=== This is the CP2K Performance-Test === Updating 8ad26fd92..77a50e61e Fast-forward src/qs_linres_current_utils.F | 2 +- tools/toolchain/install_cp2k_toolchain.sh | 26 ---------- tools/toolchain/scripts/common_vars.sh | 3 -- tools/toolchain/scripts/get_openblas_arch.sh | 3 +- tools/toolchain/scripts/stage0/install_cmake.sh | 3 +- tools/toolchain/scripts/stage0/install_gcc.sh | 17 +++--- tools/toolchain/scripts/stage1/install_mpich.sh | 3 +- tools/toolchain/scripts/stage1/install_openmpi.sh | 3 +- tools/toolchain/scripts/stage2/install_openblas.sh | 3 +- tools/toolchain/scripts/stage3/install_fftw.sh | 3 +- tools/toolchain/scripts/stage3/install_libint.sh | 3 +- tools/toolchain/scripts/stage3/install_libxc.sh | 3 +- tools/toolchain/scripts/stage4/install_cosma.sh | 4 +- tools/toolchain/scripts/stage4/install_libxsmm.sh | 23 +++------ .../toolchain/scripts/stage4/install_scalapack.sh | 3 +- tools/toolchain/scripts/stage5/install_elpa.sh | 3 +- tools/toolchain/scripts/stage5/install_pexsi.sh | 3 +- tools/toolchain/scripts/stage5/install_ptscotch.sh | 3 +- tools/toolchain/scripts/stage5/install_superlu.sh | 3 +- tools/toolchain/scripts/stage6/install_gsl.sh | 3 +- tools/toolchain/scripts/stage6/install_plumed.sh | 3 +- tools/toolchain/scripts/stage6/install_quip.sh | 3 +- tools/toolchain/scripts/stage7/install_hdf5.sh | 3 +- tools/toolchain/scripts/stage7/install_libvdwxc.sh | 17 +----- tools/toolchain/scripts/stage7/install_libvori.sh | 4 +- tools/toolchain/scripts/stage7/install_spglib.sh | 4 +- tools/toolchain/scripts/stage8/install_sirius.sh | 4 +- tools/toolchain/scripts/stage8/install_spfft.sh | 4 +- tools/toolchain/scripts/stage8/install_spla.sh | 4 +- tools/toolchain/scripts/tool_kit.sh | 60 +++------------------- 30 files changed, 44 insertions(+), 179 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 77a50e61e3cb5b09a01571247af7daff50d68f23 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/01 job id: 41786531 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/02 job id: 41786534 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/03 job id: 41786537 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/04 job id: 41786540 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/05 job id: 41786543 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/06 job id: 41786545 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/07 job id: 41786546 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/08 job id: 41786547 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/09 job id: 41786548 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/10 job id: 41786550 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/11 job id: 41786551 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/12 job id: 41786552 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/13 job id: 41786554 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/14 job id: 41786556 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/15 job id: 41786557 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/16 job id: 41786558 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/17 job id: 41786559 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/18 job id: 41786560 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/19 job id: 41786561 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/20 job id: 41786562 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/21 job id: 41786563 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/22 job id: 41786565 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/23 job id: 41786567 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/24 job id: 41786568 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/25 job id: 41786569 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/26 job id: 41786571 --- 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/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 134.543 134.544 farming_run 1 2.0 133.309 133.310 134.516 134.519 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.454379E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.027 115.883 115.883 qs_energies 1 2.0 0.000 0.000 115.613 115.616 mp2_main 1 3.0 0.000 0.000 113.762 113.765 mp2_gpw_main 1 4.0 0.020 0.026 112.920 112.922 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.982 94.402 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.534 55.954 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.171 41.822 47.207 get_2c_integrals 1 6.0 0.000 0.000 37.567 38.277 integrate_v_rspace 273 8.0 0.439 0.453 25.129 30.209 pw_transfer 6555 10.6 0.374 0.384 27.432 28.336 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.082 26.825 grid_integrate_task_list 273 9.0 20.932 26.482 20.932 26.482 fft_wrap_pw1pw2_100 2178 12.4 1.214 1.455 23.640 24.374 compute_2c_integrals 1 7.0 0.003 0.003 19.891 19.892 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.947 19.597 mp2_eri_2c_integrate_gpw 1 9.0 2.376 2.430 18.944 19.594 rpa_ri_compute_en 1 5.0 0.013 0.018 18.826 19.004 cp_fm_cholesky_decompose 12 8.2 17.632 18.362 17.632 18.362 cholesky_decomp 1 7.0 0.000 0.000 16.528 17.249 fft3d_s 5443 13.4 16.164 16.713 16.186 16.734 ao_to_mo_and_store_B_mult_1 272 7.0 10.873 15.598 10.873 15.598 calculate_wavefunction 272 8.0 5.460 5.608 12.596 13.206 rpa_num_int 1 6.0 0.000 0.000 10.542 10.543 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.499 10.543 calc_mat_Q 8 8.0 0.000 0.000 9.348 9.445 contract_S_to_Q 8 9.0 0.000 0.000 8.770 8.869 calc_potential_gpw 544 9.5 0.005 0.006 8.302 8.738 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.256 8.649 potential_pw2rs 545 10.0 0.108 0.110 7.704 8.497 parallel_gemm_fm 14 9.1 0.000 0.000 8.358 8.451 parallel_gemm_fm_cosma 14 10.1 8.358 8.451 8.358 8.451 collocate_single_gaussian 272 10.0 0.040 0.043 7.491 7.852 create_integ_mat 1 6.0 0.026 0.045 7.755 7.769 array2fm 1 7.0 0.000 0.000 6.614 7.166 pw_scatter_s 2720 13.7 4.445 4.612 4.445 4.612 pw_gather_s 2722 13.2 3.886 4.246 3.886 4.246 array2fm_buffer_send 1 8.0 2.960 3.512 2.960 3.512 pw_poisson_solve 545 10.5 1.136 1.182 2.192 2.386 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.917918, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.058 397.814 397.822 farming_run 1 2.0 397.027 397.071 397.756 397.764 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223967E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 3951177. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.051 210.337 210.337 qs_energies 1 2.0 0.000 0.000 210.064 210.077 scf_env_do_scf 1 3.0 0.000 0.000 107.231 107.231 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.378 106.388 rebuild_ks_matrix 4 6.0 0.000 0.000 106.377 106.387 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.064 106.377 106.387 hfx_ks_matrix 4 8.0 0.001 0.001 106.001 106.004 integrate_four_center 4 9.0 0.143 0.461 106.000 106.003 mp2_main 1 3.0 0.002 0.017 102.550 102.563 mp2_gpw_main 1 4.0 0.041 0.099 101.693 101.707 integrate_four_center_main 4 10.0 0.079 0.513 96.770 99.824 integrate_four_center_bin 266 11.0 96.691 99.807 96.691 99.807 init_scf_loop 1 4.0 0.000 0.000 92.875 92.875 mp2_ri_gpw_compute_in 1 5.0 0.070 0.097 74.971 76.013 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.475 55.513 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 42.263 47.279 integrate_v_rspace 95 8.0 0.398 0.569 28.627 33.443 pw_transfer 2240 10.6 0.144 0.165 29.883 30.378 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.899 29.426 ao_to_mo_and_store_B_mult_1 91 7.0 10.522 28.984 10.522 28.984 grid_integrate_task_list 95 9.0 23.918 28.951 23.918 28.951 mp2_ri_gpw_compute_en 1 5.0 0.058 0.070 26.520 28.241 fft_wrap_pw1pw2_100 730 12.4 1.280 1.450 26.626 27.024 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.858 1.915 24.833 24.842 get_2c_integrals 1 6.0 0.002 0.021 20.403 20.436 compute_2c_integrals 1 7.0 0.012 0.063 19.381 19.395 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 18.857 19.199 mp2_eri_2c_integrate_gpw 1 9.0 1.737 1.866 18.855 19.199 fft3d_s 1823 13.4 18.414 18.838 18.427 18.851 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.354 14.354 calculate_wavefunction 91 8.0 2.024 2.054 9.748 9.991 potential_pw2rs 186 10.0 0.034 0.035 8.638 9.161 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.578 8.712 9.050 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.241 8.655 local_gemm 172 8.0 8.157 8.490 8.157 8.490 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.522 7.867 8.218 collocate_single_gaussian 91 10.0 0.017 0.025 7.873 8.145 calc_potential_gpw 182 9.5 0.002 0.002 7.909 8.134 mp_sync 38 10.4 3.889 7.248 3.889 7.248 mp2_ri_gpw_compute_en_ener 172 7.0 6.351 6.431 6.351 6.431 mp_sendrecv_dm3 2068 8.0 5.894 6.250 5.894 6.250 pw_gather_s 912 13.2 4.913 5.451 4.913 5.451 pw_scatter_s 910 13.7 3.928 4.265 3.928 4.265 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.685943, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 449.699840E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.121 52.120 52.125 qs_mol_dyn_low 1 2.0 0.008 0.022 51.741 51.754 qs_forces 11 3.9 0.002 0.003 51.632 51.633 qs_energies 11 4.9 0.003 0.007 50.188 50.199 scf_env_do_scf 11 5.9 0.001 0.002 42.263 42.264 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 40.124 40.125 dbcsr_multiply_generic 2286 12.5 0.091 0.098 33.064 33.476 qs_scf_new_mos 108 7.5 0.000 0.001 30.175 30.456 qs_scf_loop_do_ot 108 8.5 0.001 0.001 30.174 30.456 ot_scf_mini 108 9.5 0.002 0.002 28.528 28.706 multiply_cannon 2286 13.5 0.188 0.195 26.548 27.948 multiply_cannon_loop 2286 14.5 1.495 1.568 25.850 27.245 velocity_verlet 10 3.0 0.002 0.002 24.659 24.662 ot_mini 108 10.5 0.001 0.001 18.134 18.401 qs_ot_get_derivative 108 11.5 0.018 0.074 15.153 15.358 mp_waitall_1 267858 16.1 8.015 14.328 8.015 14.328 multiply_cannon_metrocomm3 54864 15.5 0.066 0.072 5.572 12.551 multiply_cannon_multrec 54864 15.5 4.266 6.743 8.595 12.129 rebuild_ks_matrix 119 8.3 0.000 0.000 7.834 7.976 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.015 7.833 7.975 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.934 7.067 multiply_cannon_sync_h2d 54864 15.5 5.970 6.820 5.970 6.820 init_scf_run 11 5.9 0.000 0.001 6.633 6.633 scf_env_initial_rho_setup 11 6.9 0.039 0.157 6.632 6.633 mp_sum_l 7207 12.9 4.624 6.236 4.624 6.236 qs_ot_get_p 119 10.4 0.001 0.001 5.933 6.188 dbcsr_mm_accdrv_process 76910 16.1 1.163 1.808 4.251 5.716 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.217 5.690 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.920 5.039 sum_up_and_integrate 119 10.3 0.013 0.016 4.493 4.501 integrate_v_rspace 119 11.3 0.002 0.003 4.480 4.489 qs_rho_update_rho_low 119 7.7 0.036 0.143 4.294 4.448 calculate_rho_elec 119 8.7 0.011 0.017 4.258 4.448 calculate_dm_sparse 119 9.5 0.000 0.000 4.071 4.224 calculate_first_density_matrix 1 7.0 0.001 0.008 4.199 4.205 jit_kernel_multiply 13 15.8 3.028 3.830 3.028 3.830 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.332 3.377 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.869 3.072 apply_single 119 13.6 0.000 0.000 2.869 3.072 rs_pw_transfer 974 11.9 0.012 0.014 2.826 2.931 ot_diis_step 108 11.5 0.017 0.064 2.729 2.730 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.626 2.627 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.585 2.600 mp_sum_d 4129 12.0 1.669 2.421 1.669 2.421 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.544 2.419 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.416 2.416 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.322 2.401 cp_fm_redistribute_end 50 14.0 2.195 2.394 2.200 2.396 acc_transpose_blocks 54864 15.5 0.223 0.243 1.965 2.388 cp_fm_diag_elpa_base 50 14.0 0.194 2.349 0.195 2.362 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.309 2.311 density_rs2pw 119 9.7 0.004 0.004 2.183 2.264 init_scf_loop 11 6.9 0.004 0.031 2.122 2.122 grid_integrate_task_list 119 12.3 2.018 2.101 2.018 2.101 wfi_extrapolate 11 7.9 0.001 0.002 2.084 2.085 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.031 2.074 potential_pw2rs 119 12.3 0.004 0.005 1.872 1.886 pw_transfer 1439 11.6 0.051 0.055 1.628 1.703 mp_alltoall_d11v 2130 13.8 1.483 1.703 1.483 1.703 make_m2s 4572 13.5 0.053 0.055 1.637 1.698 dbcsr_dot_sd 1205 11.9 0.047 0.058 1.116 1.633 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.553 1.629 make_images 4572 14.5 0.132 0.137 1.556 1.616 mp_waitany 12084 13.8 1.280 1.462 1.280 1.462 fft3d_ps 1201 14.6 0.353 0.456 1.330 1.402 grid_collocate_task_list 119 9.7 1.292 1.360 1.292 1.360 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.195 1.270 acc_transpose_blocks_kernels 54864 16.5 0.248 0.382 0.994 1.265 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.125000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 486.526976E+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 1255006. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.108 39.733 39.734 qs_mol_dyn_low 1 2.0 0.003 0.005 39.103 39.304 qs_forces 11 3.9 0.002 0.003 38.949 38.950 qs_energies 11 4.9 0.001 0.003 37.287 37.290 scf_env_do_scf 11 5.9 0.001 0.001 31.370 31.371 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.903 28.903 dbcsr_multiply_generic 2286 12.5 0.098 0.104 22.526 22.905 qs_scf_new_mos 108 7.5 0.001 0.001 20.101 20.337 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.100 20.337 ot_scf_mini 108 9.5 0.002 0.003 19.194 19.368 multiply_cannon 2286 13.5 0.210 0.221 17.375 18.914 velocity_verlet 10 3.0 0.003 0.017 18.038 18.040 multiply_cannon_loop 2286 14.5 0.906 0.986 16.256 17.813 ot_mini 108 10.5 0.001 0.001 12.158 12.393 mp_waitall_1 217478 16.2 6.181 11.566 6.181 11.566 multiply_cannon_metrocomm3 27432 15.5 0.068 0.072 4.455 9.998 qs_ot_get_derivative 108 11.5 0.001 0.001 9.679 9.855 multiply_cannon_multrec 27432 15.5 1.964 4.426 6.299 9.395 rebuild_ks_matrix 119 8.3 0.000 0.000 7.050 7.195 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.028 7.050 7.194 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.234 6.365 dbcsr_mm_accdrv_process 47894 16.0 3.329 5.687 4.264 6.345 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.820 4.881 init_scf_run 11 5.9 0.000 0.001 4.640 4.640 scf_env_initial_rho_setup 11 6.9 0.003 0.020 4.640 4.640 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.264 4.382 apply_single 119 13.6 0.000 0.000 3.264 4.382 qs_ot_get_p 119 10.4 0.001 0.001 4.115 4.346 mp_sum_l 7207 12.9 2.328 4.319 2.328 4.319 sum_up_and_integrate 119 10.3 0.025 0.029 4.077 4.082 integrate_v_rspace 119 11.3 0.002 0.003 4.051 4.058 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.529 3.564 calculate_rho_elec 119 8.7 0.021 0.024 3.529 3.563 calculate_first_density_matrix 1 7.0 0.000 0.001 3.136 3.139 multiply_cannon_sync_h2d 27432 15.5 2.176 2.845 2.176 2.845 calculate_dm_sparse 119 9.5 0.000 0.001 2.686 2.763 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.701 2.721 make_m2s 4572 13.5 0.051 0.053 2.451 2.678 rs_pw_transfer 974 11.9 0.010 0.011 2.463 2.594 make_images 4572 14.5 0.200 0.237 2.366 2.592 init_scf_loop 11 6.9 0.000 0.000 2.448 2.448 ot_diis_step 108 11.5 0.011 0.018 2.427 2.428 jit_kernel_multiply 9 16.2 0.884 2.318 0.884 2.318 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.260 2.261 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.099 2.188 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.067 2.069 density_rs2pw 119 9.7 0.004 0.004 1.882 1.975 grid_integrate_task_list 119 12.3 1.838 1.950 1.838 1.950 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.929 1.930 cp_fm_redistribute_end 50 14.0 1.592 1.906 1.596 1.907 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.866 1.905 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.886 1.894 cp_fm_diag_elpa_base 50 14.0 0.302 1.853 0.310 1.884 acc_transpose_blocks 27432 15.5 0.107 0.112 1.557 1.865 potential_pw2rs 119 12.3 0.006 0.007 1.822 1.831 pw_transfer 1439 11.6 0.063 0.067 1.722 1.757 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.632 1.668 prepare_preconditioner 11 7.9 0.000 0.000 1.548 1.575 make_preconditioner 11 8.9 0.000 0.000 1.548 1.575 make_images_data 4572 15.5 0.045 0.052 1.140 1.546 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.449 1.506 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.988 1.444 wfi_extrapolate 11 7.9 0.001 0.003 1.439 1.439 mp_alltoall_d11v 2130 13.8 1.186 1.393 1.186 1.393 fft3d_ps 1201 14.6 0.496 0.551 1.339 1.369 mp_allgather_i34 2286 14.5 0.562 1.328 0.562 1.328 fft_wrap_pw1pw2_140 487 13.2 0.077 0.085 1.286 1.321 grid_collocate_task_list 119 9.7 1.242 1.314 1.242 1.314 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 1.028 1.250 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.246 mp_sum_d 4129 12.0 0.591 1.046 0.591 1.046 jit_kernel_transpose 5 15.5 0.847 0.989 0.847 0.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.986 0.987 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.872 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.734000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.181818, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 522.838016E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.073 35.632 35.634 qs_mol_dyn_low 1 2.0 0.003 0.004 35.247 35.255 qs_forces 11 3.9 0.005 0.029 35.180 35.187 qs_energies 11 4.9 0.002 0.003 33.566 33.568 scf_env_do_scf 11 5.9 0.001 0.001 27.132 27.132 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.165 24.167 dbcsr_multiply_generic 2286 12.5 0.176 0.573 19.229 19.381 qs_scf_new_mos 108 7.5 0.001 0.001 16.141 16.167 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.140 16.166 multiply_cannon 2286 13.5 0.195 0.203 14.964 15.807 ot_scf_mini 108 9.5 0.002 0.003 15.402 15.425 velocity_verlet 10 3.0 0.002 0.003 15.377 15.378 multiply_cannon_loop 2286 14.5 0.639 0.663 14.132 15.012 ot_mini 108 10.5 0.047 0.373 9.902 9.920 multiply_cannon_multrec 18288 15.5 2.067 3.109 8.018 8.606 qs_ot_get_derivative 108 11.5 0.001 0.002 8.006 8.030 dbcsr_mm_accdrv_process 38222 16.0 4.567 6.756 5.867 7.274 rebuild_ks_matrix 119 8.3 0.000 0.000 6.348 6.363 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.347 6.362 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.602 5.616 init_scf_run 11 5.9 0.000 0.001 5.239 5.239 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.238 5.239 mp_waitall_1 169478 16.3 3.336 4.431 3.336 4.431 calculate_first_density_matrix 1 7.0 0.000 0.000 3.986 3.988 sum_up_and_integrate 119 10.3 0.031 0.032 3.956 3.964 integrate_v_rspace 119 11.3 0.002 0.003 3.925 3.937 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.180 3.799 qs_ot_get_p 119 10.4 0.007 0.051 3.425 3.452 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.301 3.307 calculate_rho_elec 119 8.7 0.030 0.032 3.301 3.306 jit_kernel_multiply 11 16.0 1.249 3.155 1.249 3.155 init_scf_loop 11 6.9 0.087 0.233 2.948 2.950 calculate_dm_sparse 119 9.5 0.000 0.000 2.810 2.822 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.824 2.727 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.368 2.726 apply_single 119 13.6 0.000 0.000 2.368 2.726 make_m2s 4572 13.5 0.044 0.045 2.270 2.472 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.463 2.465 rs_pw_transfer 974 11.9 0.009 0.010 2.307 2.400 make_images 4572 14.5 0.191 0.202 2.186 2.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.281 2.292 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.174 2.179 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.164 2.167 density_rs2pw 119 9.7 0.004 0.004 1.883 1.968 mp_sum_l 7207 12.9 1.562 1.943 1.562 1.943 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.905 1.906 grid_integrate_task_list 119 12.3 1.802 1.901 1.802 1.901 ot_diis_step 108 11.5 0.011 0.014 1.824 1.870 prepare_preconditioner 11 7.9 0.000 0.000 1.796 1.801 make_preconditioner 11 8.9 0.000 0.001 1.796 1.801 pw_transfer 1439 11.6 0.063 0.066 1.778 1.792 potential_pw2rs 119 12.3 0.007 0.008 1.737 1.748 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.647 1.731 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.688 1.704 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.640 1.641 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.623 1.633 cp_fm_redistribute_end 50 14.0 1.216 1.615 1.217 1.615 cp_fm_diag_elpa_base 50 14.0 0.381 1.544 0.396 1.588 acc_transpose_blocks 18288 15.5 0.074 0.076 1.523 1.585 multiply_cannon_sync_h2d 18288 15.5 1.376 1.564 1.376 1.564 fft3d_ps 1201 14.6 0.505 0.522 1.375 1.391 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.300 1.317 grid_collocate_task_list 119 9.7 1.205 1.299 1.205 1.299 wfi_extrapolate 11 7.9 0.001 0.001 1.193 1.193 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 1.086 1.147 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.352 1.042 make_images_data 4572 15.5 0.045 0.048 0.838 0.986 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.968 0.969 jit_kernel_transpose 5 15.6 0.877 0.942 0.877 0.942 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.907 0.933 make_images_sizes 4572 15.5 0.005 0.005 0.749 0.924 hybrid_alltoall_any 4725 16.4 0.055 0.115 0.728 0.923 mp_alltoall_i44 4572 16.5 0.744 0.919 0.744 0.919 mp_alltoall_d11v 2130 13.8 0.732 0.905 0.732 0.905 dbcsr_complete_redistribute 329 12.2 0.097 0.150 0.681 0.866 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.816 0.820 mp_alltoall_z22v 1201 16.6 0.699 0.773 0.699 0.773 cp_fm_cholesky_invert 11 10.9 0.759 0.763 0.759 0.763 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.638 0.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.634000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.636364, yerr=3.364865 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 558.571520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.157 37.199 37.200 qs_mol_dyn_low 1 2.0 0.007 0.026 36.493 36.499 qs_forces 11 3.9 0.031 0.061 36.326 36.340 qs_energies 11 4.9 0.003 0.014 34.580 34.594 scf_env_do_scf 11 5.9 0.001 0.001 28.018 28.019 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.172 24.200 dbcsr_multiply_generic 2286 12.5 0.098 0.101 18.830 18.953 velocity_verlet 10 3.0 0.002 0.003 17.471 17.482 qs_scf_new_mos 108 7.5 0.001 0.001 16.145 16.226 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.144 16.225 multiply_cannon 2286 13.5 0.224 0.249 15.268 15.636 ot_scf_mini 108 9.5 0.003 0.003 15.216 15.293 multiply_cannon_loop 2286 14.5 0.936 0.961 14.347 14.705 multiply_cannon_multrec 27432 15.5 2.335 3.014 9.236 9.621 ot_mini 108 10.5 0.001 0.002 9.438 9.532 dbcsr_mm_accdrv_process 47916 15.9 5.707 7.962 6.807 8.016 qs_ot_get_derivative 108 11.5 0.001 0.001 7.599 7.675 rebuild_ks_matrix 119 8.3 0.000 0.000 6.422 6.477 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.016 6.422 6.476 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.707 5.755 init_scf_run 11 5.9 0.000 0.001 4.942 4.942 scf_env_initial_rho_setup 11 6.9 0.002 0.002 4.941 4.942 init_scf_loop 11 6.9 0.001 0.001 3.828 3.888 qs_ot_get_p 119 10.4 0.001 0.001 3.630 3.729 sum_up_and_integrate 119 10.3 0.036 0.040 3.700 3.706 integrate_v_rspace 119 11.3 0.002 0.003 3.664 3.672 calculate_first_density_matrix 1 7.0 0.001 0.001 3.575 3.577 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.060 3.489 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.230 3.261 calculate_rho_elec 119 8.7 0.040 0.046 3.230 3.260 mp_waitall_1 145218 16.4 2.107 2.728 2.107 2.728 prepare_preconditioner 11 7.9 0.000 0.000 2.542 2.551 make_preconditioner 11 8.9 0.011 0.031 2.542 2.551 calculate_dm_sparse 119 9.5 0.000 0.000 2.460 2.517 jit_kernel_multiply 10 15.9 1.040 2.517 1.040 2.517 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.140 2.475 rs_pw_transfer 974 11.9 0.009 0.009 2.069 2.270 make_m2s 4572 13.5 0.053 0.055 2.133 2.240 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.029 2.153 apply_single 119 13.6 0.000 0.000 2.028 2.153 make_images 4572 14.5 0.274 0.335 2.026 2.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.033 2.035 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.944 1.952 grid_integrate_task_list 119 12.3 1.824 1.920 1.824 1.920 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.882 1.908 density_rs2pw 119 9.7 0.004 0.004 1.722 1.870 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.849 1.864 ot_diis_step 108 11.5 0.022 0.053 1.799 1.799 pw_transfer 1439 11.6 0.063 0.066 1.732 1.760 acc_transpose_blocks 27432 15.5 0.111 0.118 1.589 1.752 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.642 1.674 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.648 1.658 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.634 1.634 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.882 1.573 potential_pw2rs 119 12.3 0.008 0.010 1.485 1.489 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.406 1.407 cp_fm_redistribute_end 50 14.0 0.927 1.384 0.929 1.384 fft_wrap_pw1pw2_140 487 13.2 0.085 0.094 1.347 1.380 cp_fm_diag_elpa_base 50 14.0 0.432 1.321 0.453 1.362 fft3d_ps 1201 14.6 0.534 0.584 1.322 1.347 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.332 1.346 mp_sum_l 7207 12.9 1.038 1.322 1.038 1.322 wfi_extrapolate 11 7.9 0.001 0.002 1.319 1.319 grid_collocate_task_list 119 9.7 1.222 1.317 1.222 1.317 mp_sum_d 4127 12.0 0.801 1.141 0.801 1.141 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.967 1.116 cp_fm_upper_to_full 72 13.5 0.805 1.111 0.805 1.111 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.089 1.109 multiply_cannon_sync_h2d 27432 15.5 1.003 1.067 1.003 1.067 dbcsr_complete_redistribute 329 12.2 0.119 0.142 0.769 1.045 make_images_data 4572 15.5 0.045 0.049 0.830 0.946 cp_fm_cholesky_decompose 22 10.9 0.706 0.928 0.706 0.928 arnoldi_extremal 119 11.4 0.018 0.089 0.843 0.916 parallel_gemm_fm 81 9.0 0.000 0.000 0.668 0.911 parallel_gemm_fm_cosma 81 10.0 0.668 0.911 0.668 0.911 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.706 0.889 arnoldi_normal_ev 119 12.4 0.055 0.170 0.824 0.886 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.885 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.741 0.874 make_basis_sm 11 9.8 0.000 0.000 0.866 0.867 mp_alltoall_d11v 2130 13.8 0.718 0.849 0.718 0.849 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.577 0.846 jit_kernel_transpose 5 15.6 0.700 0.844 