=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 49cb2407dde98ee02ed4a8f151770698be669539 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, 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 (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/01 job id: 45075035 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/02 job id: 45075036 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/03 job id: 45075037 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/04 job id: 45075038 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/05 job id: 45075039 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/06 job id: 45075040 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/07 job id: 45075041 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/08 job id: 45075042 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/09 job id: 45075043 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/10 job id: 45075044 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/11 job id: 45075045 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/12 job id: 45075046 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/13 job id: 45075047 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/14 job id: 45075048 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/15 job id: 45075049 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/16 job id: 45075051 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/17 job id: 45075052 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/18 job id: 45075053 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/19 job id: 45075054 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/20 job id: 45075055 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/21 job id: 45075057 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/22 job id: 45075058 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/23 job id: 45075059 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/24 job id: 45075060 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/25 job id: 45075061 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/26 job id: 45075062 --- 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/49cb2407dde98ee02ed4a8f151770698be669539_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 0.029 0.034 139.792 139.794 farming_run 1 2.0 138.336 138.341 139.751 139.754 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460564E+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 231 1098811. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 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 0.011 0.025 116.223 116.224 qs_energies 1 2.0 0.000 0.000 116.019 116.021 mp2_main 1 3.0 0.000 0.000 113.706 113.708 mp2_gpw_main 1 4.0 0.020 0.027 112.558 112.561 mp2_ri_gpw_compute_in 1 5.0 0.192 0.201 93.945 94.060 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.404 55.518 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.707 46.707 get_2c_integrals 1 6.0 0.000 0.001 37.697 38.340 integrate_v_rspace 273 8.0 0.438 0.449 25.062 29.723 pw_transfer 6555 10.6 0.369 0.393 27.518 27.936 fft_wrap_pw1pw2 5465 11.4 0.044 0.048 26.115 26.397 grid_integrate_task_list 273 9.0 20.886 25.982 20.886 25.982 fft_wrap_pw1pw2_100 2178 12.4 1.276 1.436 23.633 23.946 compute_2c_integrals 1 7.0 0.002 0.003 19.727 19.749 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 19.100 19.434 mp2_eri_2c_integrate_gpw 1 9.0 2.342 2.424 19.098 19.433 rpa_ri_compute_en 1 5.0 0.000 0.001 18.507 18.610 cp_fm_cholesky_decompose 12 8.2 17.948 18.603 17.948 18.603 cholesky_decomp 1 7.0 0.000 0.000 16.814 17.466 fft3d_s 5443 13.4 16.166 16.414 16.188 16.436 ao_to_mo_and_store_B_mult_1 272 7.0 10.864 15.600 10.864 15.600 calculate_wavefunction 272 8.0 5.463 5.547 12.567 13.215 rpa_num_int 1 6.0 0.000 0.001 10.575 10.584 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.561 10.565 calc_mat_Q 8 8.0 0.000 0.000 9.384 9.469 contract_S_to_Q 8 9.0 0.000 0.000 8.805 8.892 calc_potential_gpw 544 9.5 0.005 0.006 8.330 8.709 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.347 8.570 parallel_gemm_fm 14 9.1 0.000 0.000 8.391 8.459 parallel_gemm_fm_cosma 14 10.1 8.391 8.459 8.391 8.459 potential_pw2rs 545 10.0 0.107 0.109 7.731 8.335 collocate_single_gaussian 272 10.0 0.039 0.042 7.583 7.776 create_integ_mat 1 6.0 0.014 0.026 7.724 7.733 array2fm 1 7.0 0.000 0.000 6.721 7.141 pw_scatter_s 2720 13.7 4.411 4.682 4.411 4.682 pw_gather_s 2722 13.2 3.844 4.286 3.844 4.286 array2fm_buffer_send 1 8.0 2.974 3.155 2.974 3.155 pw_poisson_solve 545 10.5 1.132 1.194 2.248 2.439 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.558302, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 0.039 401.845 401.850 farming_run 1 2.0 398.392 398.398 401.802 401.807 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.224651E+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 744 385880. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.053 210.436 210.439 qs_energies 1 2.0 0.001 0.008 210.193 210.208 scf_env_do_scf 1 3.0 0.000 0.000 106.796 106.797 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.857 105.865 rebuild_ks_matrix 4 6.0 0.000 0.000 105.856 105.864 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.856 105.864 hfx_ks_matrix 4 8.0 0.001 0.001 105.457 105.461 integrate_four_center 4 9.0 0.144 0.462 105.456 105.460 mp2_main 1 3.0 0.004 0.034 103.091 103.108 mp2_gpw_main 1 4.0 0.050 0.094 102.097 102.119 integrate_four_center_main 4 10.0 0.096 0.449 96.674 99.443 integrate_four_center_bin 265 11.0 96.578 99.437 96.578 99.437 init_scf_loop 1 4.0 0.000 0.000 92.429 92.429 mp2_ri_gpw_compute_in 1 5.0 0.068 0.073 74.966 76.086 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.426 55.562 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.160 42.104 47.174 integrate_v_rspace 95 8.0 0.400 0.576 28.499 33.482 pw_transfer 2240 10.6 0.147 0.187 29.869 30.328 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.889 29.373 ao_to_mo_and_store_B_mult_1 91 7.0 10.643 29.029 10.643 29.029 grid_integrate_task_list 95 9.0 23.777 28.888 23.777 28.888 mp2_ri_gpw_compute_en 1 5.0 0.066 0.083 26.925 28.774 fft_wrap_pw1pw2_100 730 12.4 1.255 1.500 26.584 27.117 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.900 25.154 25.164 get_2c_integrals 1 6.0 0.004 0.009 20.434 20.492 compute_2c_integrals 1 7.0 0.017 0.033 19.413 19.443 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.787 19.265 mp2_eri_2c_integrate_gpw 1 9.0 1.747 1.901 18.786 19.264 fft3d_s 1823 13.4 18.395 18.654 18.408 18.667 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.365 14.365 calculate_wavefunction 91 8.0 2.024 2.048 9.737 9.953 potential_pw2rs 186 10.0 0.033 0.036 8.622 9.294 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.598 8.753 9.260 local_gemm 172 8.0 8.198 8.703 8.198 8.703 mp2_ri_gpw_compute_en_comm 22 7.0 0.492 0.514 8.174 8.600 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.196 8.580 calc_potential_gpw 182 9.5 0.002 0.003 7.888 8.213 collocate_single_gaussian 91 10.0 0.017 0.030 7.850 8.141 mp_sendrecv_dm3 2068 8.0 6.223 6.647 6.223 6.647 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.429 6.345 6.429 mp_sync 37 10.5 3.430 5.901 3.430 5.901 pw_gather_s 912 13.2 4.917 5.419 4.917 5.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.084921, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 452.194304E+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 272. MP_Sync 530 MP_Alltoall 2083 589622. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.028 58.323 58.325 qs_mol_dyn_low 1 2.0 0.003 0.004 57.392 58.114 qs_forces 11 3.9 0.002 0.003 54.293 54.295 qs_energies 11 4.9 0.002 0.002 52.811 52.822 scf_env_do_scf 11 5.9 0.001 0.001 46.252 46.252 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.087 44.088 dbcsr_multiply_generic 2286 12.5 0.092 0.097 33.563 33.964 qs_scf_new_mos 108 7.5 0.000 0.001 33.464 33.760 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.463 33.759 ot_scf_mini 108 9.5 0.002 0.002 31.785 31.940 velocity_verlet 10 3.0 0.001 0.002 28.533 28.568 multiply_cannon 2286 13.5 0.183 0.190 26.211 27.652 multiply_cannon_loop 2286 14.5 1.460 1.550 25.490 26.990 ot_mini 108 10.5 0.001 0.001 19.047 19.292 qs_ot_get_derivative 108 11.5 0.001 0.001 16.105 16.280 mp_waitall_1 245248 16.5 8.552 14.871 8.552 14.871 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 5.954 13.032 multiply_cannon_multrec 54864 15.5 4.276 6.678 7.731 11.397 qs_ot_get_p 119 10.4 0.001 0.001 8.113 8.454 rebuild_ks_matrix 119 8.3 0.000 0.000 8.151 8.271 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 8.151 8.271 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.196 7.304 mp_sum_l 7207 12.9 5.266 6.974 5.266 6.974 multiply_cannon_sync_h2d 54864 15.5 5.970 6.921 5.970 6.921 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.540 5.982 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.368 5.439 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.278 5.392 init_scf_run 11 5.9 0.000 0.001 5.329 5.329 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.329 5.329 dbcsr_mm_accdrv_process 76910 16.1 1.144 1.830 3.376 5.003 sum_up_and_integrate 119 10.3 0.012 0.015 4.711 4.718 integrate_v_rspace 119 11.3 0.002 0.003 4.699 4.708 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.517 4.644 calculate_rho_elec 119 8.7 0.012 0.017 4.516 4.643 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.611 4.611 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.389 4.390 cp_fm_redistribute_end 50 14.0 2.237 4.362 2.244 4.365 cp_fm_diag_elpa_base 50 14.0 2.115 4.264 2.119 4.273 calculate_first_density_matrix 1 7.0 0.000 0.001 3.087 3.090 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.900 3.089 apply_single 119 13.6 0.000 0.000 2.900 3.089 jit_kernel_multiply 13 15.8 2.171 3.086 2.171 3.086 calculate_dm_sparse 119 9.5 0.000 0.001 2.895 3.050 rs_pw_transfer 974 11.9 0.012 0.013 2.972 3.043 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.812 3.010 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.897 2.899 ot_diis_step 108 11.5 0.006 0.006 2.692 2.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.593 2.643 density_rs2pw 119 9.7 0.004 0.005 2.478 2.568 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.466 2.562 mp_sum_dm 438 4.9 2.368 2.432 2.368 2.432 acc_transpose_blocks 54864 15.5 0.226 0.251 1.855 2.410 update_particle_set 20 4.0 0.000 0.000 2.327 2.362 md_write_output 11 3.9 0.025 2.339 0.036 2.353 md_output 10 3.0 0.000 0.000 0.747 2.350 wfi_extrapolate 11 7.9 0.001 0.001 2.183 2.183 init_scf_loop 11 6.9 0.000 0.000 2.144 2.145 grid_integrate_task_list 119 12.3 2.020 2.121 2.020 2.121 pw_transfer 1439 11.6 0.053 0.060 1.986 2.070 potential_pw2rs 119 12.3 0.004 0.004 2.035 2.052 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.909 1.997 make_m2s 4572 13.5 0.054 0.056 1.774 1.831 mp_sum_d 4131 12.0 1.271 1.825 1.271 1.825 fft3d_ps 1201 14.6 0.369 0.475 1.680 1.761 make_images 4572 14.5 0.132 0.137 1.692 1.749 mp_alltoall_d11v 2130 13.8 1.409 1.583 1.409 1.583 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.471 1.559 write_restart 10 4.0 0.000 0.005 0.721 1.443 update_input 1 5.0 0.000 0.000 0.721 1.443 update_subsys 1 6.0 0.720 1.442 0.721 1.443 mp_max_l 33 2.8 0.722 1.443 0.722 1.443 mp_waitany 12084 13.8 1.277 1.417 1.277 1.417 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.386 1.406 grid_collocate_task_list 119 9.7 1.296 1.374 1.296 1.374 acc_transpose_blocks_kernels 54864 16.5 0.242 0.379 0.895 1.292 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.325000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 489.357312E+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 54 MP_Alltoall 2060 1293040. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.055 41.651 41.652 qs_mol_dyn_low 1 2.0 0.003 0.005 41.377 41.384 qs_forces 11 3.9 0.002 0.003 40.759 40.760 qs_energies 11 4.9 0.002 0.006 39.055 39.058 scf_env_do_scf 11 5.9 0.026 0.206 32.982 32.983 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.198 30.198 dbcsr_multiply_generic 2286 12.5 0.099 0.103 22.621 22.992 qs_scf_new_mos 108 7.5 0.001 0.001 20.712 20.961 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.711 20.961 ot_scf_mini 108 9.5 0.002 0.003 19.823 19.999 velocity_verlet 10 3.0 0.001 0.002 19.205 19.218 multiply_cannon 2286 13.5 0.207 0.214 17.299 18.798 multiply_cannon_loop 2286 14.5 0.892 0.966 16.173 17.732 ot_mini 108 10.5 0.001 0.001 12.252 12.491 mp_waitall_1 200699 16.5 6.065 11.836 6.065 11.836 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.535 10.391 qs_ot_get_derivative 108 11.5 0.001 0.001 9.768 9.950 multiply_cannon_multrec 27432 15.5 1.978 4.314 6.506 9.638 rebuild_ks_matrix 119 8.3 0.000 0.000 7.381 7.518 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.381 7.517 dbcsr_mm_accdrv_process 47894 16.0 3.422 6.126 4.459 6.862 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.528 6.653 qs_ot_get_p 119 10.4 0.001 0.001 4.675 4.911 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.727 4.847 init_scf_run 11 5.9 0.000 0.001 4.733 4.733 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.733 4.733 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.329 4.442 apply_single 119 13.6 0.000 0.000 3.328 4.442 mp_sum_l 7207 12.9 2.351 4.356 2.351 4.356 sum_up_and_integrate 119 10.3 0.024 0.027 4.340 4.346 integrate_v_rspace 119 11.3 0.002 0.003 4.315 4.322 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.982 4.012 calculate_rho_elec 119 8.7 0.021 0.024 3.981 4.011 rs_pw_transfer 974 11.9 0.010 0.011 2.766 3.201 qs_ot_p2m_diag 50 11.0 0.009 0.012 3.165 3.185 calculate_first_density_matrix 1 7.0 0.000 0.002 3.167 3.169 multiply_cannon_sync_h2d 27432 15.5 2.196 2.880 2.196 2.880 jit_kernel_multiply 10 16.1 0.986 2.778 0.986 2.778 make_m2s 4572 13.5 0.053 0.055 2.528 2.764 density_rs2pw 119 9.7 0.004 0.005 2.303 2.756 cp_dbcsr_syevd 50 12.0 0.003 0.006 2.729 2.730 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.689 2.691 make_images 4572 14.5 0.199 0.236 2.439 2.674 init_scf_loop 11 6.9 0.001 0.004 2.555 2.