435 ot_mini 117 10.6 0.001 0.001 66.147 66.412 multiply_cannon_multrec 60168 15.6 33.375 35.841 41.679 43.929 qs_ot_get_derivative 117 11.6 0.001 0.001 40.970 41.222 rebuild_ks_matrix 128 8.3 0.001 0.001 33.235 33.722 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.234 33.721 mp_waitall_1 291448 16.2 28.819 31.278 28.819 31.278 multiply_cannon_sync_h2d 60168 15.6 27.546 30.399 27.546 30.399 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.925 30.366 qs_ot_get_p 128 10.4 0.001 0.001 26.582 26.959 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.536 25.294 apply_single 128 13.6 0.001 0.001 24.536 25.294 ot_diis_step 117 11.6 0.007 0.008 24.814 24.816 init_scf_loop 11 6.9 0.000 0.001 20.633 20.634 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.893 19.961 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.729 18.930 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 15.717 18.403 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.350 17.351 prepare_preconditioner 11 7.9 0.000 0.000 16.040 16.097 make_preconditioner 11 8.9 0.000 0.000 16.040 16.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.256 15.417 make_m2s 5014 13.6 0.103 0.112 14.110 14.521 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.506 14.512 cp_fm_redistribute_end 83 14.4 11.457 14.438 11.467 14.441 make_images 5014 14.6 0.402 0.421 13.930 14.353 cp_fm_diag_elpa_base 83 14.4 2.930 14.161 2.961 14.266 sum_up_and_integrate 128 10.3 0.090 0.109 13.901 13.919 integrate_v_rspace 128 11.3 0.003 0.004 13.811 13.831 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.588 12.680 calculate_rho_elec 128 8.7 0.045 0.063 12.588 12.679 init_scf_run 11 5.9 0.000 0.001 12.144 12.145 scf_env_initial_rho_setup 11 6.9 0.001 0.004 12.144 12.144 mp_sum_l 7870 13.0 8.218 9.893 8.218 9.893 cp_fm_cholesky_invert 11 10.9 9.049 9.057 9.049 9.057 wfi_extrapolate 11 7.9 0.001 0.001 8.954 8.954 calculate_dm_sparse 128 9.5 0.001 0.001 8.393 8.477 dbcsr_mm_accdrv_process 124484 16.2 3.201 3.404 7.865 8.456 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.282 8.430 make_images_data 5014 15.6 0.068 0.074 6.954 7.944 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.818 7.895 multiply_cannon_metrocomm1 60168 15.6 0.086 0.092 6.090 7.691 grid_integrate_task_list 128 12.3 7.067 7.507 7.067 7.507 hybrid_alltoall_any 5200 16.5 0.289 2.247 6.079 7.293 density_rs2pw 128 9.7 0.006 0.007 6.387 7.128 pw_transfer 1547 11.6 0.074 0.105 6.822 7.089 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.620 6.862 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.613 6.620 rs_pw_transfer 1046 11.9 0.017 0.019 5.512 6.206 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.761 5.944 fft_wrap_pw1pw2_140 523 13.2 0.445 0.509 5.692 5.863 fft3d_ps 1291 14.7 2.091 2.573 5.411 5.599 mp_alltoall_d11v 2415 14.1 4.284 5.265 4.284 5.265 grid_collocate_task_list 128 9.7 4.727 5.125 4.727 5.125 cp_fm_cholesky_decompose 22 10.9 4.773 4.787 4.773 4.787 mp_sum_d 4464 12.1 3.651 4.465 3.651 4.465 potential_pw2rs 128 12.3 0.009 0.011 4.388 4.413 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.165000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.727273, yerr=6.383087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 826.429440E+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 5034345. 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 0.114 0.317 189.682 189.684 qs_mol_dyn_low 1 2.0 0.003 0.005 188.836 188.849 qs_forces 11 3.9 0.003 0.005 188.757 188.758 qs_energies 11 4.9 0.003 0.006 182.060 182.073 scf_env_do_scf 11 5.9 0.001 0.002 165.578 165.588 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.820 132.822 velocity_verlet 10 3.0 0.001 0.002 118.910 118.911 dbcsr_multiply_generic 2507 12.6 0.184 0.193 97.658 98.843 qs_scf_new_mos 117 7.6 0.001 0.001 94.152 94.619 qs_scf_loop_do_ot 117 8.6 0.001 0.002 94.152 94.618 ot_scf_mini 117 9.6 0.004 0.005 89.364 89.905 multiply_cannon 2507 13.6 0.477 0.528 77.647 81.770 multiply_cannon_loop 2507 14.6 1.258 1.291 74.256 76.778 ot_mini 117 10.6 0.001 0.001 49.823 50.295 mp_waitall_1 226760 16.4 24.807 39.654 24.807 39.654 multiply_cannon_multrec 30084 15.6 22.055 26.779 31.854 36.990 rebuild_ks_matrix 128 8.3 0.001 0.001 32.802 33.329 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.022 32.801 33.328 init_scf_loop 11 6.9 0.001 0.005 32.668 32.669 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.580 30.045 multiply_cannon_metrocomm3 30084 15.6 0.095 0.101 15.592 28.736 qs_ot_get_derivative 117 11.6 0.001 0.002 27.881 28.408 prepare_preconditioner 11 7.9 0.000 0.000 28.282 28.340 make_preconditioner 11 8.9 0.001 0.003 28.282 28.340 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.994 27.521 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.980 23.090 apply_single 128 13.6 0.001 0.001 21.979 23.089 multiply_cannon_sync_h2d 30084 15.6 19.361 22.088 19.361 22.088 qs_ot_get_p 128 10.4 0.001 0.001 21.221 21.774 ot_diis_step 117 11.6 0.014 0.015 21.766 21.768 cp_fm_cholesky_invert 11 10.9 16.557 16.569 