0.700 0.844 rs_pw_transfer_RS2PW_140 130 11.5 0.109 0.114 0.624 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.809 mp_waitany 7280 13.7 0.618 0.791 0.618 0.791 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.200000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.636364, yerr=3.523240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 605.995008E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.044 29.345 29.346 qs_mol_dyn_low 1 2.0 0.003 0.003 28.647 28.654 qs_forces 11 3.9 0.011 0.014 28.254 28.255 qs_energies 11 4.9 0.008 0.008 26.567 26.570 scf_env_do_scf 11 5.9 0.000 0.001 20.998 20.998 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.519 18.520 velocity_verlet 10 3.0 0.003 0.012 14.353 14.377 dbcsr_multiply_generic 2286 12.5 0.101 0.114 13.145 13.208 qs_scf_new_mos 108 7.5 0.001 0.001 11.047 11.071 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.046 11.071 multiply_cannon 2286 13.5 0.230 0.236 10.552 11.019 ot_scf_mini 108 9.5 0.002 0.002 10.375 10.401 multiply_cannon_loop 2286 14.5 0.330 0.341 9.648 9.830 multiply_cannon_multrec 9144 15.5 1.699 1.971 6.248 6.789 ot_mini 108 10.5 0.001 0.001 5.908 5.939 rebuild_ks_matrix 119 8.3 0.000 0.000 5.736 5.754 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.735 5.753 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.118 5.135 dbcsr_mm_accdrv_process 12550 15.8 3.267 3.943 4.448 4.876 qs_ot_get_derivative 108 11.5 0.001 0.001 4.597 4.634 init_scf_run 11 5.9 0.000 0.001 4.087 4.087 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.086 4.087 sum_up_and_integrate 119 10.3 0.038 0.042 3.494 3.496 integrate_v_rspace 119 11.3 0.003 0.003 3.455 3.459 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.173 3.182 calculate_rho_elec 119 8.7 0.059 0.061 3.173 3.181 calculate_first_density_matrix 1 7.0 0.000 0.000 2.969 2.970 qs_ot_get_p 119 10.4 0.001 0.001 2.691 2.739 mp_waitall_1 121218 16.5 1.995 2.506 1.995 2.506 init_scf_loop 11 6.9 0.000 0.000 2.458 2.459 calculate_dm_sparse 119 9.5 0.000 0.000 2.281 2.300 jit_kernel_multiply 10 15.7 1.143 2.133 1.143 2.133 make_m2s 4572 13.5 0.034 0.035 1.793 1.961 grid_integrate_task_list 119 12.3 1.848 1.918 1.848 1.918 make_images 4572 14.5 0.268 0.303 1.705 1.871 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.839 1.841 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.767 1.770 prepare_preconditioner 11 7.9 0.000 0.000 1.735 1.739 make_preconditioner 11 8.9 0.000 0.000 1.735 1.739 pw_transfer 1439 11.6 0.063 0.065 1.704 1.712 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.702 1.707 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.657 1.666 rs_pw_transfer 974 11.9 0.008 0.009 1.586 1.658 density_rs2pw 119 9.7 0.004 0.004 1.590 1.649 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.622 1.646 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.613 1.621 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.606 1.618 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.574 1.575 grid_collocate_task_list 119 9.7 1.271 1.351 1.271 1.351 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.314 1.322 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.306 1.307 ot_diis_step 108 11.5 0.012 0.013 1.290 1.290 fft3d_ps 1201 14.6 0.539 0.551 1.280 1.289 cp_fm_redistribute_end 50 14.0 0.650 1.286 0.650 1.286 potential_pw2rs 119 12.3 0.010 0.011 1.284 1.286 cp_fm_diag_elpa_base 50 14.0 0.590 1.215 0.634 1.271 qs_energies_init_hamiltonians 11 5.9 0.020 0.029 1.263 1.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.162 1.179 apply_single 119 13.6 0.000 0.000 1.162 1.179 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.126 1.142 acc_transpose_blocks 9144 15.5 0.039 0.040 1.063 1.110 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.532 1.106 wfi_extrapolate 11 7.9 0.001 0.001 1.070 1.070 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.782 1.034 make_images_data 4572 15.5 0.038 0.042 0.819 1.018 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.920 mp_alltoall_d11v 2130 13.8 0.774 0.878 0.774 0.878 cp_fm_cholesky_invert 11 10.9 0.869 0.872 0.869 0.872 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.823 0.866 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.487 0.835 multiply_cannon_sync_h2d 9144 15.5 0.717 0.802 0.717 0.802 jit_kernel_transpose 5 15.6 0.707 0.749 0.707 0.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.741 0.742 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.679 0.726 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.713 0.722 mp_allgather_i34 2286 14.5 0.237 0.664 0.237 0.664 mp_alltoall_z22v 1201 16.6 0.615 0.647 0.615 0.647 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.346000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.636364, yerr=7.327070 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 745.857024E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.078 41.752 41.753 qs_mol_dyn_low 1 2.0 0.003 0.003 41.454 41.462 qs_forces 11 3.9 0.023 0.070 41.272 41.272 qs_energies 11 4.9 0.006 0.012 39.161 39.188 scf_env_do_scf 11 5.9 0.001 0.001 32.252 32.252 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.470 24.472 velocity_verlet 10 3.0 0.002 0.002 22.905 22.926 dbcsr_multiply_generic 2286 12.5 0.098 0.100 17.877 17.990 qs_scf_new_mos 108 7.5 0.001 0.001 15.741 15.841 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.740 15.840 multiply_cannon 2286 13.5 0.305 0.314 14.104 14.936 ot_scf_mini 108 9.5 0.002 0.002 14.662 14.766 multiply_cannon_loop 2286 14.5 0.343 0.350 12.836 13.697 ot_mini 108 10.5 0.001 0.001 8.740 8.859 multiply_cannon_multrec 9144 15.5 3.413 4.979 8.680 8.775 init_scf_loop 11 6.9 0.000 0.000 7.752 7.759 rebuild_ks_matrix 119 8.3 0.000 0.000 6.930 7.076 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.930 7.076 qs_ot_get_derivative 108 11.5 0.001 0.001 6.706 6.812 prepare_preconditioner 11 7.9 0.000 0.000 6.794 6.806 make_preconditioner 11 8.9 0.000 0.000 6.793 6.806 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.384 6.687 dbcsr_mm_accdrv_process 12550 15.8 4.145 5.665 5.143 6.481 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.133 6.265 init_scf_run 11 5.9 0.000 0.001 4.771 4.771 scf_env_initial_rho_setup 11 6.9 0.009 0.018 4.770 4.771 cp_fm_upper_to_full 72 14.2 3.164 4.515 3.164 4.515 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.674 3.750 calculate_rho_elec 119 8.7 0.118 0.121 3.674 3.749 sum_up_and_integrate 119 10.3 0.064 0.066 3.674 3.681 integrate_v_rspace 119 11.3 0.003 0.004 3.610 3.617 mp_waitall_1 97218 16.6 2.625 3.565 2.625 3.565 calculate_first_density_matrix 1 7.0 0.000 0.000 3.294 3.380 qs_ot_get_p 119 10.4 0.001 0.001 3.187 3.326 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.470 2.856 calculate_dm_sparse 119 9.5 0.000 0.000 2.680 2.750 dbcsr_complete_redistribute 329 12.2 0.289 0.295 1.937 2.735 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.160 2.458 apply_single 119 13.6 0.000 0.000 2.160 2.458 make_m2s 4572 13.5 0.037 0.038 2.255 2.439 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.631 2.424 make_images 4572 14.5 0.353 0.385 2.135 2.320 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.434 2.283 mp_alltoall_i22 627 13.8 1.424 2.222 1.424 2.222 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.404 2.192 pw_transfer 1439 11.6 0.066 0.067 2.045 2.047 grid_integrate_task_list 119 12.3 1.991 2.014 1.991 2.014 ot_diis_step 108 11.5 0.014 0.014 2.009 2.009 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.962 1.964 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.949 1.952 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.864 1.886 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.874 1.875 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.752 1.823 jit_kernel_multiply 10 15.6 0.971 1.788 0.971 1.788 mp_sum_l 7207 12.9 0.995 1.759 0.995 1.759 density_rs2pw 119 9.7 0.003 0.004 1.722 1.737 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.659 1.713 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.691 1.691 fft_wrap_pw1pw2_140 487 13.2 0.088 0.089 1.651 1.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.583 1.588 fft3d_ps 1201 14.6 0.566 0.578 1.583 1.586 cp_fm_cholesky_invert 11 10.9 1.465 1.468 1.465 1.468 rs_pw_transfer 974 11.9 0.009 0.009 1.406 1.460 grid_collocate_task_list 119 9.7 1.437 1.451 1.437 1.451 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.406 1.408 cp_fm_diag_elpa_base 50 14.0 1.261 1.312 1.404 1.404 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.109 1.371 wfi_extrapolate 11 7.9 0.001 0.001 1.334 1.334 mp_alltoall_d11v 2130 13.8 1.213 1.301 1.213 1.301 make_images_data 4572 15.5 0.042 0.046 1.065 1.293 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.204 1.241 potential_pw2rs 119 12.3 0.014 0.015 1.218 1.221 acc_transpose_blocks 9144 15.5 0.038 0.039 1.184 1.196 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.136 1.158 multiply_cannon_sync_h2d 9144 15.5 1.040 1.042 1.040 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.027 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.008 1.022 qs_create_task_list 11 7.9 0.021 0.066 1.006 1.017 generate_qs_task_list 11 8.9 0.372 0.392 0.985 1.010 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.935 0.946 make_basis_sm 11 9.8 0.000 0.000 0.943 0.944 mp_alltoall_z22v 1201 16.6 0.881 0.902 0.881 0.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.753000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=698.181818, yerr=14.995867 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.043776E+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 2251817. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.195 84.250 84.252 qs_mol_dyn_low 1 2.0 0.003 0.005 83.648 83.686 qs_forces 11 3.9 0.007 0.040 83.521 83.527 qs_energies 11 4.9 0.009 0.067 80.368 80.378 scf_env_do_scf 11 5.9 0.001 0.002 70.891 70.899 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.075 65.076 dbcsr_multiply_generic 2055 12.4 0.101 0.104 52.663 52.958 qs_scf_new_mos 99 7.5 0.000 0.001 48.161 48.276 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.161 48.276 ot_scf_mini 99 9.5 0.002 0.003 45.737 45.832 multiply_cannon 2055 13.4 0.182 0.187 43.027 44.129 multiply_cannon_loop 2055 14.4 1.561 1.607 42.040 43.057 velocity_verlet 10 3.0 0.013 0.047 41.986 41.993 ot_mini 99 10.5 0.001 0.002 27.676 27.747 qs_ot_get_derivative 99 11.5 0.012 0.150 20.879 20.954 multiply_cannon_multrec 49320 15.4 12.100 12.873 17.366 18.040 rebuild_ks_matrix 110 8.3 0.000 0.001 14.556 14.641 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.021 14.556 14.640 mp_waitall_1 241148 16.1 11.976 13.042 11.976 13.042 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.839 12.927 multiply_cannon_sync_h2d 49320 15.4 9.925 10.520 9.925 10.520 qs_ot_get_p 110 10.4 0.001 0.001 9.399 9.498 multiply_cannon_metrocomm3 49320 15.4 0.077 0.081 7.048 8.313 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.092 7.641 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.122 7.628 apply_single 110 13.6 0.000 0.000 7.122 7.628 init_scf_run 11 5.9 0.000 0.001 7.310 7.311 scf_env_initial_rho_setup 11 6.9 0.003 0.021 7.310 7.310 sum_up_and_integrate 110 10.3 0.038 0.045 7.008 7.025 integrate_v_rspace 110 11.3 0.003 0.003 6.970 6.997 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.848 6.899 ot_diis_step 99 11.5 0.005 0.007 6.525 6.525 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.217 6.406 calculate_rho_elec 110 8.6 0.020 0.024 6.217 6.405 mp_sum_l 6514 12.8 5.454 6.251 5.454 6.251 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.091 6.122 init_scf_loop 11 6.9 0.002 0.012 5.788 5.794 dbcsr_mm_accdrv_process 87628 16.1 2.070 2.200 5.145 5.483 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.096 5.097 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.619 4.621 cp_fm_redistribute_end 48 14.0 4.024 4.594 4.028 4.595 cp_fm_diag_elpa_base 48 14.0 0.560 4.493 0.564 4.526 calculate_dm_sparse 110 9.5 0.000 0.001 4.081 4.170 wfi_extrapolate 11 7.9 0.001 0.003 4.031 4.031 rs_pw_transfer 902 11.9 0.012 0.014 3.820 3.978 make_m2s 4110 13.4 0.059 0.063 3.828 3.969 make_images 4110 14.4 0.177 0.191 3.734 3.877 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.817 3.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.681 3.686 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.442 3.506 density_rs2pw 110 9.6 0.004 0.004 3.269 3.416 grid_integrate_task_list 110 12.3 3.249 3.411 3.249 3.411 prepare_preconditioner 11 7.9 0.000 0.000 3.387 3.409 make_preconditioner 11 8.9 0.006 0.048 3.387 3.409 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.322 3.362 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.160 3.207 calculate_first_density_matrix 1 7.0 0.001 0.005 3.181 3.190 pw_transfer 1331 11.6 0.054 0.064 2.808 2.870 jit_kernel_multiply 13 15.9 2.793 2.859 2.793 2.859 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.720 2.784 potential_pw2rs 110 12.3 0.006 0.007 2.596 2.620 mp_alltoall_d11v 2046 13.8 2.148 2.584 2.148 2.584 mp_sum_d 3883 11.9 2.025 2.566 2.025 2.566 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 2.288 2.355 acc_transpose_blocks 49320 15.4 0.213 0.222 2.237 2.326 fft3d_ps 1111 14.6 0.744 0.828 2.256 2.311 grid_collocate_task_list 110 9.6 2.089 2.230 2.089 2.230 mp_waitany 14300 13.8 1.857 2.122 1.857 2.122 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.936 1.