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.481 2.519 ot_diis_step 108 11.5 0.011 0.011 2.429 2.430 cp_fm_diag_elpa 50 13.0 0.000 0.002 2.361 2.362 cp_fm_redistribute_end 50 14.0 1.194 2.328 1.197 2.331 cp_fm_diag_elpa_base 50 14.0 1.100 2.237 1.129 2.275 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.174 2.265 calculate_dm_sparse 119 9.5 0.000 0.000 2.106 2.184 pw_transfer 1439 11.6 0.065 0.072 2.130 2.176 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.037 2.087 potential_pw2rs 119 12.3 0.006 0.006 2.066 2.080 grid_integrate_task_list 119 12.3 1.829 1.948 1.829 1.948 fft3d_ps 1201 14.6 0.515 0.566 1.738 1.786 prepare_preconditioner 11 7.9 0.000 0.000 1.614 1.643 make_preconditioner 11 8.9 0.000 0.002 1.614 1.643 fft_wrap_pw1pw2_140 487 13.2 0.079 0.087 1.549 1.594 make_images_data 4572 15.5 0.044 0.050 1.160 1.592 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.512 1.568 acc_transpose_blocks 27432 15.5 0.109 0.114 1.231 1.554 wfi_extrapolate 11 7.9 0.001 0.002 1.496 1.497 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.018 1.484 grid_collocate_task_list 119 9.7 1.229 1.366 1.229 1.366 mp_alltoall_d11v 2130 13.8 1.221 1.347 1.221 1.347 mp_allgather_i34 2286 14.5 0.571 1.344 0.571 1.344 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.329 1.337 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.178 1.228 mp_sum_d 4131 12.0 0.661 1.149 0.661 1.149 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.064 1.064 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.547 0.990 mp_waitany 5720 13.7 0.526 0.982 0.526 0.982 acc_transpose_blocks_kernels 27432 16.5 0.180 0.271 0.702 0.933 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.913 0.926 mp_alltoall_z22v 1201 16.6 0.784 0.889 0.784 0.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.652000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.671343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 524.075008E+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 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.021 0.040 38.842 38.845 qs_mol_dyn_low 1 2.0 0.003 0.003 38.484 38.490 qs_forces 11 3.9 0.002 0.003 34.293 34.294 qs_energies 11 4.9 0.002 0.002 32.693 32.695 scf_env_do_scf 11 5.9 0.002 0.002 27.047 27.047 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.405 24.406 velocity_verlet 10 3.0 0.001 0.001 19.718 19.849 dbcsr_multiply_generic 2286 12.5 0.094 0.099 17.977 18.082 qs_scf_new_mos 108 7.5 0.001 0.001 15.829 15.842 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.828 15.841 multiply_cannon 2286 13.5 0.195 0.201 14.507 15.343 ot_scf_mini 108 9.5 0.002 0.002 15.101 15.110 multiply_cannon_loop 2286 14.5 0.630 0.657 13.682 14.552 ot_mini 108 10.5 0.001 0.001 9.350 9.361 multiply_cannon_multrec 18288 15.5 2.006 3.361 7.728 8.364 qs_ot_get_derivative 108 11.5 0.001 0.001 7.818 7.828 rebuild_ks_matrix 119 8.3 0.000 0.000 6.610 6.634 qs_ks_build_kohn_sham_matrix 119 9.3 0.018 0.019 6.609 6.634 dbcsr_mm_accdrv_process 38222 16.0 4.788 6.227 5.638 6.423 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.853 5.877 init_scf_run 11 5.9 0.000 0.001 4.479 4.479 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.478 4.478 mp_waitall_1 158411 16.6 3.135 4.353 3.135 4.353 mp_sum_dm 438 4.9 4.064 4.197 4.064 4.197 md_write_output 11 3.9 0.106 3.371 0.133 4.177 update_particle_set 20 4.0 0.000 0.000 4.042 4.173 sum_up_and_integrate 119 10.3 0.030 0.031 4.139 4.145 integrate_v_rspace 119 11.3 0.003 0.003 4.108 4.117 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.653 3.664 calculate_rho_elec 119 8.7 0.031 0.031 3.653 3.663 qs_ot_get_p 119 10.4 0.001 0.001 3.543 3.566 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.851 3.470 calculate_first_density_matrix 1 7.0 0.001 0.001 3.195 3.196 jit_kernel_multiply 11 16.2 0.799 2.935 0.799 2.935 rs_pw_transfer 974 11.9 0.009 0.010 2.475 2.701 init_scf_loop 11 6.9 0.000 0.000 2.621 2.621 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.548 2.550 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.527 2.547 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.402 2.411 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.385 2.392 density_rs2pw 119 9.7 0.004 0.004 2.163 2.387 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.061 2.377 apply_single 119 13.6 0.000 0.000 2.061 2.377 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.125 2.126 pw_transfer 1439 11.6 0.066 0.072 2.071 2.089 make_m2s 4572 13.5 0.045 0.046 1.895 2.027 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.977 1.999 calculate_dm_sparse 119 9.5 0.000 0.000 1.937 1.946 make_images 4572 14.5 0.188 0.199 1.810 1.941 grid_integrate_task_list 119 12.3 1.800 1.900 1.800 1.900 potential_pw2rs 119 12.3 0.007 0.008 1.890 1.899 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.828 1.842 cp_fm_diag_elpa_base 50 14.0 1.802 1.817 1.826 1.839 prepare_preconditioner 11 7.9 0.000 0.000 1.808 1.811 make_preconditioner 11 8.9 0.000 0.000 1.808 1.811 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.661 1.745 fft3d_ps 1201 14.6 0.527 0.550 1.657 1.680 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.642 1.649 mp_sum_l 7207 12.9 1.184 1.560 1.184 1.560 multiply_cannon_sync_h2d 18288 15.5 1.376 1.558 1.376 1.558 fft_wrap_pw1pw2_140 487 13.2 0.090 0.093 1.540 1.558 ot_diis_step 108 11.5 0.011 0.012 1.509 1.509 acc_transpose_blocks 18288 15.5 0.076 0.077 1.354 1.377 grid_collocate_task_list 119 9.7 1.214 1.363 1.214 1.363 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.304 1.308 wfi_extrapolate 11 7.9 0.001 0.001 1.231 1.231 make_images_data 4572 15.5 0.044 0.048 0.856 1.026 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.667 0.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.940 0.941 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 0.914 0.933 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.737 0.930 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.902 0.927 mp_alltoall_d11v 2130 13.8 0.723 0.843 0.723 0.843 mp_alltoall_z22v 1201 16.6 0.722 0.835 0.722 0.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.828 0.829 md_output 10 3.0 0.000 0.000 0.030 0.815 write_trajectory 44 4.9 0.003 0.044 0.026 0.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.845000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.545455, yerr=2.016461 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 556.797952E+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 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.028 0.069 36.811 36.811 qs_mol_dyn_low 1 2.0 0.003 0.005 36.474 36.482 qs_forces 11 3.9 0.003 0.009 36.410 36.411 qs_energies 11 4.9 0.002 0.008 34.660 34.665 scf_env_do_scf 11 5.9 0.053 0.419 29.370 29.372 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.439 25.440 dbcsr_multiply_generic 2286 12.5 0.100 0.105 18.425 18.556 velocity_verlet 10 3.0 0.003 0.004 18.233 18.235 qs_scf_new_mos 108 7.5 0.001 0.001 16.677 16.728 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.676 16.727 ot_scf_mini 108 9.5 0.002 0.003 15.733 15.785 multiply_cannon 2286 13.5 0.230 0.283 14.762 15.539 multiply_cannon_loop 2286 14.5 0.928 0.962 13.778 14.406 ot_mini 108 10.5 0.001 0.001 9.423 9.487 multiply_cannon_multrec 27432 15.5 2.328 2.992 8.736 9.110 qs_ot_get_derivative 108 11.5 0.001 0.001 7.606 7.661 dbcsr_mm_accdrv_process 47916 15.9 5.413 7.102 6.314 7.498 rebuild_ks_matrix 119 8.3 0.000 0.000 6.768 6.814 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.018 6.768 6.813 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.006 6.047 sum_up_and_integrate 119 10.3 0.035 0.037 3.954 3.965 integrate_v_rspace 119 11.3 0.003 0.003 3.919 3.930 init_scf_run 11 5.9 0.000 0.001 3.875 3.876 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.875 3.875 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.733 3.765 calculate_rho_elec 119 8.7 0.040 0.046 3.733 3.764 qs_ot_get_p 119 10.4 0.001 0.001 3.670 3.741 init_scf_loop 11 6.9 0.001 0.005 3.491 3.492 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.988 3.309 prepare_preconditioner 11 7.9 0.000 0.000 2.618 2.626 make_preconditioner 11 8.9 0.000 0.001 2.618 2.626 mp_waitall_1 137007 16.6 2.070 2.551 2.070 2.551 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.231 2.551 calculate_first_density_matrix 1 7.0 0.001 0.003 2.420 2.421 rs_pw_transfer 974 11.9 0.009 0.009 2.211 2.411 make_m2s 4572 13.5 0.055 0.058 2.305 2.410 density_rs2pw 119 9.7 0.004 0.009 2.132 2.327 make_images 4572 14.5 0.270 0.330 2.198 2.301 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.280 2.290 pw_transfer 1439 11.6 0.066 0.072 2.204 2.243 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.976 2.163 apply_single 119 13.6 0.000 0.000 1.976 2.163 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.110 2.154 calculate_dm_sparse 119 9.5 0.000 0.001 2.075 2.126 jit_kernel_multiply 10 15.9 0.841 2.013 0.841 2.013 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.969 1.996 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.960 1.962 cp_dbcsr_syevd 50 12.0 0.004 0.011 1.959 1.960 grid_integrate_task_list 119 12.3 1.819 1.916 1.819 1.916 fft3d_ps 1201 14.6 0.563 0.613 1.784 1.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.766 1.779 ot_diis_step 108 11.5 0.012 0.012 1.778 1.778 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.715 1.763 potential_pw2rs 119 12.3 0.008 0.009 1.702 1.709 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.684 1.694 cp_fm_diag_elpa_base 50 14.0 1.643 1.660 1.675 1.684 acc_transpose_blocks 27432 15.5 0.112 0.115 1.510 1.600 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.923 1.512 wfi_extrapolate 11 7.9 0.001 0.001 1.388 1.388 grid_collocate_task_list 119 9.7 1.222 1.331 1.222 1.331 mp_sum_l 7207 12.9 0.952 1.274 0.952 1.274 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.255 1.264 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.123 1.141 cp_fm_upper_to_full 72 14.2 0.803 1.139 0.803 1.139 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.137 1.138 mp_alltoall_i22 627 13.8 0.793 1.129 0.793 1.129 dbcsr_complete_redistribute 329 12.2 0.121 0.148 0.826 1.097 multiply_cannon_sync_h2d 27432 15.5 0.989 1.053 0.989 1.053 make_images_data 4572 15.5 0.045 0.049 0.918 1.033 mp_alltoall_d11v 2130 13.8 0.794 0.987 0.794 0.987 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.800 0.985 acc_transpose_blocks_kernels 27432 16.5 0.266 0.276 0.887 0.967 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.871 copy_dbcsr_to_fm 153 11.3 0.002 0.003 0.795 0.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.854 0.858 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.586 0.848 mp_alltoall_z22v 1201 16.6 0.804 0.838 0.804 0.838 cp_fm_cholesky_invert 11 10.9 0.803 0.806 0.803 0.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.811000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.454545, yerr=3.201239 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 600.174592E+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 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 29.209 29.210 qs_mol_dyn_low 1 2.0 0.003 0.003 29.036 29.045 qs_forces 11 3.9 0.002 0.003 28.976 28.977 qs_energies 11 4.9 0.002 0.002 27.242 27.245 scf_env_do_scf 11 5.9 0.002 0.002 22.056 22.056 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.448 19.449 velocity_verlet 10 3.0 0.001 0.002 14.857 14.860 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.670 12.748 qs_scf_new_mos 108 7.5 0.001 0.001 11.190 11.215 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.189 11.214 ot_scf_mini 108 9.5 0.002 0.002 10.507 10.533 multiply_cannon 2286 13.5 0.231 0.241 10.002 10.484 multiply_cannon_loop 2286 14.5 0.328 0.340 9.067 9.233 multiply_cannon_multrec 9144 15.5 1.645 1.887 6.176 6.419 rebuild_ks_matrix 119 8.3 0.000 0.000 6.140 6.160 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.018 6.140 6.159 ot_mini 108 10.5 0.001 0.001 5.791 5.820 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.472 5.490 dbcsr_mm_accdrv_process 12550 15.8 3.207 4.437 4.430 4.533 qs_ot_get_derivative 108 11.5 0.001 0.001 4.464 4.490 sum_up_and_integrate 119 10.3 0.038 0.042 3.770 3.775 integrate_v_rspace 119 11.3 0.003 0.003 3.732 3.738 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.676 3.682 calculate_rho_elec 119 8.7 0.060 0.061 3.676 3.682 init_scf_run 11 5.9 0.000 0.001 3.656 3.656 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.655 3.655 qs_ot_get_p 119 10.4 0.001 0.001 2.890 2.929 init_scf_loop 11 6.9 0.000 0.000 2.581 2.582 calculate_first_density_matrix 1 7.0 0.001 0.001 2.490 2.491 pw_transfer 1439 11.6 0.066 0.070 2.215 2.226 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.121 2.135 jit_kernel_multiply 10 15.7 1.184 2.128 1.184 2.128 density_rs2pw 119 9.7 0.004 0.004 1.977 2.115 mp_waitall_1 115863 16.7 1.511 2.036 1.511 2.036 make_m2s 4572 13.5 0.035 0.036 1.842 1.990 grid_integrate_task_list 119 12.3 1.843 1.961 1.843 1.961 rs_pw_transfer 974 11.9 0.008 0.008 1.776 1.913 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.905 1.907 make_images 4572 14.5 0.268 0.301 1.753 1.