=== 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: bbe71a2287d23928c5ebe2640c2d74e0055badb1 ################# 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 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/01 job id: 43028840 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/02 job id: 43028842 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/03 job id: 43028843 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/04 job id: 43028845 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/05 job id: 43028846 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/06 job id: 43028847 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/07 job id: 43028848 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/08 job id: 43028849 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/09 job id: 43028850 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/10 job id: 43028851 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/11 job id: 43028852 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/12 job id: 43028853 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/13 job id: 43028854 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/14 job id: 43028855 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/15 job id: 43028856 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/16 job id: 43028857 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/17 job id: 43028858 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/18 job id: 43028859 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/19 job id: 43028860 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/20 job id: 43028861 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/21 job id: 43028862 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/22 job id: 43028863 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/23 job id: 43028864 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/24 job id: 43028867 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/25 job id: 43028868 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/26 job id: 43028869 --- 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/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.037 133.597 133.598 farming_run 1 2.0 133.102 133.103 133.568 133.573 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457852E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 115.870 115.870 qs_energies 1 2.0 0.000 0.000 115.628 115.630 mp2_main 1 3.0 0.000 0.000 113.596 113.599 mp2_gpw_main 1 4.0 0.019 0.025 112.650 112.653 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.640 94.081 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.452 55.893 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.759 47.145 get_2c_integrals 1 6.0 0.000 0.000 37.282 38.018 integrate_v_rspace 273 8.0 0.439 0.452 25.123 30.227 pw_transfer 6555 10.6 0.374 0.394 27.356 27.861 grid_integrate_task_list 273 9.0 20.936 26.511 20.936 26.511 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.018 26.453 fft_wrap_pw1pw2_100 2178 12.4 1.189 1.408 23.575 24.024 compute_2c_integrals 1 7.0 0.002 0.002 19.666 19.667 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.891 19.389 mp2_eri_2c_integrate_gpw 1 9.0 2.387 2.440 18.888 19.386 rpa_ri_compute_en 1 5.0 0.000 0.000 18.902 19.078 cp_fm_cholesky_decompose 12 8.2 17.572 18.276 17.572 18.276 cholesky_decomp 1 7.0 0.000 0.000 16.470 17.188 fft3d_s 5443 13.4 16.143 16.598 16.165 16.619 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.551 10.851 15.551 calculate_wavefunction 272 8.0 5.430 5.582 12.557 13.192 rpa_num_int 1 6.0 0.000 0.000 10.568 10.569 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.531 10.552 calc_mat_Q 8 8.0 0.000 0.000 9.384 9.476 contract_S_to_Q 8 9.0 0.000 0.000 8.806 8.895 calc_potential_gpw 544 9.5 0.005 0.006 8.264 8.695 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.223 8.552 parallel_gemm_fm 14 9.1 0.000 0.000 8.387 8.467 parallel_gemm_fm_cosma 14 10.1 8.387 8.467 8.387 8.467 potential_pw2rs 545 10.0 0.107 0.110 7.679 8.329 create_integ_mat 1 6.0 0.022 0.028 7.798 7.799 collocate_single_gaussian 272 10.0 0.041 0.043 7.459 7.728 array2fm 1 7.0 0.000 0.000 6.742 7.250 pw_scatter_s 2720 13.7 4.441 4.679 4.441 4.679 pw_gather_s 2722 13.2 3.870 4.248 3.870 4.248 array2fm_buffer_send 1 8.0 3.084 3.239 3.084 3.239 pw_poisson_solve 545 10.5 1.122 1.175 2.165 2.420 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.652870, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 0.029 0.038 395.843 395.843 farming_run 1 2.0 394.986 395.011 395.799 395.803 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222742E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 209.755 209.755 qs_energies 1 2.0 0.000 0.000 209.545 209.550 scf_env_do_scf 1 3.0 0.000 0.000 106.419 106.419 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.530 105.537 rebuild_ks_matrix 4 6.0 0.000 0.000 105.529 105.536 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 105.529 105.536 hfx_ks_matrix 4 8.0 0.001 0.001 105.149 105.153 integrate_four_center 4 9.0 0.144 0.453 105.149 105.152 mp2_main 1 3.0 0.000 0.000 102.842 102.847 mp2_gpw_main 1 4.0 0.032 0.048 101.914 101.922 integrate_four_center_main 4 10.0 0.098 0.523 96.518 99.211 integrate_four_center_bin 265 11.0 96.420 99.145 96.420 99.145 init_scf_loop 1 4.0 0.000 0.000 92.191 92.191 mp2_ri_gpw_compute_in 1 5.0 0.071 0.135 74.954 76.038 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.012 54.537 55.590 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 42.147 47.201 integrate_v_rspace 95 8.0 0.398 0.563 28.515 33.592 pw_transfer 2240 10.6 0.146 0.162 29.838 30.236 ao_to_mo_and_store_B_mult_1 91 7.0 10.703 29.658 10.703 29.658 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.862 29.263 grid_integrate_task_list 95 9.0 23.800 29.060 23.800 29.060 mp2_ri_gpw_compute_en 1 5.0 0.057 0.094 26.792 28.585 fft_wrap_pw1pw2_100 730 12.4 1.252 1.475 26.602 27.028 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.905 25.065 25.074 get_2c_integrals 1 6.0 0.001 0.008 20.325 20.348 compute_2c_integrals 1 7.0 0.004 0.012 19.309 19.315 compute_2c_integrals_loop_lm 1 8.0 0.003 0.013 18.780 19.177 mp2_eri_2c_integrate_gpw 1 9.0 1.756 1.844 18.778 19.176 fft3d_s 1823 13.4 18.417 18.672 18.430 18.687 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.226 14.226 calculate_wavefunction 91 8.0 2.024 2.057 9.748 9.985 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.581 8.802 9.637 potential_pw2rs 186 10.0 0.033 0.037 8.622 9.247 local_gemm 172 8.0 8.245 9.062 8.245 9.062 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.192 8.644 mp2_ri_gpw_compute_en_comm 22 7.0 0.496 0.511 8.045 8.406 collocate_single_gaussian 91 10.0 0.018 0.039 7.835 8.196 calc_potential_gpw 182 9.5 0.002 0.002 7.883 8.027 mp_sendrecv_dm3 2068 8.0 6.089 6.449 6.089 6.449 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.403 6.340 6.403 mp_sync 38 10.4 3.254 5.712 3.254 5.712 pw_gather_s 912 13.2 4.910 5.442 4.910 5.442 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.910253, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.964928E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 1419947. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.088 54.251 54.251 qs_mol_dyn_low 1 2.0 0.003 0.004 53.945 53.953 qs_forces 11 3.9 0.002 0.002 53.875 53.876 qs_energies 11 4.9 0.002 0.007 52.396 52.409 scf_env_do_scf 11 5.9 0.001 0.003 46.519 46.520 scf_env_do_scf_inner_loop 108 6.5 0.002 0.010 44.391 44.391 dbcsr_multiply_generic 2286 12.5 0.094 0.096 35.287 35.665 qs_scf_new_mos 108 7.5 0.000 0.001 34.131 34.460 qs_scf_loop_do_ot 108 8.5 0.001 0.001 34.131 34.459 ot_scf_mini 108 9.5 0.002 0.003 32.413 32.622 multiply_cannon 2286 13.5 0.192 0.204 26.731 28.315 multiply_cannon_loop 2286 14.5 1.510 1.581 25.977 27.529 velocity_verlet 10 3.0 0.001 0.001 26.338 26.339 ot_mini 108 10.5 0.001 0.001 20.836 21.065 qs_ot_get_derivative 108 11.5 0.001 0.001 17.821 18.001 mp_waitall_1 267858 16.1 9.803 16.079 9.803 16.079 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 6.606 13.424 multiply_cannon_multrec 54864 15.5 4.118 6.275 7.406 10.698 mp_sum_l 7207 12.9 6.508 8.354 6.508 8.354 rebuild_ks_matrix 119 8.3 0.000 0.000 7.978 8.119 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.977 8.119 multiply_cannon_sync_h2d 54864 15.5 5.804 7.900 5.804 7.900 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.042 7.167 qs_ot_get_p 119 10.4 0.001 0.001 6.720 6.995 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.277 6.387 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.864 6.308 init_scf_run 11 5.9 0.000 0.001 4.622 4.622 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.621 4.622 sum_up_and_integrate 119 10.3 0.012 0.014 4.540 4.547 integrate_v_rspace 119 11.3 0.002 0.002 4.528 4.537 dbcsr_mm_accdrv_process 76910 16.1 1.186 1.829 3.210 4.475 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.226 4.342 calculate_rho_elec 119 8.7 0.011 0.016 4.226 4.341 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 2.222 3.982 qs_ot_p2m_diag 50 11.0 0.004 0.007 3.657 3.743 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.916 3.128 apply_single 119 13.6 0.000 0.000 2.916 3.128 calculate_dm_sparse 119 9.5 0.000 0.000 2.943 3.112 rs_pw_transfer 974 11.9 0.011 0.012 2.847 2.987 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.669 2.754 ot_diis_step 108 11.5 0.006 0.006 2.719 2.719 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.653 2.653 jit_kernel_multiply 13 15.8 1.961 2.562 1.961 2.562 calculate_first_density_matrix 1 7.0 0.000 0.003 2.457 2.462 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.380 2.381 density_rs2pw 119 9.7 0.004 0.004 2.256 2.376 cp_fm_redistribute_end 50 14.0 2.157 2.355 2.161 2.356 cp_fm_diag_elpa_base 50 14.0 0.193 2.319 0.194 2.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.197 2.199 acc_transpose_blocks 54864 15.5 0.230 0.243 1.727 2.131 grid_integrate_task_list 119 12.3 2.014 2.119 2.014 2.119 init_scf_loop 11 6.9 0.000 0.003 2.108 2.108 wfi_extrapolate 11 7.9 0.001 0.001 2.103 2.103 mp_sum_d 4125 12.0 1.441 2.023 1.441 2.023 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.943 1.989 potential_pw2rs 119 12.3 0.004 0.004 1.907 1.924 make_m2s 4572 13.5 0.053 0.055 1.775 1.824 pw_transfer 1439 11.6 0.051 0.055 1.739 1.815 make_images 4572 14.5 0.133 0.139 1.693 1.742 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.664 1.741 mp_alltoall_d11v 2130 13.8 1.455 1.624 1.455 1.624 fft3d_ps 1201 14.6 0.361 0.463 1.441 1.504 mp_waitany 12084 13.8 1.249 1.439 1.249 1.439 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.386 1.409 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.278 1.359 grid_collocate_task_list 119 9.7 1.290 1.355 1.290 1.355 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.789 1.133 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.251000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.230912E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1031941. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.056 38.260 38.261 qs_mol_dyn_low 1 2.0 0.003 0.004 37.958 37.966 qs_forces 11 3.9 0.002 0.002 37.900 37.901 qs_energies 11 4.9 0.001 0.001 36.224 36.227 scf_env_do_scf 11 5.9 0.000 0.001 31.207 31.208 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.737 28.737 dbcsr_multiply_generic 2286 12.5 0.097 0.101 21.272 21.629 qs_scf_new_mos 108 7.5 0.001 0.001 19.724 19.969 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.723 19.968 ot_scf_mini 108 9.5 0.002 0.003 18.817 18.993 velocity_verlet 10 3.0 0.001 0.001 18.204 18.206 multiply_cannon 2286 13.5 0.210 0.219 16.230 17.734 multiply_cannon_loop 2286 14.5 0.905 0.977 15.113 16.461 ot_mini 108 10.5 0.001 0.001 11.742 11.981 mp_waitall_1 217478 16.2 6.014 11.234 6.014 11.234 qs_ot_get_derivative 108 11.5 0.001 0.002 9.287 9.468 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.092 9.380 multiply_cannon_multrec 27432 15.5 1.971 4.294 5.858 8.663 rebuild_ks_matrix 119 8.3 0.000 0.000 7.138 7.270 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.138 7.270 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.306 6.426 dbcsr_mm_accdrv_process 47894 16.0 3.030 5.098 3.815 5.640 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.590 4.442 qs_ot_get_p 119 10.4 0.001 0.001 4.169 4.403 sum_up_and_integrate 119 10.3 0.024 0.027 4.179 4.184 integrate_v_rspace 119 11.3 0.002 0.002 4.155 4.163 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.042 4.110 apply_single 119 13.6 0.000 0.000 3.042 4.110 mp_sum_l 7207 12.9 2.075 3.995 2.075 3.995 init_scf_run 11 5.9 0.000 0.001 3.773 3.773 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.772 3.773 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.657 3.687 calculate_rho_elec 119 8.7 0.021 0.024 3.656 3.687 multiply_cannon_sync_h2d 27432 15.5 2.199 2.808 2.199 2.808 make_m2s 4572 13.5 0.052 0.054 2.548 2.768 rs_pw_transfer 974 11.9 0.010 0.011 2.621 2.718 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.696 2.715 make_images 4572 14.5 0.200 0.237 2.460 2.679 init_scf_loop 11 6.9 0.000 0.000 2.449 2.450 ot_diis_step 108 11.5 0.011 0.011 2.405 2.406 calculate_first_density_matrix 1 7.0 0.000 0.001 2.276 2.278 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.261 2.261 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.144 2.237 calculate_dm_sparse 119 9.5 0.000 0.001 2.050 2.127 density_rs2pw 119 9.7 0.004 0.004 2.014 2.102 jit_kernel_multiply 10 16.1 0.732 1.945 0.732 1.945 potential_pw2rs 119 12.3 0.006 0.006 1.930 1.942 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.925 1.925 grid_integrate_task_list 119 12.3 1.831 1.921 1.831 1.921 cp_fm_redistribute_end 50 14.0 1.583 1.899 1.586 1.900 cp_fm_diag_elpa_base 50 14.0 0.305 1.846 0.313 1.