=== This is the CP2K Performance-Test === Updating f76a00403..671523f8f Fast-forward src/emd/rt_propagation_methods.F | 4 ++ tests/QS/regtest-rtp-1/H2-hfx-emd.inp | 95 +++++++++++++++++++++++++++++++++++ tests/QS/regtest-rtp-1/H2-hfx-rtp.inp | 87 ++++++++++++++++++++++++++++++++ tests/QS/regtest-rtp-1/TEST_FILES | 2 + 4 files changed, 188 insertions(+) create mode 100644 tests/QS/regtest-rtp-1/H2-hfx-emd.inp create mode 100644 tests/QS/regtest-rtp-1/H2-hfx-rtp.inp Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 671523f8fd0da1af59ea309227fb54ab803f3eeb ################# 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 2022.11.001.rc2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.0, 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 (21.12.2022) # # \ 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 -j${maxtasks} --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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001.rc2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.0 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/01 job id: 43695020 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/02 job id: 43695022 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/03 job id: 43695024 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/04 job id: 43695026 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/05 job id: 43695029 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/06 job id: 43695035 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/07 job id: 43695043 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/08 job id: 43695045 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/09 job id: 43695046 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/10 job id: 43695047 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/11 job id: 43695048 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/12 job id: 43695049 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/13 job id: 43695050 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/14 job id: 43695052 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/15 job id: 43695053 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/16 job id: 43695055 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/17 job id: 43695057 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/18 job id: 43695060 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/19 job id: 43695062 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/20 job id: 43695065 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/21 job id: 43695067 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/22 job id: 43695070 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/23 job id: 43695074 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/24 job id: 43695077 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/25 job id: 43695079 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/26 job id: 43695082 --- 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 137.743 137.743 farming_run 1 2.0 136.095 136.096 137.713 137.716 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.462505E+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 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 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 0.012 0.060 116.673 116.673 qs_energies 1 2.0 0.000 0.000 116.125 116.125 mp2_main 1 3.0 0.000 0.000 113.939 113.940 mp2_gpw_main 1 4.0 0.019 0.025 113.020 113.021 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.738 94.296 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.378 55.936 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.677 47.141 get_2c_integrals 1 6.0 0.000 0.000 37.573 38.188 integrate_v_rspace 273 8.0 0.438 0.450 25.096 30.360 pw_transfer 6555 10.6 0.375 0.388 27.397 27.975 grid_integrate_task_list 273 9.0 20.915 26.668 20.915 26.668 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.050 26.644 fft_wrap_pw1pw2_100 2178 12.4 1.199 1.446 23.603 24.198 compute_2c_integrals 1 7.0 0.002 0.002 19.764 19.764 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.920 19.476 mp2_eri_2c_integrate_gpw 1 9.0 2.368 2.426 18.918 19.472 rpa_ri_compute_en 1 5.0 0.000 0.000 19.174 19.323 cp_fm_cholesky_decompose 12 8.2 17.746 18.371 17.746 18.371 cholesky_decomp 1 7.0 0.000 0.000 16.661 17.275 fft3d_s 5443 13.4 16.158 16.691 16.181 16.714 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.559 10.856 15.559 calculate_wavefunction 272 8.0 5.378 5.525 12.495 13.150 rpa_num_int 1 6.0 0.000 0.000 10.614 10.623 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.512 10.578 calc_mat_Q 8 8.0 0.000 0.000 9.383 9.525 contract_S_to_Q 8 9.0 0.000 0.000 8.803 8.951 calc_potential_gpw 544 9.5 0.005 0.006 8.294 8.713 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.247 8.596 parallel_gemm_fm 14 9.1 0.000 0.000 8.399 8.483 parallel_gemm_fm_cosma 14 10.1 8.399 8.483 8.399 8.483 potential_pw2rs 545 10.0 0.107 0.109 7.676 8.343 create_integ_mat 1 6.0 0.014 0.027 7.908 7.918 collocate_single_gaussian 272 10.0 0.040 0.042 7.484 7.794 array2fm 1 7.0 0.000 0.000 6.694 7.243 pw_scatter_s 2720 13.7 4.443 4.567 4.443 4.567 pw_gather_s 2722 13.2 3.869 4.211 3.869 4.211 array2fm_buffer_send 1 8.0 2.915 3.127 2.915 3.127 pw_poisson_solve 545 10.5 1.123 1.174 2.177 2.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.018183, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2734.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 0.033 0.047 402.398 402.399 farming_run 1 2.0 400.171 400.178 402.348 402.353 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.229242E+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 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 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 0.052 209.681 209.681 qs_energies 1 2.0 0.000 0.000 209.446 209.459 scf_env_do_scf 1 3.0 0.000 0.000 106.314 106.314 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.391 105.398 rebuild_ks_matrix 4 6.0 0.000 0.000 105.389 105.396 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.065 105.389 105.396 hfx_ks_matrix 4 8.0 0.001 0.001 105.004 105.008 integrate_four_center 4 9.0 0.144 0.453 105.004 105.007 mp2_main 1 3.0 0.000 0.000 102.840 102.852 mp2_gpw_main 1 4.0 0.034 0.053 101.909 101.924 integrate_four_center_main 4 10.0 0.079 0.465 96.613 99.652 integrate_four_center_bin 264 11.0 96.534 99.627 96.534 99.627 init_scf_loop 1 4.0 0.000 0.000 91.985 91.985 mp2_ri_gpw_compute_in 1 5.0 0.065 0.079 74.993 76.092 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.426 55.523 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.158 47.108 integrate_v_rspace 95 8.0 0.397 0.559 28.545 33.360 pw_transfer 2240 10.6 0.146 0.188 29.915 30.289 ao_to_mo_and_store_B_mult_1 91 7.0 10.588 29.605 10.588 29.605 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.918 29.287 grid_integrate_task_list 95 9.0 23.840 28.802 23.840 28.802 mp2_ri_gpw_compute_en 1 5.0 0.064 0.077 26.757 28.482 fft_wrap_pw1pw2_100 730 12.4 1.287 1.522 26.647 27.060 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.888 25.003 25.015 get_2c_integrals 1 6.0 0.005 0.043 20.452 20.506 compute_2c_integrals 1 7.0 0.004 0.018 19.417 19.426 compute_2c_integrals_loop_lm 1 8.0 0.002 0.011 18.903 19.261 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.899 18.901 19.261 fft3d_s 1823 13.4 18.435 18.719 18.448 18.733 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.327 14.327 calculate_wavefunction 91 8.0 2.007 2.030 9.732 9.943 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.571 8.823 9.424 potential_pw2rs 186 10.0 0.033 0.036 8.643 9.259 local_gemm 172 8.0 8.269 8.864 8.269 8.864 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.268 8.605 mp2_ri_gpw_compute_en_comm 22 7.0 0.494 0.516 7.964 8.294 calc_potential_gpw 182 9.5 0.002 0.003 7.925 8.214 collocate_single_gaussian 91 10.0 0.017 0.037 7.897 8.143 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.429 6.340 6.429 mp_sendrecv_dm3 2068 8.0 6.018 6.335 6.018 6.335 mp_sync 38 10.4 3.104 5.548 3.104 5.548 pw_gather_s 912 13.2 4.894 5.373 4.894 5.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.893841, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 451.907584E+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 636002. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 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 0.031 0.070 56.855 56.857 qs_mol_dyn_low 1 2.0 0.003 0.004 56.147 56.418 qs_forces 11 3.9 0.002 0.002 54.776 54.778 qs_energies 11 4.9 0.005 0.033 53.298 53.309 scf_env_do_scf 11 5.9 0.000 0.001 46.136 46.136 scf_env_do_scf_inner_loop 108 6.5 0.003 0.017 44.055 44.055 dbcsr_multiply_generic 2286 12.5 0.093 0.097 35.894 36.340 qs_scf_new_mos 108 7.5 0.000 0.000 33.772 34.073 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.772 34.073 ot_scf_mini 108 9.5 0.002 0.002 32.067 32.271 multiply_cannon 2286 13.5 0.184 0.206 27.802 29.536 multiply_cannon_loop 2286 14.5 1.497 1.575 27.026 28.769 velocity_verlet 10 3.0 0.001 0.002 27.680 27.694 ot_mini 108 10.5 0.001 0.001 20.508 20.756 qs_ot_get_derivative 108 11.5 0.001 0.001 17.452 17.616 mp_waitall_1 245248 16.5 9.593 16.070 9.593 16.070 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.710 14.225 multiply_cannon_multrec 54864 15.5 4.152 6.603 8.138 11.766 rebuild_ks_matrix 119 8.3 0.000 0.000 7.793 7.923 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.792 7.922 mp_sum_l 7207 12.9 5.849 7.605 5.849 7.605 multiply_cannon_sync_h2d 54864 15.5 5.814 7.202 5.814 7.202 qs_ot_get_p 119 10.4 0.001 0.001 6.777 7.059 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.874 6.992 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.101 6.671 init_scf_run 11 5.9 0.000 0.001 5.874 5.874 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.874 5.874 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.489 5.593 dbcsr_mm_accdrv_process 76910 16.1 1.199 1.845 3.909 5.507 sum_up_and_integrate 119 10.3 0.012 0.014 4.439 4.445 integrate_v_rspace 119 11.3 0.002 0.002 4.427 4.435 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.113 4.215 calculate_rho_elec 119 8.7 0.011 0.016 4.113 4.214 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 2.147 3.923 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.798 3.840 calculate_first_density_matrix 1 7.0 0.000 0.001 3.680 3.697 jit_kernel_multiply 13 15.8 2.646 3.621 2.646 3.621 calculate_dm_sparse 119 9.5 0.000 0.000 3.364 3.499 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.078 3.325 apply_single 119 13.6 0.000 0.001 3.078 3.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.969 2.971 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.950 2.950 rs_pw_transfer 974 11.9 0.012 0.013 2.743 2.848 ot_diis_step 108 11.5 0.006 0.006 2.766 2.767 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.686 2.688 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.622 2.673 cp_fm_redistribute_end 50 14.0 2.446 2.664 2.451 2.667 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.562 2.623 acc_transpose_blocks 54864 15.5 0.231 0.251 2.023 2.611 cp_fm_diag_elpa_base 50 14.0 0.212 2.561 0.213 2.569 density_rs2pw 119 9.7 0.004 0.004 2.153 2.239 grid_integrate_task_list 119 12.3 2.031 2.132 2.031 2.132 wfi_extrapolate 11 7.9 0.001 0.001 2.126 2.126 init_scf_loop 11 6.9 0.000 0.000 2.064 2.064 mp_sum_d 4125 12.0 1.412 2.048 1.412 2.048 make_m2s 4572 13.5 0.054 0.056 1.933 1.980 make_images 4572 14.5 0.135 0.140 1.850 1.896 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.819 1.853 potential_pw2rs 119 12.3 0.004 0.004 1.815 1.826 pw_transfer 1439 11.6 0.051 0.054 1.599 1.665 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.524 1.592 mp_alltoall_d11v 2130 13.8 1.372 1.504 1.372 1.504 acc_transpose_blocks_kernels 54864 16.5 0.254 0.385 1.025 1.465 md_write_output 11 3.9 0.005 0.455 0.017 1.443 mp_sum_dm 438 4.9 1.381 1.425 1.381 1.425 fft3d_ps 1201 14.6 0.356 0.463 1.302 1.366 mp_waitany 12084 13.8 1.246 1.359 1.246 1.359 update_particle_set 20 4.0 0.000 0.000 1.335 1.350 grid_collocate_task_list 119 9.7 1.288 1.348 1.288 1.348 md_output 10 3.0 0.000 0.000 0.016 1.299 fft_wrap_pw1pw2_140 487 13.2 0.080 0.092 1.146 1.214 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.784 1.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.857000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 488.837120E+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 1047341. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 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 0.027 0.028 46.637 46.638 qs_mol_dyn_low 1 2.0 0.003 0.004 45.897 46.203 qs_forces 11 3.9 0.015 0.016 44.290 44.290 qs_energies 11 4.9 0.001 0.002 42.534 42.539 scf_env_do_scf 11 5.9 0.000 0.001 35.747 35.748 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 33.155 33.155 dbcsr_multiply_generic 2286 12.5 0.102 0.105 26.216 26.618 qs_scf_new_mos 108 7.5 0.001 0.001 23.659 23.904 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.658 23.903 ot_scf_mini 108 9.5 0.002 0.003 22.734 22.908 multiply_cannon 2286 13.5 0.208 0.219 20.036 22.069 multiply_cannon_loop 2286 14.5 0.896 0.975 18.795 20.970 velocity_verlet 10 3.0 0.001 0.002 20.703 20.736 ot_mini 108 10.5 0.001 0.001 14.966 15.198 mp_waitall_1 200699 16.5 7.937 14.460 7.937 14.460 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 6.271 12.651 qs_ot_get_derivative 108 11.5 0.001 0.001 12.421 12.592 multiply_cannon_multrec 27432 15.5 2.024 4.766 6.932 9.959 dbcsr_mm_accdrv_process 47894 16.0 3.650 6.687 4.840 7.454 rebuild_ks_matrix 119 8.3 0.000 0.000 7.256 7.398 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.255 7.397 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.398 6.523 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.054 6.491 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.773 5.685 apply_single 119 13.6 0.000 0.000 3.773 5.685 init_scf_run 11 5.9 0.000 0.001 5.567 5.567 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.567 5.567 mp_sum_l 7207 12.9 2.947 5.412 2.947 5.412 qs_ot_get_p 119 10.4 0.001 0.001 4.762 4.998 sum_up_and_integrate 119 10.3 0.024 0.027 4.242 4.246 integrate_v_rspace 119 11.3 0.002 0.002 4.218 4.226 calculate_first_density_matrix 1 7.0 0.000 0.000 3.938 3.947 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.711 3.781 calculate_rho_elec 119 8.7 0.021 0.024 3.710 3.780 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.165 3.187 jit_kernel_multiply 10 16.1 1.138 3.108 1.138 3.108 make_m2s 4572 13.5 0.052 0.054 2.778 3.036 make_images 4572 14.5 0.200 0.235 2.688 2.946 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.892 2.895 calculate_dm_sparse 119 9.5 0.000 0.000 2.760 2.838 rs_pw_transfer 974 11.9 0.010 0.011 2.700 2.816 multiply_cannon_sync_h2d 27432 15.5 2.177 2.793 2.177 2.793 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.674 2.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.543 2.585 init_scf_loop 11 6.9 0.000 0.000 2.570 2.570 ot_diis_step 108 11.5 0.010 0.011 2.479 2.480 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.315 2.403 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.275 2.276 cp_fm_redistribute_end 50 14.0 1.877 2.242 1.882 2.243 acc_transpose_blocks 27432 15.5 0.107 0.111 1.648 2.237 cp_fm_diag_elpa_base 50 14.0 0.346 2.119 0.360 2.174 density_rs2pw 119 9.7 0.004 0.004 2.066 2.163 potential_pw2rs 119 12.3 0.006 0.006 1.979 1.988 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.946 1.961 grid_integrate_task_list 119 12.3 1.831 1.931 1.831 1.931 pw_transfer 1439 11.6 0.063 0.066 1.777 1.808 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.687 1.721 make_images_data 4572 15.5 0.044 0.051 1.235 1.700 md_write_output 11 3.9 0.012 0.535 0.038 1.684 mp_sum_dm 438 4.9 1.615 1.672 1.615 1.672 prepare_preconditioner 11 7.9 0.000 0.000 1.646 1.672 make_preconditioner 11 8.9 0.000 0.000 1.646 1.672 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 1.120 1.652 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.531 1.587 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.076 1.586 update_particle_set 20 4.0 0.000 0.000 1.551 1.585 wfi_extrapolate 11 7.9 0.001 0.001 1.572 1.572 md_output 10 3.0 0.000 0.000 0.034 1.511 mp_allgather_i34 2286 14.5 0.684 1.464 0.684 1.464 mp_alltoall_d11v 2130 13.8 1.299 1.451 1.299 1.451 fft3d_ps 1201 14.6 0.496 0.551 1.395 1.426 jit_kernel_transpose 5 15.5 0.939 1.392 0.939 1.392 grid_collocate_task_list 119 9.7 1.241 1.293 1.241 1.293 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.240 1.289 fft_wrap_pw1pw2_140 487 13.2 0.075 0.084 1.253 1.285 write_trajectory 44 4.9 0.025 1.122 0.025 1.148 mp_sum_d 4125 12.0 0.627 1.071 0.627 1.071 make_images_sizes 4572 15.5 0.005 0.005 0.701 0.980 mp_alltoall_i44 4572 16.5 0.697 0.975 0.697 0.975 qs_energies_init_hamiltonians 11 5.9 0.011 0.026 0.947 0.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.916 0.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=46.638000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.256960E+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_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.054 0.083 40.783 40.785 qs_mol_dyn_low 1 2.0 0.003 0.004 40.005 40.327 qs_forces 11 3.9 0.002 0.002 38.519 38.520 qs_energies 11 4.9 0.001 0.004 36.834 36.837 scf_env_do_scf 11 5.9 0.002 0.015 30.862 30.862 scf_env_do_scf_inner_loop 108 6.5 0.011 0.077 27.763 27.766 dbcsr_multiply_generic 2286 12.5 0.096 0.099 20.164 20.395 velocity_verlet 10 3.0 0.001 0.002 18.750 18.795 qs_scf_new_mos 108 7.5 0.001 0.001 18.522 18.541 qs_scf_loop_do_ot 108 8.5 0.002 0.010 18.521 18.541 ot_scf_mini 108 9.5 0.003 0.005 17.696 17.714 multiply_cannon 2286 13.5 0.197 0.204 15.368 16.213 multiply_cannon_loop 2286 14.5 0.635 0.667 14.379 15.184 ot_mini 108 10.5 0.001 0.001 11.172 11.208 qs_ot_get_derivative 108 11.5 0.001 0.001 9.533 9.553 multiply_cannon_multrec 18288 15.5 2.023 3.181 7.609 7.961 rebuild_ks_matrix 119 8.3 0.000 0.000 7.075 7.120 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 7.075 7.119 dbcsr_mm_accdrv_process 38222 16.0 4.808 6.523 5.504 6.581 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.289 6.331 mp_waitall_1 158411 16.6 4.277 5.440 4.277 5.440 init_scf_run 11 5.9 0.000 0.001 4.706 4.706 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.706 4.706 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.649 4.684 sum_up_and_integrate 119 10.3 0.031 0.032 4.549 4.558 integrate_v_rspace 119 11.3 0.002 0.003 4.518 4.529 qs_ot_get_p 119 10.4 0.001 0.001 4.038 4.074 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.706 3.725 calculate_rho_elec 119 8.7 0.030 0.031 3.706 3.724 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 2.354 3.344 calculate_first_density_matrix 1 7.0 0.000 0.002 3.264 3.265 rs_pw_transfer 974 11.9 0.009 0.010 2.996 3.103 init_scf_loop 11 6.9 0.003 0.023 3.070 3.071 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.497 2.996 apply_single 119 13.6 0.000 0.000 2.496 2.996 mp_sum_l 7207 12.9 2.070 2.594 2.070 2.594 make_m2s 4572 13.5 0.045 0.046 2.359 2.572 jit_kernel_multiply 10 16.2 0.644 2.522 0.644 2.522 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.500 2.512 make_images 4572 14.5 0.192 0.204 2.272 2.486 density_rs2pw 119 9.7 0.004 0.004 2.274 2.371 calculate_dm_sparse 119 9.5 0.000 0.000 2.331 2.343 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.317 2.318 potential_pw2rs 119 12.3 0.007 0.008 2.286 2.294 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.156 2.157 prepare_preconditioner 11 7.9 0.000 0.000 2.139 2.142 make_preconditioner 11 8.9 0.005 0.042 2.139 2.142 pw_transfer 1439 11.6 0.064 0.066 2.066 2.086 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.974 2.064 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.009 2.030 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.975 1.994 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.950 1.958 grid_integrate_task_list 119 12.3 1.799 1.907 1.799 1.907 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.829 1.831 cp_fm_redistribute_end 50 14.0 1.360 1.798 1.362 1.800 cp_fm_diag_elpa_base 50 14.0 0.420 1.697 0.434 1.748 fft3d_ps 1201 14.6 0.508 0.524 1.662 1.684 acc_transpose_blocks 18288 15.5 0.074 0.076 1.571 1.649 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.602 1.607 ot_diis_step 108 11.5 0.011 0.012 1.597 1.598 md_write_output 11 3.9 0.036 1.126 0.050 1.559 mp_sum_dm 438 4.9 1.480 1.531 1.480 1.531 multiply_cannon_sync_h2d 18288 15.5 1.303 1.514 1.303 1.514 update_particle_set 20 4.0 0.000 0.000 1.414 1.460 md_output 10 3.0 0.000 0.000 0.047 1.424 wfi_extrapolate 11 7.9 0.001 0.001 1.390 1.390 fft_wrap_pw1pw2_140 487 13.2 0.085 0.089 1.365 1.383 make_images_data 4572 15.5 0.045 0.049 1.056 1.288 grid_collocate_task_list 119 9.7 1.203 1.282 1.203 1.282 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 1.132 1.208 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.116 1.147 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.894 1.139 mp_alltoall_z22v 1201 16.6 0.951 1.014 0.951 1.014 mp_alltoall_d11v 2130 13.8 0.915 1.014 0.915 1.014 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.002 1.003 jit_kernel_transpose 5 15.6 0.922 1.002 0.922 1.002 cp_fm_cholesky_invert 11 10.9 0.936 0.941 0.936 0.941 mp_allgather_i34 2286 14.5 0.449 0.903 0.449 0.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.876 0.884 make_images_sizes 4572 15.5 0.005 0.005 0.615 0.828 mp_alltoall_i44 4572 16.5 0.610 0.823 0.610 0.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=40.785000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.945540 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 554.123264E+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_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.129 0.219 38.182 38.183 qs_mol_dyn_low 1 2.0 0.003 0.005 37.680 37.687 qs_forces 11 3.9 0.002 0.002 37.619 37.619 qs_energies 11 4.9 0.016 0.063 35.773 35.780 scf_env_do_scf 11 5.9 0.003 0.015 29.636 29.638 scf_env_do_scf_inner_loop 108 6.5 0.004 0.016 25.503 25.507 dbcsr_multiply_generic 2286 12.5 0.102 0.107 18.709 18.829 velocity_verlet 10 3.0 0.002 0.002 18.372 18.374 qs_scf_new_mos 108 7.5 0.001 0.001 16.797 16.850 qs_scf_loop_do_ot 108 8.5 0.005 0.019 16.796 16.850 ot_scf_mini 108 9.5 0.009 0.028 15.825 15.884 multiply_cannon 2286 13.5 0.228 0.264 14.685 15.559 multiply_cannon_loop 2286 14.5 0.934 0.966 13.669 14.377 ot_mini 108 10.5 0.001 0.001 9.650 9.727 multiply_cannon_multrec 27432 15.5 2.350 3.045 8.597 8.957 qs_ot_get_derivative 108 11.5 0.001 0.001 7.825 7.889 dbcsr_mm_accdrv_process 47916 15.9 5.245 6.849 6.153 7.366 rebuild_ks_matrix 119 8.3 0.000 0.000 7.109 7.170 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.109 7.169 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.279 6.320 init_scf_run 11 5.9 0.001 0.006 4.525 4.535 scf_env_initial_rho_setup 11 6.9 0.013 0.051 4.524 4.535 sum_up_and_integrate 119 10.3 0.035 0.038 4.153 4.172 integrate_v_rspace 119 11.3 0.002 0.003 4.117 4.137 init_scf_loop 11 6.9 0.009 0.035 4.102 4.112 qs_ot_get_p 119 10.4 0.002 0.008 3.755 3.837 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.600 3.632 calculate_rho_elec 119 8.7 0.040 0.046 3.599 3.631 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.962 3.296 mp_waitall_1 137007 16.6 2.441 3.055 2.441 3.055 calculate_first_density_matrix 1 7.0 0.004 0.018 3.001 3.024 prepare_preconditioner 11 7.9 0.000 0.000 2.778 2.785 make_preconditioner 11 8.9 0.006 0.049 2.778 2.785 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.375 2.706 make_m2s 4572 13.5 0.055 0.056 2.539 2.673 rs_pw_transfer 974 11.9 0.009 0.009 2.501 2.645 make_images 4572 14.5 0.273 0.337 2.431 2.564 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.975 2.163 apply_single 119 13.6 0.000 0.000 1.975 2.163 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.132 2.137 calculate_dm_sparse 119 9.5 0.000 0.000 2.062 2.116 density_rs2pw 119 9.7 0.004 0.004 1.996 2.110 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.091 2.102 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.065 2.091 pw_transfer 1439 11.6 0.063 0.067 1.985 2.025 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.895 1.938 potential_pw2rs 119 12.3 0.008 0.009 1.903 1.921 grid_integrate_task_list 119 12.3 1.809 1.878 1.809 1.878 jit_kernel_multiply 10 15.8 0.847 1.816 0.847 1.816 ot_diis_step 108 11.5 0.012 0.012 1.785 1.785 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.752 1.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.731 1.749 fft3d_ps 1201 14.6 0.537 0.588 1.577 1.613 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.521 1.566 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.959 1.548 mp_sum_l 7207 12.9 1.085 1.529 1.085 1.529 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.467 1.471 acc_transpose_blocks 27432 15.5 0.111 0.113 1.453 1.471 cp_fm_redistribute_end 50 14.0 0.973 1.441 0.975 1.442 cp_fm_diag_elpa_base 50 14.0 0.442 1.359 0.462 1.399 wfi_extrapolate 11 7.9 0.001 0.001 1.386 1.387 dbcsr_complete_redistribute 329 12.2 0.252 0.328 1.064 1.338 grid_collocate_task_list 119 9.7 1.230 1.314 1.230 1.314 qs_energies_init_hamiltonians 11 5.9 0.002 0.012 1.306 1.307 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.219 1.229 make_images_data 4572 15.5 0.045 0.049 1.021 1.200 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.141 1.161 cp_fm_upper_to_full 72 13.5 0.822 1.132 0.822 1.132 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.863 1.087 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.816 1.085 mp_sum_d 4123 12.0 0.773 1.077 0.773 1.077 mp_alltoall_d11v 2130 13.8 0.920 1.052 0.920 1.052 multiply_cannon_sync_h2d 27432 15.5 0.977 1.044 0.977 1.044 mp_alltoall_z22v 1201 16.6 0.922 0.963 0.922 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.933 0.939 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.872 cp_fm_cholesky_invert 11 10.9 0.862 0.867 0.862 0.867 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.829 0.842 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.760 0.812 arnoldi_extremal 119 11.4 0.004 0.026 0.775 0.798 arnoldi_normal_ev 119 12.4 0.017 0.120 0.771 0.797 mp_waitany 7280 13.7 0.621 0.781 0.621 0.781 mp_alltoall_i22 627 13.8 0.470 0.771 0.470 0.771 rs_pw_transfer_RS2PW_140 130 11.5 0.107 0.112 0.620 0.767 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.183000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.363636, yerr=2.422348 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 607.440896E+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_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.013 0.033 28.760 28.761 qs_mol_dyn_low 1 2.0 0.003 0.004 28.480 28.487 qs_forces 11 3.9 0.002 