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.877 1.878 prepare_preconditioner 11 7.9 0.000 0.000 1.825 1.828 make_preconditioner 11 8.9 0.000 0.000 1.825 1.828 calculate_dm_sparse 119 9.5 0.000 0.000 1.790 1.808 fft3d_ps 1201 14.6 0.566 0.580 1.779 1.788 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.742 1.758 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.711 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.697 1.703 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.701 1.702 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.570 1.585 potential_pw2rs 119 12.3 0.010 0.011 1.490 1.493 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.383 1.391 grid_collocate_task_list 119 9.7 1.273 1.390 1.273 1.390 cp_fm_diag_elpa_base 50 14.0 1.355 1.374 1.381 1.390 ot_diis_step 108 11.5 0.012 0.013 1.314 1.314 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.224 1.224 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.219 1.224 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.184 1.204 apply_single 119 13.6 0.000 0.000 1.184 1.204 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.180 1.191 wfi_extrapolate 11 7.9 0.001 0.001 1.115 1.115 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.853 1.097 make_images_data 4572 15.5 0.039 0.042 0.876 1.063 mp_alltoall_d11v 2130 13.8 0.848 0.988 0.848 0.988 acc_transpose_blocks 9144 15.5 0.038 0.039 0.980 0.986 cp_fm_cholesky_invert 11 10.9 0.941 0.944 0.941 0.944 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.919 multiply_cannon_sync_h2d 9144 15.5 0.715 0.794 0.715 0.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.791 0.794 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.743 0.752 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.740 0.742 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.677 0.729 mp_allgather_i34 2286 14.5 0.269 0.724 0.269 0.724 mp_alltoall_z22v 1201 16.6 0.660 0.716 0.660 0.716 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.312 0.700 yz_to_x 606 15.1 0.266 0.278 0.608 0.627 jit_kernel_transpose 5 15.6 0.624 0.626 0.624 0.626 x_to_yz 595 16.2 0.276 0.288 0.594 0.603 dbcsr_complete_redistribute 329 12.2 0.183 0.191 0.568 0.602 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.210000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.181818, yerr=3.761484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 747.069440E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 45.963 45.964 qs_mol_dyn_low 1 2.0 0.003 0.004 45.521 45.528 qs_forces 11 3.9 0.002 0.003 45.457 45.458 qs_energies 11 4.9 0.002 0.002 43.451 43.454 scf_env_do_scf 11 5.9 0.002 0.002 37.425 37.426 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.989 28.990 velocity_verlet 10 3.0 0.002 0.002 23.945 23.951 dbcsr_multiply_generic 2286 12.5 0.099 0.100 20.146 20.405 qs_scf_new_mos 108 7.5 0.001 0.001 18.433 18.534 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.432 18.533 ot_scf_mini 108 9.5 0.002 0.002 17.344 17.451 multiply_cannon 2286 13.5 0.301 0.312 16.126 17.234 multiply_cannon_loop 2286 14.5 0.342 0.348 14.836 15.971 ot_mini 108 10.5 0.001 0.001 11.169 11.294 multiply_cannon_multrec 9144 15.5 5.578 6.902 11.087 11.243 qs_ot_get_derivative 108 11.5 0.001 0.001 9.093 9.199 init_scf_loop 11 6.9 0.000 0.000 8.404 8.404 rebuild_ks_matrix 119 8.3 0.000 0.001 7.955 8.110 qs_ks_build_kohn_sham_matrix 119 9.3 0.423 0.425 7.955 8.109 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.228 7.370 prepare_preconditioner 11 7.9 0.000 0.000 6.983 6.997 make_preconditioner 11 8.9 0.000 0.000 6.983 6.997 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.514 6.875 dbcsr_mm_accdrv_process 12550 15.8 4.463 5.974 5.385 6.768 cp_fm_upper_to_full 72 14.2 3.196 4.606 3.196 4.606 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.395 4.407 calculate_rho_elec 119 8.7 0.118 0.121 4.394 4.407 sum_up_and_integrate 119 10.3 0.066 0.067 4.170 4.176 integrate_v_rspace 119 11.3 0.004 0.004 4.104 4.111 init_scf_run 11 5.9 0.000 0.001 3.965 3.965 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.965 3.965 qs_ot_get_p 119 10.4 0.001 0.001 3.400 3.547 mp_waitall_1 94719 16.7 2.434 3.523 2.434 3.523 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.458 2.931 pw_transfer 1439 11.6 0.069 0.069 2.921 2.927 dbcsr_complete_redistribute 329 12.2 0.287 0.295 2.031 2.861 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.822 2.828 calculate_dm_sparse 119 9.5 0.000 0.000 2.530 2.555 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.707 2.536 make_m2s 4572 13.5 0.038 0.039 2.336 2.531 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.274 2.500 apply_single 119 13.6 0.000 0.000 2.274 2.500 calculate_first_density_matrix 1 7.0 0.001 0.001 2.448 2.451 fft3d_ps 1201 14.6 0.602 0.611 2.446 2.451 make_images 4572 14.5 0.352 0.383 2.215 2.410 density_rs2pw 119 9.7 0.004 0.004 2.321 2.372 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.362 2.370 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.319 2.308 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.464 2.287 mp_alltoall_i22 627 13.8 1.457 2.286 1.457 2.286 grid_integrate_task_list 119 12.3 2.045 2.115 2.045 2.115 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.074 2.076 ot_diis_step 108 11.5 0.014 0.014 2.047 2.047 mp_sum_l 7207 12.9 1.142 1.850 1.142 1.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.828 1.830 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.796 1.796 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.791 1.792 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.735 1.790 rs_pw_transfer 974 11.9 0.009 0.009 1.558 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.590 1.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.547 1.581 cp_fm_cholesky_invert 11 10.9 1.555 1.558 1.555 1.558 grid_collocate_task_list 119 9.7 1.479 1.524 1.479 1.524 potential_pw2rs 119 12.3 0.014 0.014 1.518 1.523 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.488 1.488 cp_fm_diag_elpa_base 50 14.0 1.342 1.395 1.486 1.486 wfi_extrapolate 11 7.9 0.001 0.001 1.442 1.443 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.138 1.410 make_images_data 4572 15.5 0.043 0.046 1.108 1.361 mp_alltoall_d11v 2130 13.8 1.160 1.247 1.160 1.247 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.155 1.178 qs_env_update_s_mstruct 11 6.9 0.025 0.028 1.129 1.149 multiply_cannon_sync_h2d 9144 15.5 1.046 1.048 1.046 1.048 jit_kernel_multiply 6 15.7 0.895 1.047 0.895 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.980 1.028 yz_to_x 606 15.1 0.458 0.471 0.979 0.989 qs_create_task_list 11 7.9 0.001 0.001 0.940 0.953 generate_qs_task_list 11 8.9 0.370 0.390 0.939 0.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.939 0.952 mp_alltoall_z22v 1201 16.6 0.896 0.938 0.896 0.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.964000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=703.636364, yerr=11.826570 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 503.218176E+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 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.012 0.030 83.311 83.313 qs_mol_dyn_low 1 2.0 0.003 0.003 82.964 82.974 qs_forces 11 3.9 0.011 0.024 82.889 82.891 qs_energies 11 4.9 0.002 0.003 79.988 80.005 scf_env_do_scf 11 5.9 0.001 0.002 70.831 70.835 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.048 65.050 dbcsr_multiply_generic 2055 12.4 0.104 0.108 51.090 51.368 qs_scf_new_mos 99 7.5 0.000 0.001 47.300 47.442 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.299 47.441 ot_scf_mini 99 9.5 0.002 0.002 44.982 45.087 multiply_cannon 2055 13.4 0.180 0.186 43.120 43.914 velocity_verlet 10 3.0 0.001 0.002 43.077 43.078 multiply_cannon_loop 2055 14.4 1.491 1.527 42.135 42.968 ot_mini 99 10.5 0.001 0.001 26.611 26.719 qs_ot_get_derivative 99 11.5 0.001 0.001 19.851 19.950 multiply_cannon_multrec 49320 15.4 12.541 13.232 17.919 18.855 rebuild_ks_matrix 110 8.3 0.000 0.001 14.835 14.998 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.835 14.997 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.017 13.151 mp_waitall_1 220248 16.4 10.924 12.365 10.924 12.365 multiply_cannon_sync_h2d 49320 15.4 10.428 11.063 10.428 11.063 qs_ot_get_p 110 10.4 0.001 0.001 10.033 10.163 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.097 8.672 multiply_cannon_metrocomm3 49320 15.4 0.077 0.080 6.648 8.341 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.251 7.787 apply_single 110 13.6 0.000 0.001 7.251 7.787 sum_up_and_integrate 110 10.3 0.036 0.043 7.292 7.309 integrate_v_rspace 110 11.3 0.003 0.004 7.255 7.281 init_scf_run 11 5.9 0.000 0.001 7.093 7.093 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.092 7.093 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.803 6.914 calculate_rho_elec 110 8.6 0.021 0.026 6.802 6.914 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.831 6.846 ot_diis_step 99 11.5 0.006 0.006 6.587 6.588 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.013 6.013 dbcsr_mm_accdrv_process 87628 16.1 1.921 1.983 5.252 5.768 init_scf_loop 11 6.9 0.000 0.000 5.751 5.751 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.408 5.440 cp_fm_diag_elpa_base 48 14.0 5.393 5.426 5.406 5.438 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.220 5.266 rs_pw_transfer 902 11.9 0.012 0.013 3.905 4.423 mp_sum_l 6514 12.8 3.564 4.274 3.564 4.274 density_rs2pw 110 9.6 0.004 0.005 3.617 4.193 wfi_extrapolate 11 7.9 0.001 0.001 4.096 4.096 make_m2s 4110 13.4 0.061 0.065 3.948 4.056 make_images 4110 14.4 0.176 0.191 3.852 3.964 calculate_dm_sparse 110 9.5 0.001 0.001 3.847 3.939 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.322 3.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.710 3.714 prepare_preconditioner 11 7.9 0.000 0.000 3.544 3.568 make_preconditioner 11 8.9 0.000 0.000 3.544 3.568 pw_transfer 1331 11.6 0.055 0.072 3.382 3.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.365 3.424 grid_integrate_task_list 110 12.3 3.230 3.386 3.230 3.386 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.294 3.369 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.305 3.358 jit_kernel_multiply 13 15.9 3.057 3.349 3.057 3.349 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.056 3.092 calculate_first_density_matrix 1 7.0 0.000 0.001 2.905 2.910 fft3d_ps 1111 14.6 0.777 0.889 2.822 2.890 fft_wrap_pw1pw2_140 451 13.1 0.170 0.191 2.763 2.839 potential_pw2rs 110 12.3 0.006 0.007 2.800 2.827 mp_alltoall_d11v 2046 13.8 2.140 2.602 2.140 2.602 mp_waitany 14300 13.8 1.838 2.537 1.838 2.537 acc_transpose_blocks 49320 15.4 0.208 0.214 2.236 2.486 grid_collocate_task_list 110 9.6 2.081 2.290 2.081 2.290 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.959 1.985 cp_fm_cholesky_invert 11 10.9 1.903 1.907 1.903 1.907 make_images_data 4110 15.4 0.042 0.045 1.778 1.904 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.809 1.838 hybrid_alltoall_any 4261 16.3 0.082 0.480 1.555 1.824 mp_sum_d 3885 11.9 1.295 1.807 1.295 1.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.313000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.909091, yerr=2.810841 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.919936E+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 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2462973. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.023 0.064 71.556 71.558 qs_mol_dyn_low 1 2.0 0.004 0.009 71.191 71.202 qs_forces 11 3.9 0.005 0.016 71.035 71.037 qs_energies 11 4.9 0.002 0.008 67.645 67.651 scf_env_do_scf 11 5.9 0.001 0.002 58.527 58.531 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.651 50.654 dbcsr_multiply_generic 2055 12.4 0.114 0.119 38.461 38.662 velocity_verlet 10 3.0 0.001 0.001 37.035 37.039 qs_scf_new_mos 99 7.5 0.001 0.001 33.775 33.906 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.775 33.905 multiply_cannon 2055 13.4 0.223 0.243 31.707 32.762 ot_scf_mini 99 9.5 0.003 0.004 32.109 32.226 multiply_cannon_loop 2055 14.4 0.921 0.952 30.376 31.228 ot_mini 99 10.5 0.001 0.001 18.560 18.685 multiply_cannon_multrec 24660 15.4 7.670 9.329 13.838 15.510 rebuild_ks_matrix 110 8.3 0.000 0.001 14.090 14.175 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 14.090 14.175 qs_ot_get_derivative 99 11.5 0.001 0.001 12.710 12.833 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.407 12.481 mp_waitall_1 176588 16.5 8.369 10.867 8.369 10.867 multiply_cannon_metrocomm3 24660 15.4 0.069 0.072 5.596 8.555 multiply_cannon_sync_h2d 24660 15.4 7.093 8.099 7.093 8.099 init_scf_loop 11 6.9 0.001 0.004 7.833 7.840 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.567 7.270 apply_single 110 13.6 0.000 0.001 6.567 7.270 qs_ot_get_p 110 10.4 0.001 0.002 6.883 7.027 sum_up_and_integrate 110 10.3 0.053 0.060 6.834 6.848 integrate_v_rspace 110 11.3 0.003 0.003 6.781 6.794 init_scf_run 11 5.9 0.000 0.001 6.636 6.637 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.636 6.636 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.404 6.421 calculate_rho_elec 110 8.6 0.040 0.048 6.403 6.420 dbcsr_mm_accdrv_process 52282 16.1 4.680 5.443 6.008 6.315 ot_diis_step 99 11.5 0.010 0.010 5.795 5.795 prepare_preconditioner 11 7.9 0.000 0.000 5.752 5.766 make_preconditioner 11 8.9 0.000 0.002 5.752 5.766 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.783 5.521 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.304 5.459 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.937 4.959 make_m2s 4110 13.4 0.058 0.061 4.284 4.727 make_images 4110 14.4 0.397 0.441 4.175 4.614 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.499 4.499 pw_transfer 1331 11.6 0.067 0.074 3.910 4.069 density_rs2pw 110 9.6 0.004 0.005 3.501 3.970 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.802 3.963 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.889 3.906 cp_fm_diag_elpa_base 48 14.0 3.840 3.859 3.887 3.902 rs_pw_transfer 902 11.9 0.012 0.014 3.176 3.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.630 3.632 wfi_extrapolate 11 7.9 0.001 0.001 3.571 3.571 fft_wrap_pw1pw2_140 451 13.1 0.207 0.254 3.187 3.353 grid_integrate_task_list 110 12.3 3.139 3.342 3.139 3.342 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.273 3.307 fft3d_ps 1111 14.6 1.106 1.331 3.149 3.298 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.204 3.273 calculate_dm_sparse 110 9.5 0.001 0.001 3.071 3.097 calculate_first_density_matrix 1 7.0 0.001 0.003 2.956 2.959 make_images_data 4110 15.4 0.046 0.050 2.410 2.865 hybrid_alltoall_any 4261 16.3 0.102 0.439 2.132 2.842 cp_fm_cholesky_invert 11 10.9 2.700 2.707 2.700 2.707 potential_pw2rs 110 12.3 0.008 0.009 2.599 2.617 grid_collocate_task_list 110 9.6 2.050 2.494 2.050 2.494 mp_sum_l 6514 12.8 1.874 2.491 1.874 2.491 mp_alltoall_d11v 2046 13.8 1.890 2.211 1.890 2.211 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.973 1.994 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.954 1.957 jit_kernel_multiply 10 16.3 0.979 1.898 0.979 1.898 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.850 1.863 mp_waitany 10164 13.8 1.270 1.806 1.270 1.806 mp_allgather_i34 2055 14.4 0.660 1.730 0.660 1.730 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.654 acc_transpose_blocks 24660 15.4 0.112 0.115 1.533 1.628 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.778 1.551 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.219 0.992 1.513 dbcsr_complete_redistribute 325 12.2 0.244 0.304 1.221 1.490 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.474 mp_alltoall_z22v 1111 16.6 1.351 1.469 1.351 1.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.558000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.000000, yerr=7.071068 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 660.652032E+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 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.027 0.068 61.909 61.911 qs_mol_dyn_low 1 2.0 0.003 0.004 61.537 61.546 qs_forces 11 3.9 0.004 0.010 61.467 61.467 qs_energies 11 4.9 0.016 0.022 58.213 58.216 scf_env_do_scf 11 5.9 0.003 0.006 50.124 50.125 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.337 41.338 velocity_verlet 10 3.0 0.001 0.002 33.235 33.237 dbcsr_multiply_generic 2055 12.4 0.108 0.111 29.345 29.639 qs_scf_new_mos 99 7.5 0.001 0.001 25.983 26.097 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.983 26.096 ot_scf_mini 99 9.5 0.002 0.003 24.757 24.887 multiply_cannon 2055 13.4 0.209 0.221 22.801 23.968 multiply_cannon_loop 2055 14.4 0.609 0.628 21.560 22.644 ot_mini 99 10.5 0.001 0.001 14.249 14.384 rebuild_ks_matrix 110 8.3 0.000 0.000 12.604 12.780 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.604 12.780 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.103 11.260 multiply_cannon_multrec 16440 15.4 3.788 4.768 9.996 10.861 mp_waitall_1 139946 16.5 7.489 10.722 7.489 10.722 qs_ot_get_derivative 99 11.5 0.001 0.001 9.708 9.839 init_scf_loop 11 6.9 0.000 0.000 8.748 8.751 multiply_cannon_metrocomm3 16440 15.4 0.044 0.046 4.712 7.756 prepare_preconditioner 11 7.9 0.000 0.000 6.928 6.946 make_preconditioner 11 8.9 0.000 0.000 6.928 6.946 sum_up_and_integrate 110 10.3 0.061 0.062 6.675 6.691 integrate_v_rspace 110 11.3 0.003 0.004 6.614 6.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.249 6.593 dbcsr_mm_accdrv_process 34862 16.1 4.696 5.637 6.062 6.272 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.122 6.135 calculate_rho_elec 110 8.6 0.059 0.059 6.121 6.134 qs_ot_get_p 110 10.4 0.001 0.001 5.637 5.818 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.164 5.710 apply_single 110 13.6 0.000 0.000 5.164 5.710 init_scf_run 11 5.9 0.000 0.001 5.675 5.675 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.674 5.675 make_m2s 4110 13.4 0.050 0.051 4.271 4.660 make_images 4110 14.4 0.391 0.505 4.157 4.545 ot_diis_step 99 11.5 0.011 0.011 4.510 4.510 density_rs2pw 110 9.6 0.004 0.005 3.209 4.435 multiply_cannon_sync_h2d 16440 15.4 3.635 4.311 3.635 4.311 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.985 4.003 rs_pw_transfer 902 11.9 0.011 0.011 2.755 3.969 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.177 3.927 pw_transfer 1331 11.6 0.066 0.073 3.800 3.809 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.692 3.703 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.630 3.630 grid_integrate_task_list 110 12.3 3.161 3.367 3.161 3.367 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.173 3.184 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.059 3.067 cp_fm_diag_elpa_base 48 14.0 2.992 3.024 3.057 3.065 wfi_extrapolate 11 7.9 0.001 0.001 3.060 3.060 make_images_data 4110 15.4 0.043 0.048 2.506 3.014 fft3d_ps 1111 14.6 1.089 1.101 2.991 2.999 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.231 2.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.789 2.791 cp_fm_cholesky_invert 11 10.9 2.752 2.758 2.752 2.758 calculate_dm_sparse 110 9.5 0.001 0.001 2.674 2.710 calculate_first_density_matrix 1 7.0 0.001 0.004 2.526 2.528 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.906 2.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.495 grid_collocate_task_list 110 9.6 2.080 2.472 2.080 2.472 mp_waitany 17072 13.8 1.185 2.461 1.185 2.461 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.357 2.427 potential_pw2rs 110 12.3 0.011 0.011 2.388 2.401 mp_irecv_dv 48980 15.7 0.835 2.376 0.835 2.376 mp_sum_l 6514 12.8 1.683 2.348 1.683 2.348 mp_alltoall_d11v 2046 13.8 1.866 2.250 1.866 2.250 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.886 2.104 dbcsr_complete_redistribute 325 12.2 0.326 0.347 1.529 1.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.960 1.960 jit_kernel_multiply 8 16.6 0.980 1.931 0.980 1.931 mp_allgather_i34 2055 14.4 0.571 1.763 0.571 1.763 cp_fm_upper_to_full 70 14.2 1.390 1.757 1.390 1.757 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.700 1.719 cp_fm_cholesky_decompose 22 10.9 1.630 1.649 1.630 1.649 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.630 1.645 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.378 1.473 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.001 1.451 rs_gather_matrices 110 12.3 0.235 0.262 0.983 1.292 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.273 1.291 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.911000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.181818, yerr=8.331956 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 730.546176E+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 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.058 0.075 69.750 69.751 qs_mol_dyn_low 1 2.0 0.003 0.007 69.204 69.386 qs_forces 11 3.9 0.003 0.003 68.872 68.873 qs_energies 11 4.9 0.002 0.007 65.432 65.438 scf_env_do_scf 11 5.9 0.001 0.002 54.553 54.556 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.733 42.735 velocity_verlet 10 3.0 0.001 0.001 37.364 37.369 dbcsr_multiply_generic 2055 12.4 0.114 0.118 31.999 32.257 qs_scf_new_mos 99 7.5 0.001 0.001 27.349 27.465 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.348 27.464 multiply_cannon 2055 13.4 0.249 0.276 24.996 25.929 ot_scf_mini 99 9.5 0.003 0.004 25.718 25.813 multiply_cannon_loop 2055 14.4 0.878 0.895 23.628 24.264 multiply_cannon_multrec 24660 15.4 4.244 6.913 13.914 15.008 ot_mini 99 10.5 0.001 0.001 14.522 14.642 rebuild_ks_matrix 110 8.3 0.000 0.001 12.379 12.465 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.378 12.465 init_scf_loop 11 6.9 0.001 0.006 11.776 11.779 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.940 11.021 dbcsr_mm_accdrv_process 52304 16.0 7.529 9.056 9.523 10.468 qs_ot_get_derivative 99 11.5 0.001 0.001 10.358 10.458 prepare_preconditioner 11 7.9 0.000 0.000 9.982 9.996 make_preconditioner 11 8.9 0.000 0.002 9.982 9.996 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.195 9.630 init_scf_run 11 5.9 0.000 0.001 8.085 8.086 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.085 8.085 sum_up_and_integrate 110 10.3 0.067 0.071 6.651 6.663 integrate_v_rspace 110 11.3 0.003 0.003 6.583 6.600 qs_ot_get_p 110 10.4 0.001 0.005 6.245 6.370 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.321 6.330 calculate_rho_elec 110 8.6 0.078 0.081 6.321 6.329 mp_waitall_1 121746 16.5 4.419 6.209 4.419 6.209 make_m2s 4110 13.4 0.060 0.061 5.416 5.652 make_images 4110 14.4 0.575 0.696 5.276 5.509 calculate_dm_sparse 110 9.5 0.001 0.001 5.436 5.488 calculate_first_density_matrix 1 7.0 0.023 0.043 4.971 4.973 cp_fm_upper_to_full 70 14.2 3.319 4.803 3.319 4.803 qs_ot_p2m_diag 48 11.0 0.054 0.064 4.418 4.434 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.038 4.137 apply_single 110 13.6 0.000 0.000 4.038 4.137 ot_diis_step 99 11.5 0.011 0.011 4.127 4.127 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.044 4.083 pw_transfer 1331 11.6 0.065 0.076 4.032 4.066 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.926 3.963 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.944 3.945 dbcsr_complete_redistribute 325 12.2 0.419 0.467 2.723 3.843 density_rs2pw 110 9.6 0.004 0.004 3.232 3.748 grid_integrate_task_list 110 12.3 3.262 3.424 3.262 3.424 fft_wrap_pw1pw2_140 451 13.1 0.203 0.215 3.362 3.401 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.348 3.357 cp_fm_diag_elpa_base 48 14.0 3.196 3.258 3.345 3.355 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.298 3.345 multiply_cannon_sync_h2d 24660 15.4 3.183 3.328 3.183 3.328 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.257 3.309 jit_kernel_multiply 11 15.8 1.665 3.298 1.665 3.298 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.172 3.293 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.544 3.286 fft3d_ps 1111 14.6 1.094 1.127 3.218 3.241 mp_alltoall_i22 605 13.7 1.848 3.054 1.848 3.054 rs_pw_transfer 902 11.9 0.010 0.011 2.513 3.049 make_images_data 4110 15.4 0.046 0.050 2.741 3.033 wfi_extrapolate 11 7.9 0.001 0.001 3.013 3.013 hybrid_alltoall_any 4261 16.3 0.119 0.454 2.363 2.950 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.778 2.860 acc_transpose_blocks 24660 15.4 0.106 0.109 2.809 2.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.826 2.828 cp_fm_cholesky_invert 11 10.9 2.791 2.799 2.791 2.799 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.463 2.494 grid_collocate_task_list 110 9.6 2.183 2.442 2.183 2.442 potential_pw2rs 110 12.3 0.012 0.013 2.289 2.302 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.291 2.292 acc_transpose_blocks_kernels 24660 16.4 0.301 0.306 2.167 2.190 mp_alltoall_d11v 2046 13.8 1.830 2.121 1.830 2.121 jit_kernel_transpose 5 15.6 1.866 1.887 1.866 1.887 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.706 1.739 cp_fm_cholesky_decompose 22 10.9 1.692 1.732 1.692 1.732 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.703 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.686 1.696 mp_waitany 13376 13.8 1.082 1.602 1.082 1.602 mp_sum_l 6514 12.8 0.905 1.539 0.905 1.539 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.837 1.