16.557 16.569 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.409 16.444 make_m2s 5014 13.6 0.090 0.096 14.367 16.063 make_images 5014 14.6 1.162 1.360 14.157 15.851 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.237 15.238 sum_up_and_integrate 128 10.3 0.116 0.133 14.332 14.364 integrate_v_rspace 128 11.3 0.003 0.004 14.216 14.249 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.946 12.973 calculate_rho_elec 128 8.7 0.088 0.105 12.945 12.972 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.168 12.177 cp_fm_redistribute_end 83 14.4 7.123 12.111 7.138 12.117 cp_fm_diag_elpa_base 83 14.4 4.745 11.633 4.960 12.002 init_scf_run 11 5.9 0.000 0.001 11.518 11.520 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.518 11.520 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.911 11.317 multiply_cannon_metrocomm4 27577 15.6 0.096 0.109 3.748 10.745 make_images_data 5014 15.6 0.066 0.073 8.712 10.671 mp_irecv_dv 69486 16.3 3.554 10.361 3.554 10.361 dbcsr_mm_accdrv_process 62242 16.2 4.683 5.587 9.255 9.853 hybrid_alltoall_any 5200 16.5 0.340 1.499 7.419 9.811 wfi_extrapolate 11 7.9 0.001 0.001 8.304 8.304 pw_transfer 1547 11.6 0.084 0.101 7.764 7.813 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.541 7.590 grid_integrate_task_list 128 12.3 7.174 7.588 7.174 7.588 density_rs2pw 128 9.7 0.006 0.006 6.715 7.143 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.203 7.069 cp_fm_cholesky_decompose 22 10.9 6.843 6.939 6.843 6.939 calculate_dm_sparse 128 9.5 0.001 0.001 6.484 6.650 fft_wrap_pw1pw2_140 523 13.2 0.470 0.512 6.552 6.606 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.211 6.223 rs_pw_transfer 1046 11.9 0.014 0.016 5.644 6.108 fft3d_ps 1291 14.7 2.775 2.918 5.911 5.949 mp_sum_l 7870 13.0 3.897 5.777 3.897 5.777 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.314 5.454 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.322 5.384 grid_collocate_task_list 128 9.7 4.915 5.290 4.915 5.290 mp_allgather_i34 2507 14.6 1.963 5.001 1.963 5.001 potential_pw2rs 128 12.3 0.015 0.016 4.763 4.778 mp_alltoall_d11v 2415 14.1 4.141 4.543 4.141 4.543 dbcsr_complete_redistribute 395 12.7 0.816 0.965 3.205 4.072 mp_sum_d 4464 12.1 2.666 3.903 2.666 3.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.684000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.454545, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 936.034304E+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 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. 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 0.069 0.104 177.726 177.727 qs_mol_dyn_low 1 2.0 0.003 0.003 177.154 177.167 qs_forces 11 3.9 0.003 0.003 176.950 176.955 qs_energies 11 4.9 0.002 0.007 170.473 170.483 scf_env_do_scf 11 5.9 0.001 0.002 154.698 154.699 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 119.414 119.415 velocity_verlet 10 3.0 0.001 0.002 113.266 113.271 qs_scf_new_mos 117 7.6 0.001 0.001 82.846 83.240 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.845 83.240 dbcsr_multiply_generic 2507 12.6 0.181 0.185 81.784 82.885 ot_scf_mini 117 9.6 0.004 0.004 78.719 79.156 multiply_cannon 2507 13.6 0.499 0.520 62.000 66.871 multiply_cannon_loop 2507 14.6 0.861 0.899 58.706 61.494 ot_mini 117 10.6 0.001 0.001 42.707 43.137 init_scf_loop 11 6.9 0.001 0.004 35.178 35.179 mp_waitall_1 178456 16.5 24.836 34.293 24.836 34.293 prepare_preconditioner 11 7.9 0.000 0.000 31.195 31.251 make_preconditioner 11 8.9 0.000 0.002 31.195 31.251 rebuild_ks_matrix 128 8.3 0.001 0.001 30.036 30.607 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.035 30.607 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.802 30.182 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.141 27.662 multiply_cannon_multrec 20056 15.6 13.506 16.977 22.542 26.050 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 14.751 24.340 qs_ot_get_derivative 117 11.6 0.001 0.002 22.963 23.410 qs_ot_get_p 128 10.4 0.001 0.001 21.475 21.993 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.840 20.852 apply_single 128 13.6 0.001 0.001 19.839 20.852 ot_diis_step 117 11.6 0.018 0.018 19.642 19.642 qs_ot_p2m_diag 83 11.4 0.265 0.271 17.111 17.119 multiply_cannon_sync_h2d 20056 15.6 14.401 16.119 14.401 16.119 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.069 16.069 make_m2s 5014 13.6 0.081 0.085 14.690 15.624 make_images 5014 14.6 1.178 1.278 14.456 15.390 cp_fm_cholesky_invert 11 10.9 14.610 14.620 14.610 14.620 sum_up_and_integrate 128 10.3 0.134 0.145 14.276 14.303 integrate_v_rspace 128 11.3 0.003 0.004 14.142 14.169 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.059 13.099 calculate_rho_elec 128 8.7 0.132 0.146 13.058 13.098 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.916 12.918 cp_fm_redistribute_end 83 14.4 4.865 12.836 4.883 12.839 cp_fm_diag_elpa_base 83 14.4 7.517 12.222 7.932 12.710 init_scf_run 11 5.9 0.000 0.001 10.758 10.759 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.758 10.759 make_images_data 5014 15.6 0.060 0.068 8.934 10.263 