0.002 28.420 28.421 qs_energies 11 4.9 0.001 0.001 26.738 26.741 scf_env_do_scf 11 5.9 0.000 0.001 21.917 21.918 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.284 19.285 velocity_verlet 10 3.0 0.002 0.002 14.521 14.524 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.748 12.869 qs_scf_new_mos 108 7.5 0.001 0.001 11.647 11.667 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.647 11.666 ot_scf_mini 108 9.5 0.002 0.002 10.932 10.955 multiply_cannon 2286 13.5 0.230 0.241 10.065 10.465 multiply_cannon_loop 2286 14.5 0.329 0.340 9.095 9.330 multiply_cannon_multrec 9144 15.5 1.685 2.108 6.022 6.403 ot_mini 108 10.5 0.001 0.001 6.124 6.157 rebuild_ks_matrix 119 8.3 0.000 0.000 5.878 5.898 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.877 5.898 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.259 5.277 qs_ot_get_derivative 108 11.5 0.001 0.001 4.798 4.821 dbcsr_mm_accdrv_process 12550 15.8 3.181 4.237 4.235 4.310 sum_up_and_integrate 119 10.3 0.037 0.041 3.570 3.574 integrate_v_rspace 119 11.3 0.003 0.003 3.532 3.537 init_scf_run 11 5.9 0.000 0.001 3.383 3.383 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.382 3.382 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.232 3.239 calculate_rho_elec 119 8.7 0.059 0.061 3.232 3.238 qs_ot_get_p 119 10.4 0.001 0.001 3.041 3.092 init_scf_loop 11 6.9 0.000 0.000 2.611 2.612 mp_waitall_1 115863 16.7 1.730 2.287 1.730 2.287 calculate_first_density_matrix 1 7.0 0.000 0.000 2.247 2.248 make_m2s 4572 13.5 0.035 0.036 1.851 2.067 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.986 1.988 make_images 4572 14.5 0.269 0.303 1.762 1.975 jit_kernel_multiply 10 15.8 1.016 1.931 1.016 1.931 grid_integrate_task_list 119 12.3 1.844 1.909 1.844 1.909 calculate_dm_sparse 119 9.5 0.000 0.000 1.819 1.838 rs_pw_transfer 974 11.9 0.008 0.008 1.726 1.797 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.781 1.781 prepare_preconditioner 11 7.9 0.000 0.000 1.770 1.774 make_preconditioner 11 8.9 0.000 0.000 1.770 1.774 density_rs2pw 119 9.7 0.003 0.003 1.657 1.731 pw_transfer 1439 11.6 0.063 0.066 1.700 1.707 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.662 1.690 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.644 1.657 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.639 1.640 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.609 1.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.460 1.470 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.465 1.466 cp_fm_redistribute_end 50 14.0 0.731 1.441 0.732 1.442 cp_fm_diag_elpa_base 50 14.0 0.664 1.369 0.709 1.426 grid_collocate_task_list 119 9.7 1.272 1.377 1.272 1.377 potential_pw2rs 119 12.3 0.010 0.010 1.359 1.362 ot_diis_step 108 11.5 0.012 0.013 1.311 1.311 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.282 1.289 fft3d_ps 1201 14.6 0.535 0.547 1.275 1.284 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.221 1.222 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.202 1.207 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.176 1.199 apply_single 119 13.6 0.000 0.000 1.176 1.199 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.165 1.177 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.848 1.153 make_images_data 4572 15.5 0.039 0.042 0.884 1.130 wfi_extrapolate 11 7.9 0.001 0.001 1.092 1.092 acc_transpose_blocks 9144 15.5 0.038 0.039 0.944 0.954 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.921 cp_fm_cholesky_invert 11 10.9 0.917 0.920 0.917 0.920 mp_alltoall_d11v 2130 13.8 0.815 0.911 0.815 0.911 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.432 0.814 multiply_cannon_sync_h2d 9144 15.5 0.712 0.802 0.712 0.802 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.760 0.768 mp_allgather_i34 2286 14.5 0.304 0.763 0.304 0.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.741 0.743 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.677 0.724 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.705 0.709 mp_alltoall_z22v 1201 16.6 0.614 0.650 0.614 0.650 jit_kernel_transpose 5 15.6 0.588 0.590 0.588 0.590 qs_create_task_list 11 7.9 0.000 0.000 0.556 0.579 generate_qs_task_list 11 8.9 0.190 0.212 0.556 0.578 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.761000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.727273, yerr=7.580968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 773.554176E+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_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 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 0.014 0.033 41.845 41.845 qs_mol_dyn_low 1 2.0 0.003 0.004 41.543 41.550 qs_forces 11 3.9 0.009 0.014 41.479 41.482 qs_energies 11 4.9 0.001 0.002 39.528 39.541 scf_env_do_scf 11 5.9 0.001 0.001 33.369 33.370 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.322 25.324 velocity_verlet 10 3.0 0.002 0.002 23.283 23.289 dbcsr_multiply_generic 2286 12.5 0.112 0.114 17.936 18.181 qs_scf_new_mos 108 7.5 0.001 0.001 16.322 16.415 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.322 16.415 ot_scf_mini 108 9.5 0.002 0.002 15.226 15.321 multiply_cannon 2286 13.5 0.300 0.305 13.822 14.852 multiply_cannon_loop 2286 14.5 0.342 0.346 12.481 13.485 ot_mini 108 10.5 0.001 0.001 9.025 9.140 multiply_cannon_multrec 9144 15.5 3.433 4.894 8.590 8.717 init_scf_loop 11 6.9 0.000 0.000 8.020 8.024 rebuild_ks_matrix 119 8.3 0.000 0.000 6.949 7.087 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.948 7.087 qs_ot_get_derivative 108 11.5 0.001 0.001 6.989 7.086 prepare_preconditioner 11 7.9 0.000 0.000 7.061 7.074 make_preconditioner 11 8.9 0.000 0.000 7.061 7.074 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.638 6.951 dbcsr_mm_accdrv_process 12550 15.8 3.961 5.644 5.030 6.442 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.296 6.422 cp_fm_upper_to_full 72 14.2 3.204 4.567 3.204 4.567 init_scf_run 11 5.9 0.000 0.001 3.992 3.992 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.992 3.992 sum_up_and_integrate 119 10.3 0.064 0.065 3.755 3.761 integrate_v_rspace 119 11.3 0.003 0.004 3.691 3.698 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.693 3.697 calculate_rho_elec 119 8.7 0.118 0.121 3.692 3.696 mp_waitall_1 94719 16.7 2.600 3.597 2.600 3.597 qs_ot_get_p 119 10.4 0.001 0.001 3.412 3.543 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.467 2.934 dbcsr_complete_redistribute 329 12.2 0.288 0.293 1.990 2.827 make_m2s 4572 13.5 0.038 0.039 2.414 2.633 calculate_first_density_matrix 1 7.0 0.000 0.000 2.543 2.545 make_images 4572 14.5 0.354 0.387 2.293 2.511 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.250 2.506 apply_single 119 13.6 0.000 0.000 2.250 2.505 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.665 2.493 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.400 2.339 mp_alltoall_i22 627 13.8 1.399 2.273 1.399 2.273 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.418 2.239 calculate_dm_sparse 119 9.5 0.000 0.000 2.220 2.237 pw_transfer 1439 11.6 0.067 0.067 2.060 2.063 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.055 2.056 grid_integrate_task_list 119 12.3 2.014 2.043 2.014 2.043 ot_diis_step 108 11.5 0.014 0.014 2.009 2.010 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.989 1.990 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.965 1.967 qs_energies_init_hamiltonians 11 5.9 0.008 0.019 1.893 1.903 jit_kernel_multiply 10 15.5 1.041 1.846 1.041 1.846 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.780 1.829 density_rs2pw 119 9.7 0.003 0.003 1.794 1.813 mp_sum_l 7207 12.9 1.147 1.785 1.147 1.785 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.768 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.721 1.768 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.637 1.642 fft3d_ps 1201 14.6 0.567 0.576 1.599 1.602 cp_fm_cholesky_invert 11 10.9 1.585 1.589 1.585 1.589 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.208 1.515 grid_collocate_task_list 119 9.7 1.454 1.469 1.454 1.469 rs_pw_transfer 974 11.9 0.009 0.009 1.441 1.464 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.454 1.455 make_images_data 4572 15.5 0.043 0.046 1.169 1.453 cp_fm_diag_elpa_base 50 14.0 1.307 1.362 1.453 1.453 wfi_extrapolate 11 7.9 0.001 0.001 1.385 1.385 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.271 1.277 potential_pw2rs 119 12.3 0.014 0.015 1.262 1.266 mp_alltoall_d11v 2130 13.8 1.228 1.253 1.228 1.253 qs_env_update_s_mstruct 11 6.9 0.011 0.021 1.211 1.248 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.166 1.189 multiply_cannon_sync_h2d 9144 15.5 1.043 1.048 1.043 1.048 qs_create_task_list 11 7.9 0.053 0.091 1.034 1.043 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.986 1.031 generate_qs_task_list 11 8.9 0.374 0.393 0.981 1.024 acc_transpose_blocks 9144 15.5 0.037 0.037 0.922 0.963 mp_alltoall_z22v 1201 16.6 0.896 0.917 0.896 0.917 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.845000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=722.909091, yerr=18.377133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.493760E+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 822667. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.028 0.050 85.865 85.867 qs_mol_dyn_low 1 2.0 0.006 0.029 85.289 85.299 qs_forces 11 3.9 0.002 0.003 85.154 85.155 qs_energies 11 4.9 0.004 0.023 82.281 82.301 scf_env_do_scf 11 5.9 0.000 0.001 72.848 72.850 scf_env_do_scf_inner_loop 99 6.5 0.005 0.024 67.015 67.016 dbcsr_multiply_generic 2055 12.4 0.109 0.114 53.165 53.538 qs_scf_new_mos 99 7.5 0.000 0.001 49.677 49.817 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.676 49.816 ot_scf_mini 99 9.5 0.002 0.003 47.233 47.313 multiply_cannon 2055 13.4 0.174 0.181 43.090 43.995 multiply_cannon_loop 2055 14.4 1.540 1.574 42.050 42.977 velocity_verlet 10 3.0 0.021 0.041 42.880 42.881 ot_mini 99 10.5 0.001 0.001 28.722 28.826 qs_ot_get_derivative 99 11.5 0.001 0.001 21.833 21.922 multiply_cannon_multrec 49320 15.4 12.084 12.917 17.295 18.141 rebuild_ks_matrix 110 8.3 0.000 0.001 14.618 14.833 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.617 14.832 mp_waitall_1 220248 16.4 12.063 13.275 12.063 13.275 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.849 13.041 multiply_cannon_sync_h2d 49320 15.4 9.902 10.743 9.902 10.743 qs_ot_get_p 110 10.4 0.001 0.007 9.896 10.069 multiply_cannon_metrocomm3 49320 15.4 0.078 0.084 7.117 8.398 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.385 7.908 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.170 7.685 apply_single 110 13.6 0.000 0.001 7.170 7.685 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 7.181 7.244 sum_up_and_integrate 110 10.3 0.036 0.043 7.150 7.161 init_scf_run 11 5.9 0.000 0.001 7.140 7.140 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.139 7.140 integrate_v_rspace 110 11.3 0.002 0.003 7.114 7.133 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.461 6.609 calculate_rho_elec 110 8.6 0.020 0.024 6.461 6.608 ot_diis_step 99 11.5 0.005 0.006 6.580 6.580 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.522 6.553 mp_sum_l 6514 12.8 5.442 6.356 5.442 6.356 init_scf_loop 11 6.9 0.000 0.002 5.799 5.799 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.479 5.479 dbcsr_mm_accdrv_process 87628 16.1 2.104 2.225 5.092 5.427 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.958 4.960 cp_fm_redistribute_end 48 14.0 4.325 4.930 4.329 4.931 cp_fm_diag_elpa_base 48 14.0 0.595 4.785 0.599 4.814 make_m2s 4110 13.4 0.062 0.068 4.185 4.278 rs_pw_transfer 902 11.9 0.012 0.014 4.034 4.205 make_images 4110 14.4 0.179 0.193 4.087 4.181 wfi_extrapolate 11 7.9 0.001 0.001 4.086 4.086 calculate_dm_sparse 110 9.5 0.000 0.001 3.908 4.018 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.894 3.892 density_rs2pw 110 9.6 0.004 0.004 3.576 3.767 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.680 3.685 prepare_preconditioner 11 7.9 0.000 0.000 3.582 3.610 make_preconditioner 11 8.9 0.000 0.002 3.582 3.610 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.370 3.430 grid_integrate_task_list 110 12.3 3.250 3.423 3.250 3.423 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.347 3.400 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.261 3.333 pw_transfer 1331 11.6 0.054 0.064 3.121 3.188 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.034 3.104 calculate_first_density_matrix 1 7.0 0.000 0.002 2.961 2.967 potential_pw2rs 110 12.3 0.006 0.007 2.778 2.801 jit_kernel_multiply 13 15.9 2.699 2.734 2.699 2.734 fft3d_ps 1111 14.6 0.752 0.838 2.572 2.630 fft_wrap_pw1pw2_140 451 13.1 0.168 0.189 2.513 2.585 mp_alltoall_d11v 2046 13.8 2.111 2.463 2.111 2.463 acc_transpose_blocks 49320 15.4 0.223 0.234 2.236 2.306 grid_collocate_task_list 110 9.6 2.086 2.238 2.086 2.238 mp_sum_d 3879 11.9 1.563 2.095 1.563 2.095 mp_waitany 14300 13.8 1.812 1.972 1.812 1.972 cp_fm_cholesky_invert 11 10.9 1.958 1.962 1.958 1.962 make_images_data 4110 15.4 0.043 0.046 1.828 1.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.912 1.943 hybrid_alltoall_any 4261 16.3 0.082 0.477 1.581 1.813 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.743 1.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.867000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=2.065058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 586.317824E+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 1773891. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.015 0.040 69.228 69.229 qs_mol_dyn_low 1 2.0 0.003 0.004 68.998 69.017 qs_forces 11 3.9 0.002 0.002 68.918 68.921 qs_energies 11 4.9 0.001 0.002 65.621 65.627 scf_env_do_scf 11 5.9 0.000 0.001 56.998 57.001 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.085 49.085 dbcsr_multiply_generic 2055 12.4 0.115 0.120 37.938 38.205 velocity_verlet 10 3.0 0.001 0.002 36.472 36.475 qs_scf_new_mos 99 7.5 0.001 0.001 32.998 33.142 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.998 33.141 multiply_cannon 2055 13.4 0.222 0.247 31.114 32.259 ot_scf_mini 99 9.5 0.003 0.003 31.330 31.463 multiply_cannon_loop 2055 14.4 0.924 0.952 29.778 30.588 ot_mini 99 10.5 0.001 0.001 18.356 18.522 multiply_cannon_multrec 24660 15.4 7.577 9.677 13.802 16.105 rebuild_ks_matrix 110 8.3 0.000 0.001 13.798 13.974 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.797 13.973 qs_ot_get_derivative 99 11.5 0.001 0.002 12.500 12.638 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.191 12.347 mp_waitall_1 176588 16.5 7.876 10.241 7.876 10.241 multiply_cannon_sync_h2d 24660 15.4 7.039 8.237 7.039 8.237 multiply_cannon_metrocomm3 24660 15.4 0.069 0.072 5.277 8.057 init_scf_loop 11 6.9 0.000 0.000 7.872 7.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.564 7.303 apply_single 110 13.6 0.000 0.001 6.564 7.303 sum_up_and_integrate 110 10.3 0.052 0.060 6.650 6.676 integrate_v_rspace 110 11.3 0.002 0.002 6.598 6.627 qs_ot_get_p 110 10.4 0.001 0.001 6.299 6.469 dbcsr_mm_accdrv_process 52282 16.1 4.729 5.533 6.068 6.329 init_scf_run 11 5.9 0.000 0.001 6.216 6.216 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.216 6.216 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.878 5.893 calculate_rho_elec 110 8.6 0.039 0.047 5.877 5.893 prepare_preconditioner 11 7.9 0.000 0.000 5.796 5.828 make_preconditioner 11 8.9 0.000 0.000 5.796 5.828 ot_diis_step 99 11.5 0.010 0.010 5.804 5.804 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.366 5.545 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.680 5.439 make_m2s 4110 13.4 0.057 0.060 4.288 4.729 make_images 4110 14.4 0.400 0.445 4.178 4.616 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.308 4.335 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.859 3.860 pw_transfer 1331 11.6 0.066 0.071 3.384 3.534 wfi_extrapolate 11 7.9 0.001 0.001 3.531 3.531 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.277 3.428 grid_integrate_task_list 110 12.3 3.134 3.330 3.134 3.330 density_rs2pw 110 9.6 0.004 0.004 3.159 3.318 rs_pw_transfer 902 11.9 0.012 0.013 3.091 3.318 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.239 3.305 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.286 3.289 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.284 3.285 cp_fm_redistribute_end 48 14.0 2.455 3.256 2.457 3.256 cp_fm_diag_elpa_base 48 14.0 0.764 3.104 0.795 3.192 calculate_dm_sparse 110 9.5 0.001 0.001 3.021 3.053 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.914 2.979 fft_wrap_pw1pw2_140 451 13.1 0.199 0.217 2.721 2.869 make_images_data 4110 15.4 0.047 0.050 2.401 2.839 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.103 2.813 fft3d_ps 1111 14.6 1.073 1.268 2.637 2.770 cp_fm_cholesky_invert 11 10.9 2.706 2.713 2.706 2.713 calculate_first_density_matrix 1 7.0 0.000 0.000 2.588 2.591 mp_sum_l 6514 12.8 1.917 2.590 1.917 2.590 potential_pw2rs 110 12.3 0.008 0.008 2.505 2.533 jit_kernel_multiply 12 16.4 0.988 2.336 0.988 2.336 grid_collocate_task_list 110 9.6 2.119 2.253 2.119 2.253 mp_alltoall_d11v 2046 13.8 1.851 2.195 1.851 2.195 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.978 2.002 multiply_cannon_metrocomm4 22605 15.4 0.072 0.077 0.785 1.945 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.877 1.880 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.770 1.787 mp_irecv_dv 57340 16.2 0.664 1.732 0.664 1.732 mp_allgather_i34 2055 14.4 0.666 1.725 0.666 1.725 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.587 1.600 acc_transpose_blocks 24660 15.4 0.107 0.110 1.524 1.552 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.476 dbcsr_complete_redistribute 325 12.2 0.250 0.302 1.196 1.471 cp_fm_cholesky_decompose 22 10.9 1.439 1.446 1.439 1.446 mp_sum_d 3879 11.9 0.950 1.397 0.950 1.397 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.229000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.090909, yerr=6.515081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 659.398656E+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_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.033 0.073 60.853 60.854 qs_mol_dyn_low 1 2.0 0.003 0.003 60.492 60.502 qs_forces 11 3.9 0.002 0.002 60.409 60.413 qs_energies 11 4.9 0.003 0.011 57.238 57.245 scf_env_do_scf 11 5.9 0.001 0.001 48.932 48.932 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.920 39.921 velocity_verlet 10 3.0 0.004 0.005 32.945 32.947 dbcsr_multiply_generic 2055 12.4 0.109 0.113 28.719 28.988 qs_scf_new_mos 99 7.5 0.001 0.001 25.330 25.434 qs_scf_loop_do_ot 99 8.5 0.001 0.002 25.329 25.434 ot_scf_mini 99 9.5 0.002 0.003 24.095 24.203 multiply_cannon 2055 13.4 0.212 0.224 22.223 23.424 multiply_cannon_loop 2055 14.4 0.613 0.628 20.971 22.053 ot_mini 99 10.5 0.001 0.001 13.822 13.939 rebuild_ks_matrix 110 8.3 0.000 0.000 12.250 12.391 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.249 12.391 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.825 10.952 multiply_cannon_multrec 16440 15.4 3.913 5.025 9.718 10.672 mp_waitall_1 139946 16.5 7.079 9.904 7.079 9.904 qs_ot_get_derivative 99 11.5 0.001 0.001 9.378 9.488 init_scf_loop 11 6.9 0.000 0.001 8.975 8.975 prepare_preconditioner 11 7.9 0.000 0.000 7.111 7.130 make_preconditioner 11 8.9 0.000 0.001 7.111 7.130 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.316 7.118 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.462 6.816 sum_up_and_integrate 110 10.3 0.060 0.061 6.480 6.494 integrate_v_rspace 110 11.3 0.002 0.003 6.420 6.434 init_scf_run 11 5.9 0.000 0.001 5.818 5.818 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.817 5.818 dbcsr_mm_accdrv_process 34862 16.1 4.646 5.337 5.660 5.773 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.604 5.615 calculate_rho_elec 110 8.6 0.058 0.058 5.603 5.615 qs_ot_get_p 110 10.4 0.001 0.001 5.458 5.597 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.923 5.389 apply_single 110 13.6 0.000 0.000 4.923 5.389 make_m2s 4110 13.4 0.050 0.052 4.224 4.562 make_images 4110 14.4 0.394 0.514 4.107 4.445 ot_diis_step 99 11.5 0.010 0.011 4.404 4.404 multiply_cannon_sync_h2d 16440 15.4 3.694 4.286 3.694 4.286 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.167 3.797 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.721 3.726 grid_integrate_task_list 110 12.3 3.190 3.421 3.190 3.421 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.373 3.373 pw_transfer 1331 11.6 0.065 0.072 3.212 3.220 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.106 3.117 density_rs2pw 110 9.6 0.004 0.004 2.876 3.093 rs_pw_transfer 902 11.9 0.011 0.012 2.829 3.064 wfi_extrapolate 11 7.9 0.001 0.001 3.006 3.007 cp_fm_cholesky_invert 11 10.9 2.859 2.865 2.859 2.865 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.843 2.844 make_images_data 4110 15.4 0.043 0.048 2.395 2.826 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.810 2.812 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.132 2.791 cp_fm_redistribute_end 48 14.0 1.755 2.782 1.757 2.783 cp_fm_diag_elpa_base 48 14.0 0.964 2.627 1.021 2.730 calculate_first_density_matrix 1 7.0 0.001 0.004 2.725 2.727 fft_wrap_pw1pw2_140 451 13.1 0.209 0.212 2.614 2.624 calculate_dm_sparse 110 9.5 0.001 0.001 2.577 2.609 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.466 2.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.468 2.513 fft3d_ps 1111 14.6 1.063 1.075 2.422 2.433 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.867 2.403 grid_collocate_task_list 110 9.6 2.179 2.348 2.179 2.348 mp_sum_l 6514 12.8 1.658 2.344 1.658 2.344 potential_pw2rs 110 12.3 0.010 0.010 2.303 2.311 mp_irecv_dv 48980 15.7 0.798 2.274 0.798 2.274 mp_alltoall_d11v 2046 13.8 1.861 2.042 1.861 2.042 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.016 2.021 dbcsr_complete_redistribute 325 12.2 0.332 0.364 1.527 1.978 jit_kernel_multiply 10 16.4 0.622 1.951 0.622 1.951 cp_fm_upper_to_full 70 13.6 1.377 1.842 1.377 1.842 cp_fm_cholesky_decompose 22 10.9 1.716 1.743 1.716 1.743 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.641 mp_allgather_i34 2055 14.4 0.576 1.583 0.576 1.583 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.524 1.536 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.485 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.999 1.450 mp_waitany 17072 13.8 1.188 1.367 1.188 1.367 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.308 1.316 acc_transpose_blocks 16440 15.4 0.069 0.071 1.213 1.226 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.854000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.000000, yerr=8.399134 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 730.095616E+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_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.103 0.204 66.254 66.255 qs_mol_dyn_low 1 2.0 0.003 0.005 65.458 65.467 qs_forces 11 3.9 0.002 0.002 65.388 65.389 qs_energies 11 4.9 0.007 0.027 61.980 61.985 scf_env_do_scf 11 5.9 0.004 0.026 52.862 52.868 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.891 40.892 velocity_verlet 10 3.0 0.001 0.002 36.700 36.702 dbcsr_multiply_generic 2055 12.4 0.131 0.188 29.725 30.022 qs_scf_new_mos 99 7.5 0.001 0.001 26.493 26.598 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.493 26.598 ot_scf_mini 99 9.5 0.003 0.003 24.869 24.969 multiply_cannon 2055 13.4 0.244 0.263 22.665 23.444 multiply_cannon_loop 2055 14.4 0.882 0.898 21.135 21.823 ot_mini 99 10.5 0.001 0.001 14.325 14.427 multiply_cannon_multrec 24660 15.4 4.222 6.846 12.625 13.652 rebuild_ks_matrix 110 8.3 0.000 0.000 12.041 12.132 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.047 12.041 12.132 init_scf_loop 11 6.9 0.007 0.050 11.928 11.934 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.681 10.762 qs_ot_get_derivative 99 11.5 0.001 0.002 10.172 10.265 prepare_preconditioner 11 7.9 0.000 0.000 10.039 10.056 make_preconditioner 11 8.9 0.000 0.002 10.039 10.056 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.289 9.740 dbcsr_mm_accdrv_process 52304 16.0 6.813 8.718 8.257 9.217 mp_waitall_1 121746 16.5 4.465 6.596 4.465 6.596 sum_up_and_integrate 110 10.3 0.068 0.071 6.380 6.392 integrate_v_rspace 110 11.3 0.003 0.003 6.311 6.328 init_scf_run 11 5.9 0.000 0.001 6.088 6.089 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.088 6.089 make_m2s 4110 13.4 0.060 0.063 5.428 5.872 make_images 4110 14.4 0.582 0.698 5.287 5.730 qs_ot_get_p 110 10.4 0.001 0.001 5.549 5.668 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.603 5.611 calculate_rho_elec 110 8.6 0.077 0.081 5.602 5.611 cp_fm_upper_to_full 70 13.8 3.253 4.629 3.253 4.629 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.033 4.124 apply_single 110 13.6 0.000 0.000 4.033 4.123 ot_diis_step 99 11.5 0.011 0.011 4.110 4.110 dbcsr_complete_redistribute 325 12.2 0.447 0.565 2.759 3.910 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.775 3.790 grid_integrate_task_list 110 12.3 3.276 3.372 3.276 3.372 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.216 3.368 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.277 3.322 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.594 3.320 multiply_cannon_sync_h2d 24660 15.4 3.188 3.311 3.188 3.311 pw_transfer 1331 11.6 0.065 0.072 3.285 3.309 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.296 3.296 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.179 3.205 make_images_data 4110 15.4 0.047 0.051 2.682 3.134 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.097 3.099 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.360 3.071 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.009 3.057 calculate_first_density_matrix 1 7.0 0.001 0.003 3.043 3.046 wfi_extrapolate 11 7.9 0.001 0.001 2.956 2.956 density_rs2pw 110 9.6 0.004 0.004 2.779 2.950 calculate_dm_sparse 110 9.5 0.001 0.001 2.906 2.935 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.741 2.857 mp_alltoall_i22 605 13.7 1.684 2.840 1.684 2.840 cp_fm_cholesky_invert 11 10.9 2.816 2.825 2.816 2.825 rs_pw_transfer 902 11.9 0.010 0.011 2.558 2.777 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.769 2.771 cp_fm_redistribute_end 48 14.0 1.385 2.743 1.386 2.744 cp_fm_diag_elpa_base 48 14.0 1.277 2.611 1.355 2.719 fft_wrap_pw1pw2_140 451 13.1 0.201 0.212 2.669 2.