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.866 1.869 pw_transfer 1439 11.6 0.063 0.066 1.808 1.838 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.719 1.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.666 1.705 make_images_data 4572 15.5 0.045 0.051 1.160 1.578 prepare_preconditioner 11 7.9 0.000 0.000 1.548 1.575 make_preconditioner 11 8.9 0.000 0.000 1.548 1.575 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.449 1.507 acc_transpose_blocks 27432 15.5 0.107 0.112 1.181 1.479 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.012 1.465 fft3d_ps 1201 14.6 0.500 0.555 1.426 1.453 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.447 fft_wrap_pw1pw2_140 487 13.2 0.075 0.083 1.318 1.349 grid_collocate_task_list 119 9.7 1.240 1.319 1.240 1.319 mp_alltoall_d11v 2130 13.8 1.208 1.313 1.208 1.313 mp_allgather_i34 2286 14.5 0.555 1.305 0.555 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.260 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.169 1.217 mp_sum_d 4125 12.0 0.568 1.019 0.568 1.019 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.978 0.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.888 0.901 acc_transpose_blocks_kernels 27432 16.5 0.181 0.272 0.653 0.856 make_images_sizes 4572 15.5 0.005 0.005 0.572 0.827 mp_alltoall_i44 4572 16.5 0.567 0.822 0.567 0.822 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.610 0.751 0.803 mp_alltoall_z22v 1201 16.6 0.698 0.768 0.698 0.768 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.261000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.909091, yerr=1.564059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.727424E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.115 32.585 32.587 qs_mol_dyn_low 1 2.0 0.003 0.005 32.077 32.085 qs_forces 11 3.9 0.002 0.002 32.017 32.018 qs_energies 11 4.9 0.005 0.014 30.420 30.422 scf_env_do_scf 11 5.9 0.001 0.002 25.338 25.338 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 22.748 22.749 dbcsr_multiply_generic 2286 12.5 0.092 0.095 16.795 16.894 velocity_verlet 10 3.0 0.001 0.001 15.295 15.296 qs_scf_new_mos 108 7.5 0.001 0.001 14.783 14.800 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.782 14.799 multiply_cannon 2286 13.5 0.195 0.204 13.465 14.347 ot_scf_mini 108 9.5 0.002 0.003 14.063 14.074 multiply_cannon_loop 2286 14.5 0.636 0.666 12.666 13.569 ot_mini 108 10.5 0.001 0.001 8.785 8.795 multiply_cannon_multrec 18288 15.5 1.935 2.830 7.064 7.317 qs_ot_get_derivative 108 11.5 0.001 0.001 7.287 7.299 rebuild_ks_matrix 119 8.3 0.000 0.000 6.367 6.382 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.366 6.381 dbcsr_mm_accdrv_process 38222 16.0 4.300 5.880 5.043 5.938 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.618 5.633 mp_waitall_1 169478 16.3 2.953 4.037 2.953 4.037 sum_up_and_integrate 119 10.3 0.031 0.032 3.968 3.977 integrate_v_rspace 119 11.3 0.002 0.003 3.937 3.950 init_scf_run 11 5.9 0.000 0.001 3.862 3.864 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.862 3.864 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.771 3.398 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.255 3.264 calculate_rho_elec 119 8.7 0.030 0.031 3.254 3.264 qs_ot_get_p 119 10.4 0.001 0.001 3.100 3.121 calculate_first_density_matrix 1 7.0 0.001 0.003 2.626 2.628 init_scf_loop 11 6.9 0.001 0.003 2.572 2.573 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.418 2.407 rs_pw_transfer 974 11.9 0.009 0.010 2.277 2.371 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.039 2.353 apply_single 119 13.6 0.000 0.000 2.039 2.353 jit_kernel_multiply 10 16.2 0.691 2.209 0.691 2.209 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.006 2.012 calculate_dm_sparse 119 9.5 0.000 0.000 1.975 1.987 make_m2s 4572 13.5 0.044 0.045 1.834 1.979 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.919 1.921 density_rs2pw 119 9.7 0.004 0.004 1.835 1.920 make_images 4572 14.5 0.191 0.202 1.749 1.893 grid_integrate_task_list 119 12.3 1.801 1.889 1.801 1.889 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.769 1.774 pw_transfer 1439 11.6 0.063 0.065 1.750 1.764 potential_pw2rs 119 12.3 0.007 0.008 1.748 1.759 prepare_preconditioner 11 7.9 0.000 0.000 1.744 1.746 make_preconditioner 11 8.9 0.000 0.001 1.744 1.746 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.739 1.740 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.598 1.682 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.660 1.674 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.559 1.565 multiply_cannon_sync_h2d 18288 15.5 1.390 1.559 1.390 1.559 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.508 1.509 cp_fm_redistribute_end 50 14.0 1.118 1.487 1.119 1.488 mp_sum_l 7207 12.9 1.138 1.484 1.138 1.484 ot_diis_step 108 11.5 0.011 0.011 1.475 1.475 cp_fm_diag_elpa_base 50 14.0 0.353 1.431 0.367 1.469 acc_transpose_blocks 18288 15.5 0.074 0.076 1.353 1.379 fft3d_ps 1201 14.6 0.507 0.524 1.347 1.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.350 1.357 grid_collocate_task_list 119 9.7 1.205 1.335 1.205 1.335 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.300 1.313 wfi_extrapolate 11 7.9 0.001 0.001 1.191 1.191 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.406 1.043 make_images_data 4572 15.5 0.045 0.049 0.819 0.988 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.955 0.956 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 0.914 0.933 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.700 0.905 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.878 0.901 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.818 0.820 mp_alltoall_d11v 2130 13.8 0.703 0.795 0.703 0.795 mp_alltoall_z22v 1201 16.6 0.680 0.748 0.680 0.748 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.653 0.731 cp_fm_cholesky_invert 11 10.9 0.726 0.729 0.726 0.729 jit_kernel_transpose 5 15.6 0.706 0.717 0.706 0.717 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.587000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=2.596883 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.848256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.085 36.356 36.357 qs_mol_dyn_low 1 2.0 0.003 0.004 35.826 35.965 qs_forces 11 3.9 0.002 0.003 35.762 35.762 qs_energies 11 4.9 0.002 0.006 33.988 33.994 scf_env_do_scf 11 5.9 0.001 0.003 28.691 28.692 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 25.024 25.025 dbcsr_multiply_generic 2286 12.5 0.099 0.100 18.645 18.758 velocity_verlet 10 3.0 0.001 0.001 18.186 18.187 qs_scf_new_mos 108 7.5 0.001 0.001 16.719 16.781 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.719 16.781 ot_scf_mini 108 9.5 0.003 0.004 15.767 15.825 multiply_cannon 2286 13.5 0.234 0.272 14.821 15.232 multiply_cannon_loop 2286 14.5 0.941 0.972 13.804 14.169 ot_mini 108 10.5 0.001 0.001 9.593 9.663 multiply_cannon_multrec 27432 15.5 2.370 3.074 8.895 9.294 qs_ot_get_derivative 108 11.5 0.001 0.001 7.722 7.782 dbcsr_mm_accdrv_process 47916 15.9 5.529 7.241 6.426 7.728 rebuild_ks_matrix 119 8.3 0.000 0.000 6.626 6.683 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.626 6.682 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.845 5.895 sum_up_and_integrate 119 10.3 0.035 0.038 3.900 3.910 init_scf_run 11 5.9 0.000 0.001 3.897 3.898 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.897 3.897 integrate_v_rspace 119 11.3 0.002 0.003 3.864 3.876 init_scf_loop 11 6.9 0.001 0.003 3.643 3.645 qs_ot_get_p 119 10.4 0.001 0.001 3.523 3.605 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.371 3.404 calculate_rho_elec 119 8.7 0.040 0.046 3.370 3.402 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.941 3.339 mp_waitall_1 145218 16.4 2.354 2.977 2.354 2.977 prepare_preconditioner 11 7.9 0.000 0.000 2.763 2.772 make_preconditioner 11 8.9 0.000 0.002 2.763 2.772 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.363 2.693 calculate_first_density_matrix 1 7.0 0.001 0.003 2.478 2.480 make_m2s 4572 13.5 0.053 0.055 2.365 2.471 make_images 4572 14.5 0.272 0.334 2.259 2.362 rs_pw_transfer 974 11.9 0.009 0.009 2.179 2.323 calculate_dm_sparse 119 9.5 0.000 0.000 2.162 2.223 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.063 2.198 apply_single 119 13.6 0.000 0.000 2.063 2.197 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.139 2.148 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.992 2.022 density_rs2pw 119 9.7 0.003 0.004 1.869 2.010 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.954 1.956 grid_integrate_task_list 119 12.3 1.810 1.918 1.810 1.918 pw_transfer 1439 11.6 0.063 0.066 1.826 1.862 ot_diis_step 108 11.5 0.016 0.042 1.831 1.832 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.806 1.807 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.736 1.774 jit_kernel_multiply 8 15.9 0.836 1.774 0.836 1.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.726 1.741 potential_pw2rs 119 12.3 0.008 0.009 1.698 1.711 acc_transpose_blocks 27432 15.5 0.110 0.112 1.493 1.518 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.513 1.514 cp_fm_redistribute_end 50 14.0 0.991 1.478 0.992 1.479 cp_fm_diag_elpa_base 50 14.0 0.463 1.411 0.483 1.453 fft3d_ps 1201 14.6 0.530 0.582 1.417 1.449 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.793 1.443 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.378 1.417 wfi_extrapolate 11 7.9 0.001 0.001 1.369 1.369 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.352 1.362 mp_sum_l 7207 12.9 1.068 1.343 1.068 1.343 grid_collocate_task_list 119 9.7 1.222 1.341 1.222 1.341 dbcsr_complete_redistribute 329 12.2 0.136 0.178 0.904 1.180 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.147 1.165 cp_fm_upper_to_full 72 13.5 0.821 1.134 0.821 1.134 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.087 1.088 make_images_data 4572 15.5 0.045 0.048 0.919 1.062 multiply_cannon_sync_h2d 27432 15.5 0.989 1.058 0.989 1.058 hybrid_alltoall_any 4725 16.4 0.062 0.153 0.790 0.972 mp_alltoall_d11v 2130 13.8 0.817 0.969 0.817 0.969 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.659 0.927 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.870 0.887 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.872 0.877 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 cp_fm_cholesky_invert 11 10.9 0.854 0.858 0.854 0.858 mp_alltoall_z22v 1201 16.6 0.771 0.801 0.771 0.801 mp_alltoall_i22 627 13.8 0.442 0.731 0.442 0.