hybrid_alltoall_any 5200 16.5 0.431 1.972 7.830 9.578 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.965 9.265 multiply_cannon_metrocomm4 17549 15.6 0.062 0.071 3.411 9.121 mp_irecv_dv 50230 16.2 3.287 8.862 3.287 8.862 dbcsr_mm_accdrv_process 41502 16.2 4.575 5.595 8.482 8.594 pw_transfer 1547 11.6 0.084 0.104 7.833 7.939 grid_integrate_task_list 128 12.3 7.317 7.797 7.317 7.797 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.611 7.723 cp_fm_cholesky_decompose 22 10.9 7.525 7.548 7.525 7.548 cp_fm_upper_to_full 105 14.5 5.608 7.422 5.608 7.422 wfi_extrapolate 11 7.9 0.001 0.001 7.395 7.395 density_rs2pw 128 9.7 0.006 0.006 6.529 7.124 fft_wrap_pw1pw2_140 523 13.2 0.477 0.528 6.639 6.754 dbcsr_complete_redistribute 395 12.7 1.170 1.197 4.609 6.329 fft3d_ps 1291 14.7 2.697 2.900 5.880 5.946 calculate_dm_sparse 128 9.5 0.001 0.001 5.824 5.913 rs_pw_transfer 1046 11.9 0.014 0.014 5.159 5.812 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.741 5.750 grid_collocate_task_list 128 9.7 5.071 5.702 5.071 5.702 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.629 5.317 mp_alltoall_d11v 2415 14.1 4.244 5.128 4.244 5.128 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.373 5.098 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.802 4.981 mp_sum_l 7870 13.0 3.288 4.743 3.288 4.743 potential_pw2rs 128 12.3 0.020 0.023 4.629 4.645 mp_allgather_i34 2507 14.6 1.814 4.591 1.814 4.591 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.054 4.089 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.373 4.081 mp_alltoall_i22 716 14.1 2.001 3.915 2.001 3.915 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.725 3.726 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.727000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.636364, yerr=10.020640 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/18/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 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.158713E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922098320 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509574. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.038 189.602 189.603 qs_mol_dyn_low 1 2.0 0.003 0.003 189.157 189.169 qs_forces 11 3.9 0.003 0.003 189.040 189.052 qs_energies 11 4.9 0.002 0.005 182.054 182.069 scf_env_do_scf 11 5.9 0.001 0.001 165.178 165.194 velocity_verlet 10 3.0 0.001 0.001 125.901 125.903 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.338 118.339 qs_scf_new_mos 118 7.6 0.001 0.001 82.448 82.786 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.447 82.785 dbcsr_multiply_generic 2534 12.6 0.191 0.196 80.697 81.386 ot_scf_mini 118 9.6 0.003 0.004 77.914 78.234 multiply_cannon 2534 13.6 0.560 0.590 55.260 58.128 multiply_cannon_loop 2534 14.6 1.190 1.218 51.510 53.192 init_scf_loop 11 6.9 0.000 0.002 46.717 46.718 ot_mini 118 10.6 0.001 0.001 43.032 43.378 prepare_preconditioner 11 7.9 0.000 0.000 42.643 42.676 make_preconditioner 11 8.9 0.000 0.001 42.643 42.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.226 41.285 multiply_cannon_multrec 30408 15.6 14.063 19.143 26.227 30.996 rebuild_ks_matrix 129 8.3 0.001 0.001 29.265 29.488 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.264 29.488 mp_waitall_1 155386 16.5 18.229 28.617 18.229 28.617 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.502 26.707 qs_ot_get_derivative 118 11.6 0.002 0.002 23.134 23.466 make_m2s 5068 13.6 0.095 0.100 20.915 22.012 make_images 5068 14.6 1.993 2.326 20.607 21.703 qs_ot_get_p 129 10.4 0.001 0.001 20.895 21.236 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.264 19.788 apply_single 129 13.6 0.001 0.001 19.263 19.787 ot_diis_step 118 11.6 0.018 0.019 19.777 19.778 qs_ot_p2m_diag 84 11.4 0.346 0.394 16.641 16.695 cp_fm_upper_to_full 106 14.7 11.137 16.521 11.137 16.521 cp_fm_cholesky_invert 11 10.9 16.368 16.377 16.368 16.377 multiply_cannon_metrocomm3 30408 15.6 0.047 0.050 6.569 16.123 cp_dbcsr_syevd 84 12.4 0.005 0.005 15.335 15.336 sum_up_and_integrate 129 10.3 0.142 0.154 14.133 14.158 integrate_v_rspace 129 11.3 0.004 0.004 13.990 14.019 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.192 13.234 calculate_rho_elec 129 8.7 0.176 0.192 13.192 13.234 dbcsr_complete_redistribute 397 12.7 1.527 1.658 9.111 12.896 multiply_cannon_sync_h2d 30408 15.6 11.831 12.850 11.831 12.850 make_images_data 5068 15.6 0.063 0.067 11.390 12.786 dbcsr_mm_accdrv_process 62840 16.2 7.590 8.520 11.740 12.270 cp_fm_diag_elpa 84 13.4 0.000 0.001 12.164 12.166 cp_fm_redistribute_end 84 14.4 2.102 12.084 2.117 12.090 cp_fm_diag_elpa_base 84 14.4 9.346 11.453 9.939 11.943 hybrid_alltoall_any 5255 16.5 0.530 2.223 10.027 11.869 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.740 11.504 init_scf_run 11 5.9 0.000 0.001 10.922 10.923 scf_env_initial_rho_setup 11 6.9 0.002 0.003 10.922 10.923 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.395 10.078 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.659 9.890 mp_alltoall_i22 720 14.1 5.636 9.357 5.636 9.357 