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.629 2.671 fft3d_ps 1111 14.6 1.063 1.099 2.481 2.496 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.491 2.493 grid_collocate_task_list 110 9.6 2.220 2.377 2.220 2.377 jit_kernel_multiply 12 15.8 1.110 2.184 1.110 2.184 potential_pw2rs 110 12.3 0.012 0.013 2.154 2.164 mp_alltoall_d11v 2046 13.8 1.790 1.936 1.790 1.936 cp_fm_cholesky_decompose 22 10.9 1.708 1.755 1.708 1.755 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.698 1.737 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.605 1.731 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.642 mp_sum_l 6514 12.8 0.974 1.630 0.974 1.630 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.518 1.536 acc_transpose_blocks 24660 15.4 0.104 0.107 1.507 1.536 qs_env_update_s_mstruct 11 6.9 0.038 0.060 1.375 1.521 mp_waitany 13376 13.8 1.162 1.453 1.162 1.453 mp_sum_d 3877 11.9 0.940 1.444 0.940 1.444 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.843 1.431 mp_irecv_dv 62702 16.1 0.746 1.350 0.746 1.350 mp_allgather_i34 2055 14.4 0.634 1.327 0.634 1.327 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.255000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.636364, yerr=8.772610 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 835.956736E+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_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.054 0.089 55.510 55.510 qs_mol_dyn_low 1 2.0 0.003 0.004 55.013 55.022 qs_forces 11 3.9 0.002 0.002 54.943 54.944 qs_energies 11 4.9 0.002 0.004 51.349 51.354 scf_env_do_scf 11 5.9 0.000 0.001 42.615 42.615 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.620 34.621 velocity_verlet 10 3.0 0.001 0.002 30.872 30.875 dbcsr_multiply_generic 2055 12.4 0.110 0.111 23.084 23.213 qs_scf_new_mos 99 7.5 0.001 0.001 20.531 20.592 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.530 20.591 ot_scf_mini 99 9.5 0.002 0.002 19.280 19.332 multiply_cannon 2055 13.4 0.250 0.259 17.518 18.785 multiply_cannon_loop 2055 14.4 0.319 0.334 16.164 16.410 rebuild_ks_matrix 110 8.3 0.000 0.000 11.510 11.552 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.509 11.551 ot_mini 99 10.5 0.001 0.001 10.597 10.641 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.257 10.293 multiply_cannon_multrec 8220 15.4 3.279 4.743 7.496 8.804 mp_waitall_1 103326 16.6 6.304 8.012 6.304 8.012 init_scf_loop 11 6.9 0.000 0.000 7.948 7.950 qs_ot_get_derivative 99 11.5 0.001 0.001 6.743 6.794 prepare_preconditioner 11 7.9 0.000 0.000 6.250 6.257 make_preconditioner 11 8.9 0.000 0.000 6.250 6.257 sum_up_and_integrate 110 10.3 0.079 0.081 6.212 6.224 integrate_v_rspace 110 11.3 0.003 0.003 6.133 6.146 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.801 5.870 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.587 5.603 calculate_rho_elec 110 8.6 0.115 0.115 5.586 5.603 init_scf_run 11 5.9 0.000 0.001 5.559 5.559 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.559 5.559 dbcsr_mm_accdrv_process 17442 15.9 2.845 3.833 4.086 5.013 qs_ot_get_p 110 10.4 0.001 0.001 4.765 4.816 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.167 4.600 make_m2s 4110 13.4 0.039 0.040 4.275 4.547 make_images 4110 14.4 0.639 0.691 4.145 4.416 ot_diis_step 99 11.5 0.012 0.012 3.830 3.831 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.746 3.786 apply_single 110 13.6 0.000 0.000 3.746 3.785 grid_integrate_task_list 110 12.3 3.343 3.547 3.343 3.547 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.413 3.418 pw_transfer 1331 11.6 0.065 0.069 3.254 3.267 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.147 3.164 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.109 3.109 cp_fm_cholesky_invert 11 10.9 3.037 3.041 3.037 3.041 multiply_cannon_sync_h2d 8220 15.4 2.900 3.040 2.900 3.040 make_images_data 4110 15.4 0.038 0.043 2.427 2.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.800 2.802 density_rs2pw 110 9.6 0.004 0.004 2.644 2.780 calculate_first_density_matrix 1 7.0 0.000 0.000 2.771 2.772 hybrid_alltoall_any 4261 16.3 0.200 0.867 2.350 2.738 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.693 2.710 wfi_extrapolate 11 7.9 0.001 0.001 2.701 2.702 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.681 2.682 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.600 2.601 cp_fm_redistribute_end 48 14.0 0.661 2.576 0.665 2.577 cp_fm_diag_elpa_base 48 14.0 1.728 2.392 1.904 2.550 calculate_dm_sparse 110 9.5 0.001 0.001 2.472 2.516 grid_collocate_task_list 110 9.6 2.324 2.514 2.324 2.514 fft3d_ps 1111 14.6 1.119 1.161 2.404 2.429 rs_pw_transfer 902 11.9 0.010 0.010 2.219 2.374 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.190 2.203 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.980 potential_pw2rs 110 12.3 0.015 0.016 1.958 1.971 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.896 1.919 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.860 1.889 mp_alltoall_d11v 2046 13.8 1.690 1.888 1.690 1.888 dbcsr_complete_redistribute 325 12.2 0.694 0.833 1.732 1.832 cp_fm_cholesky_decompose 22 10.9 1.728 1.746 1.728 1.746 jit_kernel_multiply 10 15.7 0.928 1.732 0.928 1.732 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.520 1.645 mp_allgather_i34 2055 14.4 0.540 1.636 0.540 1.636 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.583 1.589 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.430 1.440 qs_create_task_list 11 7.9 0.000 0.001 1.221 1.320 generate_qs_task_list 11 8.9 0.380 0.447 1.220 1.320 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.480 1.310 mp_irecv_dv 24056 15.7 0.457 1.268 0.457 1.268 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.826 1.241 mp_waitany 9240 13.8 1.079 1.237 1.079 1.237 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.139 1.170 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.069 1.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.510000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.909091, yerr=14.311182 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 1.355981E+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_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.040 86.622 86.622 qs_mol_dyn_low 1 2.0 0.003 0.004 86.244 86.252 qs_forces 11 3.9 0.013 0.026 86.174 86.175 qs_energies 11 4.9 0.007 0.012 82.135 82.148 scf_env_do_scf 11 5.9 0.001 0.001 71.968 71.969 velocity_verlet 10 3.0 0.004 0.005 55.237 55.243 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.393 43.396 dbcsr_multiply_generic 2055 12.4 0.126 0.128 28.968 29.046 init_scf_loop 11 6.9 0.000 0.000 28.502 28.508 qs_scf_new_mos 99 7.5 0.001 0.001 26.793 26.816 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.792 26.816 prepare_preconditioner 11 7.9 0.000 0.000 26.553 26.559 make_preconditioner 11 8.9 0.000 0.000 26.553 26.559 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.792 26.026 ot_scf_mini 99 9.5 0.002 0.002 24.994 25.014 multiply_cannon 2055 13.4 0.346 0.360 21.640 22.373 multiply_cannon_loop 2055 14.4 0.340 0.346 19.829 20.262 cp_fm_upper_to_full 70 14.2 12.564 18.044 12.564 18.044 ot_mini 99 10.5 0.001 0.001 13.852 13.870 rebuild_ks_matrix 110 8.3 0.001 0.001 13.205 13.244 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.204 13.243 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.020 12.055 dbcsr_complete_redistribute 325 12.2 1.043 1.071 7.416 10.567 multiply_cannon_multrec 8220 15.4 4.366 4.567 9.627 9.763 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.356 9.509 qs_ot_get_derivative 99 11.5 0.001 0.001 9.277 9.299 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.745 8.876 mp_alltoall_i22 605 13.7 5.376 8.519 5.376 8.519 mp_waitall_1 84994 16.7 7.734 8.511 7.734 8.511 sum_up_and_integrate 110 10.3 0.151 0.152 6.679 6.695 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.621 6.658 calculate_rho_elec 110 8.6 0.227 0.227 6.620 6.657 integrate_v_rspace 110 11.3 0.004 0.004 6.527 6.542 make_m2s 4110 13.4 0.043 0.044 5.477 5.981 qs_ot_get_p 110 10.4 0.001 0.001 5.881 5.909 init_scf_run 11 5.9 0.000 0.001 5.869 5.869 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.868 5.869 make_images 4110 14.4 0.878 0.916 5.289 5.792 cp_fm_cholesky_invert 11 10.9 5.583 5.588 5.583 5.588 dbcsr_mm_accdrv_process 11614 15.7 3.367 3.698 5.117 5.418 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.841 5.305 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.666 5.115 apply_single 110 13.6 0.000 0.000 4.666 5.115 ot_diis_step 99 11.5 0.015 0.015 4.550 4.551 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.284 4.291 multiply_cannon_sync_h2d 8220 15.4 3.951 3.958 3.951 3.958 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.837 3.837 pw_transfer 1331 11.6 0.074 0.074 3.817 3.825 grid_integrate_task_list 110 12.3 3.662 3.711 3.662 3.711 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.701 3.709 hybrid_alltoall_any 4261 16.3 0.256 0.555 2.990 3.688 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.672 3.684 make_images_data 4110 15.4 0.042 0.045 3.000 3.638 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.884 3.333 wfi_extrapolate 11 7.9 0.001 0.001 3.275 3.276 calculate_dm_sparse 110 9.5 0.001 0.001 3.210 3.234 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.232 3.233 cp_fm_diag_elpa_base 48 14.0 2.702 2.896 3.231 3.231 fft_wrap_pw1pw2_140 451 13.1 0.214 0.218 3.170 3.176 density_rs2pw 110 9.6 0.004 0.004 3.080 3.101 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.986 2.990 fft3d_ps 1111 14.6 1.265 1.277 2.911 2.921 grid_collocate_task_list 110 9.6 2.631 2.659 2.631 2.659 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.521 2.536 calculate_first_density_matrix 1 7.0 0.000 0.000 2.485 2.488 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.450 2.466 rs_pw_transfer 902 11.9 0.010 0.011 2.237 2.288 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.211 2.274 mp_alltoall_d11v 2046 13.8 2.150 2.197 2.150 2.197 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.175 cp_fm_cholesky_decompose 22 10.9 2.086 2.103 2.086 2.103 potential_pw2rs 110 12.3 0.021 0.021 2.092 2.097 qs_create_task_list 11 7.9 0.011 0.028 1.935 1.979 generate_qs_task_list 11 8.9 0.738 0.792 1.924 1.962 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.824 1.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.795 jit_kernel_multiply 10 15.2 1.547 1.793 1.547 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.622000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1219.636364, yerr=59.121251 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 634.761216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 56701. MP_Allreduce 11263 785. MP_Sync 168 MP_Alltoall 2210 2550995. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.057 202.408 202.410 qs_mol_dyn_low 1 2.0 0.006 0.022 201.747 201.760 qs_forces 11 3.9 0.003 0.003 201.604 201.605 qs_energies 11 4.9 0.002 0.003 196.129 196.144 scf_env_do_scf 11 5.9 0.001 0.001 179.396 179.399 scf_env_do_scf_inner_loop 116 6.6 0.005 0.019 158.470 158.472 dbcsr_multiply_generic 2485 12.5 0.189 0.206 123.457 124.134 velocity_verlet 10 3.0 0.001 0.002 121.202 121.203 qs_scf_new_mos 116 7.6 0.001 0.001 120.168 120.324 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.167 120.323 ot_scf_mini 116 9.6 0.003 0.003 113.601 113.741 multiply_cannon 2485 13.5 0.232 0.238 100.124 101.722 multiply_cannon_loop 2485 14.5 2.083 2.154 97.950 99.802 ot_mini 116 10.6 0.001 0.001 65.370 65.489 multiply_cannon_multrec 59640 15.5 32.865 34.497 41.495 43.400 qs_ot_get_derivative 116 11.6 0.001 0.001 40.772 40.883 rebuild_ks_matrix 127 8.3 0.001 0.001 32.908 33.121 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 32.907 33.120 mp_waitall_1 264810 16.5 27.896 30.797 27.896 30.797 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.595 29.783 multiply_cannon_sync_h2d 59640 15.5 27.143 28.438 27.143 28.438 qs_ot_get_p 127 10.4 0.001 0.001 26.594 26.787 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.113 24.938 apply_single 127 13.6 0.001 0.001 24.113 24.938 ot_diis_step 116 11.6 0.007 0.008 24.226 24.228 init_scf_loop 11 6.9 0.000 0.001 20.850 20.851 qs_ot_p2m_diag 82 11.4 0.076 0.090 19.855 19.945 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.924 19.084 multiply_cannon_metrocomm3 59640 15.5 0.114 0.119 15.445 17.299 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.246 17.248 prepare_preconditioner 11 7.9 0.000 0.000 16.235 16.284 make_preconditioner 11 8.9 0.000 0.000 16.235 16.284 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.442 15.588 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.352 14.358 make_m2s 4970 13.5 0.105 0.112 13.873 14.326 cp_fm_redistribute_end 82 14.4 11.347 14.279 11.358 14.283 make_images 4970 14.5 0.398 0.417 13.691 14.151 cp_fm_diag_elpa_base 82 14.4 2.880 13.901 2.911 14.017 sum_up_and_integrate 127 10.3 0.089 0.107 13.878 13.892 integrate_v_rspace 127 11.3 0.003 0.004 13.789 13.809 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.537 12.658 calculate_rho_elec 127 8.7 0.044 0.063 12.537 12.657 init_scf_run 11 5.9 0.000 0.001 12.555 12.555 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.555 12.555 mp_sum_l 7804 13.0 8.195 9.492 8.195 9.492 cp_fm_cholesky_invert 11 10.9 9.185 9.193 9.185 9.193 wfi_extrapolate 11 7.9 0.001 0.001 9.012 9.013 dbcsr_mm_accdrv_process 123452 16.2 3.299 3.453 8.196 8.764 calculate_dm_sparse 127 9.5 0.001 0.001 8.483 8.566 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.984 8.103 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.788 7.884 make_images_data 4970 15.5 0.066 0.071 6.787 7.760 grid_integrate_task_list 127 12.3 7.089 7.487 7.089 7.487 multiply_cannon_metrocomm1 59640 15.5 0.087 0.091 5.853 7.413 hybrid_alltoall_any 5155 16.4 0.287 2.232 5.944 7.248 density_rs2pw 127 9.7 0.006 0.007 6.422 7.094 pw_transfer 1535 11.6 0.074 0.104 6.728 7.010 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.789 6.797 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 6.526 6.781 rs_pw_transfer 1038 11.9 0.017 0.020 5.680 6.366 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.848 5.968 fft_wrap_pw1pw2_140 519 13.2 0.441 0.506 5.531 5.696 fft3d_ps 1281 14.7 2.078 2.552 5.331 5.523 mp_alltoall_d11v 2401 14.1 4.158 5.417 4.158 5.417 grid_collocate_task_list 127 9.7 4.688 5.088 4.688 5.088 cp_fm_cholesky_decompose 22 10.9 4.834 4.848 4.834 4.848 potential_pw2rs 127 12.3 0.009 0.010 4.429 4.448 mp_sum_d 4444 12.2 3.585 4.159 3.585 4.