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.357000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.181818, yerr=3.214122 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 603.848704E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.058 29.375 29.376 qs_mol_dyn_low 1 2.0 0.003 0.006 29.072 29.079 qs_forces 11 3.9 0.002 0.003 29.008 29.010 qs_energies 11 4.9 0.001 0.001 27.303 27.307 scf_env_do_scf 11 5.9 0.000 0.001 21.404 21.404 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.841 18.842 velocity_verlet 10 3.0 0.001 0.001 14.238 14.241 dbcsr_multiply_generic 2286 12.5 0.091 0.092 13.389 13.461 qs_scf_new_mos 108 7.5 0.001 0.001 11.300 11.327 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.300 11.326 multiply_cannon 2286 13.5 0.228 0.241 10.804 11.293 ot_scf_mini 108 9.5 0.002 0.002 10.623 10.650 multiply_cannon_loop 2286 14.5 0.330 0.343 9.874 10.080 multiply_cannon_multrec 9144 15.5 1.673 1.905 6.451 6.754 ot_mini 108 10.5 0.001 0.001 5.973 6.006 rebuild_ks_matrix 119 8.3 0.000 0.000 5.821 5.843 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.820 5.843 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.185 5.204 dbcsr_mm_accdrv_process 12550 15.8 3.316 4.452 4.675 4.844 qs_ot_get_derivative 108 11.5 0.001 0.001 4.652 4.680 init_scf_run 11 5.9 0.000 0.001 4.468 4.468 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.468 4.468 sum_up_and_integrate 119 10.3 0.037 0.041 3.536 3.545 integrate_v_rspace 119 11.3 0.002 0.003 3.499 3.508 calculate_first_density_matrix 1 7.0 0.000 0.000 3.353 3.354 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.227 3.235 calculate_rho_elec 119 8.7 0.059 0.061 3.227 3.234 qs_ot_get_p 119 10.4 0.001 0.001 2.864 2.904 calculate_dm_sparse 119 9.5 0.000 0.000 2.523 2.544 init_scf_loop 11 6.9 0.000 0.000 2.542 2.543 jit_kernel_multiply 10 15.7 1.321 2.369 1.321 2.369 mp_waitall_1 121218 16.5 1.813 2.302 1.813 2.302 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.945 1.951 grid_integrate_task_list 119 12.3 1.849 1.950 1.849 1.950 make_m2s 4572 13.5 0.034 0.035 1.771 1.937 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.893 1.895 make_images 4572 14.5 0.268 0.301 1.683 1.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.813 1.814 prepare_preconditioner 11 7.9 0.000 0.000 1.758 1.762 make_preconditioner 11 8.9 0.000 0.000 1.758 1.762 pw_transfer 1439 11.6 0.063 0.065 1.728 1.744 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.696 1.719 rs_pw_transfer 974 11.9 0.008 0.008 1.642 1.714 density_rs2pw 119 9.7 0.003 0.004 1.632 1.714 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.691 1.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.644 1.672 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.637 1.654 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.590 1.599 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.386 1.386 cp_fm_redistribute_end 50 14.0 0.689 1.362 0.690 1.363 grid_collocate_task_list 119 9.7 1.272 1.345 1.272 1.345 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.327 1.343 cp_fm_diag_elpa_base 50 14.0 0.628 1.285 0.671 1.342 acc_transpose_blocks 9144 15.5 0.038 0.039 1.330 1.338 potential_pw2rs 119 12.3 0.010 0.010 1.322 1.333 fft3d_ps 1201 14.6 0.542 0.555 1.304 1.321 ot_diis_step 108 11.5 0.012 0.013 1.308 1.309 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.190 1.244 apply_single 119 13.6 0.000 0.000 1.190 1.244 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.210 1.212 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.159 1.174 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.091 1.093 wfi_extrapolate 11 7.9 0.001 0.001 1.071 1.071 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.817 1.051 make_images_data 4572 15.5 0.039 0.041 0.827 1.034 jit_kernel_transpose 5 15.6 0.974 0.976 0.974 0.976 mp_alltoall_d11v 2130 13.8 0.835 0.928 0.835 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.864 0.925 cp_fm_cholesky_invert 11 10.9 0.895 0.897 0.895 0.897 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.413 0.818 multiply_cannon_sync_h2d 9144 15.5 0.715 0.790 0.715 0.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.758 0.761 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.745 0.753 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.667 0.718 mp_allgather_i34 2286 14.5 0.265 0.690 0.265 0.690 mp_alltoall_z22v 1201 16.6 0.636 0.672 0.636 0.672 dbcsr_complete_redistribute 329 12.2 0.187 0.233 0.599 0.627 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.376000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.454545, yerr=5.821022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 759.889920E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.036 40.845 40.846 qs_mol_dyn_low 1 2.0 0.003 0.005 40.636 40.642 qs_forces 11 3.9 0.002 0.002 40.578 40.579 qs_energies 11 4.9 0.001 0.001 38.634 38.638 scf_env_do_scf 11 5.9 0.000 0.001 32.959 32.959 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.883 24.886 velocity_verlet 10 3.0 0.001 0.001 23.196 23.201 dbcsr_multiply_generic 2286 12.5 0.097 0.098 17.453 17.636 qs_scf_new_mos 108 7.5 0.001 0.001 15.994 16.092 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.993 16.091 ot_scf_mini 108 9.5 0.002 0.002 14.911 15.017 multiply_cannon 2286 13.5 0.302 0.315 13.578 14.488 multiply_cannon_loop 2286 14.5 0.342 0.347 12.303 13.260 ot_mini 108 10.5 0.001 0.001 8.814 8.935 multiply_cannon_multrec 9144 15.5 3.390 4.797 8.537 8.587 init_scf_loop 11 6.9 0.000 0.000 8.046 8.052 prepare_preconditioner 11 7.9 0.000 0.000 7.128 7.141 make_preconditioner 11 8.9 0.000 0.000 7.128 7.141 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.571 7.022 rebuild_ks_matrix 119 8.3 0.000 0.000 6.827 6.973 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.827 6.973 qs_ot_get_derivative 108 11.5 0.001 0.001 6.791 6.897 dbcsr_mm_accdrv_process 12550 15.8 3.932 5.285 5.021 6.327 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.168 6.301 cp_fm_upper_to_full 72 14.2 3.246 4.750 3.246 4.750 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.747 3.837 calculate_rho_elec 119 8.7 0.118 0.121 3.746 3.836 sum_up_and_integrate 119 10.3 0.064 0.065 3.677 3.683 init_scf_run 11 5.9 0.000 0.001 3.655 3.655 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.655 3.655 integrate_v_rspace 119 11.3 0.003 0.003 3.613 3.619 mp_waitall_1 97218 16.6 2.562 3.573 2.562 3.573 qs_ot_get_p 119 10.4 0.001 0.001 3.350 3.490 dbcsr_complete_redistribute 329 12.2 0.289 0.294 2.123 2.986 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.505 2.912 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.793 2.662 make_m2s 4572 13.5 0.038 0.038 2.295 2.464 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.168 2.442 apply_single 119 13.6 0.000 0.000 2.168 2.442 mp_alltoall_i22 627 13.8 1.533 2.436 1.533 2.436 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.551 2.413 make_images 4572 14.5 0.354 0.384 2.175 2.343 calculate_dm_sparse 119 9.5 0.000 0.000 2.248 2.308 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.337 2.266 calculate_first_density_matrix 1 7.0 0.000 0.000 2.165 2.230 pw_transfer 1439 11.6 0.066 0.067 2.070 2.072 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.043 2.044 grid_integrate_task_list 119 12.3 2.010 2.036 2.010 2.036 ot_diis_step 108 11.5 0.014 0.014 1.996 1.997 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.975 1.977 mp_sum_l 7207 12.9 1.050 1.821 1.050 1.821 density_rs2pw 119 9.7 0.003 0.004 1.791 1.812 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.809 1.810 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.765 1.766 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.749 1.750 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.689 1.742 jit_kernel_multiply 10 15.6 1.062 1.723 1.062 1.723 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.638 1.641 fft3d_ps 1201 14.6 0.566 0.577 1.608 1.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.535 1.577 cp_fm_cholesky_invert 11 10.9 1.523 1.527 1.523 1.527 grid_collocate_task_list 119 9.7 1.451 1.464 1.451 1.464 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.457 1.458 cp_fm_diag_elpa_base 50 14.0 1.308 1.363 1.456 1.456 rs_pw_transfer 974 11.9 0.009 0.009 1.371 1.390 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.315 1.376 wfi_extrapolate 11 7.9 0.001 0.001 1.372 1.372 mp_alltoall_d11v 2130 13.8 1.223 1.329 1.223 1.329 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.125 1.320 make_images_data 4572 15.5 0.042 0.045 1.065 1.286 potential_pw2rs 119 12.3 0.014 0.014 1.215 1.218 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.153 1.173 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.088 1.104 multiply_cannon_sync_h2d 9144 15.5 1.044 1.048 1.044 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.982 1.024 qs_create_task_list 11 7.9 0.001 0.001 0.942 0.954 generate_qs_task_list 11 8.9 0.373 0.392 0.942 0.953 mp_alltoall_z22v 1201 16.6 0.907 0.927 0.907 0.927 acc_transpose_blocks 9144 15.5 0.037 0.038 0.881 0.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.871 0.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.846000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.545455, yerr=15.185900 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.350400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.072 88.376 88.378 qs_mol_dyn_low 1 2.0 0.005 0.022 87.954 87.964 qs_forces 11 3.9 0.002 0.002 87.813 87.814 qs_energies 11 4.9 0.011 0.077 84.851 84.872 scf_env_do_scf 11 5.9 0.000 0.001 75.365 75.367 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 69.095 69.096 dbcsr_multiply_generic 2055 12.4 0.106 0.110 54.353 54.683 qs_scf_new_mos 99 7.5 0.000 0.001 50.851 50.999 qs_scf_loop_do_ot 99 8.5 0.001 0.001 50.851 50.998 ot_scf_mini 99 9.5 0.002 0.002 48.315 48.426 multiply_cannon 2055 13.4 0.180 0.188 43.914 44.868 velocity_verlet 10 3.0 0.001 0.001 44.319 44.321 multiply_cannon_loop 2055 14.4 1.561 1.598 42.762 43.789 ot_mini 99 10.5 0.001 0.001 29.201 29.309 qs_ot_get_derivative 99 11.5 0.002 0.010 22.289 22.443 multiply_cannon_multrec 49320 15.4 11.987 12.776 17.439 18.140 rebuild_ks_matrix 110 8.3 0.000 0.001 15.155 15.265 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.154 15.264 mp_waitall_1 241148 16.1 13.685 14.549 13.685 14.549 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.328 13.421 multiply_cannon_sync_h2d 49320 15.4 9.822 10.524 9.822 10.524 qs_ot_get_p 110 10.4 0.001 0.001 10.275 10.379 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 7.467 9.018 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.370 7.918 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.198 7.706 apply_single 110 13.6 0.000 0.000 7.198 7.706 sum_up_and_integrate 110 10.3 0.036 0.042 7.634 7.647 integrate_v_rspace 110 11.3 0.002 0.003 7.598 7.620 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.381 7.435 init_scf_run 11 5.9 0.000 0.001 7.213 7.214 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.213 7.213 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.980 7.133 calculate_rho_elec 110 8.6 0.020 0.024 6.980 7.132 ot_diis_step 99 11.5 0.006 0.006 6.598 6.598 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.506 6.542 mp_sum_l 6514 12.8 5.692 6.452 5.692 6.452 init_scf_loop 11 6.9 0.000 0.000 6.230 6.231 dbcsr_mm_accdrv_process 87628 16.1 2.093 2.188 5.332 5.699 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.377 5.377 rs_pw_transfer 902 11.9 0.011 0.013 4.626 4.838 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.683 4.685 cp_fm_redistribute_end 48 14.0 4.059 4.636 4.062 4.638 cp_fm_diag_elpa_base 48 14.0 0.568 4.550 0.572 4.578 make_m2s 4110 13.4 0.061 0.065 4.324 4.435 make_images 4110 14.4 0.178 0.191 4.228 4.342 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 3.138 4.273 wfi_extrapolate 11 7.9 0.001 0.001 4.205 4.205 density_rs2pw 110 9.6 0.004 0.004 4.032 4.203 calculate_dm_sparse 110 9.5 0.000 0.001 4.051 4.128 prepare_preconditioner 11 7.9 0.000 0.000 3.971 3.983 make_preconditioner 11 8.9 0.000 0.000 3.971 3.983 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.738 3.786 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 3.720 3.725 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.581 3.637 pw_transfer 1331 11.6 0.054 0.062 3.477 3.569 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.389 3.483 grid_integrate_task_list 110 12.3 3.224 3.460 3.224 3.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.365 3.398 potential_pw2rs 110 12.3 0.006 0.007 3.223 3.275 jit_kernel_multiply 13 15.9 2.954 3.038 2.954 3.038 fft3d_ps 1111 14.6 0.751 0.847 2.928 3.008 calculate_first_density_matrix 1 7.0 0.000 0.002 2.918 2.922 fft_wrap_pw1pw2_140 451 13.1 0.167 0.187 2.704 2.795 mp_alltoall_d11v 2046 13.8 2.274 2.736 2.274 2.736 acc_transpose_blocks 49320 15.4 0.221 0.231 2.269 2.337 grid_collocate_task_list 110 9.6 2.090 2.200 2.090 2.200 mp_sum_d 3879 11.9 1.644 2.198 1.644 2.198 cp_fm_cholesky_invert 11 10.9 2.178 2.183 2.178 2.183 mp_waitany 14300 13.8 1.817 2.068 1.817 2.068 make_images_data 4110 15.4 0.043 0.046 1.857 2.005 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.970 1.987 mp_alltoall_z22v 1111 16.6 1.780 1.952 1.780 1.952 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.809 1.839 hybrid_alltoall_any 4261 16.3 0.082 0.478 1.616 1.