grid_integrate_task_list 129 12.3 7.617 8.010 7.617 8.010 pw_transfer 1559 11.6 0.085 0.099 7.903 7.986 cp_fm_cholesky_decompose 22 10.9 7.658 7.765 7.658 7.765 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.677 7.764 wfi_extrapolate 11 7.9 0.001 0.001 7.666 7.666 multiply_cannon_metrocomm4 25340 15.6 0.076 0.086 2.816 7.283 mp_irecv_dv 76866 16.2 2.669 7.019 2.669 7.019 fft_wrap_pw1pw2_140 527 13.2 0.485 0.492 6.766 6.873 density_rs2pw 129 9.7 0.005 0.006 6.294 6.749 calculate_dm_sparse 129 9.5 0.001 0.001 6.309 6.396 fft3d_ps 1301 14.7 2.813 2.883 5.896 5.966 grid_collocate_task_list 129 9.7 5.240 5.658 5.240 5.658 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.570 5.617 mp_alltoall_d11v 2429 14.1 4.776 5.165 4.776 5.165 rs_pw_transfer 1054 12.0 0.013 0.015 4.603 5.044 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.443 4.534 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.429 4.478 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.431 4.433 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.225 4.285 potential_pw2rs 129 12.3 0.022 0.023 4.269 4.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.603000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.818182, yerr=22.830674 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.522762E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838186. 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 0.017 0.034 170.291 170.291 qs_mol_dyn_low 1 2.0 0.003 0.003 169.907 169.926 qs_forces 11 3.9 0.003 0.003 169.809 169.812 qs_energies 11 4.9 0.002 0.002 162.503 162.515 scf_env_do_scf 11 5.9 0.001 0.001 145.382 145.392 velocity_verlet 10 3.0 0.001 0.001 111.545 111.549 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.848 109.849 dbcsr_multiply_generic 2507 12.6 0.180 0.184 74.214 74.724 qs_scf_new_mos 117 7.6 0.001 0.001 74.366 74.445 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.365 74.444 ot_scf_mini 117 9.6 0.004 0.004 69.961 70.015 multiply_cannon 2507 13.6 0.588 0.642 54.568 59.108 multiply_cannon_loop 2507 14.6 0.448 0.457 49.768 50.787 ot_mini 117 10.6 0.001 0.001 39.680 39.736 init_scf_loop 11 6.9 0.000 0.000 35.377 35.378 mp_waitall_1 129618 16.6 25.959 33.350 25.959 33.350 prepare_preconditioner 11 7.9 0.000 0.000 31.527 31.553 make_preconditioner 11 8.9 0.000 0.000 31.527 31.553 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.301 29.571 rebuild_ks_matrix 128 8.3 0.001 0.001 28.470 28.549 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.470 28.549 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.016 26.091 multiply_cannon_multrec 10028 15.6 10.446 14.153 17.827 21.528 ot_diis_step 117 11.6 0.020 0.020 19.915 19.915 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.470 19.803 apply_single 128 13.6 0.001 0.001 19.470 19.803 qs_ot_get_derivative 117 11.6 0.002 0.002 19.691 19.744 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.387 19.243 make_m2s 5014 13.6 0.066 0.069 16.089 18.430 make_images 5014 14.6 2.303 2.637 15.782 18.123 cp_fm_cholesky_invert 11 10.9 17.924 17.930 17.924 17.930 qs_ot_get_p 128 10.4 0.001 0.001 17.348 17.413 sum_up_and_integrate 128 10.3 0.180 0.190 14.036 14.084 integrate_v_rspace 128 11.3 0.004 0.004 13.856 13.913 qs_ot_p2m_diag 83 11.4 0.495 0.500 13.613 13.628 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.092 13.142 calculate_rho_elec 128 8.7 0.258 0.269 13.091 13.141 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.481 12.482 make_images_data 5014 15.6 0.052 0.060 9.762 12.349 hybrid_alltoall_any 5200 16.5 0.845 3.844 9.582 12.123 multiply_cannon_sync_h2d 10028 15.6 11.647 12.085 11.647 12.085 init_scf_run 11 5.9 0.000 0.001 10.299 10.299 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.298 10.299 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.501 9.506 cp_fm_diag_elpa_base 83 14.4 9.255 9.332 9.494 9.498 grid_integrate_task_list 128 12.3 7.760 8.090 7.760 8.090 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.941 7.980 cp_fm_cholesky_decompose 22 10.9 7.842 7.925 7.842 7.925 pw_transfer 1547 11.6 0.083 0.089 7.758 7.772 dbcsr_mm_accdrv_process 20762 16.1 2.574 3.270 7.013 7.627 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.539 7.557 mp_allgather_i34 2507 14.6 2.998 7.484 2.998 7.484 wfi_extrapolate 11 7.9 0.001 0.001 7.290 7.290 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.327 7.214 fft_wrap_pw1pw2_140 523 13.2 0.501 0.523 6.618 6.638 calculate_dm_sparse 128 9.5 0.001 0.001 6.118 6.222 density_rs2pw 128 9.7 0.005 0.006 5.907 6.174 grid_collocate_task_list 128 9.7 5.485 5.817 5.485 5.817 dbcsr_complete_redistribute 395 12.7 2.124 2.180 5.351 5.753 fft3d_ps 1291 14.7 2.717 2.783 5.703 5.731 mp_alltoall_d11v 2415 14.1 4.786 5.396 4.786 5.396 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.285 5.286 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.175 5.185 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.846 4.704 mp_irecv_dv 28860 15.9 1.808 4.630 1.808 4.630 rs_pw_transfer 1046 11.9 0.012 0.013 4.082 4.379 