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.410000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.818182, yerr=8.122004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/16/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 2.183246E+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 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 828.452864E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+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 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5800847. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 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 0.022 0.061 193.000 193.000 qs_mol_dyn_low 1 2.0 0.003 0.004 192.521 192.534 qs_forces 11 3.9 0.003 0.003 192.410 192.411 qs_energies 11 4.9 0.013 0.024 185.643 185.655 scf_env_do_scf 11 5.9 0.001 0.001 169.088 169.098 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.905 134.908 velocity_verlet 10 3.0 0.022 0.033 121.403 121.404 dbcsr_multiply_generic 2507 12.6 0.190 0.214 98.324 99.557 qs_scf_new_mos 117 7.6 0.001 0.001 96.155 96.571 qs_scf_loop_do_ot 117 8.6 0.001 0.002 96.155 96.570 ot_scf_mini 117 9.6 0.004 0.005 91.290 91.673 multiply_cannon 2507 13.6 0.480 0.541 77.876 82.520 multiply_cannon_loop 2507 14.6 1.248 1.285 74.443 76.780 ot_mini 117 10.6 0.001 0.001 50.320 50.750 mp_waitall_1 214728 16.6 24.879 39.844 24.879 39.844 multiply_cannon_multrec 30084 15.6 22.072 26.554 31.688 36.456 init_scf_loop 11 6.9 0.000 0.002 34.086 34.093 rebuild_ks_matrix 128 8.3 0.001 0.001 32.974 33.435 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.974 33.434 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.698 30.137 prepare_preconditioner 11 7.9 0.000 0.000 29.689 29.745 make_preconditioner 11 8.9 0.000 0.001 29.689 29.745 multiply_cannon_metrocomm3 30084 15.6 0.094 0.099 15.915 29.139 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.375 28.898 qs_ot_get_derivative 117 11.6 0.001 0.002 28.151 28.512 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.135 23.315 apply_single 128 13.6 0.001 0.001 22.135 23.314 qs_ot_get_p 128 10.4 0.001 0.002 22.490 22.961 ot_diis_step 117 11.6 0.014 0.015 21.983 21.985 multiply_cannon_sync_h2d 30084 15.6 19.404 21.509 19.404 21.509 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.638 17.673 cp_fm_cholesky_invert 11 10.9 17.470 17.483 17.470 17.483 cp_dbcsr_syevd 83 12.4 0.006 0.006 16.416 16.417 make_m2s 5014 13.6 0.091 0.096 14.584 16.232 make_images 5014 14.6 1.160 1.342 14.372 16.017 sum_up_and_integrate 128 10.3 0.116 0.132 14.440 14.476 integrate_v_rspace 128 11.3 0.003 0.003 14.323 14.364 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.176 13.188 cp_fm_redistribute_end 83 14.4 7.694 13.102 7.709 13.106 cp_fm_diag_elpa_base 83 14.4 5.150 12.543 5.375 12.934 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.885 12.915 calculate_rho_elec 128 8.7 0.087 0.105 12.884 12.914 init_scf_run 11 5.9 0.000 0.001 11.739 11.741 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.739 11.740 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.086 11.373 make_images_data 5014 15.6 0.065 0.071 8.732 10.658 multiply_cannon_metrocomm4 27577 15.6 0.095 0.108 3.723 10.456 mp_irecv_dv 69486 16.3 3.533 10.078 3.533 10.078 hybrid_alltoall_any 5200 16.5 0.341 1.494 7.374 10.002 dbcsr_mm_accdrv_process 62242 16.2 4.448 5.170 9.078 9.682 wfi_extrapolate 11 7.9 0.001 0.001 8.425 8.425 pw_transfer 1547 11.6 0.085 0.100 7.750 7.822 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.525 7.602 grid_integrate_task_list 128 12.3 7.152 7.596 7.152 7.596 cp_fm_cholesky_decompose 22 10.9 7.328 7.425 7.328 7.425 density_rs2pw 128 9.7 0.006 0.007 6.738 7.257 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.254 6.978 calculate_dm_sparse 128 9.5 0.001 0.001 6.559 6.682 fft_wrap_pw1pw2_140 523 13.2 0.470 0.509 6.491 6.572 rs_pw_transfer 1046 11.9 0.014 0.016 5.753 6.247 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.204 6.212 mp_sum_l 7870 13.0 4.098 6.036 4.098 6.036 fft3d_ps 1291 14.7 2.779 2.943 5.893 5.933 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.457 5.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.338 5.488 grid_collocate_task_list 128 9.7 4.912 5.427 4.912 5.427 mp_allgather_i34 2507 14.6 2.003 5.171 2.003 5.171 potential_pw2rs 128 12.3 0.014 0.016 4.966 4.993 mp_alltoall_d11v 2415 14.1 4.062 4.814 4.062 4.814 mp_sum_d 4455 12.2 2.770 4.084 2.770 4.084 dbcsr_complete_redistribute 395 12.7 0.776 0.857 3.210 4.047 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.000000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.000000, yerr=2.558409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 968.658944E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.043 0.055 180.549 180.550 qs_mol_dyn_low 1 2.0 0.003 0.006 179.969 179.982 qs_forces 11 3.9 0.003 0.003 179.873 179.878 qs_energies 11 4.9 0.013 0.029 173.325 173.338 scf_env_do_scf 11 5.9 0.001 0.002 157.582 157.583 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.111 121.112 velocity_verlet 10 3.0 0.001 0.002 115.421 115.423 qs_scf_new_mos 117 7.6 0.001 0.001 84.050 84.356 qs_scf_loop_do_ot 117 8.6 0.001 0.002 84.049 84.355 dbcsr_multiply_generic 2507 12.6 0.182 0.187 82.148 83.249 ot_scf_mini 117 9.6 0.004 0.004 79.906 80.229 multiply_cannon 2507 13.6 0.501 0.524 62.197 66.342 multiply_cannon_loop 2507 14.6 0.853 0.886 58.798 61.245 ot_mini 117 10.6 0.001 0.001 43.236 43.567 init_scf_loop 11 6.9 0.001 0.004 36.357 36.358 mp_waitall_1 170520 16.6 25.128 34.782 25.128 34.782 prepare_preconditioner 11 7.9 0.000 0.000 32.300 32.345 make_preconditioner 11 8.9 0.000 0.001 32.300 32.345 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.962 31.317 rebuild_ks_matrix 128 8.3 0.001 0.001 30.336 30.784 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.336 30.784 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.403 27.814 multiply_cannon_multrec 20056 15.6 13.362 16.551 22.127 25.175 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.371 25.065 qs_ot_get_derivative 117 11.6 0.001 0.002 23.407 23.727 qs_ot_get_p 128 10.4 0.001 0.001 21.984 22.402 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.974 21.011 apply_single 128 13.6 0.001 0.001 19.974 21.011 ot_diis_step 117 11.6 0.018 0.018 19.716 19.716 qs_ot_p2m_diag 83 11.4 0.265 0.271 17.624 17.638 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.560 16.562 make_m2s 5014 13.6 0.082 0.086 14.668 15.598 multiply_cannon_sync_h2d 20056 15.6 14.359 15.560 14.359 15.560 cp_fm_cholesky_invert 11 10.9 15.514 15.524 15.514 15.524 make_images 5014 14.6 1.180 1.282 14.434 15.365 sum_up_and_integrate 128 10.3 0.134 0.145 14.440 14.465 integrate_v_rspace 128 11.3 0.003 0.004 14.306 14.335 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.383 13.416 calculate_rho_elec 128 8.7 0.132 0.147 13.383 13.415 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.177 13.179 cp_fm_redistribute_end 83 14.4 4.982 13.080 5.001 13.083 cp_fm_diag_elpa_base 83 14.4 7.649 12.436 8.058 12.923 init_scf_run 11 5.9 0.000 0.001 10.691 10.692 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.691 10.691 make_images_data 5014 15.6 0.061 0.069 8.872 10.175 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.436 9.510 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.183 9.422 hybrid_alltoall_any 5200 16.5 0.433 1.987 7.779 9.380 mp_irecv_dv 50230 16.2 3.315 9.265 3.315 9.265 dbcsr_mm_accdrv_process 41502 16.2 4.579 5.251 8.227 8.452 pw_transfer 1547 11.6 0.085 0.104 7.954 8.078 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.729 7.863 cp_fm_cholesky_decompose 22 10.9 7.736 7.789 7.736 7.789 grid_integrate_task_list 128 12.3 7.372 7.784 7.372 7.784 wfi_extrapolate 11 7.9 0.001 0.001 7.488 7.488 cp_fm_upper_to_full 105 14.5 5.658 7.362 5.658 7.362 density_rs2pw 128 9.7 0.006 0.006 6.839 7.195 fft_wrap_pw1pw2_140 523 13.2 0.477 0.532 6.652 6.799 dbcsr_complete_redistribute 395 12.7 1.194 1.258 4.787 6.528 fft3d_ps 1291 14.7 2.693 2.891 5.997 6.062 rs_pw_transfer 1046 11.9 0.014 0.015 5.513 5.870 calculate_dm_sparse 128 9.5 0.001 0.001 5.766 5.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.600 5.606 grid_collocate_task_list 128 9.7 5.073 5.435 5.073 5.435 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.700 5.427 mp_alltoall_d11v 2415 14.1 4.406 5.306 4.406 5.306 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.523 5.262 mp_sum_l 7870 13.0 3.465 4.896 3.465 4.896 potential_pw2rs 128 12.3 0.020 0.022 4.783 4.813 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.609 4.729 mp_allgather_i34 2507 14.6 1.920 4.559 1.920 4.559 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.117 4.166 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.318 4.013 mp_alltoall_i22 716 14.1 1.933 3.816 1.933 3.816 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.729 3.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.550000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.454545, yerr=21.330148 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/18/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 4.387242E+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 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.140802E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 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 12509604. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 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 0.215 0.378 191.737 191.747 qs_mol_dyn_low 1 2.0 0.003 0.005 190.614 190.630 qs_forces 11 3.9 0.003 0.003 190.471 190.481 qs_energies 11 4.9 0.015 0.042 183.308 183.321 scf_env_do_scf 11 5.9 0.001 0.002 165.881 165.893 velocity_verlet 10 3.0 0.002 0.002 126.745 126.749 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.568 118.570 qs_scf_new_mos 118 7.6 0.001 0.001 82.024 82.312 qs_scf_loop_do_ot 118 8.6 0.001 0.002 82.023 82.311 dbcsr_multiply_generic 2529 12.6 0.192 0.199 80.907 81.655 ot_scf_mini 118 9.6 0.004 0.005 77.489 77.810 multiply_cannon 2529 13.6 0.556 0.593 55.756 58.572 multiply_cannon_loop 2529 14.6 1.188 1.214 51.747 53.586 init_scf_loop 11 6.9 0.004 0.011 47.182 47.184 ot_mini 118 10.6 0.001 0.001 43.077 43.406 prepare_preconditioner 11 7.9 0.000 0.000 43.016 43.048 make_preconditioner 11 8.9 0.001 0.002 43.016 43.048 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.625 41.603 multiply_cannon_multrec 30348 15.6 14.280 19.460 26.554 31.189 rebuild_ks_matrix 129 8.3 0.001 0.001 29.742 29.992 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 29.741 29.992 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.839 27.070 mp_waitall_1 149172 16.7 17.775 26.909 17.775 26.909 qs_ot_get_derivative 118 11.6 0.001 0.002 23.320 23.634 make_m2s 5058 13.6 0.098 0.102 20.599 21.818 make_images 5058 14.6 1.967 2.323 20.288 21.507 qs_ot_get_p 129 10.4 0.001 0.002 20.495 20.868 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.149 19.721 apply_single 129 13.6 0.001 0.001 19.149 19.721 ot_diis_step 118 11.6 0.018 0.019 19.620 19.622 cp_fm_cholesky_invert 11 10.9 16.679 16.688 16.679 16.688 cp_fm_upper_to_full 106 14.7 11.019 16.329 11.019 16.329 qs_ot_p2m_diag 84 11.4 0.346 0.393 16.181 16.234 multiply_cannon_metrocomm3 30348 15.6 0.047 0.050 6.403 15.409 cp_dbcsr_syevd 84 12.4 0.006 0.006 14.868 14.870 sum_up_and_integrate 129 10.3 0.142 0.155 14.584 14.630 integrate_v_rspace 129 11.3 0.003 0.004 14.441 14.493 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.634 13.657 calculate_rho_elec 129 8.7 0.176 0.192 13.633 13.656 make_images_data 5058 15.6 0.065 0.069 11.072 12.977 dbcsr_complete_redistribute 397 12.7 1.522 1.628 9.143 12.899 multiply_cannon_sync_h2d 30348 15.6 11.794 12.471 11.794 12.471 dbcsr_mm_accdrv_process 62780 16.2 7.393 8.223 11.849 12.380 hybrid_alltoall_any 5245 16.5 0.530 2.228 9.963 11.767 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.684 11.687 cp_fm_redistribute_end 84 14.4 2.034 11.603 2.053 11.609 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.703 11.434 cp_fm_diag_elpa_base 84 14.4 8.912 10.956 9.512 11.429 init_scf_run 11 5.9 0.000 0.001 11.308 11.309 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.308 11.309 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.368 10.033 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.743 9.966 mp_alltoall_i22 720 14.1 5.578 9.413 5.578 9.413 pw_transfer 1559 11.6 0.086 0.101 8.326 8.417 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.099 8.194 grid_integrate_task_list 129 12.3 7.566 8.026 7.566 8.026 cp_fm_cholesky_decompose 22 10.9 7.825 7.911 7.825 7.911 wfi_extrapolate 11 7.9 0.001 0.001 7.691 7.691 fft_wrap_pw1pw2_140 527 13.2 0.485 0.490 7.092 7.208 density_rs2pw 129 9.7 0.005 0.006 6.605 6.917 multiply_cannon_metrocomm4 25290 15.6 0.076 0.085 2.775 6.725 mp_irecv_dv 76751 16.2 2.632 6.470 2.632 6.470 calculate_dm_sparse 129 9.5 0.001 0.001 6.335 6.437 fft3d_ps 1301 14.7 2.829 2.910 6.315 6.395 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.825 5.884 mp_alltoall_d11v 2429 14.1 5.242 5.750 5.242 5.750 grid_collocate_task_list 129 9.7 5.254 5.626 5.254 5.626 rs_pw_transfer 1054 12.0 0.013 0.014 4.985 5.