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.876 1.901 make_images_data 4110 15.4 0.044 0.048 1.757 1.888 cp_fm_cholesky_invert 11 10.9 1.828 1.832 1.828 1.832 hybrid_alltoall_any 4261 16.3 0.083 0.485 1.532 1.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.252000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.636364, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.682944E+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 3387049. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.107 0.243 69.327 69.328 qs_mol_dyn_low 1 2.0 0.006 0.014 68.815 68.839 qs_forces 11 3.9 0.019 0.077 68.568 68.577 qs_energies 11 4.9 0.005 0.021 65.102 65.111 scf_env_do_scf 11 5.9 0.001 0.001 56.040 56.043 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.451 48.456 dbcsr_multiply_generic 2055 12.4 0.109 0.113 38.391 38.638 velocity_verlet 10 3.0 0.004 0.011 35.775 35.785 multiply_cannon 2055 13.4 0.222 0.244 31.621 32.955 qs_scf_new_mos 99 7.5 0.001 0.001 32.759 32.898 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.758 32.898 multiply_cannon_loop 2055 14.4 0.929 0.950 30.361 31.518 ot_scf_mini 99 9.5 0.003 0.004 31.111 31.238 ot_mini 99 10.5 0.001 0.002 18.499 18.625 multiply_cannon_multrec 24660 15.4 7.665 9.684 14.212 16.173 rebuild_ks_matrix 110 8.3 0.000 0.001 13.527 13.697 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.023 13.526 13.697 qs_ot_get_derivative 99 11.5 0.001 0.002 12.714 12.842 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.931 12.082 mp_waitall_1 186928 16.3 8.002 10.397 8.002 10.397 multiply_cannon_sync_h2d 24660 15.4 7.036 8.178 7.036 8.178 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.306 7.835 init_scf_loop 11 6.9 0.000 0.002 7.553 7.558 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.734 7.292 apply_single 110 13.6 0.000 0.001 6.733 7.292 dbcsr_mm_accdrv_process 52282 16.1 4.937 6.217 6.386 6.842 init_scf_run 11 5.9 0.000 0.001 6.614 6.615 scf_env_initial_rho_setup 11 6.9 0.002 0.006 6.614 6.615 sum_up_and_integrate 110 10.3 0.054 0.063 6.457 6.468 integrate_v_rspace 110 11.3 0.003 0.003 6.403 6.419 qs_ot_get_p 110 10.4 0.001 0.001 6.013 6.159 ot_diis_step 99 11.5 0.010 0.012 5.737 5.737 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.821 5.735 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.699 5.710 calculate_rho_elec 110 8.6 0.039 0.047 5.698 5.709 prepare_preconditioner 11 7.9 0.000 0.000 5.517 5.537 make_preconditioner 11 8.9 0.000 0.001 5.517 5.537 make_full_inverse_cholesky 11 9.9 0.001 0.005 5.101 5.256 make_m2s 4110 13.4 0.056 0.059 4.106 4.571 make_images 4110 14.4 0.402 0.448 4.004 4.465 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.133 4.151 cp_dbcsr_syevd 48 12.0 0.003 0.006 3.709 3.709 calculate_dm_sparse 110 9.5 0.001 0.001 3.486 3.515 wfi_extrapolate 11 7.9 0.001 0.002 3.479 3.479 pw_transfer 1331 11.6 0.066 0.071 3.234 3.367 grid_integrate_task_list 110 12.3 3.156 3.321 3.156 3.321 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.128 3.262 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.258 3.259 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.140 3.212 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.174 3.177 density_rs2pw 110 9.6 0.004 0.005 2.993 3.163 cp_fm_redistribute_end 48 14.0 2.365 3.146 2.368 3.147 rs_pw_transfer 902 11.9 0.012 0.014 2.907 3.137 cp_fm_diag_elpa_base 48 14.0 0.744 3.025 0.776 3.109 mp_sum_l 6514 12.8 2.145 3.055 2.145 3.055 calculate_first_density_matrix 1 7.0 0.000 0.001 3.036 3.040 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.900 2.955 fft_wrap_pw1pw2_140 451 13.1 0.201 0.222 2.654 2.788 make_images_data 4110 15.4 0.048 0.052 2.321 2.778 hybrid_alltoall_any 4261 16.3 0.103 0.448 2.014 2.715 jit_kernel_multiply 12 16.3 1.102 2.617 1.102 2.617 fft3d_ps 1111 14.6 1.072 1.266 2.484 2.606 cp_fm_cholesky_invert 11 10.9 2.559 2.565 2.559 2.565 potential_pw2rs 110 12.3 0.009 0.009 2.335 2.350 grid_collocate_task_list 110 9.6 2.124 2.320 2.124 2.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.069 2.081 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.935 1.956 mp_alltoall_d11v 2046 13.8 1.709 1.936 1.709 1.936 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.875 1.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.760 1.775 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.787 1.675 mp_allgather_i34 2055 14.4 0.589 1.592 0.589 1.592 mp_irecv_dv 57340 16.2 0.664 1.558 0.664 1.558 acc_transpose_blocks 24660 15.4 0.107 0.111 1.524 1.555 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.388 1.526 mp_waitany 10164 13.8 1.207 1.492 1.207 1.492 mp_sum_d 3883 11.9 1.052 1.437 1.052 1.437 dbcsr_complete_redistribute 325 12.2 0.246 0.307 1.152 1.417 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.328000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.000000, yerr=7.298817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 659.206144E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 61.761 61.761 qs_mol_dyn_low 1 2.0 0.003 0.003 61.437 61.446 qs_forces 11 3.9 0.002 0.002 60.750 60.751 qs_energies 11 4.9 0.001 0.001 57.586 57.588 scf_env_do_scf 11 5.9 0.000 0.001 49.141 49.142 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.392 40.392 velocity_verlet 10 3.0 0.013 0.025 33.298 33.319 dbcsr_multiply_generic 2055 12.4 0.105 0.107 29.713 29.965 qs_scf_new_mos 99 7.5 0.001 0.001 25.815 25.942 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.815 25.941 ot_scf_mini 99 9.5 0.002 0.002 24.586 24.719 multiply_cannon 2055 13.4 0.212 0.223 23.057 24.233 multiply_cannon_loop 2055 14.4 0.619 0.630 21.834 22.893 ot_mini 99 10.5 0.001 0.001 14.189 14.322 rebuild_ks_matrix 110 8.3 0.000 0.001 12.217 12.376 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.217 12.376 multiply_cannon_multrec 16440 15.4 3.805 4.755 10.236 11.129 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.787 10.929 mp_waitall_1 146766 16.3 7.601 10.541 7.601 10.541 qs_ot_get_derivative 99 11.5 0.001 0.001 9.635 9.770 init_scf_loop 11 6.9 0.000 0.000 8.713 8.713 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.651 7.479 prepare_preconditioner 11 7.9 0.000 0.000 6.939 6.956 make_preconditioner 11 8.9 0.000 0.000 6.939 6.956 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.244 6.601 sum_up_and_integrate 110 10.3 0.060 0.062 6.437 6.453 dbcsr_mm_accdrv_process 34862 16.1 5.064 6.049 6.285 6.442 integrate_v_rspace 110 11.3 0.003 0.003 6.376 6.393 init_scf_run 11 5.9 0.000 0.001 6.077 6.077 scf_env_initial_rho_setup 11 6.9 0.004 0.005 6.076 6.077 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.218 5.767 apply_single 110 13.6 0.000 0.000 5.217 5.766 qs_ot_get_p 110 10.4 0.001 0.001 5.555 5.700 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.565 5.576 calculate_rho_elec 110 8.6 0.058 0.058 5.564 5.575 make_m2s 4110 13.4 0.049 0.051 4.315 4.663 make_images 4110 14.4 0.395 0.510 4.203 4.551 ot_diis_step 99 11.5 0.010 0.011 4.522 4.522 multiply_cannon_sync_h2d 16440 15.4 3.628 4.196 3.628 4.196 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.336 4.127 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.917 3.922 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.571 3.571 grid_integrate_task_list 110 12.3 3.188 3.368 3.188 3.368 pw_transfer 1331 11.6 0.065 0.071 3.215 3.225 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.108 3.121 density_rs2pw 110 9.6 0.004 0.004 2.846 3.096 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.045 3.046 calculate_dm_sparse 110 9.5 0.001 0.001 3.002 3.028 cp_fm_redistribute_end 48 14.0 1.893 3.016 1.895 3.017 calculate_first_density_matrix 1 7.0 0.000 0.000 3.012 3.016 cp_fm_diag_elpa_base 48 14.0 1.058 2.877 1.118 2.988 wfi_extrapolate 11 7.9 0.001 0.001 2.979 2.979 make_images_data 4110 15.4 0.044 0.049 2.523 2.972 rs_pw_transfer 902 11.9 0.011 0.012 2.686 2.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.931 2.932 hybrid_alltoall_any 4261 16.3 0.106 0.377 2.244 2.860 cp_fm_cholesky_invert 11 10.9 2.786 2.793 2.786 2.793 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.605 2.659 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.619 2.635 mp_sum_l 6514 12.8 1.724 2.522 1.724 2.522 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.917 2.492 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.388 2.458 fft3d_ps 1111 14.6 1.070 1.086 2.419 2.430 mp_irecv_dv 48980 15.7 0.845 2.368 0.845 2.368 grid_collocate_task_list 110 9.6 2.174 2.353 2.174 2.353 potential_pw2rs 110 12.3 0.010 0.010 2.257 2.267 jit_kernel_multiply 8 16.5 0.835 2.051 0.835 2.051 mp_alltoall_d11v 2046 13.8 1.831 2.015 1.831 2.015 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.956 1.968 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.928 1.929 dbcsr_complete_redistribute 325 12.2 0.318 0.346 1.428 1.907 cp_fm_upper_to_full 70 13.6 1.399 1.872 1.399 1.872 mp_allgather_i34 2055 14.4 0.545 1.712 0.545 1.712 cp_fm_cholesky_decompose 22 10.9 1.644 1.666 1.644 1.666 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.649 acc_transpose_blocks 16440 15.4 0.072 0.074 1.293 1.521 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.343 1.474 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.006 1.462 mp_waitany 17072 13.8 1.131 1.356 1.131 1.356 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.265 1.268 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.761000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.818182, yerr=7.565690 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 732.684288E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.047 66.178 66.178 qs_mol_dyn_low 1 2.0 0.003 0.003 65.933 65.943 qs_forces 11 3.9 0.002 0.002 65.867 65.868 qs_energies 11 4.9 0.002 0.002 62.504 62.507 scf_env_do_scf 11 5.9 0.000 0.001 53.740 53.743 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.961 41.961 velocity_verlet 10 3.0 0.001 0.002 36.783 36.785 dbcsr_multiply_generic 2055 12.4 0.113 0.117 30.236 30.454 qs_scf_new_mos 99 7.5 0.001 0.001 27.421 27.524 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.421 27.523 ot_scf_mini 99 9.5 0.003 0.003 25.775 25.873 multiply_cannon 2055 13.4 0.244 0.266 22.967 23.943 multiply_cannon_loop 2055 14.4 0.882 0.909 21.566 22.212 ot_mini 99 10.5 0.001 0.001 14.569 14.682 multiply_cannon_multrec 24660 15.4 4.214 6.628 12.778 14.225 rebuild_ks_matrix 110 8.3 0.000 0.000 11.964 12.071 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.964 12.070 init_scf_loop 11 6.9 0.000 0.000 11.734 11.735 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.610 10.702 qs_ot_get_derivative 99 11.5 0.001 0.001 10.357 10.455 prepare_preconditioner 11 7.9 0.000 0.000 10.026 10.042 make_preconditioner 11 8.9 0.000 0.000 10.026 10.042 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.257 9.703 dbcsr_mm_accdrv_process 52304 16.0 6.884 8.592 8.414 9.397 mp_waitall_1 126806 16.4 5.053 7.072 5.053 7.072 sum_up_and_integrate 110 10.3 0.069 0.073 6.397 6.411 integrate_v_rspace 110 11.3 0.003 0.004 6.328 6.343 qs_ot_get_p 110 10.4 0.001 0.001 6.162 6.280 init_scf_run 11 5.9 0.000 0.001 6.051 6.052 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.051 6.051 make_m2s 4110 13.4 0.058 0.059 5.560 5.770 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.686 5.698 calculate_rho_elec 110 8.6 0.077 0.081 5.685 5.697 make_images 4110 14.4 0.582 0.696 5.420 5.626 cp_fm_upper_to_full 70 13.8 3.299 4.677 3.299 4.677 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.295 4.311 ot_diis_step 99 11.5 0.011 0.011 4.174 4.174 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.057 4.144 apply_single 110 13.6 0.000 0.000 4.057 4.144 dbcsr_complete_redistribute 325 12.2 0.556 0.650 2.919 4.042 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.663 3.836 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.798 3.799 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.347 3.457 grid_integrate_task_list 110 12.3 3.278 3.421 3.278 3.421 calculate_dm_sparse 110 9.5 0.001 0.001 3.356 3.383 multiply_cannon_sync_h2d 24660 15.4 3.182 3.379 3.182 3.379 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.302 3.355 pw_transfer 1331 11.6 0.064 0.072 3.314 3.339 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.193 3.253 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.208 3.237 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.185 3.188 cp_fm_redistribute_end 48 14.0 1.588 3.148 1.590 3.149 cp_fm_diag_elpa_base 48 14.0 1.475 3.001 1.556 3.117 make_images_data 4110 15.4 0.046 0.050 2.805 3.078 density_rs2pw 110 9.6 0.004 0.004 2.832 3.057 calculate_first_density_matrix 1 7.0 0.000 0.000 2.995 2.998 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.982 2.984 wfi_extrapolate 11 7.9 0.001 0.001 2.967 2.968 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.422 2.959 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.760 2.855 mp_alltoall_i22 605 13.7 1.646 2.809 1.646 2.809 cp_fm_cholesky_invert 11 10.9 2.771 2.780 2.771 2.780 fft_wrap_pw1pw2_140 451 13.1 0.204 0.218 2.680 2.712 rs_pw_transfer 902 11.9 0.010 0.011 2.472 2.706 jit_kernel_multiply 13 15.8 1.200 2.626 1.200 2.626 fft3d_ps 1111 14.6 1.072 1.107 2.508 2.522 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.409 2.456 grid_collocate_task_list 110 9.6 2.221 2.366 2.221 2.366 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.220 2.220 potential_pw2rs 110 12.3 0.013 0.013 2.162 2.174 mp_alltoall_d11v 2046 13.8 1.881 2.163 1.881 2.163 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.949 1.967 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.746 1.777 cp_fm_cholesky_decompose 22 10.9 1.720 1.767 1.720 1.767 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.706 mp_sum_l 6514 12.8 1.046 1.646 1.046 1.646 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.602 1.614 acc_transpose_blocks 24660 15.4 0.104 0.106 1.487 1.508 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.853 1.472 mp_irecv_dv 62702 16.1 0.755 1.394 0.755 1.394 mp_waitany 13376 13.8 1.047 1.368 1.047 1.