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.087 4.104 potential_pw2rs 128 12.3 0.027 0.028 4.085 4.099 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.573 3.905 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.524 3.858 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.733 3.766 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.544 3.619 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.431 3.446 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.291000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1414.000000, yerr=54.729085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.056443E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 58536. MP_Allreduce 11109 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 0.024 0.048 289.467 289.467 qs_mol_dyn_low 1 2.0 0.003 0.003 288.947 288.958 qs_forces 11 3.9 0.003 0.003 288.804 288.808 qs_energies 11 4.9 0.002 0.002 280.207 280.219 scf_env_do_scf 11 5.9 0.001 0.001 258.653 258.664 velocity_verlet 10 3.0 0.001 0.001 210.583 210.596 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.515 132.517 init_scf_loop 11 6.9 0.000 0.000 125.884 125.888 prepare_preconditioner 11 7.9 0.000 0.000 121.248 121.282 make_preconditioner 11 8.9 0.000 0.000 121.248 121.281 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.125 118.347 qs_scf_new_mos 118 7.6 0.001 0.001 90.486 90.622 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.485 90.621 ot_scf_mini 118 9.6 0.004 0.004 85.701 85.798 dbcsr_multiply_generic 2529 12.6 0.213 0.220 83.186 83.870 cp_fm_upper_to_full 106 14.8 53.887 77.987 53.887 77.987 multiply_cannon 2529 13.6 0.708 0.781 58.610 59.197 multiply_cannon_loop 2529 14.6 0.475 0.484 54.919 56.469 dbcsr_complete_redistribute 397 12.7 4.011 4.048 30.769 44.714 ot_mini 118 10.6 0.001 0.001 44.386 44.485 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.402 41.329 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 25.077 38.903 mp_alltoall_i22 720 14.1 22.907 36.691 22.907 36.691 rebuild_ks_matrix 129 8.3 0.001 0.001 33.592 33.672 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.592 33.671 cp_fm_cholesky_invert 11 10.9 33.290 33.296 33.290 33.296 mp_waitall_1 106386 16.7 27.959 32.131 27.959 32.131 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.363 31.441 qs_ot_get_p 129 10.4 0.001 0.001 25.813 25.903 qs_ot_get_derivative 118 11.6 0.002 0.002 24.116 24.207 qs_ot_p2m_diag 84 11.4 0.890 0.896 21.794 21.824 make_m2s 5058 13.6 0.076 0.078 19.907 20.913 make_images 5058 14.6 3.802 3.969 19.425 20.435 ot_diis_step 118 11.6 0.022 0.023 20.241 20.241 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.074 20.075 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 18.520 19.822 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.400 19.614 apply_single 129 13.6 0.001 0.001 19.400 19.614 multiply_cannon_multrec 10116 15.6 10.552 12.329 18.074 18.129 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.872 16.874 cp_fm_diag_elpa_base 84 14.4 12.337 13.982 16.868 16.871 multiply_cannon_sync_h2d 10116 15.6 15.777 15.796 15.777 15.796 sum_up_and_integrate 129 10.3 0.323 0.327 15.620 15.722 integrate_v_rspace 129 11.3 0.004 0.004 15.297 15.398 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.918 14.936 calculate_rho_elec 129 8.7 0.486 0.487 14.917 14.935 hybrid_alltoall_any 5245 16.5 1.315 3.102 10.856 12.756 make_images_data 5058 15.6 0.060 0.064 10.611 12.399 init_scf_run 11 5.9 0.000 0.001 11.674 11.674 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.674 11.674 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.409 9.485 dbcsr_mm_accdrv_process 20934 16.1 3.883 5.780 7.284 9.153 cp_fm_cholesky_decompose 22 10.9 8.814 8.828 8.814 8.828 grid_integrate_task_list 129 12.3 8.618 8.801 8.618 8.801 wfi_extrapolate 11 7.9 0.001 0.001 8.652 8.652 pw_transfer 1559 11.6 0.091 0.092 8.287 8.304 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.050 8.068 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.909 7.910 fft_wrap_pw1pw2_140 527 13.2 0.545 0.564 7.158 7.172 mp_alltoall_d11v 2429 14.1 6.759 6.900 6.759 6.900 calculate_dm_sparse 129 9.5 0.001 0.001 6.621 6.702 grid_collocate_task_list 129 9.7 6.384 6.420 6.384 6.420 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.335 6.375 fft3d_ps 1301 14.7 2.783 2.825 6.072 6.100 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.018 6.079 density_rs2pw 129 9.7 0.005 0.005 5.953 5.984 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.467000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2703.909091, yerr=191.793191 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 1.261486E+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 43782264. MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 85.063 85.064 qs_energies 1 2.0 0.000 0.000 84.652 84.659 ls_scf 1 3.0 0.000 0.000 83.748 83.755 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.913 73.112 multiply_cannon 111 7.7 0.018 0.021 56.151 57.564 multiply_cannon_loop 111 8.7 0.201 0.219 52.612 54.011 ls_scf_main 1 4.0 0.000 0.000 