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=88.378000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.496278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 586.575872E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2290155. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.094 68.116 68.117 qs_mol_dyn_low 1 2.0 0.003 0.005 67.802 67.811 qs_forces 11 3.9 0.002 0.002 67.731 67.732 qs_energies 11 4.9 0.001 0.004 64.419 64.422 scf_env_do_scf 11 5.9 0.001 0.001 56.000 56.003 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.359 48.360 dbcsr_multiply_generic 2055 12.4 0.111 0.117 37.559 37.745 velocity_verlet 10 3.0 0.001 0.001 35.921 35.922 qs_scf_new_mos 99 7.5 0.001 0.001 32.515 32.659 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.514 32.659 multiply_cannon 2055 13.4 0.222 0.243 30.866 31.894 ot_scf_mini 99 9.5 0.003 0.004 30.862 30.996 multiply_cannon_loop 2055 14.4 0.926 0.949 29.591 30.437 ot_mini 99 10.5 0.001 0.001 18.139 18.275 multiply_cannon_multrec 24660 15.4 7.657 9.369 13.859 15.562 rebuild_ks_matrix 110 8.3 0.000 0.001 13.606 13.845 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.606 13.845 qs_ot_get_derivative 99 11.5 0.001 0.001 12.359 12.493 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.995 12.204 mp_waitall_1 186928 16.3 7.806 10.229 7.806 10.229 multiply_cannon_sync_h2d 24660 15.4 7.018 8.314 7.018 8.314 init_scf_loop 11 6.9 0.000 0.002 7.604 7.607 multiply_cannon_metrocomm3 24660 15.4 0.068 0.070 5.064 7.475 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.521 7.204 apply_single 110 13.6 0.000 0.001 6.521 7.204 sum_up_and_integrate 110 10.3 0.052 0.059 6.551 6.564 integrate_v_rspace 110 11.3 0.002 0.003 6.498 6.513 dbcsr_mm_accdrv_process 52282 16.1 4.722 5.649 6.038 6.343 qs_ot_get_p 110 10.4 0.001 0.001 6.117 6.282 init_scf_run 11 5.9 0.000 0.001 6.057 6.058 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.057 6.058 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.792 5.802 calculate_rho_elec 110 8.6 0.039 0.048 5.791 5.802 ot_diis_step 99 11.5 0.010 0.010 5.733 5.734 prepare_preconditioner 11 7.9 0.000 0.000 5.567 5.589 make_preconditioner 11 8.9 0.000 0.001 5.567 5.589 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.657 5.385 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.156 5.313 make_m2s 4110 13.4 0.056 0.059 4.207 4.646 make_images 4110 14.4 0.400 0.447 4.099 4.534 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.133 4.153 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.704 3.704 wfi_extrapolate 11 7.9 0.001 0.001 3.510 3.510 pw_transfer 1331 11.6 0.065 0.070 3.305 3.447 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.199 3.344 grid_integrate_task_list 110 12.3 3.147 3.317 3.147 3.317 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.194 3.268 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.234 3.236 density_rs2pw 110 9.6 0.004 0.004 3.085 3.231 rs_pw_transfer 902 11.9 0.012 0.013 3.012 3.193 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.146 3.148 cp_fm_redistribute_end 48 14.0 2.340 3.116 2.342 3.117 cp_fm_diag_elpa_base 48 14.0 0.741 3.005 0.772 3.090 calculate_dm_sparse 110 9.5 0.001 0.001 2.950 2.989 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.881 2.956 make_images_data 4110 15.4 0.047 0.051 2.371 2.834 fft_wrap_pw1pw2_140 451 13.1 0.201 0.220 2.677 2.822 hybrid_alltoall_any 4261 16.3 0.103 0.447 2.052 2.726 fft3d_ps 1111 14.6 1.072 1.267 2.555 2.684 cp_fm_cholesky_invert 11 10.9 2.593 2.600 2.593 2.600 mp_sum_l 6514 12.8 1.900 2.505 1.900 2.505 calculate_first_density_matrix 1 7.0 0.000 0.000 2.460 2.462 potential_pw2rs 110 12.3 0.008 0.009 2.424 2.441 grid_collocate_task_list 110 9.6 2.123 2.257 2.123 2.257 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.948 1.970 mp_alltoall_d11v 2046 13.8 1.738 1.964 1.738 1.964 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.853 1.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.775 1.801 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.789 1.759 jit_kernel_multiply 9 16.2 0.963 1.667 0.963 1.667 mp_irecv_dv 57340 16.2 0.664 1.643 0.664 1.643 mp_allgather_i34 2055 14.4 0.604 1.627 0.604 1.627 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.528 1.539 acc_transpose_blocks 24660 15.4 0.109 0.111 1.510 1.539 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.366 1.469 dbcsr_complete_redistribute 325 12.2 0.244 0.301 1.147 1.411 mp_waitany 10164 13.8 1.198 1.388 1.198 1.388 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.117000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.545455, yerr=6.773795 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.222528E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.062 60.471 60.473 qs_mol_dyn_low 1 2.0 0.003 0.005 60.044 60.054 qs_forces 11 3.9 0.002 0.002 59.977 59.978 qs_energies 11 4.9 0.014 0.051 56.780 56.784 scf_env_do_scf 11 5.9 0.001 0.002 48.449 48.449 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.812 39.813 velocity_verlet 10 3.0 0.001 0.001 32.310 32.312 dbcsr_multiply_generic 2055 12.4 0.119 0.125 28.617 28.885 qs_scf_new_mos 99 7.5 0.001 0.001 25.296 25.391 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.295 25.390 ot_scf_mini 99 9.5 0.003 0.003 24.077 24.188 multiply_cannon 2055 13.4 0.210 0.224 22.299 23.462 multiply_cannon_loop 2055 14.4 0.614 0.631 21.158 22.305 ot_mini 99 10.5 0.001 0.001 13.598 13.710 rebuild_ks_matrix 110 8.3 0.000 0.001 12.187 12.349 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.186 12.349 multiply_cannon_multrec 16440 15.4 3.896 5.044 9.939 10.946 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.730 10.871 mp_waitall_1 146766 16.3 7.170 9.976 7.170 9.976 qs_ot_get_derivative 99 11.5 0.001 0.001 9.184 9.295 init_scf_loop 11 6.9 0.001 0.005 8.601 8.603 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.257 7.000 prepare_preconditioner 11 7.9 0.000 0.000 6.822 6.842 make_preconditioner 11 8.9 0.000 0.002 6.822 6.842 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.195 6.543 sum_up_and_integrate 110 10.3 0.060 0.060 6.458 6.476 integrate_v_rspace 110 11.3 0.002 0.003 6.398 6.417 dbcsr_mm_accdrv_process 34862 16.1 4.878 5.525 5.895 6.071 qs_ot_get_p 110 10.4 0.001 0.002 5.720 5.865 init_scf_run 11 5.9 0.000 0.001 5.791 5.791 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.791 5.791 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.574 5.586 calculate_rho_elec 110 8.6 0.058 0.058 5.574 5.586 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.929 5.364 apply_single 110 13.6 0.000 0.000 4.929 5.364 make_m2s 4110 13.4 0.049 0.051 4.196 4.511 make_images 4110 14.4 0.394 0.513 4.082 4.396 ot_diis_step 99 11.5 0.011 0.011 4.388 4.388 multiply_cannon_sync_h2d 16440 15.4 3.689 4.252 3.689 4.252 qs_ot_p2m_diag 48 11.0 0.041 0.044 4.017 4.020 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.138 3.757 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.678 3.679 grid_integrate_task_list 110 12.3 3.181 3.387 3.181 3.387 pw_transfer 1331 11.6 0.064 0.071 3.242 3.256 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.136 3.152 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.119 3.120 cp_fm_redistribute_end 48 14.0 1.935 3.087 1.938 3.087 density_rs2pw 110 9.6 0.004 0.004 2.857 3.070 cp_fm_diag_elpa_base 48 14.0 1.086 2.949 1.145 3.057 rs_pw_transfer 902 11.9 0.010 0.011 2.746 2.989 wfi_extrapolate 11 7.9 0.001 0.001 2.956 2.956 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.910 2.911 make_images_data 4110 15.4 0.043 0.047 2.411 2.814 calculate_first_density_matrix 1 7.0 0.001 0.004 2.752 2.753 cp_fm_cholesky_invert 11 10.9 2.714 2.720 2.714 2.720 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.118 2.707 fft_wrap_pw1pw2_140 451 13.1 0.210 0.212 2.657 2.672 calculate_dm_sparse 110 9.5 0.001 0.001 2.635 2.664 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.577 2.627 fft3d_ps 1111 14.6 1.064 1.071 2.450 2.471 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.360 2.422 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.877 2.408 grid_collocate_task_list 110 9.6 2.179 2.342 2.179 2.342 potential_pw2rs 110 12.3 0.010 0.011 2.290 2.307 mp_irecv_dv 48980 15.7 0.808 2.287 0.808 2.287 jit_kernel_multiply 10 16.5 0.624 2.172 0.624 2.172 mp_alltoall_d11v 2046 13.8 1.751 2.143 1.751 2.143 mp_sum_l 6514 12.8 1.529 2.086 1.529 2.086 qs_energies_init_hamiltonians 11 5.9 0.003 0.019 2.036 2.037 dbcsr_complete_redistribute 325 12.2 0.321 0.361 1.430 1.896 cp_fm_upper_to_full 70 13.6 1.387 1.850 1.387 1.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.659 1.679 cp_fm_cholesky_decompose 22 10.9 1.628 1.647 1.628 1.647 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.601 1.613 mp_allgather_i34 2055 14.4 0.469 1.536 0.469 1.536 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.351 1.463 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.933 1.385 mp_waitany 17072 13.8 1.140 1.314 1.140 1.314 acc_transpose_blocks 16440 15.4 0.070 0.073 1.282 1.300 rs_gather_matrices 110 12.3 0.137 0.151 0.877 1.264 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.245 1.256 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.473000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.181818, yerr=7.837186 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 735.739904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 66.262 66.263 qs_mol_dyn_low 1 2.0 0.003 0.004 66.036 66.045 qs_forces 11 3.9 0.002 0.002 65.967 65.969 qs_energies 11 4.9 0.001 0.001 62.576 62.581 scf_env_do_scf 11 5.9 0.000 0.001 54.302 54.304 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.790 41.791 velocity_verlet 10 3.0 0.001 0.001 37.803 37.806 dbcsr_multiply_generic 2055 12.4 0.113 0.116 29.637 29.835 qs_scf_new_mos 99 7.5 0.001 0.001 27.161 27.260 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.160 27.259 ot_scf_mini 99 9.5 0.003 0.003 25.532 25.631 multiply_cannon 2055 13.4 0.244 0.261 22.442 23.673 multiply_cannon_loop 2055 14.4 0.890 0.908 20.963 21.509 ot_mini 99 10.5 0.001 0.001 14.256 14.375 multiply_cannon_multrec 24660 15.4 4.189 6.779 12.619 13.711 init_scf_loop 11 6.9 0.000 0.000 12.467 12.468 rebuild_ks_matrix 110 8.3 0.000 0.000 12.050 12.149 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.049 12.148 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.672 10.758 prepare_preconditioner 11 7.9 0.000 0.000 10.729 10.747 make_preconditioner 11 8.9 0.000 0.000 10.729 10.747 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.888 10.387 qs_ot_get_derivative 99 11.5 0.001 0.001 10.101 10.202 dbcsr_mm_accdrv_process 52304 16.0 6.749 7.990 8.281 9.202 mp_waitall_1 126806 16.4 4.857 6.643 4.857 6.643 sum_up_and_integrate 110 10.3 0.068 0.071 6.460 6.474 integrate_v_rspace 110 11.3 0.002 0.003 6.392 6.405 qs_ot_get_p 110 10.4 0.001 0.001 6.225 6.369 make_m2s 4110 13.4 0.059 0.061 5.596 5.935 make_images 4110 14.4 0.576 0.696 5.456 5.791 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.747 5.765 calculate_rho_elec 110 8.6 0.077 0.081 5.747 5.765 init_scf_run 11 5.9 0.000 0.001 5.533 5.533 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.532 5.533 cp_fm_upper_to_full 70 13.8 3.381 4.808 3.381 4.808 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.326 4.341 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.025 4.134 apply_single 110 13.6 0.000 0.000 4.025 4.134 ot_diis_step 99 11.5 0.011 0.011 4.114 4.114 dbcsr_complete_redistribute 325 12.2 0.415 0.453 2.785 3.946 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.842 3.842 grid_integrate_task_list 110 12.3 3.295 3.451 3.295 3.451 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.366 3.422 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.215 3.369 multiply_cannon_sync_h2d 24660 15.4 3.195 3.346 3.195 3.346 multiply_cannon_metrocomm3 24660 15.4 0.035 0.037 1.450 3.312 pw_transfer 1331 11.6 0.064 0.071 3.267 3.296 cp_fm_cholesky_invert 11 10.9 3.201 3.209 3.201 3.209 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.201 3.204 make_images_data 4110 15.4 0.046 0.050 2.830 3.202 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.162 3.193 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.358 3.187 cp_fm_redistribute_end 48 14.0 1.590 3.164 1.593 3.165 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.097 3.154 cp_fm_diag_elpa_base 48 14.0 1.486 3.023 1.567 3.137 density_rs2pw 110 9.6 0.004 0.004 2.931 3.112 wfi_extrapolate 11 7.9 0.001 0.001 3.065 3.065 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.832 2.967 calculate_dm_sparse 