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.709 4.770 potential_pw2rs 129 12.3 0.022 0.023 4.621 4.675 qs_energies_init_hamiltonians 11 5.9 0.008 0.012 4.567 4.569 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.456 4.547 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.241 4.293 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.747000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.272727, yerr=17.914397 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.508680E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591807936 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57813. MP_Allreduce 11072 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.049 174.503 174.504 qs_mol_dyn_low 1 2.0 0.003 0.004 173.978 173.990 qs_forces 11 3.9 0.003 0.003 173.526 173.528 qs_energies 11 4.9 0.003 0.006 166.211 166.222 scf_env_do_scf 11 5.9 0.001 0.001 148.883 148.893 velocity_verlet 10 3.0 0.017 0.021 114.388 114.414 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 112.678 112.679 qs_scf_new_mos 118 7.6 0.001 0.001 76.453 76.596 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.452 76.595 dbcsr_multiply_generic 2527 12.6 0.184 0.189 75.814 76.462 ot_scf_mini 118 9.6 0.004 0.004 71.937 72.054 multiply_cannon 2527 13.6 0.588 0.611 55.436 59.474 multiply_cannon_loop 2527 14.6 0.449 0.462 50.785 52.107 ot_mini 118 10.6 0.001 0.001 40.533 40.663 init_scf_loop 11 6.9 0.000 0.000 36.048 36.049 mp_waitall_1 126780 16.7 27.022 35.073 27.022 35.073 prepare_preconditioner 11 7.9 0.000 0.000 32.173 32.210 make_preconditioner 11 8.9 0.000 0.000 32.173 32.210 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.033 30.287 rebuild_ks_matrix 129 8.3 0.001 0.001 28.989 29.144 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.989 29.143 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.506 26.645 multiply_cannon_multrec 10108 15.6 10.506 14.577 18.032 22.696 multiply_cannon_metrocomm3 10108 15.6 0.023 0.024 12.860 20.607 qs_ot_get_derivative 118 11.6 0.002 0.002 20.208 20.322 ot_diis_step 118 11.6 0.020 0.021 20.247 20.248 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.756 20.140 apply_single 129 13.6 0.001 0.001 19.755 20.140 make_m2s 5054 13.6 0.067 0.071 16.673 19.027 make_images 5054 14.6 2.342 2.852 16.365 18.710 cp_fm_cholesky_invert 11 10.9 18.406 18.413 18.406 18.413 qs_ot_get_p 129 10.4 0.001 0.001 18.160 18.319 sum_up_and_integrate 129 10.3 0.182 0.193 14.241 14.285 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.256 14.272 integrate_v_rspace 129 11.3 0.004 0.004 14.059 14.111 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.341 13.381 calculate_rho_elec 129 8.7 0.260 0.271 13.341 13.380 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.088 13.090 make_images_data 5054 15.6 0.053 0.061 10.214 12.921 hybrid_alltoall_any 5240 16.5 0.846 3.824 9.932 12.488 multiply_cannon_sync_h2d 10108 15.6 11.609 12.324 11.609 12.324 init_scf_run 11 5.9 0.000 0.001 10.524 10.524 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.524 10.524 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.030 10.040 cp_fm_diag_elpa_base 83 14.4 9.782 9.860 10.018 10.028 multiply_cannon_metrocomm1 10108 15.6 0.029 0.029 4.642 8.220 grid_integrate_task_list 129 12.3 7.824 8.218 7.824 8.218 cp_fm_cholesky_decompose 22 10.9 8.089 8.186 8.089 8.186 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.054 8.143 pw_transfer 1559 11.6 0.084 0.091 7.900 7.917 dbcsr_mm_accdrv_process 20926 16.1 2.666 3.645 7.160 7.820 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.677 7.702 wfi_extrapolate 11 7.9 0.001 0.001 7.333 7.333 mp_allgather_i34 2527 14.6 2.845 7.243 2.845 7.243 fft_wrap_pw1pw2_140 527 13.2 0.506 0.527 6.740 6.768 density_rs2pw 129 9.7 0.005 0.006 6.057 6.389 calculate_dm_sparse 129 9.5 0.001 0.001 6.169 6.277 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.886 5.945 mp_irecv_dv 29086 15.9 1.850 5.860 1.850 5.860 grid_collocate_task_list 129 9.7 5.528 5.853 5.528 5.853 fft3d_ps 1301 14.7 2.743 2.809 5.821 5.850 dbcsr_complete_redistribute 395 12.7 2.144 2.231 5.330 5.689 mp_alltoall_d11v 2423 14.1 4.867 5.359 4.867 5.359 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.314 5.321 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.233 5.235 rs_pw_transfer 1054 12.0 0.013 0.013 4.177 4.531 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.153 4.195 potential_pw2rs 129 12.3 0.027 0.028 4.155 4.173 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.559 3.862 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.799 3.839 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.496 3.801 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.656 3.697 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.587 3.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.504000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1408.181818, yerr=50.369215 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1399806074880 0.0% 0.0% 100.0% flops 32 x 32 x 32 1906965479424 0.0% 0.0% 100.0% flops 22 x 9 x 32 1943685697536 0.0% 0.0% 100.0% flops 9 x 22 x 32 1949313282048 0.0% 0.0% 100.0% flops 22 x 22 x 32 2694948675584 0.0% 0.0% 100.0% flops 32 x 32 x 9 4338990710784 0.0% 0.0% 100.0% flops 32 x 32 x 22 5303210868736 0.0% 0.0% 100.0% flops 9 x 32 x 32 5347334946816 0.0% 0.0% 100.0% flops 22 x 32 x 32 6535631601664 0.0% 0.0% 100.0% flops 9 x 32 x 9 11360521838592 0.0% 0.0% 100.0% flops 22 x 32 x 9 14909203279872 0.0% 0.0% 100.0% flops 9 x 32 x 22 14909203279872 0.0% 0.0% 100.0% flops 22 x 32 x 22 19481731219456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.080547E+12 0.0% 0.0% 100.0% flops max/rank 11.516856E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6655093856 0.0% 0.0% 100.0% number of processed stacks 1931936 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.8 marketing flops 142.438295E+12 ------------------------------------------------------------------------------- # multiplications 2464 max memory usage/rank 3.008733E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 98560 MPI messages size (bytes): total size 1.118682E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.350266E+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 44368 34435235840 4194304 < size <= 16777216 43616 373376942080 16777216 < size 9960 710867104896 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4014 58969. MP_Allreduce 11032 1514. MP_Sync 87 MP_Alltoall 1688 36934197. MP_SendRecv 1764 218624. MP_ISendRecv 1764 218624. MP_Wait 9654 MP_ISend 6312 1080635. MP_IRecv 6312 1080635. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.087 288.601 288.601 qs_mol_dyn_low 1 2.0 0.007 0.037 287.870 287.881 qs_forces 11 3.9 0.003 0.003 287.770 287.776 qs_energies 11 4.9 0.003 0.003 279.168 279.188 scf_env_do_scf 11 5.9 0.001 0.001 257.031 257.048 velocity_verlet 10 3.0 0.016 0.017 208.012 208.019 scf_env_do_scf_inner_loop 115 6.6 0.003 0.008 131.420 131.421 init_scf_loop 11 6.9 0.000 0.000 125.352 125.356 prepare_preconditioner 11 7.9 0.000 0.000 120.690 120.714 make_preconditioner 11 8.9 0.000 0.000 120.690 120.714 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.676 117.898 qs_scf_new_mos 115 7.6 0.001 0.001 90.237 90.279 qs_scf_loop_do_ot 115 8.6 0.001 0.001 90.236 90.278 ot_scf_mini 115 9.6 0.004 0.004 85.584 85.649 dbcsr_multiply_generic 2464 12.5 0.209 0.215 81.394 81.567 cp_fm_upper_to_full 105 14.8 52.924 76.000 52.924 76.000 multiply_cannon 2464 13.5 0.691 0.736 57.163 58.313 multiply_cannon_loop 2464 14.5 0.463 0.474 53.614 54.202 ot_mini 115 10.6 0.001 0.001 43.628 43.711 dbcsr_complete_redistribute 395 12.7 4.021 4.113 29.825 42.810 copy_fm_to_dbcsr 209 11.7 0.002 0.002 26.350 39.357 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.969 36.862 mp_alltoall_i22 712 14.1 21.859 35.010 21.859 35.010 cp_fm_cholesky_invert 11 10.9 34.276 34.283 34.276 34.283 rebuild_ks_matrix 126 8.3 0.001 0.001 32.847 32.967 qs_ks_build_kohn_sham_matrix 126 9.3 0.017 0.017 32.847 32.967 qs_ks_update_qs_env 126 7.6 0.001 0.001 30.585 30.690 mp_waitall_1 101902 16.8 26.855 30.040 26.855 30.040 qs_ot_get_p 126 10.4 0.001 0.001 26.937 27.077 qs_ot_get_derivative 115 11.6 0.002 0.002 23.919 23.989 qs_ot_p2m_diag 83 11.4 0.879 0.883 22.835 22.864 cp_dbcsr_syevd 83 12.4 0.006 0.007 21.078 21.081 make_m2s 4928 13.5 0.076 0.081 19.377 20.221 make_images 4928 14.5 3.696 3.827 18.905 19.750 ot_diis_step 115 11.6 0.021 0.022 19.646 19.646 apply_preconditioner_dbcsr 126 12.6 0.000 0.000 18.878 19.063 apply_single 126 13.6 0.001 0.001 18.878 19.063 multiply_cannon_metrocomm3 9856 15.5 0.023 0.024 17.806 18.219 multiply_cannon_multrec 9856 15.5 10.365 12.068 17.892 18.035 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.785 17.786 cp_fm_diag_elpa_base 83 14.4 13.389 15.037 17.780 17.781 sum_up_and_integrate 126 10.3 0.317 0.319 15.610 15.694 multiply_cannon_sync_h2d 9856 15.5 15.414 15.426 15.414 15.426 integrate_v_rspace 126 11.3 0.004 0.004 15.292 15.376 qs_rho_update_rho_low 126 7.7 0.001 0.001 14.813 14.843 calculate_rho_elec 126 8.7 0.475 0.476 14.812 14.842 init_scf_run 11 5.9 0.000 0.001 12.138 12.140 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.138 12.140 hybrid_alltoall_any 5114 16.4 1.281 2.997 10.397 12.102 make_images_data 4928 15.5 0.059 0.063 10.210 11.919 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.458 9.502 cp_fm_cholesky_decompose 22 10.9 9.246 9.289 9.246 9.289 dbcsr_mm_accdrv_process 20422 16.0 3.708 5.630 7.292 9.017 wfi_extrapolate 11 7.9 0.001 0.001 8.791 8.791 grid_integrate_task_list 126 12.3 8.410 8.588 8.410 8.588 pw_transfer 1523 11.6 0.089 0.090 8.368 8.377 fft_wrap_pw1pw2 1271 12.7 0.011 0.011 8.137 8.146 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.004 8.009 fft_wrap_pw1pw2_140 515 13.2 0.532 0.542 7.173 7.179 mp_alltoall_d11v 2399 14.1 6.993 7.144 6.993 7.144 calculate_dm_sparse 126 9.5 0.001 0.001 6.598 6.689 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.539 6.654 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.242 6.383 grid_collocate_task_list 126 9.7 6.235 6.270 6.235 6.270 fft3d_ps 1271 14.7 2.718 2.739 6.205 6.219 density_rs2pw 126 9.7 0.005 0.005 6.003 6.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.601000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2653.272727, yerr=176.756685 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 1.261588E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 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 0.181 0.365 87.239 87.257 qs_energies 1 2.0 0.003 0.018 85.982 86.011 ls_scf 1 3.0 0.007 0.032 85.007 85.027 dbcsr_multiply_generic 111 6.7 0.016 0.024 73.115 73.259 multiply_cannon 111 7.7 0.021 0.051 56.155 57.662 multiply_cannon_loop 111 8.7 0.210 0.226 52.725 54.455 ls_scf_main 1 4.0 0.000 0.006 52.683 52.698 density_matrix_trs4 2 5.0 0.013 0.046 47.084 47.221 ls_scf_init_scf 1 4.0 0.001 0.003 29.081 29.123 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.482 27.524 mp_waitall_1 11031 10.9 22.834 25.626 22.834 25.626 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.360 25.385 multiply_cannon_multrec 2664 9.7 8.161 8.963 15.557 17.397 multiply_cannon_sync_h2d 2664 9.7 13.564 14.903 13.564 14.903 make_m2s 222 7.7 0.009 0.013 13.235 13.651 make_images 222 8.7 0.099 0.113 13.212 13.630 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.981 12.978 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.429 9.280 make_images_data 222 9.7 0.004 0.006 7.745 8.259 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.633 7.011 8.024 hybrid_alltoall_any 227 10.6 0.216 1.831 6.594 7.915 dbcsr_mm_accdrv_process_sort 4760 11.4 6.296 7.220 6.296 7.220 calculate_norms 4752 9.8 5.504 6.234 5.504 6.234 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.154 5.259 mp_sum_l 807 5.4 3.282 4.361 3.282 4.361 make_images_sizes 222 9.7 0.000 0.000 0.788 3.646 mp_alltoall_i44 222 10.7 0.788 3.646 0.788 3.646 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.060 3.583 mp_irecv_dv 6231 10.9 2.043 3.552 2.043 3.552 arnoldi_extremal 4 6.8 0.000 0.003 3.514 3.545 arnoldi_normal_ev 4 7.8 0.033 0.043 3.514 3.545 build_subspace 16 8.4 0.009 0.015 3.344 3.349 ls_scf_post 1 4.0 0.002 0.012 3.236 3.262 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.329 3.130 ls_scf_store_result 1 5.0 0.000 0.000 2.928 2.967 dbcsr_special_finalize 555 9.7 0.005 0.006 2.351 2.809 dbcsr_merge_single_wm 555 10.7 0.457 0.592 2.343 2.802 make_images_pack 222 9.7 2.203 2.623 2.205 2.624 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.381 2.624 dbcsr_sort_data 658 11.4 2.145 2.554 2.145 2.554 dbcsr_matrix_vector_mult_local 304 10.0 2.059 2.447 2.061 2.449 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.236 2.348 buffer_matrices_ensure_size 222 8.7 1.757 2.112 1.757 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.885 1.930 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.921 qs_ks_build_kohn_sham_matrix 3 8.3 0.012 0.061 1.875 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.257000, 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 