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.178000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.000000, yerr=10.753435 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 828.600320E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.079 56.353 56.353 qs_mol_dyn_low 1 2.0 0.003 0.003 55.967 55.976 qs_forces 11 3.9 0.013 0.024 55.878 55.879 qs_energies 11 4.9 0.001 0.002 52.282 52.299 scf_env_do_scf 11 5.9 0.001 0.001 42.501 42.501 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.769 34.771 velocity_verlet 10 3.0 0.031 0.032 30.549 30.552 dbcsr_multiply_generic 2055 12.4 0.103 0.105 23.899 24.040 qs_scf_new_mos 99 7.5 0.001 0.001 20.711 20.763 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.710 20.763 multiply_cannon 2055 13.4 0.248 0.260 18.451 19.637 ot_scf_mini 99 9.5 0.002 0.002 19.458 19.474 multiply_cannon_loop 2055 14.4 0.323 0.337 17.095 17.292 rebuild_ks_matrix 110 8.3 0.000 0.000 11.373 11.395 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.373 11.395 ot_mini 99 10.5 0.001 0.001 10.568 10.579 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.135 10.156 multiply_cannon_multrec 8220 15.4 3.264 4.536 8.451 9.487 mp_waitall_1 106626 16.5 6.214 8.039 6.214 8.039 init_scf_loop 11 6.9 0.000 0.000 7.683 7.685 qs_ot_get_derivative 99 11.5 0.001 0.001 6.777 6.790 init_scf_run 11 5.9 0.000 0.001 6.681 6.681 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.680 6.681 sum_up_and_integrate 110 10.3 0.080 0.081 6.118 6.131 prepare_preconditioner 11 7.9 0.000 0.000 6.059 6.064 make_preconditioner 11 8.9 0.000 0.000 6.059 6.064 integrate_v_rspace 110 11.3 0.003 0.003 6.037 6.049 dbcsr_mm_accdrv_process 17442 15.9 2.791 3.802 5.056 6.004 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.638 5.711 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.623 5.660 calculate_rho_elec 110 8.6 0.115 0.115 5.622 5.659 qs_ot_get_p 110 10.4 0.001 0.001 5.000 5.030 make_m2s 4110 13.4 0.038 0.040 4.219 4.510 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.038 4.487 make_images 4110 14.4 0.639 0.691 4.092 4.382 calculate_first_density_matrix 1 7.0 0.000 0.000 3.927 3.932 ot_diis_step 99 11.5 0.012 0.012 3.770 3.770 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.694 3.725 apply_single 110 13.6 0.000 0.000 3.694 3.725 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.584 3.589 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.475 3.477 grid_integrate_task_list 110 12.3 3.370 3.476 3.370 3.476 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.282 3.283 pw_transfer 1331 11.6 0.065 0.070 3.220 3.235 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.129 3.142 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.113 3.132 multiply_cannon_sync_h2d 8220 15.4 2.922 3.057 2.922 3.057 cp_fm_cholesky_invert 11 10.9 2.942 2.946 2.942 2.946 make_images_data 4110 15.4 0.038 0.043 2.409 2.824 density_rs2pw 110 9.6 0.004 0.004 2.674 2.807 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.766 2.766 hybrid_alltoall_any 4261 16.3 0.199 0.864 2.320 2.764 cp_fm_redistribute_end 48 14.0 0.701 2.739 0.705 2.740 jit_kernel_multiply 10 15.8 1.957 2.740 1.957 2.740 cp_fm_diag_elpa_base 48 14.0 1.847 2.545 2.027 2.707 fft_wrap_pw1pw2_140 451 13.1 0.211 0.218 2.660 2.686 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.653 2.665 wfi_extrapolate 11 7.9 0.001 0.001 2.664 2.664 calculate_dm_sparse 110 9.5 0.001 0.001 2.605 2.654 grid_collocate_task_list 110 9.6 2.325 2.507 2.325 2.507 fft3d_ps 1111 14.6 1.118 1.148 2.372 2.384 rs_pw_transfer 902 11.9 0.010 0.010 2.162 2.310 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.983 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.913 1.924 potential_pw2rs 110 12.3 0.015 0.016 1.869 1.874 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.803 1.810 mp_alltoall_d11v 2046 13.8 1.615 1.751 1.615 1.751 cp_fm_cholesky_decompose 22 10.9 1.693 1.714 1.693 1.714 mp_allgather_i34 2055 14.4 0.538 1.649 0.538 1.649 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.496 1.614 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.570 1.573 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.569 dbcsr_complete_redistribute 325 12.2 0.561 0.581 1.458 1.559 qs_create_task_list 11 7.9 0.001 0.001 1.224 1.322 generate_qs_task_list 11 8.9 0.379 0.447 1.224 1.321 make_basis_sm 11 9.8 0.000 0.000 1.259 1.261 mp_waitany 9240 13.8 1.038 1.215 1.038 1.215 parallel_gemm_fm 81 9.0 0.000 0.000 1.174 1.181 parallel_gemm_fm_cosma 81 10.0 1.173 1.180 1.173 1.180 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.145 1.173 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.486 1.137 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.804 1.132 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.353000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=770.909091, yerr=12.816570 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 1.331036E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.054 86.618 86.619 qs_mol_dyn_low 1 2.0 0.003 0.003 86.281 86.289 qs_forces 11 3.9 0.002 0.002 85.927 85.928 qs_energies 11 4.9 0.004 0.004 81.889 81.891 scf_env_do_scf 11 5.9 0.001 0.001 70.573 70.573 velocity_verlet 10 3.0 0.005 0.006 54.666 54.708 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.379 42.381 dbcsr_multiply_generic 2055 12.4 0.118 0.120 29.610 29.700 init_scf_loop 11 6.9 0.000 0.000 28.122 28.125 prepare_preconditioner 11 7.9 0.000 0.000 26.191 26.201 make_preconditioner 11 8.9 0.000 0.000 26.191 26.201 qs_scf_new_mos 99 7.5 0.001 0.001 26.051 26.096 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.051 26.095 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.391 25.664 ot_scf_mini 99 9.5 0.002 0.002 24.283 24.311 multiply_cannon 2055 13.4 0.346 0.380 22.527 23.316 multiply_cannon_loop 2055 14.4 0.341 0.343 20.745 21.217 cp_fm_upper_to_full 70 14.2 12.635 18.149 12.635 18.149 ot_mini 99 10.5 0.001 0.001 13.483 13.510 rebuild_ks_matrix 110 8.3 0.001 0.001 13.142 13.177 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.141 13.176 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.949 11.981 multiply_cannon_multrec 8220 15.4 4.369 4.548 10.357 10.818 dbcsr_complete_redistribute 325 12.2 1.030 1.053 7.384 10.604 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.354 9.570 mp_waitall_1 87304 16.6 8.150 9.042 8.150 9.042 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.785 8.979 qs_ot_get_derivative 99 11.5 0.001 0.001 8.909 8.941 mp_alltoall_i22 605 13.7 5.454 8.609 5.454 8.609 init_scf_run 11 5.9 0.000 0.001 7.074 7.074 scf_env_initial_rho_setup 11 6.9 0.002 0.002 7.074 7.074 sum_up_and_integrate 110 10.3 0.151 0.152 6.600 6.614 integrate_v_rspace 110 11.3 0.004 0.004 6.449 6.463 dbcsr_mm_accdrv_process 11614 15.7 3.220 3.646 5.846 6.437 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.393 6.432 calculate_rho_elec 110 8.6 0.227 0.227 6.393 6.432 make_m2s 4110 13.4 0.043 0.043 5.388 5.892 make_images 4110 14.4 0.881 0.920 5.202 5.705 qs_ot_get_p 110 10.4 0.001 0.001 5.596 5.613 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.987 5.595 cp_fm_cholesky_invert 11 10.9 5.235 5.240 5.235 5.240 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.680 5.132 apply_single 110 13.6 0.000 0.000 4.680 5.132 ot_diis_step 99 11.5 0.015 0.017 4.548 4.548 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.068 4.071 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.063 4.070 multiply_cannon_sync_h2d 8220 15.4 3.955 3.961 3.955 3.961 calculate_first_density_matrix 1 7.0 0.000 0.000 3.774 3.774 hybrid_alltoall_any 4261 16.3 0.257 0.557 2.999 3.763 pw_transfer 1331 11.6 0.074 0.074 3.718 3.730 grid_integrate_task_list 110 12.3 3.657 3.704 3.657 3.704 make_images_data 4110 15.4 0.041 0.045 3.013 3.655 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.645 3.646 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.623 3.623 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.602 3.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.514 3.581 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.879 3.328 wfi_extrapolate 11 7.9 0.001 0.001 3.196 3.196 calculate_dm_sparse 110 9.5 0.001 0.001 3.127 3.149 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.120 3.127 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.012 3.013 cp_fm_diag_elpa_base 48 14.0 2.484 2.675 3.011 3.011 density_rs2pw 110 9.6 0.004 0.004 2.901 2.926 fft3d_ps 1111 14.6 1.262 1.280 2.812 2.827 grid_collocate_task_list 110 9.6 2.627 2.652 2.627 2.652 jit_kernel_multiply 10 15.4 2.422 2.627 2.422 2.627 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.361 2.373 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.174 2.235 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.095 2.204 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.815 2.183 rs_pw_transfer 902 11.9 0.011 0.011 2.082 2.148 mp_alltoall_d11v 2046 13.8 2.026 2.086 2.026 2.086 potential_pw2rs 110 12.3 0.022 0.022 2.042 2.045 cp_fm_cholesky_decompose 22 10.9 1.988 2.005 1.988 2.005 qs_create_task_list 11 7.9 0.001 0.001 1.903 1.947 generate_qs_task_list 11 8.9 0.740 0.795 1.902 1.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.796 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.619000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1196.727273, yerr=53.445114 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.806208E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 3626986. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 202.073 202.075 qs_mol_dyn_low 1 2.0 0.003 0.003 201.728 201.741 qs_forces 11 3.9 0.003 0.004 201.645 201.647 qs_energies 11 4.9 0.001 0.002 196.165 196.175 scf_env_do_scf 11 5.9 0.001 0.001 179.742 179.746 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 159.448 159.450 dbcsr_multiply_generic 2507 12.6 0.172 0.175 124.084 124.677 qs_scf_new_mos 117 7.6 0.001 0.001 121.144 121.381 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.143 121.380 velocity_verlet 10 3.0 0.001 0.001 120.950 120.951 ot_scf_mini 117 9.6 0.003 0.003 114.622 114.856 multiply_cannon 2507 13.6 0.238 0.246 100.854 102.417 multiply_cannon_loop 2507 14.6 2.119 2.198 98.699 100.385 ot_mini 117 10.6 0.001 0.001 65.629 65.866 multiply_cannon_multrec 60168 15.6 33.086 35.670 41.567 43.978 qs_ot_get_derivative 117 11.6 0.001 0.001 40.703 40.898 rebuild_ks_matrix 128 8.3 0.001 0.001 33.072 33.325 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.071 33.324 mp_waitall_1 291448 16.2 28.649 32.389 28.649 32.389 multiply_cannon_sync_h2d 60168 15.6 27.227 30.100 27.227 30.100 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.752 29.980 qs_ot_get_p 128 10.4 0.001 0.001 27.081 27.376 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.315 24.982 apply_single 128 13.6 0.001 0.001 24.315 24.981 ot_diis_step 117 11.6 0.007 0.008 24.582 24.583 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.420 20.508 init_scf_loop 11 6.9 0.000 0.001 20.219 20.220 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.124 19.276 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 15.667 18.192 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.814 17.815 prepare_preconditioner 11 7.9 0.000 0.000 15.682 15.711 make_preconditioner 11 8.9 0.000 0.000 15.682 15.711 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.916 15.101 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.972 14.980 cp_fm_redistribute_end 83 14.4 11.824 14.906 11.840 14.910 cp_fm_diag_elpa_base 83 14.4 3.026 14.627 3.058 14.746 make_m2s 5014 13.6 0.102 0.110 13.803 14.302 make_images 5014 14.6 0.406 0.427 13.631 14.143 sum_up_and_integrate 128 10.3 0.091 0.111 13.796 13.810 integrate_v_rspace 128 11.3 0.004 0.004 13.705 13.721 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.355 12.486 calculate_rho_elec 128 8.7 0.045 0.063 12.355 12.485 init_scf_run 11 5.9 0.000 0.001 12.330 12.330 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.330 12.330 mp_sum_l 7870 13.0 8.161 9.478 8.161 9.478 multiply_cannon_metrocomm1 60168 15.6 0.087 0.092 6.174 9.018 wfi_extrapolate 11 7.9 0.001 0.001 9.000 9.000 cp_fm_cholesky_invert 11 10.9 8.896 8.903 8.896 8.903 dbcsr_mm_accdrv_process 124484 16.2 3.313 3.457 8.041 8.574 calculate_dm_sparse 128 9.5 0.001 0.001 8.412 8.573 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.771 7.906 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.770 7.890 make_images_data 5014 15.6 0.073 0.079 6.820 7.813 grid_integrate_task_list 128 12.3 7.088 7.474 7.088 7.474 hybrid_alltoall_any 5200 16.5 0.293 2.260 5.981 7.375 density_rs2pw 128 9.7 0.006 0.007 6.204 6.897 pw_transfer 1547 11.6 0.075 0.105 6.553 6.816 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.688 6.696 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 6.349 6.587 rs_pw_transfer 1046 11.9 0.018 0.019 5.497 6.142 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.802 5.965 fft_wrap_pw1pw2_140 523 13.2 0.447 0.514 5.498 5.653 fft3d_ps 1291 14.7 2.087 2.567 5.136 5.314 mp_alltoall_d11v 2415 14.1 4.055 5.153 4.055 5.153 grid_collocate_task_list 128 9.7 4.733 5.047 4.733 5.047 cp_fm_cholesky_decompose 22 10.9 4.618 4.634 4.618 4.634 potential_pw2rs 128 12.3 0.010 0.011 4.332 4.351 mp_sum_d 4473 12.2 3.643 4.296 3.643 4.296 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.075000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.727273, yerr=6.282601 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 830.238720E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473827979536 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56635. MP_Allreduce 11357 942. MP_Sync 172 MP_Alltoall 1983 5652938. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 84 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 189.837 189.838 qs_mol_dyn_low 1 2.0 0.003 0.003 189.468 189.481 qs_forces 11 3.9 0.003 0.003 189.371 189.374 qs_energies 11 4.9 0.002 0.006 182.616 182.629 scf_env_do_scf 11 5.9 0.001 0.001 166.312 166.322 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.751 133.753 velocity_verlet 10 3.0 0.001 0.002 119.540 119.541 dbcsr_multiply_generic 2529 12.6 0.184 0.191 98.042 99.138 qs_scf_new_mos 118 7.6 0.001 0.001 94.992 95.519 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.992 95.518 ot_scf_mini 118 9.6 0.004 0.004 90.157 90.760 multiply_cannon 2529 13.6 0.484 0.534 77.971 82.420 multiply_cannon_loop 2529 14.6 1.265 1.303 74.658 77.248 ot_mini 118 10.6 0.001 0.001 50.090 50.686 mp_waitall_1 228724 16.4 24.969 39.525 24.969 39.525 multiply_cannon_multrec 30348 15.6 22.038 25.990 31.793 35.896 rebuild_ks_matrix 129 8.3 0.001 0.001 32.837 33.302 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 32.836 33.301 init_scf_loop 11 6.9 0.000 0.000 32.472 32.473 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.613 30.050 multiply_cannon_metrocomm3 30348 15.6 0.097 0.106 15.949 28.810 qs_ot_get_derivative 118 11.6 0.001 0.002 28.006 28.589 prepare_preconditioner 11 7.9 0.000 0.000 28.127 28.200 make_preconditioner 11 8.9 0.000 0.000 28.127 28.200 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.837 27.367 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.160 23.254 apply_single 129 13.6 0.001 0.001 22.160 23.253 qs_ot_get_p 129 10.4 0.001 0.001 21.714 22.425 ot_diis_step 118 11.6 0.014 0.015 21.908 21.910 multiply_cannon_sync_h2d 30348 15.6 19.332 21.280 19.332 21.280 qs_ot_p2m_diag 84 11.4 0.189 0.218 16.893 16.925 cp_fm_cholesky_invert 11 10.9 16.423 16.435 16.423 16.435 make_m2s 5058 13.6 0.091 0.096 14.242 15.903 cp_dbcsr_syevd 84 12.4 0.005 0.006 15.754 15.755 make_images 5058 14.6 1.179 1.373 14.042 15.703 sum_up_and_integrate 129 10.3 0.118 0.133 14.297 14.327 integrate_v_rspace 129 11.3 0.003 0.004 14.179 14.211 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.961 12.997 calculate_rho_elec 129 8.7 0.088 0.106 12.960 12.996 cp_fm_diag_elpa 84 13.4 0.000 0.001 12.587 12.596 cp_fm_redistribute_end 84 14.4 7.351 12.527 7.364 12.530 cp_fm_diag_elpa_base 84 14.4 4.928 12.071 5.148 12.428 init_scf_run 11 5.9 0.000 0.001 11.538 11.539 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.538 11.539 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 11.062 11.481 multiply_cannon_metrocomm4 27819 15.6 0.097 0.112 3.844 10.982 make_images_data 5058 15.6 0.069 0.076 8.663 10.633 mp_irecv_dv 70084 16.3 3.647 10.587 3.647 10.587 hybrid_alltoall_any 5245 16.5 0.346 1.523 7.442 10.327 dbcsr_mm_accdrv_process 62758 16.2 4.699 5.707 9.211 9.920 wfi_extrapolate 11 7.9 0.001 0.001 8.299 8.299 pw_transfer 1559 11.6 0.085 0.109 7.735 7.791 grid_integrate_task_list 129 12.3 7.234 7.573 7.234 7.573 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 7.509 7.568 density_rs2pw 129 9.7 0.006 0.007 6.615 7.109 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.227 6.981 cp_fm_cholesky_decompose 22 10.9 6.820 6.892 6.820 6.892 calculate_dm_sparse 129 9.5 0.001 0.001 6.673 6.816 fft_wrap_pw1pw2_140 527 13.2 0.474 0.515 6.600 6.667 mp_sum_l 7936 13.1 4.030 6.231 4.030 6.231 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.122 6.132 fft3d_ps 1301 14.7 2.799 2.946 5.864 5.898 rs_pw_transfer 1054 12.0 0.015 0.017 5.409 5.869 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.337 5.400 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.245 5.393 grid_collocate_task_list 129 9.7 4.950 5.319 4.950 5.319 mp_allgather_i34 2529 14.6 1.872 5.154 1.872 5.154 mp_alltoall_d11v 2429 14.1 4.343 4.772 4.343 4.772 potential_pw2rs 129 12.3 0.015 0.016 4.706 4.719 mp_sum_d 4503 12.2 2.706 4.043 2.706 4.043 dbcsr_complete_redistribute 397 12.7 0.776 0.870 3.123 3.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.838000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.818182, yerr=2.516064 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/17/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 2.951022E+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 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 937.385984E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4129 56602. MP_Allreduce 11361 981. MP_Sync 172 MP_Alltoall 1724 9394234. MP_ISendRecv 15996 75008. MP_Wait 29984 MP_comm_split 84 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.111 0.120 175.561 175.562 qs_mol_dyn_low 1 2.0 0.003 0.003 175.006 175.020 qs_forces 11 3.9 0.003 0.003 174.907 174.911 qs_energies 11 4.9 0.001 0.002 168.465 168.475 scf_env_do_scf 11 5.9 0.001 0.001 152.787 152.789 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.771 117.772 velocity_verlet 10 3.0 0.002 0.002 112.313 112.315 dbcsr_multiply_generic 2529 12.6 0.180 0.184 81.951 82.981 qs_scf_new_mos 118 7.6 0.001 0.001 81.596 81.955 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.595 81.954 ot_scf_mini 118 9.6 0.003 0.004 77.448 77.917 multiply_cannon 2529 13.6 0.508 0.530 62.148 65.549 multiply_cannon_loop 2529 14.6 0.868 0.899 59.098 61.704 ot_mini 118 10.6 0.001 0.001 42.579 43.039 init_scf_loop 11 6.9 0.000 0.000 34.919 34.920 mp_waitall_1 180004 16.5 24.832 33.841 24.832 33.841 prepare_preconditioner 11 7.9 0.000 0.000 30.755 30.810 make_preconditioner 11 8.9 0.000 0.000 30.755 30.810 rebuild_ks_matrix 129 8.3 0.001 0.001 29.772 30.243 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 29.771 30.243 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.413 29.811 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.905 27.336 multiply_cannon_multrec 20232 15.6 13.828 17.158 22.957 26.196 multiply_cannon_metrocomm3 20232 15.6 0.059 0.063 14.823 24.161 qs_ot_get_derivative 118 11.6 0.001 0.002 22.802 23.266 qs_ot_get_p 129 10.4 0.001 0.001 20.503 21.045 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.902 20.908 apply_single 129 13.6 0.001 0.001 19.902 20.907 ot_diis_step 118 11.6 0.018 0.026 19.676 19.676 multiply_cannon_sync_h2d 20232 15.6 14.358 16.192 14.358 16.192 qs_ot_p2m_diag 84 11.4 0.268 0.276 16.024 16.031 make_m2s 5058 13.6 0.082 0.087 14.709 15.614 make_images 5058 14.6 1.205 1.312 14.480 15.376 cp_dbcsr_syevd 84 12.4 0.005 0.006 15.021 15.022 cp_fm_cholesky_invert 11 10.9 14.319 14.327 14.319 14.327 sum_up_and_integrate 129 10.3 0.136 0.148 13.953 13.976 integrate_v_rspace 129 11.3 0.004 0.004 13.817 13.844 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.845 12.893 calculate_rho_elec 129 8.7 0.133 0.148 12.844 12.892 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.860 11.861 cp_fm_redistribute_end 84 14.4 4.489 11.801 4.502 11.804 cp_fm_diag_elpa_base 84 14.4 6.870 11.166 7.281 11.714 init_scf_run 11 5.9 0.000 0.001 10.715 10.715 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.715 10.715 make_images_data 5058 15.6 0.062 0.070 9.068 10.348 hybrid_alltoall_any 5245 16.5 0.434 1.985 7.827 9.629 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 8.955 9.294 multiply_cannon_metrocomm4 17703 15.6 0.064 0.073 3.388 9.116 mp_irecv_dv 50659 16.2 3.264 8.869 3.264 8.869 dbcsr_mm_accdrv_process 41846 16.2 4.889 5.631 8.574 8.771 grid_integrate_task_list 129 12.3 7.321 7.821 7.321 7.821 pw_transfer 1559 11.6 0.085 0.104 7.394 7.497 cp_fm_upper_to_full 106 14.5 5.752 7.464 5.752 7.464 cp_fm_cholesky_decompose 22 10.9 7.301 7.342 7.301 7.342 wfi_extrapolate 11 7.9 0.001 0.001 7.292 7.292 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 7.169 7.280 density_rs2pw 129 9.7 0.006 0.006 6.337 6.788 fft_wrap_pw1pw2_140 527 13.2 0.483 0.546 6.321 6.439 dbcsr_complete_redistribute 397 12.7 1.169 1.198 4.545 6.331 calculate_dm_sparse 129 9.5 0.001 0.001 6.162 6.267 grid_collocate_task_list 129 9.7 5.106 5.697 5.106 5.697 rs_pw_transfer 1054 12.0 0.014 0.015 5.034 5.500 fft3d_ps 1301 14.7 2.709 2.920 5.422 5.484 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.433 5.440 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.582 5.317 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.391 5.181 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.550 4.694 mp_alltoall_d11v 2429 14.1 4.100 4.657 4.100 4.657 mp_sum_l 7936 13.1 3.254 4.602 3.254 4.602 mp_allgather_i34 2529 14.6 1.557 4.399 1.557 4.399 potential_pw2rs 129 12.3 0.021 0.023 4.346 4.362 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.323 4.092 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.012 4.038 mp_alltoall_i22 720 14.1 1.968 3.937 1.968 3.937 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.701 3.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.562000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.000000, yerr=10.162319 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 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 1963542011904 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.144394E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4073 57357. MP_Allreduce 11221 1069. MP_Sync 168 MP_Alltoall 1700 12496420. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.095 186.887 186.888 qs_mol_dyn_low 1 2.0 0.003 0.003 186.024 186.037 qs_forces 11 3.9 0.012 0.029 185.739 185.747 qs_energies 11 4.9 0.001 0.002 178.765 178.783 scf_env_do_scf 11 5.9 0.001 0.001 161.463 161.475 velocity_verlet 10 3.0 0.002 0.003 122.741 122.751 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.642 114.643 qs_scf_new_mos 116 7.6 0.001 0.001 79.855 80.180 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.855 80.179 dbcsr_multiply_generic 2485 12.5 0.186 0.191 79.119 79.994 ot_scf_mini 116 9.6 0.003 0.003 75.426 75.720 multiply_cannon 2485 13.5 0.551 0.585 54.867 57.862 multiply_cannon_loop 2485 14.5 1.167 1.194 51.232 53.294 init_scf_loop 11 6.9 0.000 0.000 46.699 46.700 prepare_preconditioner 11 7.9 0.000 0.000 42.672 42.693 make_preconditioner 11 8.9 0.000 0.000 42.672 42.693 ot_mini 116 10.6 0.001 0.001 41.718 42.004 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.169 41.268 multiply_cannon_multrec 29820 15.5 14.118 19.461 26.710 31.513 rebuild_ks_matrix 127 8.3 0.001 0.001 28.472 28.772 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.472 28.772 mp_waitall_1 152434 16.5 16.849 27.089 16.849 27.089 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.739 26.008 qs_ot_get_derivative 116 11.6 0.001 0.002 22.330 22.615 make_m2s 4970 13.5 0.097 0.100 19.951 21.022 make_images 4970 14.5 1.951 2.278 19.649 20.717 qs_ot_get_p 127 10.4 0.001 0.001 20.069 20.366 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.733 19.382 apply_single 127 13.6 0.001 0.001 18.733 19.382 ot_diis_step 116 11.6 0.017 0.018 19.263 19.265 cp_fm_upper_to_full 104 14.7 11.280 16.707 11.280 16.707 cp_fm_cholesky_invert 11 10.9 16.193 16.202 16.193 16.202 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.019 16.072 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.149 14.997 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.763 14.764 sum_up_and_integrate 127 10.3 0.141 0.154 13.762 13.791 integrate_v_rspace 127 11.3 0.004 0.004 13.621 13.655 dbcsr_complete_redistribute 393 12.7 1.510 1.635 9.079 12.943 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.835 12.872 calculate_rho_elec 127 8.7 0.174 0.189 12.834 12.871 dbcsr_mm_accdrv_process 61748 16.2 7.742 9.012 12.173 12.665 multiply_cannon_sync_h2d 29820 15.5 11.643 12.493 11.643 12.493 make_images_data 4970 15.5 0.064 0.069 10.616 12.206 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.719 11.721 cp_fm_redistribute_end 82 14.4 2.012 11.652 2.027 11.658 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.725 11.576 cp_fm_diag_elpa_base 82 14.4 9.008 11.052 9.598 11.531 init_scf_run 11 5.9 0.000 0.001 11.445 11.446 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.444 11.446 hybrid_alltoall_any 5155 16.4 0.525 2.218 9.437 11.220 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.480 10.194 mp_alltoall_i22 712 14.1 5.650 9.550 5.650 9.550 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.201 9.407 grid_integrate_task_list 127 12.3 7.458 7.823 7.458 7.823 cp_fm_cholesky_decompose 22 10.9 7.624 7.710 7.624 7.710 pw_transfer 1535 11.6 0.084 0.098 7.512 7.594 wfi_extrapolate 11 7.9 0.001 0.001 7.514 7.514 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.288 7.374 calculate_dm_sparse 127 9.5 0.001 0.001 6.864 6.959 multiply_cannon_metrocomm4 24850 15.5 0.075 0.085 2.748 6.786 fft_wrap_pw1pw2_140 519 13.2 0.478 0.490 6.442 6.526 mp_irecv_dv 75445 16.2 2.603 6.508 2.603 6.508 density_rs2pw 127 9.7 0.005 0.006 6.034 6.420 fft3d_ps 1281 14.7 2.771 2.839 5.534 5.