110 9.5 0.001 0.001 2.938 2.967 rs_pw_transfer 902 11.9 0.010 0.011 2.721 2.945 mp_alltoall_i22 605 13.7 1.695 2.903 1.695 2.903 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.808 2.809 fft_wrap_pw1pw2_140 451 13.1 0.201 0.214 2.624 2.660 fft3d_ps 1111 14.6 1.063 1.095 2.464 2.483 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.430 2.471 grid_collocate_task_list 110 9.6 2.224 2.431 2.224 2.431 calculate_first_density_matrix 1 7.0 0.000 0.000 2.377 2.382 potential_pw2rs 110 12.3 0.012 0.013 2.269 2.283 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.232 2.234 jit_kernel_multiply 12 15.9 1.198 2.189 1.198 2.189 mp_alltoall_d11v 2046 13.8 1.796 2.041 1.796 2.041 cp_fm_cholesky_decompose 22 10.9 1.837 1.885 1.837 1.885 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.729 1.763 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.634 mp_allgather_i34 2055 14.4 0.584 1.617 0.584 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.551 1.566 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.847 1.532 acc_transpose_blocks 24660 15.4 0.102 0.104 1.495 1.515 mp_irecv_dv 62702 16.1 0.748 1.452 0.748 1.452 mp_sum_l 6514 12.8 0.919 1.430 0.919 1.430 mp_waitany 13376 13.8 1.060 1.367 1.060 1.367 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.263000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.818182, yerr=11.166656 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 836.718592E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.072 54.242 54.242 qs_mol_dyn_low 1 2.0 0.003 0.015 53.911 53.920 qs_forces 11 3.9 0.002 0.002 53.845 53.846 qs_energies 11 4.9 0.001 0.001 50.240 50.245 scf_env_do_scf 11 5.9 0.000 0.001 42.025 42.025 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.289 34.289 velocity_verlet 10 3.0 0.001 0.001 30.516 30.519 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.644 22.787 qs_scf_new_mos 99 7.5 0.001 0.001 20.326 20.366 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.325 20.365 ot_scf_mini 99 9.5 0.002 0.002 19.094 19.117 multiply_cannon 2055 13.4 0.243 0.267 17.273 18.539 multiply_cannon_loop 2055 14.4 0.322 0.336 15.915 16.177 rebuild_ks_matrix 110 8.3 0.000 0.000 11.361 11.380 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.361 11.380 ot_mini 99 10.5 0.001 0.001 10.369 10.381 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.106 10.124 multiply_cannon_multrec 8220 15.4 3.227 4.760 7.511 8.597 init_scf_loop 11 6.9 0.000 0.000 7.688 7.689 mp_waitall_1 106626 16.5 5.987 7.552 5.987 7.552 qs_ot_get_derivative 99 11.5 0.001 0.001 6.593 6.612 sum_up_and_integrate 110 10.3 0.079 0.081 6.150 6.164 prepare_preconditioner 11 7.9 0.000 0.000 6.085 6.089 make_preconditioner 11 8.9 0.000 0.000 6.085 6.089 integrate_v_rspace 110 11.3 0.003 0.003 6.070 6.084 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.674 5.745 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.560 5.571 calculate_rho_elec 110 8.6 0.115 0.115 5.559 5.571 init_scf_run 11 5.9 0.000 0.001 5.121 5.121 dbcsr_mm_accdrv_process 17442 15.9 2.906 3.855 4.154 5.121 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.121 5.121 qs_ot_get_p 110 10.4 0.001 0.001 4.854 4.880 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.960 4.471 make_m2s 4110 13.4 0.038 0.039 4.118 4.381 make_images 4110 14.4 0.637 0.694 3.989 4.251 ot_diis_step 99 11.5 0.012 0.012 3.757 3.757 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.682 3.705 apply_single 110 13.6 0.000 0.000 3.682 3.705 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.513 3.517 grid_integrate_task_list 110 12.3 3.365 3.498 3.365 3.498 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.214 3.214 pw_transfer 1331 11.6 0.064 0.069 3.200 3.209 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.094 3.105 multiply_cannon_sync_h2d 8220 15.4 2.919 3.022 2.919 3.022 cp_fm_cholesky_invert 11 10.9 2.905 2.909 2.905 2.909 density_rs2pw 110 9.6 0.004 0.004 2.641 2.810 make_images_data 4110 15.4 0.038 0.043 2.300 2.727 wfi_extrapolate 11 7.9 0.001 0.001 2.711 2.711 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.686 2.686 hybrid_alltoall_any 4261 16.3 0.200 0.868 2.235 2.663 cp_fm_redistribute_end 48 14.0 0.678 2.658 0.682 2.659 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.641 2.656 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.650 2.651 cp_fm_diag_elpa_base 48 14.0 1.797 2.469 1.969 2.629 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.535 2.536 calculate_dm_sparse 110 9.5 0.001 0.001 2.465 2.509 grid_collocate_task_list 110 9.6 2.324 2.403 2.324 2.403 rs_pw_transfer 902 11.9 0.010 0.010 2.190 2.390 fft3d_ps 1111 14.6 1.117 1.143 2.350 2.358 calculate_first_density_matrix 1 7.0 0.000 0.000 2.325 2.326 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.151 2.163 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.986 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.949 1.960 potential_pw2rs 110 12.3 0.015 0.015 1.915 1.921 mp_alltoall_d11v 2046 13.8 1.600 1.823 1.600 1.823 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.805 1.812 cp_fm_cholesky_decompose 22 10.9 1.726 1.750 1.726 1.750 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.497 1.611 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.588 1.590 dbcsr_complete_redistribute 325 12.2 0.575 0.615 1.472 1.570 mp_allgather_i34 2055 14.4 0.546 1.555 0.546 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.431 1.444 qs_create_task_list 11 7.9 0.001 0.001 1.220 1.318 generate_qs_task_list 11 8.9 0.379 0.447 1.219 1.318 mp_waitany 9240 13.8 1.053 1.266 1.053 1.266 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.485 1.264 mp_irecv_dv 24056 15.7 0.461 1.224 0.461 1.224 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.744 1.198 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.176 1.197 jit_kernel_multiply 6 15.7 0.935 1.098 0.935 1.098 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.242000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.090909, yerr=15.424034 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 1.339732E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.054 86.055 86.057 qs_mol_dyn_low 1 2.0 0.003 0.004 85.628 85.637 qs_forces 11 3.9 0.002 0.003 85.562 85.564 qs_energies 11 4.9 0.002 0.005 81.516 81.517 scf_env_do_scf 11 5.9 0.000 0.001 71.593 71.594 velocity_verlet 10 3.0 0.001 0.001 54.906 54.913 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.945 42.947 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.525 28.637 init_scf_loop 11 6.9 0.000 0.000 28.572 28.575 qs_scf_new_mos 99 7.5 0.001 0.001 26.638 26.698 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.637 26.697 prepare_preconditioner 11 7.9 0.000 0.000 26.598 26.611 make_preconditioner 11 8.9 0.000 0.000 26.598 26.611 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.816 26.070 ot_scf_mini 99 9.5 0.002 0.002 24.878 24.939 multiply_cannon 2055 13.4 0.343 0.366 21.427 22.220 multiply_cannon_loop 2055 14.4 0.341 0.344 19.627 19.960 cp_fm_upper_to_full 70 14.2 12.770 18.267 12.770 18.267 ot_mini 99 10.5 0.001 0.001 13.636 13.687 rebuild_ks_matrix 110 8.3 0.001 0.001 13.078 13.131 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.077 13.130 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.900 11.947 dbcsr_complete_redistribute 325 12.2 1.029 1.043 7.437 10.642 multiply_cannon_multrec 8220 15.4 4.358 4.563 9.533 9.688 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.359 9.565 qs_ot_get_derivative 99 11.5 0.001 0.001 9.019 9.081 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.767 8.934 mp_waitall_1 87304 16.6 7.806 8.621 7.806 8.621 mp_alltoall_i22 605 13.7 5.395 8.615 5.395 8.615 sum_up_and_integrate 110 10.3 0.150 0.152 6.603 6.619 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.433 6.473 calculate_rho_elec 110 8.6 0.227 0.227 6.433 6.473 integrate_v_rspace 110 11.3 0.003 0.004 6.453 6.469 qs_ot_get_p 110 10.4 0.001 0.001 6.088 6.151 make_m2s 4110 13.4 0.043 0.043 5.306 5.827 make_images 4110 14.4 0.878 0.921 5.120 5.641 init_scf_run 11 5.9 0.000 0.001 5.631 5.631 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.631 5.631 cp_fm_cholesky_invert 11 10.9 5.447 5.451 5.447 5.451 dbcsr_mm_accdrv_process 11614 15.7 3.267 3.707 5.032 5.322 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.711 5.189 apply_single 110 13.6 0.000 0.000 4.711 5.189 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.741 5.113 ot_diis_step 99 11.5 0.015 0.015 4.590 4.590 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.459 4.467 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.011 4.012 multiply_cannon_sync_h2d 8220 15.4 3.954 3.960 3.954 3.960 pw_transfer 1331 11.6 0.073 0.074 3.703 3.709 grid_integrate_task_list 110 12.3 3.658 3.709 3.658 3.709 qs_energies_init_hamiltonians 11 5.9 0.023 0.046 3.673 3.686 hybrid_alltoall_any 4261 16.3 0.256 0.551 2.893 3.658 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.588 3.594 make_images_data 4110 15.4 0.041 0.044 2.887 3.558 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.376 3.376 cp_fm_diag_elpa_base 48 14.0 2.841 3.034 3.374 3.374 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.867 3.325 wfi_extrapolate 11 7.9 0.001 0.001 3.241 3.242 calculate_dm_sparse 110 9.5 0.001 0.001 3.123 3.149 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.091 3.099 density_rs2pw 110 9.6 0.004 0.004 2.941 2.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.890 2.893 fft3d_ps 1111 14.6 1.266 1.279 2.797 2.804 grid_collocate_task_list 110 9.6 2.628 2.653 2.628 2.653 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.430 2.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.377 2.397 calculate_first_density_matrix 1 7.0 0.000 0.000 2.284 2.285 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.178 2.234 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.195 rs_pw_transfer 902 11.9 0.010 0.011 2.147 2.184 mp_alltoall_d11v 2046 13.8 2.046 2.096 2.046 2.096 cp_fm_cholesky_decompose 22 10.9 2.048 2.070 2.048 2.070 potential_pw2rs 110 12.3 0.021 0.021 2.059 2.063 qs_create_task_list 11 7.9 0.000 0.001 1.897 1.941 generate_qs_task_list 11 8.9 0.740 0.792 1.896 1.941 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.820 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.783 1.789 jit_kernel_multiply 10 15.3 1.561 1.727 1.561 1.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.057000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1204.090909, yerr=54.753090 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.449280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 3731285. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.082 202.531 202.535 qs_mol_dyn_low 1 2.0 0.005 0.025 201.798 201.811 qs_forces 11 3.9 0.006 0.027 201.675 201.678 qs_energies 11 4.9 0.003 0.014 196.185 196.199 scf_env_do_scf 11 5.9 0.001 0.001 179.733 179.738 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.076 159.078 dbcsr_multiply_generic 2507 12.6 0.176 0.185 123.634 124.410 velocity_verlet 10 3.0 0.002 0.006 121.626 121.631 qs_scf_new_mos 117 7.6 0.001 0.001 120.581 120.867 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.580 120.867 ot_scf_mini 117 9.6 0.003 0.004 114.031 114.272 multiply_cannon 2507 13.6 0.235 0.249 100.553 102.983 multiply_cannon_loop 2507 14.6 2.134 2.204 98.400 100.678 ot_mini 117 10.6 0.004 0.026 65.214 65.464 multiply_cannon_multrec 60168 15.6 33.198 35.704 41.594 43.429 qs_ot_get_derivative 117 11.6 0.001 0.002 40.365 40.648 rebuild_ks_matrix 128 8.3 0.001 0.001 33.099 33.620 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.021 33.098 33.620 mp_waitall_1 291448 16.2 28.388 31.575 28.388 31.575 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.802 30.313 multiply_cannon_sync_h2d 60168 15.6 27.260 29.696 27.260 29.696 qs_ot_get_p 128 10.4 0.001 0.001 26.945 27.322 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.272 25.147 apply_single 128 13.6 0.001 0.001 24.272 25.146 ot_diis_step 117 11.6 0.008 0.011 24.498 24.502 init_scf_loop 11 6.9 0.000 0.000 20.583 20.585 qs_ot_p2m_diag 83 11.4 0.080 0.110 20.202 20.270 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.809 19.018 multiply_cannon_metrocomm3 60168 15.6 0.116 0.121 15.438 17.862 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.652 17.657 prepare_preconditioner 11 7.9 0.000 0.000 16.007 16.052 make_preconditioner 11 8.9 0.000 0.000 16.007 16.052 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.233 15.444 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.752 14.757 cp_fm_redistribute_end 83 14.4 11.623 14.675 11.635 14.678 cp_fm_diag_elpa_base 83 14.4 3.001 14.444 3.032 14.564 make_m2s 5014 13.6 0.103 0.112 13.747 14.064 make_images 5014 14.6 0.401 0.418 13.567 13.893 sum_up_and_integrate 128 10.3 0.089 0.107 13.860 13.880 integrate_v_rspace 128 11.3 0.003 0.004 13.771 13.795 qs_rho_update_rho_low 128 7.7 0.001 0.002 12.529 12.669 calculate_rho_elec 128 8.7 0.045 0.064 12.528 12.669 init_scf_run 11 5.9 0.000 0.001 12.235 12.236 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.235 12.236 mp_sum_l 7870 13.0 8.125 9.306 8.125 9.306 cp_fm_cholesky_invert 11 10.9 9.202 9.210 9.202 9.210 wfi_extrapolate 11 7.9 0.002 0.009 8.947 8.947 dbcsr_mm_accdrv_process 124484 16.2 3.298 3.474 7.955 8.534 calculate_dm_sparse 128 9.5 0.001 0.001 8.392 8.487 multiply_cannon_metrocomm1 60168 15.6 0.087 0.092 6.011 8.282 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.928 8.059 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.797 7.924 make_images_data 5014 15.6 0.067 0.074 6.743 7.660 grid_integrate_task_list 128 12.3 7.069 7.507 7.069 7.507 density_rs2pw 128 9.7 0.006 0.007 6.364 7.133 hybrid_alltoall_any 5200 16.5 0.291 2.252 5.917 7.112 pw_transfer 1547 11.6 0.074 0.107 6.662 6.937 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.459 6.708 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.664 6.672 rs_pw_transfer 1046 11.9 0.016 0.019 5.723 6.