2.111082E+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_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 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 0.047 0.083 92.015 92.033 qs_energies 1 2.0 0.000 0.000 91.401 91.415 ls_scf 1 3.0 0.002 0.020 90.042 90.058 dbcsr_multiply_generic 111 6.7 0.019 0.053 75.610 75.928 multiply_cannon 111 7.7 0.028 0.040 53.350 57.743 ls_scf_main 1 4.0 0.000 0.013 55.324 55.334 multiply_cannon_loop 111 8.7 0.116 0.123 50.044 53.745 density_matrix_trs4 2 5.0 0.002 0.003 49.594 49.782 ls_scf_init_scf 1 4.0 0.000 0.002 31.100 31.102 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.776 29.892 mp_waitall_1 9105 10.9 21.208 29.634 21.208 29.634 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.365 27.378 multiply_cannon_multrec 1332 9.7 13.033 16.131 22.327 26.535 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.905 21.227 make_m2s 222 7.7 0.006 0.009 15.229 15.846 make_images 222 8.7 1.574 1.926 15.199 15.817 dbcsr_mm_accdrv_process 4041 10.4 0.270 0.434 8.897 10.447 dbcsr_mm_accdrv_process_sort 4041 11.4 8.491 10.017 8.491 10.017 make_images_data 222 9.7 0.004 0.004 8.648 9.495 mp_sum_l 807 5.4 5.739 9.231 5.739 9.231 hybrid_alltoall_any 227 10.6 0.520 2.443 8.042 9.141 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.249 7.581 mp_irecv_dv 3311 11.0 3.230 7.528 3.230 7.528 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.436 7.471 calculate_norms 2376 9.8 5.991 6.626 5.991 6.626 multiply_cannon_sync_h2d 1332 9.7 4.731 6.132 4.731 6.132 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.068 5.268 arnoldi_extremal 4 6.8 0.000 0.003 4.864 4.892 arnoldi_normal_ev 4 7.8 0.004 0.022 4.863 4.892 build_subspace 16 8.4 0.014 0.021 4.567 4.574 ls_scf_post 1 4.0 0.001 0.005 3.616 3.631 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.218 3.444 ls_scf_store_result 1 5.0 0.000 0.000 3.248 3.385 dbcsr_matrix_vector_mult_local 304 10.0 2.734 3.214 2.736 3.215 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.568 2.670 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.265 2.475 make_images_pack 222 9.7 2.018 2.419 2.020 2.421 mp_allgather_i34 111 8.7 1.004 2.320 1.004 2.320 dbcsr_sort_data 436 11.2 1.824 2.066 1.824 2.066 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.934 1.936 rebuild_ks_matrix 3 7.3 0.000 0.000 1.921 1.923 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.921 1.923 dbcsr_data_new 4174 10.1 1.613 1.841 1.613 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.033000, 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 2.694513E+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_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 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 0.088 0.174 95.211 95.213 qs_energies 1 2.0 0.002 0.017 94.335 94.375 ls_scf 1 3.0 0.005 0.037 92.947 92.974 dbcsr_multiply_generic 111 6.7 0.016 0.023 77.396 77.771 ls_scf_main 1 4.0 0.001 0.026 58.327 58.338 multiply_cannon 111 7.7 0.041 0.074 53.103 56.910 multiply_cannon_loop 111 8.7 0.101 0.115 49.395 54.114 density_matrix_trs4 2 5.0 0.003 0.016 52.367 52.543 mp_waitall_1 7281 11.0 23.751 34.118 23.751 34.118 ls_scf_init_scf 1 4.0 0.001 0.006 30.992 31.004 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.700 29.805 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.265 27.280 multiply_cannon_multrec 888 9.7 12.567 15.356 21.187 24.559 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.018 23.593 make_m2s 222 7.7 0.006 0.007 17.108 18.475 make_images 222 8.7 1.971 2.290 17.070 18.436 make_images_data 222 9.7 0.003 0.004 9.651 10.831 hybrid_alltoall_any 227 10.6 0.623 2.874 9.167 10.830 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.419 8.144 9.360 mp_sum_l 807 5.4 5.691 9.115 5.691 9.115 dbcsr_mm_accdrv_process_sort 3754 11.4 7.771 8.941 7.771 8.941 multiply_cannon_sync_h2d 888 9.7 6.114 7.896 6.114 7.896 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.480 7.177 mp_irecv_dv 2335 11.1 2.464 7.135 2.464 7.135 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.313 7.097 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.703 7.074 arnoldi_extremal 4 6.8 0.000 0.000 5.171 5.191 arnoldi_normal_ev 4 7.8 0.001 0.005 5.171 5.191 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.966 5.176 build_subspace 16 8.4 0.014 0.020 4.853 4.859 calculate_norms 1584 9.8 4.248 4.634 4.248 4.634 mp_allgather_i34 111 8.7 1.543 3.942 1.543 3.942 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.461 3.790 ls_scf_post 1 4.0 0.003 0.025 3.623 3.650 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.598 3.020 3.600 ls_scf_store_result 1 5.0 0.000 0.000 3.351 3.443 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.759 2.859 make_images_sizes 222 9.7 0.000 0.000 1.163 2.250 mp_alltoall_i44 222 10.7 1.163 2.250 1.163 2.250 dbcsr_sort_data 325 11.1 1.854 2.152 1.854 2.152 make_images_pack 222 9.7 1.817 2.127 1.819 2.130 dbcsr_data_release 9322 10.9 1.306 1.963 1.306 1.963 dbcsr_finalize 304 7.8 0.026 0.032 1.604 1.937 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.908 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.213000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 3.320787E+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_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 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 0.082 0.150 98.216 98.219 qs_energies 1 2.0 0.001 0.006 97.355 97.381 ls_scf 1 3.0 0.001 0.004 95.667 95.694 dbcsr_multiply_generic 111 6.7 0.016 0.018 79.126 79.389 ls_scf_main 1 4.0 0.006 0.039 59.137 59.146 multiply_cannon 111 7.7 0.054 0.194 52.042 56.893 density_matrix_trs4 2 5.0 0.003 0.016 53.042 53.155 multiply_cannon_loop 111 8.7 0.114 0.126 47.024 50.529 ls_scf_init_scf 1 4.0 0.001 0.006 33.242 33.264 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.786 31.854 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.304 29.318 mp_waitall_1 6369 11.0 22.985 29.106 22.985 29.106 multiply_cannon_multrec 1332 9.7 14.147 17.228 22.108 25.422 make_m2s 222 7.7 0.007 0.008 20.952 22.348 make_images 222 8.7 3.137 3.585 20.902 22.301 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.485 17.938 make_images_data 222 9.7 0.004 0.004 11.644 13.236 hybrid_alltoall_any 227 10.6 0.797 3.773 11.017 12.834 dbcsr_mm_accdrv_process 3641 10.4 0.216 0.439 7.601 9.144 dbcsr_mm_accdrv_process_sort 3641 11.4 7.216 8.720 7.216 8.720 mp_sum_l 807 5.4 4.541 8.677 4.541 8.677 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.564 6.896 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.079 6.074 multiply_cannon_sync_h2d 1332 9.7 5.520 6.024 5.520 6.024 mp_irecv_dv 3229 10.9 2.056 5.990 2.056 5.990 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.687 5.750 arnoldi_extremal 4 6.8 0.000 0.000 5.279 5.293 arnoldi_normal_ev 4 7.8 0.001 0.005 5.279 5.293 build_subspace 16 8.4 0.014 0.021 4.942 4.949 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.560 4.713 calculate_norms 2376 9.8 4.203 4.547 4.203 4.547 mp_allgather_i34 111 8.7 2.136 4.324 2.136 4.324 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.600 3.886 dbcsr_matrix_vector_mult_local 304 10.0 3.199 3.689 3.201 3.691 dbcsr_sort_data 658 11.4 3.082 3.425 3.082 3.425 ls_scf_post 1 4.0 0.002 0.008 3.288 3.314 dbcsr_special_finalize 555 9.7 0.006 0.007 2.818 3.159 dbcsr_merge_single_wm 555 10.7 0.535 0.651 2.809 3.150 ls_scf_store_result 1 5.0 0.000 0.000 3.022 3.072 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.929 2.990 dbcsr_data_release 10477 10.7 1.587 2.470 1.587 2.470 dbcsr_finalize 304 7.8 0.049 0.061 1.808 2.006 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.947 1.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.219000, 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/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 4.624458E+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_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 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 0.054 0.128 99.361 99.368 qs_energies 1 2.0 0.000 0.000 98.502 98.508 ls_scf 1 3.0 0.000 0.000 96.589 96.594 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.074 78.267 ls_scf_main 1 4.0 0.000 0.000 62.090 62.090 multiply_cannon 111 7.7 0.131 0.203 55.286 60.560 density_matrix_trs4 2 5.0 0.002 0.003 55.001 55.096 multiply_cannon_loop 111 8.7 0.070 0.081 50.735 52.229 mp_waitall_1 5436 11.0 26.667 32.690 26.667 32.690 ls_scf_init_scf 1 4.0 0.000 0.000 30.895 30.900 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.667 29.711 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.477 27.487 multiply_cannon_multrec 444 9.7 14.044 16.123 21.153 24.168 make_m2s 222 7.7 0.004 0.005 17.915 20.455 make_images 222 8.7 3.714 4.418 17.853 20.395 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.245 15.804 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.999 14.123 make_images_data 222 9.7 0.003 0.004 10.218 12.628 hybrid_alltoall_any 227 10.6 0.786 3.775 9.872 12.383 multiply_cannon_sync_h2d 444 9.7 6.526 8.770 6.526 8.770 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.341 6.818 7.948 dbcsr_mm_accdrv_process_sort 3003 11.4 6.504 7.608 6.504 7.608 mp_allgather_i34 111 8.7 2.781 7.010 2.781 7.010 arnoldi_extremal 4 6.8 0.000 0.000 5.972 5.984 arnoldi_normal_ev 4 7.8 0.003 0.008 5.972 5.984 mp_sum_l 807 5.4 2.954 5.591 2.954 5.591 build_subspace 16 8.4 0.015 0.019 5.571 5.578 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.634 4.831 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.226 4.448 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.638 4.386 mp_irecv_dv 1241 11.2 1.618 4.358 1.618 4.358 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.017 4.273 dbcsr_matrix_vector_mult_local 304 10.0 3.721 4.181 3.723 4.183 make_images_sizes 222 9.7 0.000 0.000 1.064 3.728 mp_alltoall_i44 222 10.7 1.064 3.728 1.064 3.728 calculate_norms 792 9.8 3.531 3.666 3.531 3.666 ls_scf_post 1 4.0 0.000 0.000 3.604 3.610 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.428 3.525 ls_scf_store_result 1 5.0 0.000 0.000 3.394 3.433 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.334 dbcsr_merge_all 275 8.9 0.475 0.543 2.054 2.168 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.005 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.368000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3696.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/671523f8fd0da1af59ea309227fb54ab803f3eeb_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 8.710124E+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_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 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 0.083 0.115 106.981 106.985 qs_energies 1 2.0 0.000 0.000 105.472 105.487 ls_scf 1 3.0 0.000 0.000 102.580 102.593 dbcsr_multiply_generic 111 6.7 0.023 0.027 76.356 76.443 ls_scf_main 1 4.0 0.000 0.000 65.256 65.257 density_matrix_trs4 2 5.0 0.002 0.003 56.356 56.404 multiply_cannon 111 7.7 0.143 0.228 49.139 50.924 multiply_cannon_loop 111 8.7 0.067 0.069 45.805 46.775 ls_scf_init_scf 1 4.0 0.000 0.000 33.688 33.688 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.363 32.380 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.576 29.589 mp_waitall_1 4527 11.1 21.277 24.717 21.277 24.717 make_m2s 222 7.7 0.005 0.005 23.685 24.496 make_images 222 8.7 4.578 4.956 23.578 24.388 multiply_cannon_multrec 444 9.7 17.889 18.491 22.571 23.160 hybrid_alltoall_any 227 10.6 1.662 3.656 12.814 15.424 make_images_data 222 9.7 0.003 0.003 12.915 15.168 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.785 10.306 multiply_cannon_sync_h2d 444 9.7 8.847 8.895 8.847 8.895 arnoldi_extremal 4 6.8 0.000 0.000 7.686 7.702 arnoldi_normal_ev 4 7.8 0.003 0.009 7.686 7.702 build_subspace 16 8.4 0.026 0.036 7.112 7.124 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.716 5.884 dbcsr_matrix_vector_mult_local 304 10.0 5.118 5.442 5.121 5.444 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.993 5.257 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.855 4.952 dbcsr_mm_accdrv_process 1814 10.4 0.216 0.340 4.505 4.628 dbcsr_mm_accdrv_process_sort 1814 11.4 4.197 4.332 4.197 4.332 ls_scf_post 1 4.0 0.000 0.000 3.636 3.650 make_images_sizes 222 9.7 0.000 0.000 1.444 3.484 mp_alltoall_i44 222 10.7 1.444 3.483 1.444 3.483 ls_scf_store_result 1 5.0 0.000 0.000 3.384 3.399 calculate_norms 792 9.8 3.241 3.275 3.241 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.131 mp_allgather_i34 111 8.7 0.950 2.973 0.950 2.973 dbcsr_complete_redistribute 5 7.6 1.478 1.522 2.805 2.957 dbcsr_merge_all 275 8.9 0.893 0.920 2.861 2.911 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.862 2.862 dbcsr_data_release 12724 10.6 2.334 2.861 2.334 2.861 matrix_ls_to_qs 2 6.0 0.000 0.000 2.449 2.618 dbcsr_sort_data 325 11.1 2.441 2.501 2.441 2.501 dbcsr_new_transposed 4 7.5 0.244 0.253 2.294 2.305 dbcsr_frobenius_norm 74 6.6 2.055 2.129 2.204 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.210 dbcsr_add_anytype 103 7.2 0.859 0.891 2.132 2.209 mp_alltoall_d11v 48 9.2 2.128 2.207 2.128 2.207 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.185 2.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.985000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6985.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 671523f8fd0da1af59ea309227fb54ab803f3eeb Summary: empty Status: OK