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.521 5.558 grid_collocate_task_list 127 9.7 5.170 5.408 5.170 5.408 mp_alltoall_d11v 2401 14.1 4.703 5.306 4.703 5.306 rs_pw_transfer 1038 11.9 0.013 0.015 4.506 4.898 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.464 4.529 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.404 4.415 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.317 4.400 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.143 4.206 potential_pw2rs 127 12.3 0.023 0.023 4.129 4.148 calculate_first_density_matrix 1 7.0 0.000 0.000 3.748 3.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.888000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.272727, yerr=20.833063 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.506378E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58095. MP_Allreduce 11022 1174. MP_Sync 87 MP_Alltoall 1712 18838222. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.039 172.637 172.638 qs_mol_dyn_low 1 2.0 0.003 0.003 171.743 171.755 qs_forces 11 3.9 0.027 0.060 171.632 171.637 qs_energies 11 4.9 0.001 0.002 164.331 164.361 scf_env_do_scf 11 5.9 0.001 0.001 146.523 146.535 velocity_verlet 10 3.0 0.020 0.023 111.688 111.692 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 110.578 110.579 qs_scf_new_mos 117 7.6 0.001 0.001 75.213 75.295 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.212 75.294 dbcsr_multiply_generic 2507 12.6 0.180 0.186 74.437 74.759 ot_scf_mini 117 9.6 0.003 0.004 70.794 70.825 multiply_cannon 2507 13.6 0.587 0.621 55.183 59.258 multiply_cannon_loop 2507 14.6 0.449 0.461 50.363 50.938 ot_mini 117 10.6 0.001 0.001 39.453 39.487 init_scf_loop 11 6.9 0.000 0.000 35.788 35.790 mp_waitall_1 129618 16.6 25.315 32.679 25.315 32.679 prepare_preconditioner 11 7.9 0.000 0.000 31.903 31.939 make_preconditioner 11 8.9 0.000 0.000 31.903 31.939 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.776 30.028 rebuild_ks_matrix 128 8.3 0.001 0.001 28.320 28.351 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.319 28.350 qs_ks_update_qs_env 128 7.6 0.001 0.001 25.878 25.907 multiply_cannon_multrec 10028 15.6 10.480 14.410 18.429 22.247 qs_ot_get_derivative 117 11.6 0.002 0.002 19.688 19.715 ot_diis_step 117 11.6 0.019 0.020 19.695 19.695 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.313 19.512 apply_single 128 13.6 0.001 0.001 19.313 19.512 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.255 18.783 qs_ot_get_p 128 10.4 0.001 0.001 18.620 18.659 make_m2s 5014 13.6 0.067 0.072 15.718 18.314 cp_fm_cholesky_invert 11 10.9 18.258 18.264 18.258 18.264 make_images 5014 14.6 2.308 2.887 15.416 18.014 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.850 14.865 sum_up_and_integrate 128 10.3 0.181 0.191 14.012 14.059 integrate_v_rspace 128 11.3 0.004 0.004 13.831 13.889 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.708 13.710 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.056 13.088 calculate_rho_elec 128 8.7 0.258 0.269 13.055 13.087 multiply_cannon_sync_h2d 10028 15.6 11.632 12.521 11.632 12.521 make_images_data 5014 15.6 0.054 0.060 9.469 12.030 hybrid_alltoall_any 5200 16.5 0.806 3.534 9.292 11.693 init_scf_run 11 5.9 0.000 0.001 11.018 11.018 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.018 11.018 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.633 10.638 cp_fm_diag_elpa_base 83 14.4 10.386 10.463 10.629 10.633 dbcsr_mm_accdrv_process 20762 16.1 2.810 3.838 7.575 8.238 grid_integrate_task_list 128 12.3 7.780 8.113 7.780 8.113 cp_fm_cholesky_decompose 22 10.9 7.998 8.086 7.998 8.086 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.888 7.907 pw_transfer 1547 11.6 0.083 0.091 7.687 7.698 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.467 7.484 wfi_extrapolate 11 7.9 0.001 0.001 7.200 7.200 mp_allgather_i34 2507 14.6 3.018 6.977 3.018 6.977 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.056 6.694 calculate_dm_sparse 128 9.5 0.001 0.001 6.547 6.609 fft_wrap_pw1pw2_140 523 13.2 0.502 0.522 6.553 6.575 density_rs2pw 128 9.7 0.005 0.006 5.922 6.161 grid_collocate_task_list 128 9.7 5.487 5.695 5.487 5.695 fft3d_ps 1291 14.7 2.724 2.803 5.624 5.654 dbcsr_complete_redistribute 395 12.7 2.126 2.170 5.224 5.556 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.448 5.454 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.251 5.275 mp_alltoall_d11v 2415 14.1 4.628 5.166 4.628 5.166 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.856 4.884 mp_irecv_dv 28860 15.9 1.819 4.800 1.819 4.800 rs_pw_transfer 1046 11.9 0.013 0.013 4.174 4.459 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.390 4.421 potential_pw2rs 128 12.3 0.027 0.028 4.110 4.129 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.562 3.882 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.723 3.750 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.391 3.706 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.575 3.634 calculate_first_density_matrix 1 7.0 0.000 0.000 3.625 3.626 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.480 3.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.638000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1401.090909, yerr=42.193180 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.064529E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 59070. MP_Allreduce 11015 1514. MP_Sync 86 MP_Alltoall 1700 36954382. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.064 291.968 291.969 qs_mol_dyn_low 1 2.0 0.003 0.003 291.411 291.422 qs_forces 11 3.9 0.003 0.003 291.287 291.291 qs_energies 11 4.9 0.002 0.002 282.720 282.725 scf_env_do_scf 11 5.9 0.001 0.001 260.949 260.966 velocity_verlet 10 3.0 0.043 0.045 209.542 209.550 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 135.524 135.525 init_scf_loop 11 6.9 0.000 0.000 125.163 125.165 prepare_preconditioner 11 7.9 0.000 0.000 120.467 120.504 make_preconditioner 11 8.9 0.000 0.000 120.467 120.504 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.356 117.603 qs_scf_new_mos 116 7.6 0.001 0.001 93.598 93.768 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.597 93.767 ot_scf_mini 116 9.6 0.004 0.004 88.728 88.845 dbcsr_multiply_generic 2485 12.5 0.210 0.218 85.222 86.063 cp_fm_upper_to_full 104 14.8 52.917 75.968 52.917 75.968 multiply_cannon 2485 13.5 0.699 0.748 59.820 60.749 multiply_cannon_loop 2485 14.5 0.467 0.476 56.076 58.052 ot_mini 116 10.6 0.001 0.001 46.147 46.286 dbcsr_complete_redistribute 393 12.7 4.003 4.032 29.776 42.914 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.365 39.448 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.065 36.987 mp_alltoall_i22 712 14.1 21.874 34.794 21.874 34.794 cp_fm_cholesky_invert 11 10.9 34.010 34.017 34.010 34.017 mp_waitall_1 104546 16.7 29.599 33.975 29.599 33.975 rebuild_ks_matrix 127 8.3 0.001 0.001 33.563 33.716 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.562 33.715 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.303 31.439 qs_ot_get_p 127 10.4 0.001 0.001 27.043 27.171 qs_ot_get_derivative 116 11.6 0.001 0.002 25.417 25.532 qs_ot_p2m_diag 82 11.4 0.868 0.873 22.789 22.823 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 19.817 21.827 cp_dbcsr_syevd 82 12.4 0.006 0.006 21.068 21.070 make_m2s 4970 13.5 0.076 0.081 20.058 21.046 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.282 20.752 apply_single 127 13.6 0.001 0.001 20.282 20.752 ot_diis_step 116 11.6 0.022 0.022 20.683 20.683 make_images 4970 14.5 3.742 3.926 19.589 20.580 multiply_cannon_multrec 9940 15.5 10.471 12.292 18.130 18.332 cp_fm_diag_elpa 82 13.4 0.000 0.001 17.849 17.851 cp_fm_diag_elpa_base 82 14.4 13.536 15.120 17.844 17.847 sum_up_and_integrate 127 10.3 0.319 0.321 15.669 15.758 multiply_cannon_sync_h2d 9940 15.5 15.544 15.569 15.544 15.569 integrate_v_rspace 127 11.3 0.004 0.004 15.350 15.440 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.917 14.969 calculate_rho_elec 127 8.7 0.479 0.480 14.917 14.969 hybrid_alltoall_any 5155 16.4 1.289 3.003 10.923 12.890 make_images_data 4970 15.5 0.061 0.065 10.766 12.680 init_scf_run 11 5.9 0.000 0.001 11.920 11.920 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.920 11.920 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.554 9.637 dbcsr_mm_accdrv_process 20590 16.0 3.798 5.783 7.425 9.421 cp_fm_cholesky_decompose 22 10.9 9.045 9.062 9.045 9.062 wfi_extrapolate 11 7.9 0.001 0.001 8.679 8.680 grid_integrate_task_list 127 12.3 8.477 8.666 8.477 8.666 pw_transfer 1535 11.6 0.090 0.091 8.357 8.376 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.125 8.145 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 7.864 7.864 fft_wrap_pw1pw2_140 519 13.2 0.534 0.537 7.173 7.202 mp_alltoall_d11v 2401 14.1 7.025 7.158 7.025 7.158 calculate_dm_sparse 127 9.5 0.001 0.001 6.905 6.976 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.316 6.360 grid_collocate_task_list 127 9.7 6.286 6.327 6.286 6.327 fft3d_ps 1281 14.7 2.734 2.755 6.184 6.197 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.143 6.184 density_rs2pw 127 9.7 0.005 0.006 6.038 6.079 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.969000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2710.363636, yerr=180.316526 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 1.259856E+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 22064128. MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.200 86.209 86.209 qs_energies 1 2.0 0.000 0.000 84.839 84.923 ls_scf 1 3.0 0.002 0.015 83.923 83.941 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.787 72.985 multiply_cannon 111 7.7 0.018 0.021 56.056 57.555 multiply_cannon_loop 111 8.7 0.214 0.225 52.673 54.264 ls_scf_main 1 4.0 0.000 0.002 52.687 52.707 density_matrix_trs4 2 5.0 0.002 0.003 47.076 47.145 ls_scf_init_scf 1 4.0 0.000 0.001 28.175 28.188 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.053 27.094 mp_waitall_1 11316 10.9 22.275 25.460 22.275 25.460 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.943 24.962 multiply_cannon_multrec 2664 9.7 8.242 8.961 15.649 17.389 multiply_cannon_sync_h2d 2664 9.7 13.895 15.612 13.895 15.612 make_m2s 222 7.7 0.009 0.012 13.162 13.642 make_images 222 8.7 0.101 0.111 13.140 13.622 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.514 12.136 make_images_data 222 9.7 0.005 0.006 7.681 8.239 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.412 8.051 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.607 7.024 8.042 hybrid_alltoall_any 227 10.6 0.218 1.845 6.631 8.008 dbcsr_mm_accdrv_process_sort 4760 11.4 6.317 7.270 6.317 7.270 calculate_norms 4752 9.8 5.568 6.203 5.568 6.203 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.067 5.188 mp_sum_l 807 5.4 3.153 4.270 3.153 4.270 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.022 3.408 make_images_sizes 222 9.7 0.000 0.000 0.723 3.402 mp_alltoall_i44 222 10.7 0.723 3.402 0.723 3.402 mp_irecv_dv 6231 10.9 2.004 3.379 2.004 3.379 arnoldi_extremal 4 6.8 0.000 0.000 3.184 3.224 arnoldi_normal_ev 4 7.8 0.001 0.002 3.184 3.224 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.244 3.087 build_subspace 16 8.4 0.009 0.012 3.081 3.083 ls_scf_post 1 4.0 0.000 0.001 3.058 3.079 ls_scf_store_result 1 5.0 0.000 0.000 2.837 2.892 dbcsr_special_finalize 555 9.7 0.005 0.007 2.395 2.794 dbcsr_merge_single_wm 555 10.7 0.455 0.593 2.386 2.785 make_images_pack 222 9.7 2.214 2.644 2.216 2.646 dbcsr_sort_data 658 11.4 2.188 2.556 2.188 2.556 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.312 2.553 dbcsr_matrix_vector_mult_local 304 10.0 2.060 2.452 2.062 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.333 2.411 buffer_matrices_ensure_size 222 8.7 1.754 2.081 1.754 2.081 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.921 1.933 rebuild_ks_matrix 3 7.3 0.000 0.000 1.911 1.924 qs_ks_build_kohn_sham_matrix 3 8.3 0.022 0.032 1.911 1.