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.805 5.885 fft_wrap_pw1pw2_140 523 13.2 0.453 0.523 5.538 5.716 mp_alltoall_d11v 2415 14.1 4.077 5.486 4.077 5.486 fft3d_ps 1291 14.7 2.095 2.559 5.243 5.443 grid_collocate_task_list 128 9.7 4.726 5.086 4.726 5.086 cp_fm_cholesky_decompose 22 10.9 4.641 4.654 4.641 4.654 mp_sum_d 4460 12.2 3.626 4.506 3.626 4.506 potential_pw2rs 128 12.3 0.009 0.010 4.437 4.472 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.535000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.000000, yerr=6.835735 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 826.253312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57283. MP_Allreduce 11237 947. MP_Sync 168 MP_Alltoall 1955 6570003. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.094 189.965 189.966 qs_mol_dyn_low 1 2.0 0.004 0.036 189.388 189.401 qs_forces 11 3.9 0.003 0.003 189.299 189.300 qs_energies 11 4.9 0.005 0.008 182.604 182.617 scf_env_do_scf 11 5.9 0.001 0.003 166.103 166.113 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 133.126 133.129 velocity_verlet 10 3.0 0.001 0.001 118.397 118.400 dbcsr_multiply_generic 2485 12.5 0.183 0.189 96.518 97.832 qs_scf_new_mos 116 7.6 0.001 0.001 94.906 95.465 qs_scf_loop_do_ot 116 8.6 0.001 0.002 94.905 95.464 ot_scf_mini 116 9.6 0.004 0.005 90.134 90.713 multiply_cannon 2485 13.5 0.473 0.524 76.562 81.579 multiply_cannon_loop 2485 14.5 1.238 1.283 73.497 76.103 ot_mini 116 10.6 0.001 0.001 49.500 50.037 mp_waitall_1 224796 16.4 25.094 38.274 25.094 38.274 multiply_cannon_multrec 29820 15.5 21.670 25.506 31.278 35.347 rebuild_ks_matrix 127 8.3 0.001 0.001 32.422 32.896 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 32.422 32.895 init_scf_loop 11 6.9 0.001 0.002 32.884 32.885 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.201 29.650 prepare_preconditioner 11 7.9 0.000 0.000 28.519 28.593 make_preconditioner 11 8.9 0.000 0.003 28.519 28.593 multiply_cannon_metrocomm3 29820 15.5 0.091 0.097 15.768 28.311 qs_ot_get_derivative 116 11.6 0.001 0.002 27.655 28.226 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.220 27.777 qs_ot_get_p 127 10.4 0.001 0.001 22.678 23.351 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.926 23.227 apply_single 127 13.6 0.001 0.001 21.925 23.226 multiply_cannon_sync_h2d 29820 15.5 18.945 21.933 18.945 21.933 ot_diis_step 116 11.6 0.014 0.015 21.680 21.682 qs_ot_p2m_diag 82 11.4 0.185 0.213 17.808 17.845 cp_fm_cholesky_invert 11 10.9 16.707 16.719 16.707 16.719 cp_dbcsr_syevd 82 12.4 0.005 0.005 16.657 16.659 make_m2s 4970 13.5 0.089 0.094 14.324 15.677 make_images 4970 14.5 1.148 1.336 14.117 15.472 sum_up_and_integrate 127 10.3 0.116 0.134 14.329 14.356 integrate_v_rspace 127 11.3 0.003 0.003 14.212 14.246 cp_fm_diag_elpa 82 13.4 0.000 0.001 13.492 13.501 cp_fm_redistribute_end 82 14.4 7.838 13.412 7.847 13.413 cp_fm_diag_elpa_base 82 14.4 5.333 12.961 5.546 13.319 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.874 12.902 calculate_rho_elec 127 8.7 0.087 0.104 12.874 12.901 init_scf_run 11 5.9 0.000 0.001 11.663 11.665 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.663 11.664 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.849 11.273 multiply_cannon_metrocomm4 27335 15.5 0.097 0.111 3.748 10.747 make_images_data 4970 15.5 0.065 0.072 8.740 10.384 mp_irecv_dv 68888 16.3 3.553 10.357 3.553 10.357 dbcsr_mm_accdrv_process 61726 16.2 4.526 5.152 9.070 9.638 hybrid_alltoall_any 5155 16.4 0.343 1.507 7.437 9.418 wfi_extrapolate 11 7.9 0.001 0.001 8.370 8.370 pw_transfer 1535 11.6 0.084 0.100 7.786 7.853 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.564 7.622 grid_integrate_task_list 127 12.3 7.056 7.470 7.056 7.470 density_rs2pw 127 9.7 0.006 0.006 6.740 7.252 cp_fm_cholesky_decompose 22 10.9 6.966 7.053 6.966 7.053 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.192 6.962 fft_wrap_pw1pw2_140 519 13.2 0.474 0.523 6.562 6.621 calculate_dm_sparse 127 9.5 0.001 0.001 6.439 6.568 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.214 6.224 rs_pw_transfer 1038 11.9 0.014 0.016 5.647 6.211 fft3d_ps 1281 14.7 2.752 2.904 5.936 5.981 mp_sum_l 7804 13.0 3.896 5.864 3.896 5.864 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.356 5.513 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.313 5.373 grid_collocate_task_list 127 9.7 4.866 5.298 4.866 5.298 mp_alltoall_d11v 2401 14.1 4.202 5.062 4.202 5.062 potential_pw2rs 127 12.3 0.015 0.017 4.841 4.862 mp_allgather_i34 2485 14.5 1.653 4.780 1.653 4.780 dbcsr_complete_redistribute 393 12.7 0.768 0.852 3.157 4.019 mp_sum_d 4439 12.1 2.630 3.993 2.630 3.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.966000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 968.802304E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.050 177.449 177.451 qs_mol_dyn_low 1 2.0 0.003 0.005 176.951 176.964 qs_forces 11 3.9 0.003 0.003 176.854 176.858 qs_energies 11 4.9 0.002 0.007 170.369 170.377 scf_env_do_scf 11 5.9 0.001 0.002 154.841 154.844 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.621 119.622 velocity_verlet 10 3.0 0.001 0.001 112.902 112.904 qs_scf_new_mos 117 7.6 0.001 0.001 83.050 83.412 qs_scf_loop_do_ot 117 8.6 0.001 0.002 83.050 83.411 dbcsr_multiply_generic 2507 12.6 0.182 0.188 81.623 82.662 ot_scf_mini 117 9.6 0.004 0.005 78.910 79.288 multiply_cannon 2507 13.6 0.495 0.525 61.439 66.010 multiply_cannon_loop 2507 14.6 0.858 0.890 58.551 61.088 ot_mini 117 10.6 0.001 0.001 42.701 43.092 mp_waitall_1 178456 16.5 25.877 35.762 25.877 35.762 init_scf_loop 11 6.9 0.001 0.003 35.120 35.121 prepare_preconditioner 11 7.9 0.000 0.000 31.111 31.162 make_preconditioner 11 8.9 0.000 0.002 31.111 31.161 rebuild_ks_matrix 128 8.3 0.001 0.001 30.110 30.653 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.109 30.652 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.689 30.086 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.204 27.706 multiply_cannon_metrocomm3 20056 15.6 0.060 0.064 15.473 25.223 multiply_cannon_multrec 20056 15.6 13.363 16.168 22.042 24.793 qs_ot_get_derivative 117 11.6 0.001 0.002 22.801 23.182 qs_ot_get_p 128 10.4 0.001 0.001 21.591 22.090 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.014 20.978 apply_single 128 13.6 0.001 0.001 20.014 20.978 ot_diis_step 117 11.6 0.018 0.019 19.796 19.797 qs_ot_p2m_diag 83 11.4 0.265 0.272 17.259 17.268 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.247 16.248 multiply_cannon_sync_h2d 20056 15.6 14.186 16.056 14.186 16.056 make_m2s 5014 13.6 0.081 0.087 15.131 15.911 make_images 5014 14.6 1.192 1.320 14.898 15.676 cp_fm_cholesky_invert 11 10.9 14.542 14.551 14.542 14.551 sum_up_and_integrate 128 10.3 0.134 0.145 14.205 14.232 integrate_v_rspace 128 11.3 0.003 0.004 14.071 14.099 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.028 13.061 calculate_rho_elec 128 8.7 0.131 0.146 13.027 13.061 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.001 13.002 cp_fm_redistribute_end 83 14.4 4.877 12.912 4.892 12.916 cp_fm_diag_elpa_base 83 14.4 7.591 12.328 7.997 12.809 make_images_data 5014 15.6 0.060 0.066 9.446 10.661 init_scf_run 11 5.9 0.000 0.001 10.449 10.450 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.449 10.449 hybrid_alltoall_any 5200 16.5 0.431 1.974 8.140 9.786 multiply_cannon_metrocomm4 17549 15.6 0.060 0.070 3.477 9.347 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.967 9.256 mp_irecv_dv 50230 16.2 3.355 9.097 3.355 9.097 dbcsr_mm_accdrv_process 41502 16.2 4.463 5.000 8.133 8.293 pw_transfer 1547 11.6 0.084 0.105 7.782 7.885 grid_integrate_task_list 128 12.3 7.262 7.748 7.262 7.748 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.559 7.667 cp_fm_cholesky_decompose 22 10.9 7.388 7.457 7.388 7.457 cp_fm_upper_to_full 105 14.5 5.708 7.451 5.708 7.451 wfi_extrapolate 11 7.9 0.001 0.001 7.384 7.384 density_rs2pw 128 9.7 0.005 0.006 6.480 6.790 fft_wrap_pw1pw2_140 523 13.2 0.477 0.528 6.619 6.734 dbcsr_complete_redistribute 395 12.7 1.184 1.231 4.765 6.492 fft3d_ps 1291 14.7 2.693 2.913 5.826 5.905 calculate_dm_sparse 128 9.5 0.001 0.001 5.734 5.834 grid_collocate_task_list 128 9.7 5.086 5.682 5.086 5.682 mp_alltoall_d11v 2415 14.1 4.452 5.676 4.452 5.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.485 5.490 rs_pw_transfer 1046 11.9 0.013 0.014 5.127 5.478 copy_fm_to_dbcsr 209 11.7 0.002 0.005 3.545 5.272 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.545 5.183 mp_sum_l 7870 13.0 3.244 4.735 3.244 4.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.523 4.653 potential_pw2rs 128 12.3 0.020 0.022 4.574 4.601 mp_allgather_i34 2507 14.6 1.419 4.561 1.419 4.561 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.402 4.109 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.040 4.067 mp_alltoall_i22 716 14.1 1.924 3.806 1.924 3.806 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 3.779 3.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.451000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=914.272727, yerr=15.196237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 4.320340E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 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 1.141895E+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 660837893360 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57191. MP_Allreduce 11251 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.118 186.884 186.885 qs_mol_dyn_low 1 2.0 0.003 0.005 186.137 186.150 qs_forces 11 3.9 0.005 0.006 186.023 186.035 qs_energies 11 4.9 0.014 0.015 179.039 179.052 scf_env_do_scf 11 5.9 0.001 0.002 162.350 162.363 velocity_verlet 10 3.0 0.001 0.001 122.946 122.952 scf_env_do_scf_inner_loop 116 6.6 0.011 0.049 115.365 115.367 qs_scf_new_mos 116 7.6 0.001 0.001 80.288 80.609 qs_scf_loop_do_ot 116 8.6 0.001 0.002 80.287 80.609 dbcsr_multiply_generic 2485 12.5 0.187 0.192 78.196 78.952 ot_scf_mini 116 9.6 0.009 0.025 75.829 76.114 multiply_cannon 2485 13.5 0.550 0.582 54.104 56.752 multiply_cannon_loop 2485 14.5 1.177 1.205 50.501 52.085 init_scf_loop 11 6.9 0.000 0.001 46.850 46.851 prepare_preconditioner 11 7.9 0.000 0.000 42.813 42.842 make_preconditioner 11 8.9 0.000 0.002 42.813 42.842 ot_mini 116 10.6 0.001 0.001 41.651 41.917 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.323 41.444 multiply_cannon_multrec 29820 15.5 14.008 18.671 25.905 30.101 rebuild_ks_matrix 127 8.3 0.001 0.001 28.641 28.865 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.641 28.865 mp_waitall_1 152434 16.5 17.117 27.063 17.117 27.063 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.896 26.089 qs_ot_get_derivative 116 11.6 0.003 0.013 22.330 22.602 qs_ot_get_p 127 10.4 0.001 0.001 20.605 20.898 make_m2s 4970 13.5 0.094 0.100 19.800 20.750 make_images 4970 14.5 1.928 2.303 19.497 20.449 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.753 19.370 apply_single 127 13.6 0.001 0.001 18.753 19.370 ot_diis_step 116 11.6 0.017 0.018 19.197 19.199 cp_fm_upper_to_full 104 14.7 11.344 16.783 11.344 16.783 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.484 16.536 cp_fm_cholesky_invert 11 10.9 16.296 16.305 16.296 16.305 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.220 15.221 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.359 15.218 sum_up_and_integrate 127 10.3 0.140 0.151 13.939 13.967 integrate_v_rspace 127 11.3 0.003 0.004 13.798 13.829 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.964 12.993 calculate_rho_elec 127 8.7 0.174 0.189 12.963 12.993 dbcsr_complete_redistribute 393 12.7 1.526 1.649 9.135 12.935 multiply_cannon_sync_h2d 29820 15.5 11.630 12.573 11.630 12.573 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.134 12.136 make_images_data 4970 15.5 0.063 0.066 10.582 12.094 cp_fm_redistribute_end 82 14.4 2.063 12.053 2.078 12.058 dbcsr_mm_accdrv_process 61748 16.2 7.299 8.339 11.472 11.974 cp_fm_diag_elpa_base 82 14.4 9.365 11.455 9.947 11.945 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.718 11.523 hybrid_alltoall_any 5155 16.4 0.521 2.180 9.633 10.961 init_scf_run 11 5.9 0.000 0.001 10.724 10.725 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.724 10.725 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.467 10.170 mp_alltoall_i22 712 14.1 5.679 9.407 5.679 9.407 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.189 9.394 pw_transfer 1535 11.6 0.084 0.097 7.790 7.861 grid_integrate_task_list 127 12.3 7.477 7.809 7.477 7.809 cp_fm_cholesky_decompose 22 10.9 7.635 7.718 7.635 7.718 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.567 7.643 wfi_extrapolate 11 7.9 0.001 0.001 7.605 7.605 multiply_cannon_metrocomm4 24850 15.5 0.075 0.085 2.734 7.459 mp_irecv_dv 75445 16.2 2.589 7.185 2.589 7.185 fft_wrap_pw1pw2_140 519 13.2 0.478 0.489 6.679 6.771 density_rs2pw 127 9.7 0.005 0.006 6.179 6.612 calculate_dm_sparse 127 9.5 0.001 0.001 6.143 6.224 fft3d_ps 1281 14.7 2.776 2.872 5.809 5.