52.300 52.301 density_matrix_trs4 2 5.0 0.002 0.003 46.774 46.860 ls_scf_init_scf 1 4.0 0.000 0.000 28.354 28.355 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.287 27.340 mp_waitall_1 11316 10.9 22.783 25.619 22.783 25.619 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.167 25.198 multiply_cannon_multrec 2664 9.7 8.162 8.885 15.557 17.259 multiply_cannon_sync_h2d 2664 9.7 13.464 15.880 13.464 15.880 make_m2s 222 7.7 0.008 0.011 13.051 13.624 make_images 222 8.7 0.099 0.108 13.029 13.604 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.919 13.171 make_images_data 222 9.7 0.004 0.005 7.632 8.273 dbcsr_mm_accdrv_process 4760 10.4 0.523 0.640 7.013 7.974 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.557 7.772 hybrid_alltoall_any 227 10.6 0.215 1.850 6.521 7.670 dbcsr_mm_accdrv_process_sort 4760 11.4 6.287 7.163 6.287 7.163 calculate_norms 4752 9.8 5.495 6.080 5.495 6.080 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.147 5.302 mp_sum_l 807 5.4 3.169 4.945 3.169 4.945 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.355 3.867 make_images_sizes 222 9.7 0.000 0.000 0.694 3.595 mp_alltoall_i44 222 10.7 0.694 3.595 0.694 3.595 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.061 3.401 mp_irecv_dv 6231 10.9 2.044 3.378 2.044 3.378 arnoldi_extremal 4 6.8 0.000 0.000 3.194 3.235 arnoldi_normal_ev 4 7.8 0.001 0.003 3.194 3.235 ls_scf_post 1 4.0 0.000 0.000 3.094 3.101 build_subspace 16 8.4 0.009 0.012 3.091 3.094 ls_scf_store_result 1 5.0 0.000 0.000 2.876 2.928 dbcsr_special_finalize 555 9.7 0.005 0.006 2.371 2.757 dbcsr_merge_single_wm 555 10.7 0.485 0.628 2.363 2.749 make_images_pack 222 9.7 2.207 2.635 2.209 2.637 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.311 2.545 dbcsr_sort_data 658 11.4 2.137 2.466 2.137 2.466 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.453 2.072 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.204 2.319 buffer_matrices_ensure_size 222 8.7 1.827 2.190 1.827 2.190 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.736 1.737 rebuild_ks_matrix 3 7.3 0.000 0.000 1.726 1.727 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.726 1.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.064000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 2.104414E+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 0.028 0.043 90.635 90.636 qs_energies 1 2.0 0.000 0.000 90.042 90.047 ls_scf 1 3.0 0.000 0.000 88.740 88.745 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.936 75.254 multiply_cannon 111 7.7 0.029 0.040 53.128 57.184 ls_scf_main 1 4.0 0.000 0.000 54.575 54.576 multiply_cannon_loop 111 8.7 0.116 0.123 49.855 53.251 density_matrix_trs4 2 5.0 0.002 0.003 48.971 49.180 ls_scf_init_scf 1 4.0 0.000 0.000 30.663 30.664 mp_waitall_1 9246 10.9 20.413 29.833 20.413 29.833 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.480 29.577 multiply_cannon_multrec 1332 9.7 13.258 17.008 22.561 27.418 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.080 27.093 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.424 20.974 make_m2s 222 7.7 0.006 0.008 15.005 15.623 make_images 222 8.7 1.575 1.935 14.975 15.595 dbcsr_mm_accdrv_process 4041 10.4 0.312 0.453 8.900 10.479 dbcsr_mm_accdrv_process_sort 4041 11.4 8.490 10.029 8.490 10.029 make_images_data 222 9.7 0.004 0.004 8.417 9.382 mp_sum_l 807 5.4 5.497 8.946 5.497 8.946 hybrid_alltoall_any 227 10.6 0.522 2.443 7.867 8.616 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.234 7.790 mp_irecv_dv 3311 11.0 3.214 7.722 3.214 7.722 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.308 7.184 calculate_norms 2376 9.8 6.023 6.677 6.023 6.677 multiply_cannon_sync_h2d 1332 9.7 4.851 6.025 4.851 6.025 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.024 5.216 arnoldi_extremal 4 6.8 0.000 0.000 4.633 4.663 arnoldi_normal_ev 4 7.8 0.001 0.005 4.633 4.663 build_subspace 16 8.4 0.014 0.021 4.373 4.376 ls_scf_post 1 4.0 0.000 0.000 3.502 3.507 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.145 3.359 ls_scf_store_result 1 5.0 0.000 0.000 3.208 3.325 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.217 2.746 3.219 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.198 2.977 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.478 2.572 make_images_pack 222 9.7 2.027 2.430 2.030 2.432 mp_allgather_i34 111 8.7 0.963 2.281 0.963 2.281 dbcsr_sort_data 436 11.2 1.820 2.107 1.820 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.847 1.849 rebuild_ks_matrix 3 7.3 0.000 0.000 1.833 1.836 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.833 1.835 dbcsr_data_new 4174 10.1 1.618 1.833 1.618 1.833 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.636000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 2.716258E+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 0.048 0.080 94.423 94.424 qs_energies 1 2.0 0.000 0.000 93.697 93.701 ls_scf 1 3.0 0.002 0.014 92.231 92.235 dbcsr_multiply_generic 111 6.7 0.015 0.022 76.969 77.299 ls_scf_main 1 4.0 0.000 0.000 57.907 57.913 multiply_cannon 111 