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.751000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.363636, yerr=8.488203 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 840.048640E+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 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.027 0.052 57.381 57.388 qs_mol_dyn_low 1 2.0 0.003 0.004 57.015 57.078 qs_forces 11 3.9 0.003 0.003 56.939 56.940 qs_energies 11 4.9 0.001 0.002 52.899 52.904 scf_env_do_scf 11 5.9 0.002 0.002 44.315 44.316 scf_env_do_scf_inner_loop 99 6.5 0.005 0.008 36.320 36.321 velocity_verlet 10 3.0 0.001 0.002 31.764 31.768 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.393 23.569 qs_scf_new_mos 99 7.5 0.001 0.001 20.898 20.972 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.898 20.972 ot_scf_mini 99 9.5 0.002 0.003 19.656 19.718 multiply_cannon 2055 13.4 0.248 0.267 17.895 19.038 multiply_cannon_loop 2055 14.4 0.318 0.331 16.544 16.780 rebuild_ks_matrix 110 8.3 0.000 0.000 12.437 12.479 qs_ks_build_kohn_sham_matrix 110 9.3 0.326 0.328 12.437 12.479 ot_mini 99 10.5 0.001 0.001 10.783 10.839 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.766 10.804 multiply_cannon_multrec 8220 15.4 3.212 4.713 7.646 8.811 init_scf_loop 11 6.9 0.000 0.000 7.945 7.946 mp_waitall_1 103326 16.6 6.118 7.887 6.118 7.887 qs_ot_get_derivative 99 11.5 0.001 0.001 6.930 6.989 sum_up_and_integrate 110 10.3 0.080 0.081 6.630 6.642 integrate_v_rspace 110 11.3 0.003 0.003 6.551 6.562 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.499 6.513 calculate_rho_elec 110 8.6 0.116 0.116 6.499 6.512 prepare_preconditioner 11 7.9 0.000 0.000 6.259 6.264 make_preconditioner 11 8.9 0.000 0.000 6.259 6.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.832 5.905 init_scf_run 11 5.9 0.000 0.001 5.388 5.388 scf_env_initial_rho_setup 11 6.9 0.003 0.004 5.388 5.388 dbcsr_mm_accdrv_process 17442 15.9 2.821 3.919 4.303 5.256 qs_ot_get_p 110 10.4 0.001 0.001 4.942 5.000 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.146 4.608 make_m2s 4110 13.4 0.039 0.040 4.229 4.497 make_images 4110 14.4 0.639 0.693 4.100 4.371 pw_transfer 1331 11.6 0.067 0.074 4.220 4.239 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.112 4.134 ot_diis_step 99 11.5 0.012 0.012 3.832 3.832 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.744 3.786 apply_single 110 13.6 0.000 0.000 3.744 3.786 density_rs2pw 110 9.6 0.004 0.004 3.212 3.589 fft_wrap_pw1pw2_140 451 13.1 0.215 0.219 3.554 3.584 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.563 3.567 grid_integrate_task_list 110 12.3 3.356 3.535 3.356 3.535 fft3d_ps 1111 14.6 1.149 1.209 3.357 3.375 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.258 3.258 multiply_cannon_sync_h2d 8220 15.4 2.898 3.072 2.898 3.072 cp_fm_cholesky_invert 11 10.9 3.053 3.057 3.053 3.057 make_images_data 4110 15.4 0.038 0.043 2.406 2.851 wfi_extrapolate 11 7.9 0.001 0.001 2.759 2.759 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.717 2.725 cp_fm_diag_elpa_base 48 14.0 2.660 2.687 2.715 2.723 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.322 2.713 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.697 2.699 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 2.694 2.695 rs_pw_transfer 902 11.9 0.010 0.010 2.244 2.645 calculate_dm_sparse 110 9.5 0.001 0.001 2.572 2.621 grid_collocate_task_list 110 9.6 2.278 2.541 2.278 2.541 calculate_first_density_matrix 1 7.0 0.001 0.001 2.529 2.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.315 2.328 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.205 2.234 potential_pw2rs 110 12.3 0.015 0.015 2.164 2.168 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.005 2.009 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.997 mp_alltoall_d11v 2046 13.8 1.686 1.991 1.686 1.991 jit_kernel_multiply 10 15.7 1.176 1.966 1.176 1.966 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.821 1.853 cp_fm_cholesky_decompose 22 10.9 1.743 1.756 1.743 1.756 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.526 1.649 mp_allgather_i34 2055 14.4 0.538 1.640 0.538 1.640 dbcsr_complete_redistribute 325 12.2 0.579 0.614 1.523 1.621 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.552 mp_waitany 9240 13.8 1.073 1.503 1.073 1.503 qs_create_task_list 11 7.9 0.001 0.001 1.229 1.331 generate_qs_task_list 11 8.9 0.379 0.447 1.228 1.331 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.480 1.291 mp_irecv_dv 24056 15.7 0.456 1.248 0.456 1.248 acc_transpose_blocks 8220 15.4 0.035 0.037 1.238 1.247 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.167 0.806 1.214 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.162 1.194 rs_gather_matrices 110 12.3 0.327 0.369 0.941 1.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.388000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=786.272727, yerr=13.612100 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 1.377821E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.061 91.157 91.159 qs_mol_dyn_low 1 2.0 0.003 0.004 90.148 90.315 qs_forces 11 3.9 0.007 0.019 90.046 90.047 qs_energies 11 4.9 0.002 0.002 85.802 85.806 scf_env_do_scf 11 5.9 0.001 0.001 75.112 75.112 velocity_verlet 10 3.0 0.002 0.002 57.282 57.288 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.813 45.815 dbcsr_multiply_generic 2055 12.4 0.131 0.134 29.136 29.254 init_scf_loop 11 6.9 0.000 0.000 29.225 29.227 prepare_preconditioner 11 7.9 0.000 0.000 27.171 27.181 make_preconditioner 11 8.9 0.000 0.000 27.170 27.181 qs_scf_new_mos 99 7.5 0.001 0.001 26.963 27.041 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.962 27.040 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.278 26.647 ot_scf_mini 99 9.5 0.002 0.002 25.182 25.251 multiply_cannon 2055 13.4 0.343 0.368 21.996 22.792 multiply_cannon_loop 2055 14.4 0.339 0.346 20.170 20.434 cp_fm_upper_to_full 70 14.2 13.132 18.752 13.132 18.752 rebuild_ks_matrix 110 8.3 0.001 0.001 14.509 14.586 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.016 14.509 14.586 ot_mini 99 10.5 0.001 0.001 14.106 14.168 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.149 13.219 dbcsr_complete_redistribute 325 12.2 1.015 1.052 7.506 10.910 multiply_cannon_multrec 8220 15.4 4.603 4.809 9.907 10.035 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.453 9.851 qs_ot_get_derivative 99 11.5 0.001 0.001 9.429 9.497 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.876 9.232 mp_alltoall_i22 605 13.7 5.512 8.933 5.512 8.933 mp_waitall_1 84994 16.7 7.722 8.760 7.722 8.760 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.977 8.013 calculate_rho_elec 110 8.6 0.227 0.228 7.976 8.013 sum_up_and_integrate 110 10.3 0.150 0.151 7.628 7.644 integrate_v_rspace 110 11.3 0.004 0.004 7.477 7.494 qs_ot_get_p 110 10.4 0.001 0.001 5.868 5.937 make_m2s 4110 13.4 0.043 0.044 5.279 5.809 init_scf_run 11 5.9 0.000 0.001 5.792 5.793 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.792 5.792 make_images 4110 14.4 0.881 0.934 5.091 5.621 cp_fm_cholesky_invert 11 10.9 5.566 5.571 5.566 5.571 pw_transfer 1331 11.6 0.075 0.076 5.433 5.444 dbcsr_mm_accdrv_process 11614 15.7 3.439 3.705 5.161 5.397 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.800 5.347 apply_single 110 13.6 0.000 0.000 4.800 5.346 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.316 5.327 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.917 5.265 ot_diis_step 99 11.5 0.016 0.016 4.658 4.658 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.613 4.619 fft3d_ps 1111 14.6 1.298 1.307 4.509 4.524 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.243 4.249 density_rs2pw 110 9.6 0.004 0.004 3.945 3.998 qs_energies_init_hamiltonians 11 5.9 0.020 0.021 3.989 3.993 multiply_cannon_sync_h2d 8220 15.4 3.953 3.961 3.953 3.961 grid_integrate_task_list 110 12.3 3.720 3.855 3.720 3.855 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.793 3.794 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.888 3.628 make_images_data 4110 15.4 0.042 0.045 2.892 3.531 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.996 3.485 wfi_extrapolate 11 7.9 0.001 0.001 3.355 3.355 calculate_dm_sparse 110 9.5 0.001 0.001 3.205 3.225 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.166 3.167 cp_fm_diag_elpa_base 48 14.0 2.616 2.819 3.163 3.164 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.881 2.887 grid_collocate_task_list 110 9.6 2.678 2.741 2.678 2.741 potential_pw2rs 110 12.3 0.021 0.021 2.596 2.602 rs_pw_transfer 902 11.9 0.011 0.011 2.509 2.549 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.480 2.548 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.399 2.435 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.377 2.399 calculate_first_density_matrix 1 7.0 0.001 0.001 2.314 2.314 mp_alltoall_d11v 2046 13.8 2.078 2.223 2.078 2.223 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.110 2.213 cp_fm_cholesky_decompose 22 10.9 2.057 2.078 2.057 2.078 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.964 1.972 qs_create_task_list 11 7.9 0.016 0.016 1.917 1.966 generate_qs_task_list 11 8.9 0.736 0.791 1.901 1.950 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.803 1.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.159000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1216.272727, yerr=64.550320 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 630.403072E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 57780. MP_Allreduce 11002 802. MP_Sync 86 MP_Alltoall 2210 2520169. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.053 210.943 210.945 qs_mol_dyn_low 1 2.0 0.004 0.012 210.474 210.488 qs_forces 11 3.9 0.005 0.005 210.373 210.374 qs_energies 11 4.9 0.002 0.003 204.695 204.706 scf_env_do_scf 11 5.9 0.001 0.001 187.209 187.213 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 166.278 166.280 qs_scf_new_mos 116 7.6 0.001 0.001 125.585 125.866 qs_scf_loop_do_ot 116 8.6 0.001 0.001 125.584 125.865 velocity_verlet 10 3.0 0.001 0.001 125.060 125.062 dbcsr_multiply_generic 2485 12.5 0.174 0.181 124.347 124.813 ot_scf_mini 116 9.6 0.003 0.003 119.041 119.311 multiply_cannon 2485 13.5 0.235 0.243 100.812 102.046 multiply_cannon_loop 2485 14.5 2.033 2.085 98.628 99.918 ot_mini 116 10.6 0.001 0.001 66.047 66.331 multiply_cannon_multrec 59640 15.5 33.053 34.840 41.538 43.569 qs_ot_get_derivative 116 11.6 0.001 0.001 41.466 41.736 rebuild_ks_matrix 127 8.3 0.001 0.001 34.053 34.343 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 34.052 34.342 mp_waitall_1 264810 16.5 28.819 31.527 28.819 31.527 qs_ot_get_p 127 10.4 0.001 0.001 31.220 31.514 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.561 30.858 multiply_cannon_sync_h2d 59640 15.5 27.199 28.802 27.199 28.802 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.117 24.785 apply_single 127 13.6 0.001 0.001 24.117 24.785 qs_ot_p2m_diag 82 11.4 0.078 0.091 24.595 24.664 ot_diis_step 116 11.6 0.007 0.008 24.333 24.334 cp_dbcsr_syevd 82 12.4 0.004 0.005 22.073 22.074 init_scf_loop 11 6.9 0.000 0.001 20.848 20.849 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.768 19.966 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.008 19.038 cp_fm_diag_elpa_base 82 14.4 18.925 18.962 19.004 19.032 multiply_cannon_metrocomm3 59640 15.5 0.111 0.115 15.915 18.358 prepare_preconditioner 11 7.9 0.000 0.000 16.151 16.195 make_preconditioner 11 8.9 0.000 0.000 16.151 16.195 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.385 15.577 sum_up_and_integrate 127 10.3 0.088 0.106 14.640 14.655 integrate_v_rspace 127 11.3 0.004 0.005 14.551 14.568 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.191 14.320 calculate_rho_elec 127 8.7 0.045 0.065 14.191 14.319 make_m2s 4970 13.5 0.104 0.111 13.776 14.179 make_images 4970 14.5 0.396 0.414 13.597 14.006 init_scf_run 11 5.9 0.000 0.001 13.301 13.302 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.301 13.302 density_rs2pw 127 9.7 0.006 0.007 7.541 10.617 mp_sum_l 7804 13.0 8.525 9.544 8.525 9.544 rs_pw_transfer 1038 11.9 0.017 0.019 6.185 9.321 cp_fm_cholesky_invert 11 10.9 9.291 9.299 9.291 9.299 wfi_extrapolate 11 7.9 0.001 0.001 9.272 9.272 calculate_dm_sparse 127 9.5 0.001 0.001 8.655 8.774 dbcsr_mm_accdrv_process 123452 16.2 3.134 3.307 8.057 8.714 pw_transfer 1535 11.6 0.077 0.092 8.262 8.518 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 8.056 8.319 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.121 8.279 multiply_cannon_metrocomm1 59640 15.5 0.089 0.092 6.188 7.910 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.781 7.870 make_images_data 4970 15.5 0.066 0.072 6.733 7.698 grid_integrate_task_list 127 12.3 6.993 7.498 6.993 7.498 fft_wrap_pw1pw2_140 519 13.2 0.439 0.505 6.894 7.204 hybrid_alltoall_any 5155 16.4 0.290 2.249 5.908 7.148 fft3d_ps 1281 14.7 2.089 2.868 6.854 7.033 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.750 6.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.876 5.972 mp_alltoall_d11v 2401 14.1 4.396 5.912 4.396 5.912 mp_waitany 15900 13.9 2.691 5.854 2.691 5.854 grid_collocate_task_list 127 9.7 4.525 5.740 4.525 5.740 rs_pw_transfer_RS2PW_140 138 11.5 0.276 0.294 2.103 5.212 potential_pw2rs 127 12.3 0.009 0.011 5.034 5.073 cp_fm_cholesky_decompose 22 10.9 4.674 4.690 4.674 4.690 mp_sum_d 4447 12.1 3.490 4.233 3.490 4.233 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.945000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.090909, yerr=7.153569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 830.390272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58220. MP_Allreduce 11003 967. MP_Sync 87 MP_Alltoall 1969 5618843. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.069 192.625 192.628 qs_mol_dyn_low 1 2.0 0.003 0.004 191.439 192.226 qs_forces 11 3.9 0.005 0.005 190.844 190.846 qs_energies 11 4.9 0.003 0.008 183.809 183.820 scf_env_do_scf 11 5.9 0.001 0.002 167.385 167.395 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 134.495 134.497 velocity_verlet 10 3.0 0.001 0.002 120.274 120.276 dbcsr_multiply_generic 2507 12.6 0.185 0.190 97.138 98.243 qs_scf_new_mos 117 7.6 0.001 0.001 93.891 94.339 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.890 94.339 ot_scf_mini 117 9.6 0.004 0.005 89.099 89.676 multiply_cannon 2507 13.6 0.479 0.530 77.321 81.679 multiply_cannon_loop 2507 14.6 1.246 1.281 73.999 76.376 ot_mini 117 10.6 0.001 0.001 49.497 50.022 mp_waitall_1 214728 16.6 24.330 38.952 24.330 38.952 multiply_cannon_multrec 30084 15.6 22.065 26.080 31.693 35.927 rebuild_ks_matrix 128 8.3 0.001 0.001 33.909 34.376 qs_ks_build_kohn_sham_matrix 128 9.3 0.103 0.190 33.909 34.376 init_scf_loop 11 6.9 0.001 0.003 32.800 32.802 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.372 30.802 multiply_cannon_metrocomm3 30084 15.6 0.093 0.098 15.552 28.706 prepare_preconditioner 11 7.9 0.000 0.000 28.320 28.380 make_preconditioner 11 8.9 0.000 0.002 28.320 28.380 qs_ot_get_derivative 117 11.6 0.001 0.002 27.658 28.197 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.001 27.519 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.942 22.979 apply_single 128 13.6 0.001 0.001 21.941 22.978 qs_ot_get_p 128 10.4 0.001 0.001 21.373 21.955 ot_diis_step 117 11.6 0.014 0.015 21.668 21.671 multiply_cannon_sync_h2d 30084 15.6 19.338 21.531 19.338 21.531 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.567 16.598 cp_fm_cholesky_invert 11 10.9 16.546 16.560 16.546 16.560 make_m2s 5014 13.6 0.091 0.096 14.224 15.898 make_images 5014 14.6 1.149 1.342 14.014 15.687 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.426 15.427 sum_up_and_integrate 128 10.3 0.117 0.134 15.000 15.033 integrate_v_rspace 128 11.3 0.004 0.004 14.883 14.921 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.342 14.370 calculate_rho_elec 128 8.7 0.088 0.106 14.341 14.370 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.263 12.294 cp_fm_diag_elpa_base 83 14.4 12.004 12.087 12.256 12.285 init_scf_run 11 5.9 0.000 0.001 11.572 11.573 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.572 11.573 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.876 11.278 multiply_cannon_metrocomm4 27577 15.6 0.098 0.111 3.738 10.666 make_images_data 5014 15.6 0.064 0.071 8.641 10.642 mp_irecv_dv 69486 16.3 3.540 10.271 3.540 10.271 hybrid_alltoall_any 5200 16.5 0.342 1.499 7.354 10.199 density_rs2pw 128 9.7 0.006 0.006 7.700 10.054 dbcsr_mm_accdrv_process 62242 16.2 4.602 5.394 9.078 9.655 pw_transfer 1547 11.6 0.088 0.111 9.225 9.301 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.997 9.077 wfi_extrapolate 11 7.9 0.001 0.001 8.398 8.398 rs_pw_transfer 1046 11.9 0.014 0.017 5.969 8.244 fft_wrap_pw1pw2_140 523 13.2 0.472 0.531 7.884 7.988 grid_integrate_task_list 128 12.3 7.170 7.642 7.170 7.642 fft3d_ps 1291 14.7 2.755 2.928 7.366 7.408 cp_fm_cholesky_decompose 22 10.9 6.880 6.969 6.880 6.969 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.166 6.899 calculate_dm_sparse 128 9.5 0.001 0.001 6.504 6.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.079 6.090 grid_collocate_task_list 128 9.7 4.703 5.890 4.703 5.890 mp_sum_l 7870 13.0 3.812 5.615 3.812 5.615 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.257 5.397 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.309 5.383 potential_pw2rs 128 12.3 0.015 0.018 5.224 5.245 mp_alltoall_d11v 2415 14.1 4.156 4.950 4.156 4.950 mp_allgather_i34 2507 14.6 1.898 4.927 1.898 4.927 mp_waitany 11748 13.9 2.605 4.857 2.605 4.857 rs_pw_transfer_RS2PW_140 139 11.5 0.350 0.378 2.121 4.454 dbcsr_complete_redistribute 395 12.7 0.766 0.856 3.158 4.006 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.945 3.984 mp_sum_d 4466 12.1 2.651 3.961 2.651 3.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.628000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.000000, yerr=2.132007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 943.910912E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58312. MP_Allreduce 10985 1007. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.089 178.630 178.653 qs_mol_dyn_low 1 2.0 0.004 0.005 178.111 178.126 qs_forces 11 3.9 0.009 0.046 177.164 177.170 qs_energies 11 4.9 0.044 0.128 170.377 170.393 scf_env_do_scf 11 5.9 0.044 0.100 153.798 153.799 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 118.343 118.345 velocity_verlet 10 3.0 0.002 0.002 113.633 113.663 dbcsr_multiply_generic 2485 12.5 0.180 0.186 81.493 82.568 qs_scf_new_mos 116 7.6 0.001 0.001 79.733 80.051 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.733 80.050 ot_scf_mini 116 9.6 0.004 0.005 75.638 76.022 multiply_cannon 2485 13.5 0.495 0.519 61.455 66.614 multiply_cannon_loop 2485 14.5 0.840 0.867 58.302 61.075 ot_mini 116 10.6 0.001 0.001 42.276 42.675 init_scf_loop 11 6.9 0.002 0.007 35.258 35.261 mp_waitall_1 169034 16.6 24.963 34.752 24.963 34.752 rebuild_ks_matrix 127 8.3 0.001 0.001 31.344 31.792 qs_ks_build_kohn_sham_matrix 127 9.3 0.187 0.293 31.344 31.792 prepare_preconditioner 11 7.9 0.000 0.000 30.877 30.932 make_preconditioner 11 8.9 0.001 0.002 30.877 30.932 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.423 29.927 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.185 28.600 multiply_cannon_multrec 19880 15.5 13.468 16.162 22.300 25.080 multiply_cannon_metrocomm3 19880 15.5 0.059 0.063 14.904 24.454 qs_ot_get_derivative 116 11.6 0.001 0.002 22.785 23.156 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.668 20.658 apply_single 127 13.6 0.001 0.001 19.668 20.657 ot_diis_step 116 11.6 0.018 0.018 19.392 19.392 qs_ot_get_p 127 10.4 0.001 0.002 18.823 19.251 make_m2s 4970 13.5 0.081 0.087 14.913 15.996 multiply_cannon_sync_h2d 19880 15.5 14.166 15.772 14.166 15.772 make_images 4970 14.5 1.160 1.262 14.682 15.755 sum_up_and_integrate 127 10.3 0.132 0.145 14.861 14.900 integrate_v_rspace 127 11.3 0.004 0.005 14.729 14.768 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.567 14.602 calculate_rho_elec 127 8.7 0.131 0.146 14.567 14.602 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.503 14.511 cp_fm_cholesky_invert 11 10.9 14.326 14.335 14.326 14.335 cp_dbcsr_syevd 82 12.4 0.006 0.008 13.512 13.514 make_images_data 4970 15.5 0.060 0.068 9.308 10.835 init_scf_run 11 5.9 0.000 0.001 10.790 10.790 scf_env_initial_rho_setup 11 6.9 0.021 0.022 10.790 10.790 cp_fm_diag_elpa 82 13.4 0.001 0.002 10.353 10.369 cp_fm_diag_elpa_base 82 14.4 9.937 10.098 10.346 10.363 hybrid_alltoall_any 5155 16.4 0.430 1.972 8.027 10.063 pw_transfer 1535 11.6 0.086 0.108 9.368 9.486 density_rs2pw 127 9.7 0.006 0.006 7.402 9.404 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.144 9.270 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.872 9.153 multiply_cannon_metrocomm4 17395 15.5 0.062 0.072 3.400 9.060 mp_irecv_dv 49801 16.2 3.277 8.816 3.277 8.816 dbcsr_mm_accdrv_process 41158 16.2 4.686 5.464 8.278 8.396 fft_wrap_pw1pw2_140 519 13.2 0.473 0.517 8.065 8.195 grid_integrate_task_list 127 12.3 7.216 7.711 7.216 7.711 fft3d_ps 1281 14.7 2.652 2.873 7.422 7.505 wfi_extrapolate 11 7.9 0.001 0.002 7.457 7.457 cp_fm_upper_to_full 104 14.8 5.809 7.429 5.809 7.429 cp_fm_cholesky_decompose 22 10.9 7.265 7.316 7.265 7.316 rs_pw_transfer 1038 11.9 0.014 0.014 5.224 7.260 dbcsr_complete_redistribute 393 12.7 1.164 1.194 4.717 6.516 calculate_dm_sparse 127 9.5 0.001 0.002 5.922 6.036 grid_collocate_task_list 127 9.7 4.872 5.749 4.872 5.749 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.401 5.407 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.536 5.337 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.580 5.289 mp_alltoall_d11v 2401 14.1 4.502 5.062 4.502 5.062 mp_sum_l 7804 13.0 3.326 5.053 3.326 5.053 potential_pw2rs 127 12.3 0.020 0.022 4.975 5.006 mp_allgather_i34 2485 14.5 1.695 4.785 1.695 4.785 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.534 4.654 mp_waitany 11660 13.9 2.358 4.445 2.358 4.445 qs_energies_init_hamiltonians 11 5.9 0.402 0.446 4.256 4.258 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.434 4.197 mp_alltoall_i22 712 14.1 2.098 3.993 2.098 3.993 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.949 3.987 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.352 1.938 3.966 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.669 3.707 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.653000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.181818, yerr=9.083633 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 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 1.136845E+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 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58281. MP_Allreduce 10990 1091. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 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 0.022 0.054 188.145 188.146 qs_mol_dyn_low 1 2.0 0.003 0.004 187.709 187.723 qs_forces 11 3.9 0.004 0.006 187.594 187.600 qs_energies 11 4.9 0.007 0.042 180.411 180.423 scf_env_do_scf 11 5.9 0.045 0.349 163.218 163.226 velocity_verlet 10 3.0 0.001 0.002 124.133 124.136 scf_env_do_scf_inner_loop 116 6.6 0.013 0.034 116.506 116.507 dbcsr_multiply_generic 2485 12.5 0.187 0.192 78.818 79.622 qs_scf_new_mos 116 7.6 0.001 0.001 78.879 79.187 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.878 79.186 ot_scf_mini 116 9.6 0.003 0.005 74.437 74.744 multiply_cannon 2485 13.5 0.553 0.587 54.396 57.594 multiply_cannon_loop 2485 14.5 1.160 1.184 50.755 52.617 init_scf_loop 11 6.9 0.001 0.003 46.243 46.244 ot_mini 116 10.6 0.001 0.001 41.933 42.232 prepare_preconditioner 11 7.9 0.000 0.000 42.075 42.103 make_preconditioner 11 8.9 0.000 0.002 42.075 42.103 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.659 40.636 multiply_cannon_multrec 29820 15.5 14.069 19.162 26.079 30.621 rebuild_ks_matrix 127 8.3 0.001 0.001 29.971 30.265 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.020 29.970 30.264 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.989 27.247 mp_waitall_1 146592 16.7 17.223 27.180 17.223 27.180 qs_ot_get_derivative 116 11.6 0.001 0.002 22.602 22.909 make_m2s 4970 13.5 0.097 0.100 20.057 21.143 make_images 4970 14.5 1.909 2.258 19.751 20.840 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.762 19.301 apply_single 127 13.6 0.001 0.001 18.762 19.301 qs_ot_get_p 127 10.4 0.001 0.001 18.955 19.261 ot_diis_step 116 11.6 0.017 0.018 19.206 19.208 cp_fm_upper_to_full 104 14.8 11.296 16.661 11.296 16.661 cp_fm_cholesky_invert 11 10.9 16.097 16.106 16.097 16.106 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.337 15.347 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.837 14.888 sum_up_and_integrate 127 10.3 0.140 0.152 14.814 14.838 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.754 14.792 calculate_rho_elec 127 8.7 0.174 0.189 14.753 14.791 integrate_v_rspace 127 11.3 0.004 0.005 14.674 14.701 cp_dbcsr_syevd 82 12.4 0.005 0.007 13.582 13.583 dbcsr_complete_redistribute 393 12.7 1.500 1.618 9.006 12.753 multiply_cannon_sync_h2d 29820 15.5 11.641 12.632 11.641 12.632 make_images_data 4970 15.5 0.064 0.069 10.842 12.406 dbcsr_mm_accdrv_process 61748 16.2 7.454 8.482 11.587 12.112 hybrid_alltoall_any 5155 16.4 0.520 2.183 9.629 11.505 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.619 11.367 init_scf_run 11 5.9 0.000 0.001 11.144 11.145 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.144 11.145 cp_fm_diag_elpa 82 13.4 0.001 0.001 10.543 10.553 cp_fm_diag_elpa_base 82 14.4 9.590 9.907 10.535 10.547 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.393 10.061 mp_alltoall_i22 712 14.1 5.892 9.800 5.892 9.800 pw_transfer 1535 11.6 0.088 0.108 9.555 9.623 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.313 9.534 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.328 9.402 density_rs2pw 127 9.7 0.006 0.006 7.220 8.509 fft_wrap_pw1pw2_140 519 13.2 0.481 0.492 8.265 8.358 grid_integrate_task_list 127 12.3 7.457 7.936 7.457 7.936 wfi_extrapolate 11 7.9 0.003 0.021 7.743 7.743 cp_fm_cholesky_decompose 22 10.9 7.547 7.628 7.547 7.628 fft3d_ps 1281 14.7 2.758 2.845 7.567 7.627 multiply_cannon_metrocomm4 24850 15.5 0.076 0.087 2.729 7.152 mp_irecv_dv 75445 16.2 2.583 6.890 2.583 6.890 calculate_dm_sparse 127 9.5 0.001 0.001 6.326 6.413 rs_pw_transfer 1038 11.9 0.013 0.014 4.693 5.996 mp_alltoall_d11v 2401 14.1 4.928 5.823 4.928 5.823 grid_collocate_task_list 127 9.7 5.040 5.757 5.040 5.