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.463 5.541 grid_collocate_task_list 127 9.7 5.163 5.463 5.163 5.463 mp_alltoall_d11v 2401 14.1 4.749 5.173 4.749 5.173 rs_pw_transfer 1038 11.9 0.013 0.015 4.565 4.970 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.459 4.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.370 4.450 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.335 4.418 potential_pw2rs 127 12.3 0.022 0.023 4.218 4.231 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.154 4.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.885000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1075.090909, yerr=20.214552 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.514652E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516879E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096702. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848050. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.044 174.391 174.391 qs_mol_dyn_low 1 2.0 0.003 0.004 173.925 173.938 qs_forces 11 3.9 0.003 0.003 173.671 173.673 qs_energies 11 4.9 0.002 0.002 166.325 166.335 scf_env_do_scf 11 5.9 0.001 0.001 149.193 149.204 velocity_verlet 10 3.0 0.001 0.001 114.212 114.225 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 113.203 113.204 qs_scf_new_mos 118 7.6 0.001 0.001 77.365 77.462 qs_scf_loop_do_ot 118 8.6 0.001 0.001 77.364 77.462 dbcsr_multiply_generic 2534 12.6 0.183 0.190 75.615 76.201 ot_scf_mini 118 9.6 0.004 0.004 72.883 72.971 multiply_cannon 2534 13.6 0.589 0.623 55.473 59.594 multiply_cannon_loop 2534 14.6 0.452 0.467 50.537 51.677 ot_mini 118 10.6 0.001 0.001 40.331 40.438 init_scf_loop 11 6.9 0.000 0.000 35.837 35.838 mp_waitall_1 130986 16.6 26.849 34.976 26.849 34.976 prepare_preconditioner 11 7.9 0.000 0.000 31.931 31.954 make_preconditioner 11 8.9 0.000 0.000 31.931 31.954 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.738 29.986 rebuild_ks_matrix 129 8.3 0.001 0.001 28.763 28.916 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.762 28.915 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.255 26.404 multiply_cannon_multrec 10136 15.6 10.523 16.443 17.885 22.252 ot_diis_step 118 11.6 0.020 0.021 20.276 20.276 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.820 20.194 apply_single 129 13.6 0.001 0.001 19.819 20.193 multiply_cannon_metrocomm3 10136 15.6 0.023 0.025 12.858 20.106 qs_ot_get_derivative 118 11.6 0.002 0.002 19.988 20.079 qs_ot_get_p 129 10.4 0.001 0.001 19.283 19.413 make_m2s 5068 13.6 0.066 0.071 16.452 19.044 make_images 5068 14.6 2.352 2.918 16.145 18.738 cp_fm_cholesky_invert 11 10.9 18.295 18.301 18.295 18.301 qs_ot_p2m_diag 84 11.4 0.502 0.507 15.356 15.376 cp_dbcsr_syevd 84 12.4 0.005 0.005 14.187 14.188 sum_up_and_integrate 129 10.3 0.182 0.192 14.095 14.137 integrate_v_rspace 129 11.3 0.004 0.004 13.913 13.966 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.191 13.228 calculate_rho_elec 129 8.7 0.260 0.271 13.190 13.227 make_images_data 5068 15.6 0.053 0.060 9.996 12.734 multiply_cannon_sync_h2d 10136 15.6 11.614 12.491 11.614 12.491 hybrid_alltoall_any 5255 16.5 0.819 3.590 9.825 12.298 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.074 11.075 cp_fm_diag_elpa_base 84 14.4 10.819 10.901 11.065 11.066 init_scf_run 11 5.9 0.000 0.001 10.354 10.354 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.354 10.354 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.089 8.156 grid_integrate_task_list 129 12.3 7.834 8.117 7.834 8.117 cp_fm_cholesky_decompose 22 10.9 7.915 8.019 7.915 8.019 multiply_cannon_metrocomm1 10136 15.6 0.028 0.028 4.519 8.011 pw_transfer 1559 11.6 0.083 0.091 7.792 7.809 dbcsr_mm_accdrv_process 20954 16.1 2.653 3.473 6.994 7.620 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.571 7.590 mp_allgather_i34 2534 14.6 3.135 7.509 3.135 7.509 wfi_extrapolate 11 7.9 0.001 0.001 7.340 7.340 fft_wrap_pw1pw2_140 527 13.2 0.506 0.525 6.652 6.679 calculate_dm_sparse 129 9.5 0.001 0.001 6.155 6.242 density_rs2pw 129 9.7 0.005 0.006 5.946 6.202 multiply_cannon_metrocomm4 7602 15.6 0.023 0.027 1.893 5.750 fft3d_ps 1301 14.7 2.740 2.802 5.717 5.748 grid_collocate_task_list 129 9.7 5.541 5.746 5.541 5.746 dbcsr_complete_redistribute 397 12.7 2.147 2.265 5.351 5.685 mp_irecv_dv 29142 15.9 1.856 5.663 1.856 5.663 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.275 5.290 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.198 5.199 mp_alltoall_d11v 2429 14.1 4.814 5.149 4.814 5.149 rs_pw_transfer 1054 12.0 0.013 0.013 4.057 4.332 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.134 4.169 potential_pw2rs 129 12.3 0.026 0.027 4.077 4.091 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.546 3.873 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.539 3.865 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.788 3.841 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.581 3.662 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.469 3.495 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.391000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1408.818182, yerr=48.926323 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237301E+12 0.0% 0.0% 100.0% flops max/rank 11.786605E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806449600 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 2.961560E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58052. MP_Allreduce 11197 1509. MP_Sync 87 MP_Alltoall 1724 36993608. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.113 292.459 292.460 qs_mol_dyn_low 1 2.0 0.004 0.013 291.703 291.715 qs_forces 11 3.9 0.003 0.005 291.615 291.619 qs_energies 11 4.9 0.001 0.002 283.055 283.060 scf_env_do_scf 11 5.9 0.001 0.001 261.144 261.155 velocity_verlet 10 3.0 0.001 0.001 210.928 210.935 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.639 135.641 init_scf_loop 11 6.9 0.000 0.000 125.239 125.241 prepare_preconditioner 11 7.9 0.000 0.000 120.519 120.547 make_preconditioner 11 8.9 0.000 0.000 120.519 120.547 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.257 117.673 qs_scf_new_mos 118 7.6 0.001 0.001 93.451 93.538 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.450 93.537 ot_scf_mini 118 9.6 0.004 0.004 88.591 88.626 dbcsr_multiply_generic 2531 12.6 0.214 0.222 84.781 85.475 cp_fm_upper_to_full 105 14.8 53.507 76.775 53.507 76.775 multiply_cannon 2531 13.6 0.700 0.724 59.443 60.081 multiply_cannon_loop 2531 14.6 0.476 0.490 55.583 57.226 ot_mini 118 10.6 0.001 0.001 45.411 45.451 dbcsr_complete_redistribute 395 12.7 3.988 4.036 30.002 43.457 copy_fm_to_dbcsr 209 11.7 0.002 0.002 26.572 39.993 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.218 37.505 mp_alltoall_i22 718 14.1 22.046 35.442 22.046 35.442 rebuild_ks_matrix 129 8.3 0.001 0.001 33.659 33.745 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 33.658 33.744 cp_fm_cholesky_invert 11 10.9 33.670 33.677 33.670 33.677 mp_waitall_1 106466 16.7 29.131 32.905 29.131 32.905 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.415 31.498 qs_ot_get_p 129 10.4 0.001 0.001 27.409 27.497 qs_ot_get_derivative 118 11.6 0.002 0.002 24.538 24.572 qs_ot_p2m_diag 83 11.4 0.878 0.883 23.150 23.180 make_m2s 5062 13.6 0.075 0.079 20.581 21.732 cp_dbcsr_syevd 83 12.4 0.005 0.006 21.404 21.405 make_images 5062 14.6 3.794 3.943 20.101 21.256 ot_diis_step 118 11.6 0.022 0.023 20.835 20.835 multiply_cannon_metrocomm3 10124 15.6 0.024 0.025 18.962 20.460 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.029 20.290 apply_single 129 13.6 0.001 0.001 20.029 20.289 multiply_cannon_multrec 10124 15.6 10.630 12.514 18.218 18.278 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.144 18.145 cp_fm_diag_elpa_base 83 14.4 13.718 15.335 18.141 18.141 multiply_cannon_sync_h2d 10124 15.6 15.783 15.815 15.783 15.815 sum_up_and_integrate 129 10.3 0.324 0.326 15.701 15.791 integrate_v_rspace 129 11.3 0.004 0.004 15.377 15.466 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.042 15.066 calculate_rho_elec 129 8.7 0.486 0.488 15.041 15.065 hybrid_alltoall_any 5248 16.5 1.308 3.051 11.344 13.315 make_images_data 5062 15.6 0.060 0.064 11.256 13.218 init_scf_run 11 5.9 0.000 0.001 11.958 11.959 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.958 11.959 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.548 9.578 dbcsr_mm_accdrv_process 20942 16.1 3.752 5.592 7.350 9.203 cp_fm_cholesky_decompose 22 10.9 8.877 8.919 8.877 8.919 grid_integrate_task_list 129 12.3 8.598 8.772 8.598 8.772 wfi_extrapolate 11 7.9 0.001 0.001 8.706 8.706 pw_transfer 1559 11.6 0.090 0.091 8.379 8.394 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.144 8.159 qs_energies_init_hamiltonians 11 5.9 0.005 0.010 7.946 7.948 fft_wrap_pw1pw2_140 527 13.2 0.542 0.545 7.210 7.231 mp_alltoall_d11v 2423 14.1 6.904 7.036 6.904 7.036 calculate_dm_sparse 129 9.5 0.001 0.001 6.776 6.