7.7 0.046 0.114 52.813 56.876 multiply_cannon_loop 111 8.7 0.099 0.105 49.254 53.278 density_matrix_trs4 2 5.0 0.002 0.003 51.946 52.157 mp_waitall_1 7374 11.0 24.117 34.642 24.117 34.642 ls_scf_init_scf 1 4.0 0.000 0.002 30.742 30.745 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.480 29.560 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.082 27.093 multiply_cannon_multrec 888 9.7 12.631 15.409 21.233 24.577 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.996 23.480 make_m2s 222 7.7 0.006 0.007 17.398 18.672 make_images 222 8.7 1.969 2.277 17.360 18.635 hybrid_alltoall_any 227 10.6 0.619 2.865 9.637 11.171 make_images_data 222 9.7 0.003 0.004 10.056 11.118 dbcsr_mm_accdrv_process 3754 10.4 0.254 0.415 8.118 9.305 dbcsr_mm_accdrv_process_sort 3754 11.4 7.747 8.891 7.747 8.891 mp_sum_l 807 5.4 5.217 8.864 5.217 8.864 multiply_cannon_sync_h2d 888 9.7 5.996 7.598 5.996 7.598 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.683 7.031 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.471 6.840 mp_irecv_dv 2335 11.1 2.456 6.798 2.456 6.798 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.909 6.740 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.997 5.238 arnoldi_extremal 4 6.8 0.000 0.000 5.096 5.104 arnoldi_normal_ev 4 7.8 0.004 0.009 5.095 5.104 build_subspace 16 8.4 0.014 0.020 4.796 4.803 calculate_norms 1584 9.8 4.268 4.542 4.268 4.542 mp_allgather_i34 111 8.7 1.394 3.949 1.394 3.949 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.455 3.762 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.589 3.023 3.591 ls_scf_post 1 4.0 0.000 0.001 3.580 3.585 ls_scf_store_result 1 5.0 0.000 0.000 3.305 3.407 ls_scf_dm_to_ks 2 5.0 0.000 0.004 2.737 2.840 make_images_sizes 222 9.7 0.000 0.000 1.014 2.140 mp_alltoall_i44 222 10.7 1.014 2.140 1.014 2.140 make_images_pack 222 9.7 1.814 2.135 1.817 2.138 dbcsr_sort_data 325 11.1 1.882 2.136 1.882 2.136 dbcsr_data_release 9322 10.9 1.325 2.065 1.325 2.065 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.928 1.930 rebuild_ks_matrix 3 7.3 0.000 0.000 1.910 1.912 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.038 1.910 1.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.424000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2194.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 3.363181E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.077 97.677 97.678 qs_energies 1 2.0 0.000 0.000 96.948 96.966 ls_scf 1 3.0 0.000 0.001 95.288 95.308 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.653 78.963 ls_scf_main 1 4.0 0.000 0.000 58.988 58.989 multiply_cannon 111 7.7 0.059 0.130 51.818 55.897 density_matrix_trs4 2 5.0 0.002 0.003 52.926 53.045 multiply_cannon_loop 111 8.7 0.114 0.123 46.694 50.221 ls_scf_init_scf 1 4.0 0.002 0.018 33.009 33.011 ls_scf_init_matrix_S 1 5.0 0.004 0.037 31.639 31.724 mp_waitall_1 6438 11.0 22.702 29.541 22.702 29.541 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.928 28.948 multiply_cannon_multrec 1332 9.7 14.153 17.345 22.069 25.116 make_m2s 222 7.7 0.007 0.007 20.963 22.478 make_images 222 8.7 3.134 3.594 20.913 22.430 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.408 17.058 make_images_data 222 9.7 0.004 0.004 11.624 13.237 hybrid_alltoall_any 227 10.6 0.794 3.756 11.024 12.793 dbcsr_mm_accdrv_process 3641 10.4 0.215 0.400 7.546 9.077 dbcsr_mm_accdrv_process_sort 3641 11.4 7.167 8.656 7.167 8.656 mp_sum_l 807 5.4 4.290 7.690 4.290 7.690 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.101 6.169 mp_irecv_dv 3229 10.9 2.079 6.124 2.079 6.124 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.289 6.080 multiply_cannon_sync_h2d 1332 9.7 5.502 6.046 5.502 6.046 arnoldi_extremal 4 6.8 0.000 0.000 5.275 5.293 arnoldi_normal_ev 4 7.8 0.001 0.005 5.275 5.292 build_subspace 16 8.4 0.015 0.021 4.917 4.925 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.605 4.764 mp_allgather_i34 111 8.7 2.245 4.748 2.245 4.748 calculate_norms 2376 9.8 4.182 4.523 4.182 4.523 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.490 4.340 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.613 3.885 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.677 3.189 3.679 dbcsr_sort_data 658 11.4 3.054 3.507 3.054 3.507 ls_scf_post 1 4.0 0.001 0.007 3.291 3.309 dbcsr_special_finalize 555 9.7 0.006 0.007 2.795 3.244 dbcsr_merge_single_wm 555 10.7 0.533 0.662 2.787 3.236 ls_scf_store_result 1 5.0 0.000 0.000 3.018 3.093 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.898 2.931 dbcsr_data_release 10477 10.7 1.600 2.467 1.600 2.467 dbcsr_finalize 304 7.8 0.049 0.061 1.807 1.997 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.970 1.972 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.678000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 4.643869E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.076 98.339 98.339 qs_energies 1 2.0 0.000 0.000 97.500 97.506 ls_scf 1 3.0 0.000 0.000 95.583 95.588 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.254 77.483 ls_scf_main 1 4.0 0.000 0.000 