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.209000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 2.097377E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.050 92.334 92.336 qs_energies 1 2.0 0.000 0.000 91.646 91.651 ls_scf 1 3.0 0.006 0.026 90.312 90.316 dbcsr_multiply_generic 111 6.7 0.117 0.440 76.265 76.626 multiply_cannon 111 7.7 0.030 0.044 53.632 57.495 ls_scf_main 1 4.0 0.001 0.005 55.403 55.404 multiply_cannon_loop 111 8.7 0.116 0.123 50.298 54.047 density_matrix_trs4 2 5.0 0.002 0.003 49.672 49.912 ls_scf_init_scf 1 4.0 0.000 0.001 31.358 31.360 mp_waitall_1 9246 10.9 21.109 30.376 21.109 30.376 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.204 30.313 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.788 27.809 multiply_cannon_multrec 1332 9.7 13.256 17.118 22.595 27.615 multiply_cannon_metrocomm3 1332 9.7 0.007 0.009 11.893 22.485 make_m2s 222 7.7 0.008 0.008 15.609 16.508 make_images 222 8.7 1.585 1.926 15.579 16.480 dbcsr_mm_accdrv_process 4041 10.4 0.313 0.475 8.936 10.439 dbcsr_mm_accdrv_process_sort 4041 11.4 8.516 10.012 8.516 10.012 make_images_data 222 9.7 0.007 0.023 8.688 9.715 hybrid_alltoall_any 227 10.6 0.522 2.444 8.044 9.377 mp_sum_l 807 5.4 5.629 9.014 5.629 9.014 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.251 7.939 mp_irecv_dv 3311 11.0 3.231 7.887 3.231 7.887 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.325 7.173 calculate_norms 2376 9.8 6.024 6.650 6.024 6.650 multiply_cannon_sync_h2d 1332 9.7 4.784 5.797 4.784 5.797 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.092 5.309 arnoldi_extremal 4 6.8 0.000 0.002 4.752 4.780 arnoldi_normal_ev 4 7.8 0.002 0.010 4.751 4.780 build_subspace 16 8.4 0.014 0.021 4.457 4.461 ls_scf_post 1 4.0 0.002 0.008 3.545 3.556 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.208 3.392 ls_scf_store_result 1 5.0 0.000 0.000 3.238 3.373 dbcsr_matrix_vector_mult_local 304 10.0 2.769 3.227 2.771 3.228 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.164 2.785 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.540 2.633 mp_allgather_i34 111 8.7 1.027 2.555 1.027 2.555 make_images_pack 222 9.7 2.026 2.431 2.029 2.433 dbcsr_sort_data 436 11.2 1.785 2.020 1.785 2.020 make_images_sizes 222 9.7 0.000 0.000 1.000 2.018 mp_alltoall_i44 222 10.7 1.000 2.017 1.000 2.017 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.336000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 2.702627E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.161 95.812 95.813 qs_energies 1 2.0 0.000 0.000 95.022 95.033 ls_scf 1 3.0 0.000 0.002 93.155 93.167 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.498 77.751 ls_scf_main 1 4.0 0.001 0.016 58.381 58.387 multiply_cannon 111 7.7 0.046 0.119 53.158 57.034 multiply_cannon_loop 111 8.7 0.100 0.107 49.549 53.979 density_matrix_trs4 2 5.0 0.002 0.003 52.512 52.686 mp_waitall_1 7374 11.0 24.005 34.540 24.005 34.540 ls_scf_init_scf 1 4.0 0.000 0.001 31.197 31.201 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.781 29.857 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.313 27.330 multiply_cannon_multrec 888 9.7 12.664 15.253 21.274 24.495 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.221 23.724 make_m2s 222 7.7 0.007 0.008 17.210 18.530 make_images 222 8.7 1.975 2.287 17.172 18.491 hybrid_alltoall_any 227 10.6 0.622 2.872 9.443 10.844 make_images_data 222 9.7 0.004 0.005 9.806 10.824 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.410 8.141 9.324 mp_sum_l 807 5.4 5.578 9.261 5.578 9.261 dbcsr_mm_accdrv_process_sort 3754 11.4 7.771 8.914 7.771 8.914 multiply_cannon_sync_h2d 888 9.7 6.073 7.602 6.073 7.602 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.326 7.446 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.486 7.266 mp_irecv_dv 2335 11.1 2.470 7.216 2.470 7.216 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.601 6.246 arnoldi_extremal 4 6.8 0.000 0.002 5.237 5.268 arnoldi_normal_ev 4 7.8 0.001 0.009 5.237 5.267 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.934 5.152 build_subspace 16 8.4 0.014 0.020 4.899 4.912 calculate_norms 1584 9.8 4.268 4.656 4.268 4.656 mp_allgather_i34 111 8.7 1.444 4.011 1.444 4.011 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.467 3.805 dbcsr_matrix_vector_mult_local 304 10.0 3.040 3.634 3.042 3.636 ls_scf_post 1 4.0 0.002 0.018 3.577 3.591 ls_scf_store_result 1 5.0 0.000 0.000 3.319 3.405 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.736 2.845 dbcsr_sort_data 325 11.1 1.891 2.152 1.891 2.152 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.138 2.145 make_images_pack 222 9.7 1.813 2.135 1.816 2.138 rebuild_ks_matrix 3 7.3 0.000 0.000 2.121 2.127 qs_ks_build_kohn_sham_matrix 3 8.3 0.035 0.141 2.120 2.127 make_images_sizes 222 9.7 0.000 0.000 1.058 2.106 mp_alltoall_i44 222 10.7 1.057 2.105 1.057 2.105 dbcsr_data_release 9322 10.9 1.312 1.953 1.312 1.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.813000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2219.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 3.334238E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.217 0.456 100.216 100.217 qs_energies 1 2.0 0.000 0.000 98.464 98.478 ls_scf 1 3.0 0.011 0.047 96.257 96.266 dbcsr_multiply_generic 111 6.7 0.018 0.023 78.605 78.813 ls_scf_main 1 4.0 0.001 0.018 59.748 59.766 multiply_cannon 111 7.7 0.048 0.144 51.803 56.092 density_matrix_trs4 2 5.0 0.002 0.003 53.613 53.747 multiply_cannon_loop 111 8.7 0.115 0.125 46.675 49.840 ls_scf_init_scf 1 4.0 0.000 0.002 33.228 33.253 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.632 31.682 mp_waitall_1 6438 11.0 22.523 30.172 22.523 30.172 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.129 29.141 multiply_cannon_multrec 1332 9.7 14.265 17.244 22.239 24.662 make_m2s 222 7.7 0.008 0.009 21.048 22.564 make_images 222 8.7 3.154 3.610 20.997 22.516 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.205 16.271 make_images_data 222 9.7 0.004 0.005 11.737 13.361 hybrid_alltoall_any 227 10.6 0.801 3.746 11.029 13.021 dbcsr_mm_accdrv_process 3641 10.4 0.213 0.416 7.602 9.152 dbcsr_mm_accdrv_process_sort 3641 11.4 7.226 8.731 7.226 8.731 mp_sum_l 807 5.4 4.197 6.965 4.197 6.965 arnoldi_extremal 4 6.8 0.016 0.070 6.294 6.311 arnoldi_normal_ev 4 7.8 0.012 0.047 6.278 6.311 multiply_cannon_sync_h2d 1332 9.7 5.603 6.237 5.603 6.237 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.080 5.935 mp_irecv_dv 3229 10.9 2.057 5.866 2.057 5.866 build_subspace 16 8.4 0.015 0.021 5.698 5.828 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.164 5.432 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.616 4.827 mp_allgather_i34 111 8.7 2.236 4.768 2.236 4.768 calculate_norms 2376 9.8 4.186 4.523 4.186 4.523 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.412 4.410 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.593 3.883 dbcsr_matrix_vector_mult_local 304 10.0 3.202 3.707 3.204 3.709 dbcsr_sort_data 658 11.4 3.067 3.448 3.067 3.448 ls_scf_post 1 4.0 0.005 0.019 3.270 3.275 dbcsr_special_finalize 555 9.7 0.006 0.007 2.813 3.240 dbcsr_merge_single_wm 555 10.7 0.535 0.666 2.804 3.232 ls_scf_store_result 1 5.0 0.000 0.000 3.018 3.071 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.920 2.988 dbcsr_data_release 10477 10.7 1.601 2.410 1.601 2.410 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 2.194 2.194 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.019 2.054 rebuild_ks_matrix 3 7.3 0.000 0.000 1.995 2.030 qs_ks_build_kohn_sham_matrix 3 8.3 0.028 0.108 1.995 2.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.217000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2694.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 4.628419E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.499 101.009 101.009 qs_energies 1 2.0 0.000 0.000 99.238 99.239 ls_scf 1 3.0 0.000 0.000 97.307 97.314 dbcsr_multiply_generic 111 6.7 0.018 0.019 78.832 79.116 ls_scf_main 1 4.0 0.000 0.000 62.604 62.614 multiply_cannon 111 7.7 0.116 0.214 55.797 60.478 density_matrix_trs4 2 5.0 0.002 0.003 55.565 55.634 multiply_cannon_loop 111 8.7 0.069 0.077 51.337 53.742 mp_waitall_1 5481 11.0 27.108 32.647 27.108 32.647 ls_scf_init_scf 1 4.0 0.000 0.000 31.063 31.069 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.767 29.809 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.497 27.512 multiply_cannon_multrec 444 9.7 14.142 16.368 21.250 24.094 make_m2s 222 7.7 0.006 0.007 17.923 20.364 make_images 222 8.7 3.734 4.425 17.859 20.302 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.549 16.294 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.137 15.443 hybrid_alltoall_any 227 10.6 0.787 3.765 10.005 12.485 make_images_data 222 9.7 0.003 0.004 10.220 12.483 multiply_cannon_sync_h2d 444 9.7 6.539 7.983 6.539 7.983 dbcsr_mm_accdrv_process 3003 10.4 0.155 0.363 6.812 7.939 dbcsr_mm_accdrv_process_sort 3003 11.4 6.493 7.596 6.493 7.596 mp_allgather_i34 111 8.7 2.679 7.002 2.679 7.002 arnoldi_extremal 4 6.8 0.000 0.000 5.756 5.772 arnoldi_normal_ev 4 7.8 0.001 0.004 5.756 5.772 build_subspace 16 8.4 0.015 0.020 5.354 5.364 mp_sum_l 807 5.4 3.228 5.004 3.228 5.004 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.619 4.784 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.157 4.334 dbcsr_matrix_vector_mult_local 304 10.0 3.709 4.149 3.711 4.151 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.657 4.044 mp_irecv_dv 1241 11.2 1.636 4.017 1.636 4.017 calculate_norms 792 9.8 3.554 3.742 3.554 3.742 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.203 3.708 ls_scf_post 1 4.0 0.000 0.000 3.640 3.641 make_images_sizes 222 9.7 0.000 0.000 1.022 3.530 mp_alltoall_i44 222 10.7 1.022 3.529 1.022 3.529 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.398 3.503 ls_scf_store_result 1 5.0 0.000 0.000 3.407 3.444 dbcsr_finalize 304 7.8 0.062 0.078 2.216 2.323 dbcsr_merge_all 275 8.9 0.476 0.524 2.059 2.153 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.030 2.032 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.009000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3632.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77a50e61e3cb5b09a01571247af7daff50d68f23_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 8.774496E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.107 0.206 108.155 108.155 qs_energies 1 2.0 0.000 0.000 106.236 106.250 ls_scf 1 3.0 0.000 0.000 102.954 102.968 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.117 77.261 ls_scf_main 1 4.0 0.000 0.000 65.484 65.485 density_matrix_trs4 2 5.0 0.002 0.003 56.643 56.694 multiply_cannon 111 7.7 0.157 0.255 49.443 50.969 multiply_cannon_loop 111 8.7 0.068 0.071 46.069 47.231 ls_scf_init_scf 1 4.0 0.000 0.000 33.802 33.803 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.483 32.498 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.652 29.663 mp_waitall_1 4569 11.1 21.925 25.825 21.925 25.825 make_m2s 222 7.7 0.007 0.007 23.999 24.862 make_images 222 8.7 4.593 4.981 23.890 24.750 multiply_cannon_multrec 444 9.7 17.880 18.543 22.531 23.246 make_images_data 222 9.7 0.004 0.004 13.248 15.524 hybrid_alltoall_any 227 10.6 1.662 3.645 13.039 15.477 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.098 10.673 multiply_cannon_sync_h2d 444 9.7 8.846 8.897 8.846 8.897 arnoldi_extremal 4 6.8 0.000 0.000 7.431 7.446 arnoldi_normal_ev 4 7.8 0.002 0.008 7.431 7.446 build_subspace 16 8.4 0.026 0.036 6.874 6.885 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.504 5.653 dbcsr_matrix_vector_mult_local 304 10.0 5.068 5.379 5.071 5.382 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.278 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.801 4.892 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.321 4.481 4.601 dbcsr_mm_accdrv_process_sort 1814 11.4 4.182 4.308 4.182 4.308 ls_scf_post 1 4.0 0.000 0.000 3.668 3.682 make_images_sizes 222 9.7 0.000 0.000 1.402 3.453 mp_alltoall_i44 222 10.7 1.402 3.453 1.402 3.453 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.427 calculate_norms 792 9.8 3.240 3.284 3.240 3.284 qs_energies_init_hamiltonians 1 3.0 0.364 0.373 3.249 3.250 dbcsr_finalize 304 7.8 0.082 0.090 3.087 3.130 mp_allgather_i34 111 8.7 0.990 2.935 0.990 2.935 dbcsr_merge_all 275 8.9 0.892 0.920 2.871 2.908 dbcsr_complete_redistribute 5 7.6 1.451 1.487 2.772 2.906 dbcsr_data_release 12724 10.6 2.328 2.871 2.328 2.871 matrix_ls_to_qs 2 6.0 0.000 0.000 2.428 2.577 dbcsr_sort_data 325 11.1 2.439 2.494 2.439 2.494 dbcsr_new_transposed 4 7.5 0.244 0.253 2.270 2.281 dbcsr_frobenius_norm 74 6.6 2.057 2.141 2.214 2.262 dbcsr_add_d 103 6.2 0.000 0.000 2.136 2.209 dbcsr_add_anytype 103 7.2 0.860 0.892 2.135 2.208 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.176 2.178 mp_sum_l 807 5.4 1.351 2.164 1.351 2.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.155000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6840.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 77a50e61e3cb5b09a01571247af7daff50d68f23 Summary: empty Status: OK