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.464 5.517 potential_pw2rs 127 12.3 0.022 0.023 4.657 4.673 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.410 4.505 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.478 4.480 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.382 4.474 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.106 4.146 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.146000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1066.727273, yerr=23.433042 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/19/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 5.820057E+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 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.525187E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58308. MP_Allreduce 10985 1176. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.035 177.038 177.039 qs_mol_dyn_low 1 2.0 0.003 0.004 176.193 176.208 qs_forces 11 3.9 0.004 0.004 176.080 176.084 qs_energies 11 4.9 0.002 0.002 168.445 168.453 scf_env_do_scf 11 5.9 0.073 0.085 151.051 151.059 velocity_verlet 10 3.0 0.002 0.002 115.137 115.141 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 114.684 114.685 qs_scf_new_mos 116 7.6 0.001 0.001 75.700 75.808 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.699 75.807 dbcsr_multiply_generic 2485 12.5 0.180 0.186 74.680 75.131 ot_scf_mini 116 9.6 0.003 0.004 71.304 71.396 multiply_cannon 2485 13.5 0.577 0.604 54.762 59.109 multiply_cannon_loop 2485 14.5 0.438 0.447 50.133 51.031 ot_mini 116 10.6 0.001 0.001 40.001 40.111 init_scf_loop 11 6.9 0.000 0.000 36.141 36.213 mp_waitall_1 124680 16.7 26.575 33.693 26.575 33.693 prepare_preconditioner 11 7.9 0.000 0.000 32.115 32.146 make_preconditioner 11 8.9 0.000 0.000 32.114 32.146 rebuild_ks_matrix 127 8.3 0.001 0.001 30.201 30.317 qs_ks_build_kohn_sham_matrix 127 9.3 0.022 0.025 30.201 30.317 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.876 30.134 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.392 27.496 multiply_cannon_multrec 9940 15.5 10.328 14.463 17.727 21.432 qs_ot_get_derivative 116 11.6 0.001 0.002 20.118 20.223 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.873 19.911 ot_diis_step 116 11.6 0.019 0.020 19.808 19.808 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.432 19.697 apply_single 127 13.6 0.001 0.001 19.431 19.697 make_m2s 4970 13.5 0.066 0.070 16.277 18.684 qs_ot_get_p 127 10.4 0.001 0.001 18.349 18.459 make_images 4970 14.5 2.252 2.622 15.975 18.372 cp_fm_cholesky_invert 11 10.9 18.341 18.348 18.341 18.348 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.555 15.601 calculate_rho_elec 127 8.7 0.257 0.267 15.554 15.600 sum_up_and_integrate 127 10.3 0.180 0.191 15.200 15.247 integrate_v_rspace 127 11.3 0.004 0.005 15.019 15.076 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.524 14.538 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.392 13.394 make_images_data 4970 15.5 0.051 0.060 9.980 12.728 hybrid_alltoall_any 5155 16.4 0.806 3.774 9.773 12.099 multiply_cannon_sync_h2d 9940 15.5 11.452 12.007 11.452 12.007 init_scf_run 11 5.9 0.000 0.001 10.561 10.561 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.561 10.561 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.367 10.377 cp_fm_diag_elpa_base 82 14.4 10.131 10.210 10.363 10.373 pw_transfer 1535 11.6 0.085 0.093 10.058 10.093 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.836 9.875 fft_wrap_pw1pw2_140 519 13.2 0.492 0.515 8.650 8.694 grid_integrate_task_list 127 12.3 7.669 8.243 7.669 8.243 cp_fm_cholesky_decompose 22 10.9 7.975 8.068 7.975 8.068 fft3d_ps 1281 14.7 2.688 2.775 8.015 8.036 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.949 8.030 density_rs2pw 127 9.7 0.005 0.005 7.160 8.024 dbcsr_mm_accdrv_process 20590 16.1 2.779 3.619 7.036 7.628 multiply_cannon_metrocomm1 9940 15.5 0.028 0.028 4.425 7.567 wfi_extrapolate 11 7.9 0.001 0.001 7.540 7.541 mp_allgather_i34 2485 14.5 2.855 7.454 2.855 7.454 mp_alltoall_d11v 2401 14.1 5.288 6.689 5.288 6.689 calculate_dm_sparse 127 9.5 0.001 0.001 6.153 6.238 grid_collocate_task_list 127 9.7 5.350 5.999 5.350 5.999 dbcsr_complete_redistribute 393 12.7 2.112 2.169 5.377 5.753 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 5.311 5.311 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.198 5.209 rs_pw_transfer 1038 11.9 0.012 0.013 4.211 5.111 potential_pw2rs 127 12.3 0.026 0.026 4.698 4.711 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.841 4.665 mp_irecv_dv 28618 15.9 1.805 4.582 1.805 4.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.100 4.129 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.569 3.865 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.520 3.843 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.749 3.775 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.632 3.767 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.724 3.747 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.710 3.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.039000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1429.909091, yerr=55.164967 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024941568 0.0% 0.0% 100.0% flops 32 x 32 x 32 1939177734144 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957880770560 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544039424 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 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 11444715307008 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019205160960 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019205160960 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.811684E+12 0.0% 0.0% 100.0% flops max/rank 11.608293E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705736096 0.0% 0.0% 100.0% number of processed stacks 1950384 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.2 marketing flops 143.522775E+12 ------------------------------------------------------------------------------- # multiplications 2492 max memory usage/rank 3.036537E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99680 MPI messages size (bytes): total size 1.127657E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.312775E+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 45048 34980495360 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716109313056 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4025 58814. MP_Allreduce 11059 1513. MP_Sync 88 MP_Alltoall 1700 36954294. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 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 0.054 0.078 306.956 306.962 qs_mol_dyn_low 1 2.0 0.003 0.004 306.324 306.331 qs_forces 11 3.9 0.005 0.005 306.178 306.186 qs_energies 11 4.9 0.002 0.002 296.718 296.725 scf_env_do_scf 11 5.9 0.018 0.022 273.480 273.497 velocity_verlet 10 3.0 0.017 0.029 220.470 220.478 scf_env_do_scf_inner_loop 116 6.6 0.021 0.025 144.335 144.337 init_scf_loop 11 6.9 0.000 0.000 128.870 128.873 prepare_preconditioner 11 7.9 0.000 0.000 123.658 123.684 make_preconditioner 11 8.9 0.000 0.000 123.658 123.684 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.974 120.792 qs_scf_new_mos 116 7.6 0.001 0.001 93.309 93.426 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.308 93.425 ot_scf_mini 116 9.6 0.004 0.004 88.473 88.535 dbcsr_multiply_generic 2492 12.5 0.213 0.220 84.792 85.582 cp_fm_upper_to_full 106 14.8 54.405 79.158 54.405 79.158 multiply_cannon 2492 13.5 0.699 0.744 59.731 60.308 multiply_cannon_loop 2492 14.5 0.466 0.479 55.856 57.613 ot_mini 116 10.6 0.001 0.001 45.599 45.674 dbcsr_complete_redistribute 397 12.7 3.988 4.037 31.375 45.498 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.966 42.072 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 25.638 39.588 rebuild_ks_matrix 127 8.3 0.001 0.001 38.660 38.748 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.020 38.659 38.748 mp_alltoall_i22 716 14.1 23.447 37.434 23.447 37.434 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.561 35.648 cp_fm_cholesky_invert 11 10.9 34.252 34.259 34.252 34.259 mp_waitall_1 103048 16.8 29.370 33.587 29.370 33.587 qs_ot_get_p 127 10.4 0.001 0.001 27.302 27.374 qs_ot_get_derivative 116 11.6 0.002 0.002 24.679 24.749 qs_ot_p2m_diag 84 11.4 0.889 0.894 23.095 23.128 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.336 21.339 make_m2s 4984 13.5 0.077 0.078 20.137 21.267 multiply_cannon_metrocomm3 9968 15.5 0.023 0.024 19.620 21.198 ot_diis_step 116 11.6 0.022 0.023 20.863 20.863 make_images 4984 14.5 3.725 3.956 19.661 20.792 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.742 20.758 calculate_rho_elec 127 8.7 0.479 0.480 20.741 20.758 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.052 20.403 apply_single 127 13.6 0.001 0.001 20.051 20.403 sum_up_and_integrate 127 10.3 0.319 0.321 19.729 19.808 integrate_v_rspace 127 11.3 0.004 0.004 19.409 19.487 multiply_cannon_multrec 9968 15.5 10.432 12.277 18.045 18.132 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.043 18.044 cp_fm_diag_elpa_base 84 14.4 13.501 15.242 18.039 18.039 multiply_cannon_sync_h2d 9968 15.5 15.550 15.568 15.550 15.568 hybrid_alltoall_any 5171 16.4 1.291 2.999 11.087 13.107 make_images_data 4984 15.5 0.060 0.065 10.926 12.877 init_scf_run 11 5.9 0.000 0.001 12.757 12.759 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.757 12.758 pw_transfer 1535 11.6 0.092 0.094 12.455 12.465 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 12.221 12.232 fft_wrap_pw1pw2_140 519 13.2 0.538 0.541 10.839 10.854 fft3d_ps 1281 14.7 2.725 2.734 10.273 10.282 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.768 9.809 cp_fm_cholesky_decompose 22 10.9 9.207 9.232 9.207 9.232 mp_alltoall_d11v 2413 14.1 8.354 9.225 8.354 9.225 wfi_extrapolate 11 7.9 0.001 0.001 9.223 9.223 dbcsr_mm_accdrv_process 20618 16.0 3.883 5.675 7.374 9.196 grid_integrate_task_list 127 12.3 8.508 8.674 8.508 8.674 density_rs2pw 127 9.7 0.005 0.005 8.402 8.534 qs_energies_init_hamiltonians 11 5.9 0.003 0.006 8.483 8.484 calculate_dm_sparse 127 9.5 0.001 0.001 6.803 6.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.486 6.571 grid_collocate_task_list 127 9.7 6.309 6.341 6.309 6.341 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.104 6.231 rs_scatter_matrices 138 9.7 3.571 4.460 6.006 6.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=306.962000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2699.181818, yerr=170.802288 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 1.262129E+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 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 85.067 85.068 qs_energies 1 2.0 0.000 0.000 84.586 84.593 ls_scf 1 3.0 0.000 0.000 83.624 83.631 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.452 72.623 multiply_cannon 111 7.7 0.018 0.021 55.766 57.155 multiply_cannon_loop 111 8.7 0.210 0.226 52.374 53.876 ls_scf_main 1 4.0 0.000 0.000 52.241 52.252 density_matrix_trs4 2 5.0 0.002 0.003 46.693 46.763 ls_scf_init_scf 1 4.0 0.000 0.001 28.281 28.293 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.083 27.145 mp_waitall_1 11031 10.9 22.431 25.874 22.431 25.874 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.955 24.969 multiply_cannon_multrec 2664 9.7 8.170 8.968 15.472 17.236 multiply_cannon_sync_h2d 2664 9.7 13.578 15.843 13.578 15.843 make_m2s 222 7.7 0.008 0.011 13.074 13.456 make_images 222 8.7 0.098 0.108 13.052 13.435 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.664 13.037 make_images_data 222 9.7 0.004 0.005 7.623 8.221 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.639 6.921 7.913 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.470 7.864 hybrid_alltoall_any 227 10.6 0.215 1.839 6.613 7.845 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.128 6.208 7.128 calculate_norms 4752 9.8 5.515 6.212 5.515 6.212 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.029 5.163 mp_sum_l 807 5.4 3.081 4.584 3.081 4.584 make_images_sizes 222 9.7 0.000 0.000 0.704 3.689 mp_alltoall_i44 222 10.7 0.703 3.689 0.703 3.689 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.091 3.586 mp_irecv_dv 6231 10.9 2.072 3.561 2.072 3.561 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.270 3.333 arnoldi_extremal 4 6.8 0.000 0.000 3.296 3.319 arnoldi_normal_ev 4 7.8 0.001 0.003 3.296 3.319 build_subspace 16 8.4 0.009 0.012 3.199 3.202 ls_scf_post 1 4.0 0.000 0.000 3.103 3.110 ls_scf_store_result 1 5.0 0.000 0.000 2.884 2.920 dbcsr_special_finalize 555 9.7 0.005 0.006 2.391 2.731 dbcsr_merge_single_wm 555 10.7 0.457 0.597 2.383 2.723 make_images_pack 222 9.7 2.211 2.626 2.213 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.331 2.599 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.482 2.068 2.484 dbcsr_sort_data 658 11.4 2.184 2.454 2.184 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.304 2.389 buffer_matrices_ensure_size 222 8.7 1.753 2.122 1.753 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.867 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.857 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.025 0.045 1.857 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.068000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 2.082738E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.064 91.679 91.680 qs_energies 1 2.0 0.000 0.001 91.065 91.083 ls_scf 1 3.0 0.000 0.000 89.484 89.499 