61.544 61.545 multiply_cannon 111 7.7 0.088 0.175 54.869 59.950 density_matrix_trs4 2 5.0 0.002 0.003 54.575 54.665 multiply_cannon_loop 111 8.7 0.069 0.078 50.323 52.001 mp_waitall_1 5481 11.0 25.764 31.858 25.764 31.858 ls_scf_init_scf 1 4.0 0.000 0.000 30.473 30.478 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.294 29.337 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.130 27.143 multiply_cannon_multrec 444 9.7 14.024 16.176 21.104 24.232 make_m2s 222 7.7 0.004 0.005 17.418 19.966 make_images 222 8.7 3.720 4.436 17.356 19.906 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.672 15.475 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.155 15.181 make_images_data 222 9.7 0.003 0.004 9.737 12.166 hybrid_alltoall_any 227 10.6 0.790 3.782 9.545 12.016 multiply_cannon_sync_h2d 444 9.7 6.521 8.544 6.521 8.544 dbcsr_mm_accdrv_process 3003 10.4 0.165 0.341 6.788 7.973 dbcsr_mm_accdrv_process_sort 3003 11.4 6.472 7.624 6.472 7.624 mp_allgather_i34 111 8.7 2.777 6.970 2.777 6.970 arnoldi_extremal 4 6.8 0.000 0.000 5.926 5.938 arnoldi_normal_ev 4 7.8 0.002 0.005 5.926 5.938 build_subspace 16 8.4 0.015 0.020 5.537 5.545 mp_sum_l 807 5.4 3.034 5.479 3.034 5.479 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.544 4.692 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.241 4.429 dbcsr_matrix_vector_mult_local 304 10.0 3.685 4.147 3.687 4.149 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.093 4.092 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.678 4.028 mp_irecv_dv 1241 11.2 1.658 4.002 1.658 4.002 calculate_norms 792 9.8 3.543 3.665 3.543 3.665 ls_scf_post 1 4.0 0.000 0.000 3.566 3.572 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.381 3.472 ls_scf_store_result 1 5.0 0.000 0.000 3.349 3.396 make_images_sizes 222 9.7 0.000 0.000 1.043 3.200 mp_alltoall_i44 222 10.7 1.043 3.199 1.043 3.199 dbcsr_finalize 304 7.8 0.062 0.078 2.195 2.274 dbcsr_merge_all 275 8.9 0.473 0.521 2.042 2.085 dbcsr_data_release 10123 10.8 1.332 2.007 1.332 2.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.339000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3608.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2940f7cc1a499533767160797f0663fb9d581809_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 8.704426E+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 0.078 0.098 107.888 107.888 qs_energies 1 2.0 0.000 0.000 106.224 106.240 ls_scf 1 3.0 0.000 0.000 103.302 103.317 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.516 77.653 ls_scf_main 1 4.0 0.000 0.000 65.719 65.720 density_matrix_trs4 2 5.0 0.002 0.003 56.908 56.966 multiply_cannon 111 7.7 0.141 0.207 49.978 51.964 multiply_cannon_loop 111 8.7 0.067 0.069 46.440 47.451 ls_scf_init_scf 1 4.0 0.000 0.000 33.939 33.940 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.608 32.629 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.778 29.786 mp_waitall_1 4569 11.1 22.306 26.318 22.306 26.318 make_m2s 222 7.7 0.005 0.005 23.934 24.967 make_images 222 8.7 4.592 4.976 23.829 24.860 multiply_cannon_multrec 444 9.7 17.790 18.462 22.427 23.087 hybrid_alltoall_any 227 10.6 1.660 3.618 12.941 15.692 make_images_data 222 9.7 0.003 0.003 13.162 15.661 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.556 11.090 multiply_cannon_sync_h2d 444 9.7 8.850 8.901 8.850 8.901 arnoldi_extremal 4 6.8 0.000 0.000 7.374 7.384 arnoldi_normal_ev 4 7.8 0.003 0.009 7.374 7.384 build_subspace 16 8.4 0.026 0.036 6.811 6.822 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.411 5.544 dbcsr_matrix_vector_mult_local 304 10.0 4.998 5.295 5.000 5.297 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.986 5.237 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.805 4.892 dbcsr_mm_accdrv_process 1814 10.4 0.280 0.324 4.466 4.589 dbcsr_mm_accdrv_process_sort 1814 11.4 4.163 4.292 4.163 4.292 ls_scf_post 1 4.0 0.000 0.000 3.644 3.659 mp_allgather_i34 111 8.7 1.166 3.655 1.166 3.655 make_images_sizes 222 9.7 0.000 0.000 1.445 3.541 mp_alltoall_i44 222 10.7 1.445 3.540 1.445 3.540 ls_scf_store_result 1 5.0 0.000 0.000 3.398 3.407 calculate_norms 792 9.8 3.237 3.276 3.237 3.276 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.122 dbcsr_merge_all 275 8.9 0.890 0.915 2.862 2.913 dbcsr_complete_redistribute 5 7.6 1.444 1.489 2.769 2.895 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.892 2.892 dbcsr_data_release 12724 10.6 2.316 2.877 2.316 2.877 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.561 dbcsr_sort_data 325 11.1 2.437 2.493 2.437 2.493 dbcsr_new_transposed 4 7.5 0.243 0.251 2.300 2.313 dbcsr_frobenius_norm 74 6.6 2.057 2.136 2.199 2.237 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.201 dbcsr_add_anytype 103 7.2 0.859 0.891 2.130 2.200 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.189 2.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.888000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7010.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 2940f7cc1a499533767160797f0663fb9d581809 Summary: empty Status: OK