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.260 75.593 multiply_cannon 111 7.7 0.030 0.083 53.464 56.758 ls_scf_main 1 4.0 0.000 0.000 54.766 54.775 multiply_cannon_loop 111 8.7 0.116 0.122 50.179 53.133 density_matrix_trs4 2 5.0 0.002 0.003 48.961 49.119 ls_scf_init_scf 1 4.0 0.000 0.002 31.149 31.150 mp_waitall_1 9105 10.9 21.134 30.200 21.134 30.200 ls_scf_init_matrix_S 1 5.0 0.000 0.006 29.879 29.971 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.444 27.453 multiply_cannon_multrec 1332 9.7 13.141 16.937 22.350 27.352 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.718 21.044 make_m2s 222 7.7 0.006 0.007 15.248 15.891 make_images 222 8.7 1.580 1.927 15.218 15.861 dbcsr_mm_accdrv_process 4041 10.4 0.302 0.476 8.814 10.335 dbcsr_mm_accdrv_process_sort 4041 11.4 8.398 9.912 8.398 9.912 make_images_data 222 9.7 0.004 0.004 8.742 9.620 hybrid_alltoall_any 227 10.6 0.520 2.430 8.248 9.212 mp_sum_l 807 5.4 5.251 7.767 5.251 7.767 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.228 7.569 mp_irecv_dv 3311 11.0 3.209 7.505 3.209 7.505 calculate_norms 2376 9.8 6.009 6.874 6.009 6.874 multiply_cannon_sync_h2d 1332 9.7 4.786 6.139 4.786 6.139 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.037 6.138 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.088 5.338 arnoldi_extremal 4 6.8 0.000 0.000 4.709 4.722 arnoldi_normal_ev 4 7.8 0.001 0.005 4.709 4.722 build_subspace 16 8.4 0.014 0.021 4.453 4.460 ls_scf_post 1 4.0 0.000 0.000 3.570 3.584 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.149 3.394 ls_scf_store_result 1 5.0 0.000 0.000 3.246 3.385 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.221 2.742 3.223 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.300 2.936 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.647 2.763 make_images_pack 222 9.7 2.028 2.382 2.030 2.384 mp_allgather_i34 111 8.7 0.984 2.325 0.984 2.325 dbcsr_sort_data 436 11.2 1.811 2.019 1.811 2.019 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.956 1.958 rebuild_ks_matrix 3 7.3 0.000 0.000 1.942 1.944 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.012 1.942 1.944 dbcsr_data_new 4174 10.1 1.616 1.839 1.616 1.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.680000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 2.694463E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.136 94.371 94.371 qs_energies 1 2.0 0.000 0.000 93.719 93.731 ls_scf 1 3.0 0.000 0.000 92.161 92.175 dbcsr_multiply_generic 111 6.7 0.016 0.018 76.747 77.055 ls_scf_main 1 4.0 0.000 0.000 57.671 57.676 multiply_cannon 111 7.7 0.043 0.130 52.809 57.212 multiply_cannon_loop 111 8.7 0.100 0.110 49.239 52.901 density_matrix_trs4 2 5.0 0.002 0.003 51.669 51.839 mp_waitall_1 7281 11.0 24.098 33.747 24.098 33.747 ls_scf_init_scf 1 4.0 0.000 0.001 30.894 30.896 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.690 29.742 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.005 27.295 27.306 multiply_cannon_multrec 888 9.7 12.518 15.160 21.043 24.277 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.205 22.903 make_m2s 222 7.7 0.006 0.007 17.258 18.465 make_images 222 8.7 1.979 2.297 17.219 18.427 make_images_data 222 9.7 0.003 0.004 9.926 10.939 hybrid_alltoall_any 227 10.6 0.617 2.824 9.523 10.890 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.481 8.060 9.297 dbcsr_mm_accdrv_process_sort 3754 11.4 7.675 8.816 7.675 8.816 mp_sum_l 807 5.4 5.135 8.387 5.135 8.387 multiply_cannon_sync_h2d 888 9.7 6.030 7.662 6.030 7.662 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.496 6.957 mp_irecv_dv 2335 11.1 2.481 6.917 2.481 6.917 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.583 6.798 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.948 6.575 arnoldi_extremal 4 6.8 0.000 0.000 5.152 5.169 arnoldi_normal_ev 4 7.8 0.001 0.005 5.152 5.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.954 5.145 build_subspace 16 8.4 0.014 0.020 4.848 4.855 calculate_norms 1584 9.8 4.269 4.620 4.269 4.620 mp_allgather_i34 111 8.7 1.416 3.944 1.416 3.944 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.447 3.801 ls_scf_post 1 4.0 0.000 0.000 3.596 3.610 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.604 3.020 3.606 ls_scf_store_result 1 5.0 0.000 0.000 3.315 3.408 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.872 2.966 dbcsr_sort_data 325 11.1 1.898 2.177 1.898 2.177 make_images_pack 222 9.7 1.818 2.136 1.821 2.139 make_images_sizes 222 9.7 0.000 0.000 0.988 2.051 mp_alltoall_i44 222 10.7 0.988 2.051 0.988 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.954 1.956 dbcsr_data_release 9322 10.9 1.304 1.946 1.304 1.946 rebuild_ks_matrix 3 7.3 0.000 0.000 1.935 1.937 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.020 1.935 1.937 dbcsr_finalize 304 7.8 0.026 0.032 1.615 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.371000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2212.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 3.320742E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.085 97.632 97.633 qs_energies 1 2.0 0.009 0.035 96.861 96.908 ls_scf 1 3.0 0.008 0.033 95.177 95.195 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.380 78.641 ls_scf_main 1 4.0 0.005 0.028 58.875 58.900 multiply_cannon 111 7.7 0.076 0.164 51.569 56.560 density_matrix_trs4 2 5.0 0.003 0.021 52.625 52.770 multiply_cannon_loop 111 8.7 0.115 0.126 46.511 49.671 ls_scf_init_scf 1 4.0 0.001 0.002 33.080 33.127 ls_scf_init_matrix_S 1 5.0 0.000 0.006 31.499 31.572 mp_waitall_1 6369 11.0 22.665 29.442 22.665 29.442 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.027 29.043 multiply_cannon_multrec 1332 9.7 14.200 18.031 22.034 25.300 make_m2s 222 7.7 0.006 0.008 21.088 22.519 make_images 222 8.7 3.140 3.594 21.038 22.472 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.217 17.334 make_images_data 222 9.7 0.004 0.004 11.765 13.299 hybrid_alltoall_any 227 10.6 0.798 3.802 11.070 12.940 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.409 7.477 8.971 dbcsr_mm_accdrv_process_sort 3641 11.4 7.103 8.557 7.103 8.557 mp_sum_l 807 5.4 4.152 7.588 4.152 7.588 multiply_cannon_sync_h2d 1332 9.7 5.532 6.175 5.532 6.175 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.088 5.979 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.232 5.937 mp_irecv_dv 3229 10.9 2.066 5.909 2.066 5.909 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.510 5.281 arnoldi_extremal 4 6.8 0.000 0.000 5.258 5.277 arnoldi_normal_ev 4 7.8 0.001 0.005 5.258 5.277 build_subspace 16 8.4 0.014 0.021 4.918 4.925 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.530 4.693 mp_allgather_i34 111 8.7 2.188 4.571 2.188 4.571 calculate_norms 2376 9.8 4.190 4.542 4.190 4.542 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.606 3.881 dbcsr_matrix_vector_mult_local 304 10.0 3.183 3.681 3.185 3.683 dbcsr_sort_data 658 11.4 3.091 3.430 3.091 3.430 ls_scf_post 1 4.0 0.005 0.020 3.214 3.232 dbcsr_special_finalize 555 9.7 0.006 0.007 2.841 3.217 dbcsr_merge_single_wm 555 10.7 0.535 0.661 2.832 3.209 ls_scf_dm_to_ks 2 5.0 0.009 0.071 3.093 3.182 ls_scf_store_result 1 5.0 0.000 0.000 2.951 3.007 dbcsr_data_release 10477 10.7 1.572 2.394 1.572 2.394 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.054 2.102 rebuild_ks_matrix 3 7.3 0.000 0.000 2.030 2.078 qs_ks_build_kohn_sham_matrix 3 8.3 0.020 0.091 2.030 2.077 dbcsr_finalize 304 7.8 0.049 0.061 1.797 1.977 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.633000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2722.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 4.630888E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.090 99.680 99.681 qs_energies 1 2.0 0.000 0.000 98.819 98.825 ls_scf 1 3.0 0.000 0.000 96.865 96.876 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.015 78.214 ls_scf_main 1 4.0 0.000 0.000 62.316 62.386 multiply_cannon 111 7.7 0.096 0.186 55.380 60.873 density_matrix_trs4 2 5.0 0.002 0.003 55.006 55.118 multiply_cannon_loop 111 8.7 0.069 0.075 50.987 52.577 mp_waitall_1 5436 11.0 26.788 32.620 26.788 32.620 ls_scf_init_scf 1 4.0 0.001 0.001 30.934 30.950 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.465 29.510 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.322 27.331 multiply_cannon_multrec 444 9.7 13.941 16.888 20.959 22.709 make_m2s 222 7.7 0.004 0.005 17.842 20.210 make_images 222 8.7 3.723 4.428 17.780 20.151 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.110 16.297 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.440 14.136 make_images_data 222 9.7 0.003 0.004 10.040 12.391 hybrid_alltoall_any 227 10.6 0.790 3.762 9.886 12.272 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.345 6.730 7.841 multiply_cannon_sync_h2d 444 9.7 6.560 7.691 6.560 7.691 dbcsr_mm_accdrv_process_sort 3003 11.4 6.415 7.499 6.415 7.499 mp_allgather_i34 111 8.7 2.624 7.023 2.624 7.023 arnoldi_extremal 4 6.8 0.000 0.000 5.820 5.829 arnoldi_normal_ev 4 7.8 0.002 0.005 5.820 5.829 build_subspace 16 8.4 0.015 0.020 5.423 5.431 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.567 4.764 mp_sum_l 807 5.4 2.824 4.397 2.824 4.397 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.688 4.374 mp_irecv_dv 1241 11.2 1.669 4.359 1.669 4.359 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.165 4.357 dbcsr_matrix_vector_mult_local 304 10.0 3.685 4.153 3.687 4.155 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.712 3.792 calculate_norms 792 9.8 3.593 3.693 3.593 3.693 ls_scf_post 1 4.0 0.000 0.000 3.615 3.621 make_images_sizes 222 9.7 0.000 0.000 1.153 3.515 mp_alltoall_i44 222 10.7 1.153 3.515 1.153 3.515 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.440 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.903 3.243 dbcsr_finalize 304 7.8 0.062 0.078 2.210 2.349 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.325 2.337 rebuild_ks_matrix 3 7.3 0.000 0.000 2.292 2.304 qs_ks_build_kohn_sham_matrix 3 8.3 0.080 0.104 2.292 2.304 dbcsr_merge_all 275 8.9 0.474 0.532 2.058 2.184 dbcsr_data_release 10123 10.8 1.332 1.997 1.332 1.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.681000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3693.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/49cb2407dde98ee02ed4a8f151770698be669539_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 8.775823E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.152 0.522 109.716 109.717 qs_energies 1 2.0 0.000 0.000 107.415 107.435 ls_scf 1 3.0 0.000 0.000 104.474 104.494 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.566 77.661 ls_scf_main 1 4.0 0.000 0.000 66.165 66.166 density_matrix_trs4 2 5.0 0.002 0.003 56.872 56.924 multiply_cannon 111 7.7 0.181 0.286 50.132 52.175 multiply_cannon_loop 111 8.7 0.067 0.070 46.644 47.458 ls_scf_init_scf 1 4.0 0.001 0.001 34.634 34.635 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.675 32.693 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.850 29.858 mp_waitall_1 4527 11.1 22.367 26.133 22.367 26.133 make_m2s 222 7.7 0.005 0.005 23.817 24.821 make_images 222 8.7 4.584 4.963 23.711 24.713 multiply_cannon_multrec 444 9.7 17.876 18.520 22.469 22.995 hybrid_alltoall_any 227 10.6 1.661 3.631 12.869 15.514 make_images_data 222 9.7 0.003 0.003 13.075 15.512 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.711 11.104 multiply_cannon_sync_h2d 444 9.7 8.846 8.896 8.846 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.385 7.394 arnoldi_normal_ev 4 7.8 0.003 0.009 7.385 7.394 build_subspace 16 8.4 0.026 0.037 6.819 6.826 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.455 5.616 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.281 5.374 dbcsr_matrix_vector_mult_local 304 10.0 5.025 5.338 5.028 5.341 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.002 5.252 dbcsr_mm_accdrv_process 1814 10.4 0.278 0.323 4.424 4.544 dbcsr_mm_accdrv_process_sort 1814 11.4 4.122 4.248 4.122 4.248 ls_scf_post 1 4.0 0.000 0.000 3.675 3.695 mp_allgather_i34 111 8.7 1.141 3.570 1.141 3.570 make_images_sizes 222 9.7 0.000 0.000 1.415 3.472 mp_alltoall_i44 222 10.7 1.415 3.471 1.415 3.471 ls_scf_store_result 1 5.0 0.000 0.000 3.424 3.431 calculate_norms 792 9.8 3.243 3.285 3.243 3.285 dbcsr_finalize 304 7.8 0.082 0.090 3.088 3.169 dbcsr_merge_all 275 8.9 0.891 0.917 2.873 2.946 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 2.911 2.911 dbcsr_complete_redistribute 5 7.6 1.428 1.467 2.769 2.873 dbcsr_data_release 12724 10.6 2.337 2.841 2.337 2.841 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.581 2.583 matrix_ls_to_qs 2 6.0 0.000 0.000 2.441 2.554 rebuild_ks_matrix 3 7.3 0.000 0.000 2.516 2.517 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.003 2.516 2.517 dbcsr_sort_data 325 11.1 2.443 2.501 2.443 2.501 dbcsr_new_transposed 4 7.5 0.243 0.251 2.299 2.317 dbcsr_frobenius_norm 74 6.6 2.057 2.140 2.202 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.206 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.717000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6892.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 49cb2407dde98ee02ed4a8f151770698be669539 Summary: empty Status: OK