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.452 6.501 grid_collocate_task_list 129 9.7 6.384 6.420 6.384 6.420 fft3d_ps 1301 14.7 2.774 2.785 6.173 6.185 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.107 6.172 density_rs2pw 129 9.7 0.005 0.005 6.055 6.083 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.460000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2635.454545, yerr=169.201528 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 1.261752E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.129 85.260 85.261 qs_energies 1 2.0 0.000 0.000 84.554 84.563 ls_scf 1 3.0 0.003 0.020 83.542 83.549 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.465 72.640 multiply_cannon 111 7.7 0.018 0.021 55.958 57.160 multiply_cannon_loop 111 8.7 0.213 0.224 52.504 53.863 ls_scf_main 1 4.0 0.000 0.002 52.309 52.327 density_matrix_trs4 2 5.0 0.002 0.003 46.800 46.883 ls_scf_init_scf 1 4.0 0.000 0.001 28.201 28.228 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.061 27.111 mp_waitall_1 11316 10.9 22.033 25.786 22.033 25.786 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.960 24.984 multiply_cannon_multrec 2664 9.7 8.185 8.866 15.601 17.309 multiply_cannon_sync_h2d 2664 9.7 13.883 15.778 13.883 15.778 make_m2s 222 7.7 0.008 0.010 12.956 13.580 make_images 222 8.7 0.099 0.108 12.934 13.559 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.489 12.780 make_images_data 222 9.7 0.004 0.005 7.525 8.102 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.601 7.033 8.064 hybrid_alltoall_any 227 10.6 0.216 1.832 6.436 7.701 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.325 7.346 dbcsr_mm_accdrv_process_sort 4760 11.4 6.317 7.250 6.317 7.250 calculate_norms 4752 9.8 5.563 6.200 5.563 6.200 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.021 5.225 mp_sum_l 807 5.4 3.043 4.411 3.043 4.411 arnoldi_extremal 4 6.8 0.000 0.000 3.363 3.399 arnoldi_normal_ev 4 7.8 0.001 0.003 3.363 3.399 make_images_sizes 222 9.7 0.000 0.000 0.744 3.307 mp_alltoall_i44 222 10.7 0.744 3.307 0.744 3.307 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.022 3.291 mp_irecv_dv 6231 10.9 2.005 3.265 2.005 3.265 build_subspace 16 8.4 0.009 0.012 3.257 3.259 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.216 3.075 ls_scf_post 1 4.0 0.000 0.001 3.029 3.036 ls_scf_store_result 1 5.0 0.000 0.000 2.850 2.891 dbcsr_special_finalize 555 9.7 0.005 0.006 2.336 2.806 dbcsr_merge_single_wm 555 10.7 0.455 0.583 2.328 2.798 make_images_pack 222 9.7 2.205 2.620 2.207 2.622 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.354 2.595 dbcsr_sort_data 658 11.4 2.131 2.538 2.131 2.538 dbcsr_matrix_vector_mult_local 304 10.0 2.059 2.450 2.061 2.451 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.224 2.319 buffer_matrices_ensure_size 222 8.7 1.742 2.033 1.742 2.033 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.793 1.821 rebuild_ks_matrix 3 7.3 0.000 0.000 1.784 1.811 qs_ks_build_kohn_sham_matrix 3 8.3 0.014 0.027 1.784 1.811 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.261000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 2.087465E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.066 91.430 91.432 qs_energies 1 2.0 0.000 0.000 90.865 90.880 ls_scf 1 3.0 0.002 0.018 89.557 89.573 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.514 75.880 multiply_cannon 111 7.7 0.028 0.042 53.711 57.679 ls_scf_main 1 4.0 0.000 0.000 55.069 55.078 multiply_cannon_loop 111 8.7 0.116 0.124 50.398 53.757 density_matrix_trs4 2 5.0 0.002 0.003 49.359 49.595 ls_scf_init_scf 1 4.0 0.000 0.002 30.965 30.966 mp_waitall_1 9246 10.9 21.374 30.129 21.374 30.129 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.675 29.799 multiply_cannon_multrec 1332 9.7 13.008 16.947 22.306 27.371 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.251 27.265 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.319 21.007 make_m2s 222 7.7 0.006 0.007 14.965 15.586 make_images 222 8.7 1.574 1.969 14.935 15.558 dbcsr_mm_accdrv_process 4041 10.4 0.272 0.441 8.901 10.421 dbcsr_mm_accdrv_process_sort 4041 11.4 8.498 9.980 8.498 9.980 make_images_data 222 9.7 0.004 0.004 8.454 9.312 hybrid_alltoall_any 227 10.6 0.522 2.487 7.875 8.754 mp_sum_l 807 5.4 5.555 8.425 5.555 8.425 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.222 7.919 mp_irecv_dv 3311 11.0 3.202 7.865 3.202 7.865 calculate_norms 2376 9.8 5.980 6.779 5.980 6.779 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.282 6.717 multiply_cannon_sync_h2d 1332 9.7 4.787 5.691 4.787 5.691 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.112 5.355 arnoldi_extremal 4 6.8 0.000 0.000 4.653 4.671 arnoldi_normal_ev 4 7.8 0.001 0.005 4.653 4.671 build_subspace 16 8.4 0.014 0.021 4.395 4.397 ls_scf_post 1 4.0 0.002 0.015 3.521 3.536 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.145 3.366 ls_scf_store_result 1 5.0 0.000 0.000 3.255 3.341 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.213 2.738 3.214 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.220 3.016 ls_scf_dm_to_ks 2 5.0 0.000 0.009 2.512 2.604 make_images_pack 222 9.7 2.024 2.407 2.027 2.409 mp_allgather_i34 111 8.7 1.011 2.256 1.011 2.256 dbcsr_sort_data 436 11.2 1.831 2.073 1.831 2.073 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.900 1.902 rebuild_ks_matrix 3 7.3 0.000 0.000 1.888 1.889 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.888 1.889 dbcsr_data_new 4174 10.1 1.611 1.866 1.611 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.432000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 2.715054E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.144 93.997 93.998 qs_energies 1 2.0 0.000 0.000 93.236 93.265 ls_scf 1 3.0 0.001 0.005 91.846 91.875 dbcsr_multiply_generic 111 6.7 0.016 0.022 76.593 76.867 ls_scf_main 1 4.0 0.000 0.000 57.553 57.558 multiply_cannon 111 7.7 0.044 0.113 52.894 56.724 multiply_cannon_loop 111 8.7 0.100 0.105 49.299 53.300 density_matrix_trs4 2 5.0 0.002 0.003 51.738 51.886 mp_waitall_1 7374 11.0 23.612 34.081 23.612 34.081 ls_scf_init_scf 1 4.0 0.000 0.002 30.704 30.706 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.443 29.503 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.019 27.035 multiply_cannon_multrec 888 9.7 12.612 15.364 21.263 24.600 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.973 23.341 make_m2s 222 7.7 0.006 0.007 16.794 18.019 make_images 222 8.7 1.968 2.298 16.756 17.980 make_images_data 222 9.7 0.004 0.004 9.608 10.708 hybrid_alltoall_any 227 10.6 0.624 2.861 9.251 10.523 dbcsr_mm_accdrv_process 3754 10.4 0.253 0.412 8.162 9.352 mp_sum_l 807 5.4 5.377 9.077 5.377 9.077 dbcsr_mm_accdrv_process_sort 3754 11.4 7.792 8.940 7.792 8.940 multiply_cannon_sync_h2d 888 9.7 6.040 7.499 6.040 7.499 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.165 7.184 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.464 7.012 mp_irecv_dv 2335 11.1 2.448 6.971 2.448 6.971 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.653 6.724 arnoldi_extremal 4 6.8 0.000 0.000 5.064 5.083 arnoldi_normal_ev 4 7.8 0.001 0.005 5.064 5.083 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.879 5.079 build_subspace 16 8.4 0.014 0.021 4.741 4.748 calculate_norms 1584 9.8 4.290 4.683 4.290 4.683 mp_allgather_i34 111 8.7 1.440 3.783 1.440 3.783 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.420 3.760 ls_scf_post 1 4.0 0.002 0.015 3.588 3.619 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 3.299 3.403 ls_scf_dm_to_ks 2 5.0 0.000 0.008 2.726 2.840 make_images_pack 222 9.7 1.812 2.135 1.815 2.138 dbcsr_sort_data 325 11.1 1.907 2.103 1.907 2.103 dbcsr_data_release 9322 10.9 1.317 2.001 1.317 2.001 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.949 1.951 rebuild_ks_matrix 3 7.3 0.000 0.000 1.931 1.933 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.931 1.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.998000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2215.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 3.305316E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.049 97.062 97.064 qs_energies 1 2.0 0.000 0.000 96.385 96.394 ls_scf 1 3.0 0.000 0.001 94.736 94.743 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.579 78.877 ls_scf_main 1 4.0 0.000 0.000 58.765 58.766 multiply_cannon 111 7.7 0.050 0.107 51.773 56.635 density_matrix_trs4 2 5.0 0.002 0.003 52.720 52.867 multiply_cannon_loop 111 8.7 0.115 0.127 46.720 49.518 ls_scf_init_scf 1 4.0 0.000 0.002 32.734 32.736 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.586 31.660 mp_waitall_1 6438 11.0 22.784 29.438 22.784 29.438 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.095 29.111 multiply_cannon_multrec 1332 9.7 14.184 17.298 22.141 25.037 make_m2s 222 7.7 0.007 0.008 21.098 22.533 make_images 222 8.7 3.143 3.591 21.048 22.485 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.291 17.308 make_images_data 222 9.7 0.004 0.004 11.715 13.330 hybrid_alltoall_any 227 10.6 0.797 3.810 11.068 12.704 dbcsr_mm_accdrv_process 3641 10.4 0.210 0.412 7.584 9.115 dbcsr_mm_accdrv_process_sort 3641 11.4 7.213 8.696 7.213 8.696 mp_sum_l 807 5.4 4.114 7.869 4.114 7.869 multiply_cannon_sync_h2d 1332 9.7 5.516 6.294 5.516 6.294 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.206 6.250 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.063 5.992 mp_irecv_dv 3229 10.9 2.041 5.915 2.041 5.915 arnoldi_extremal 4 6.8 0.000 0.000 5.285 5.302 arnoldi_normal_ev 4 7.8 0.001 0.004 5.285 5.302 build_subspace 16 8.4 0.014 0.021 4.952 4.961 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.598 4.939 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.569 4.758 calculate_norms 2376 9.8 4.176 4.517 4.176 4.517 mp_allgather_i34 111 8.7 2.182 4.469 2.182 4.469 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.592 3.897 dbcsr_matrix_vector_mult_local 304 10.0 3.181 3.683 3.183 3.685 dbcsr_sort_data 658 11.4 3.061 3.462 3.061 3.462 ls_scf_post 1 4.0 0.000 0.001 3.237 3.246 dbcsr_special_finalize 555 9.7 0.006 0.007 2.801 3.218 dbcsr_merge_single_wm 555 10.7 0.534 0.664 2.792 3.209 ls_scf_store_result 1 5.0 0.000 0.000 2.982 3.039 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.902 2.965 dbcsr_data_release 10477 10.7 1.599 2.472 1.599 2.472 dbcsr_finalize 304 7.8 0.049 0.061 1.809 1.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.064000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 4.630286E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.094 0.099 98.459 98.460 qs_energies 1 2.0 0.000 0.000 97.553 97.558 ls_scf 1 3.0 0.000 0.000 95.648 95.654 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.405 77.603 ls_scf_main 1 4.0 0.000 0.000 61.626 61.627 multiply_cannon 111 7.7 0.087 0.176 55.111 60.789 density_matrix_trs4 2 5.0 0.002 0.003 54.630 54.749 multiply_cannon_loop 111 8.7 0.069 0.077 50.480 52.404 mp_waitall_1 5481 11.0 25.892 31.400 25.892 31.400 ls_scf_init_scf 1 4.0 0.000 0.000 30.414 30.419 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.216 29.251 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.063 27.070 multiply_cannon_multrec 444 9.7 13.928 16.288 21.059 23.263 make_m2s 222 7.7 0.004 0.005 17.381 19.984 make_images 222 8.7 3.719 4.411 17.319 19.925 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.837 15.865 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.161 14.482 make_images_data 222 9.7 0.003 0.004 9.689 12.133 hybrid_alltoall_any 227 10.6 0.790 3.778 9.413 11.914 dbcsr_mm_accdrv_process 3003 10.4 0.192 0.354 6.842 7.965 dbcsr_mm_accdrv_process_sort 3003 11.4 6.522 7.619 6.522 7.619 multiply_cannon_sync_h2d 444 9.7 6.558 7.603 6.558 7.603 mp_allgather_i34 111 8.7 2.858 7.015 2.858 7.015 arnoldi_extremal 4 6.8 0.000 0.000 5.852 5.860 arnoldi_normal_ev 4 7.8 0.002 0.005 5.852 5.860 mp_sum_l 807 5.4 2.958 5.654 2.958 5.654 build_subspace 16 8.4 0.015 0.020 5.466 5.476 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.618 4.792 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.684 4.710 mp_irecv_dv 1241 11.2 1.663 4.681 1.663 4.681 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.214 4.377 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.028 4.366 dbcsr_matrix_vector_mult_local 304 10.0 3.700 4.164 3.702 4.166 calculate_norms 792 9.8 3.538 3.676 3.538 3.676 ls_scf_post 1 4.0 0.000 0.000 3.609 3.613 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.367 3.464 ls_scf_store_result 1 5.0 0.000 0.000 3.375 3.432 make_images_sizes 222 9.7 0.000 0.000 1.048 3.276 mp_alltoall_i44 222 10.7 1.048 3.276 1.048 3.276 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.319 dbcsr_merge_all 275 8.9 0.475 0.532 2.051 2.148 dbcsr_data_release 10123 10.8 1.332 2.039 1.332 2.039 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.976 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.460000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3656.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbe71a2287d23928c5ebe2640c2d74e0055badb1_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 8.737165E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.104 0.144 107.160 107.161 qs_energies 1 2.0 0.000 0.000 105.657 105.668 ls_scf 1 3.0 0.000 0.000 102.774 102.785 dbcsr_multiply_generic 111 6.7 0.023 0.026 76.976 77.073 ls_scf_main 1 4.0 0.000 0.000 65.244 65.252 density_matrix_trs4 2 5.0 0.002 0.003 56.457 56.512 multiply_cannon 111 7.7 0.152 0.234 49.537 51.673 multiply_cannon_loop 111 8.7 0.068 0.071 46.042 46.678 ls_scf_init_scf 1 4.0 0.000 0.000 33.837 33.845 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.470 32.496 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.630 29.641 mp_waitall_1 4569 11.1 21.810 25.355 21.810 25.355 make_m2s 222 7.7 0.005 0.005 23.889 24.924 make_images 222 8.7 4.590 4.974 23.783 24.816 multiply_cannon_multrec 444 9.7 17.829 18.548 22.498 23.106 hybrid_alltoall_any 227 10.6 1.661 3.617 12.858 15.947 make_images_data 222 9.7 0.003 0.003 13.132 15.636 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.091 10.818 multiply_cannon_sync_h2d 444 9.7 8.851 8.885 8.851 8.885 arnoldi_extremal 4 6.8 0.000 0.000 7.355 7.371 arnoldi_normal_ev 4 7.8 0.003 0.009 7.355 7.371 build_subspace 16 8.4 0.026 0.036 6.803 6.815 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.450 5.622 dbcsr_matrix_vector_mult_local 304 10.0 5.050 5.363 5.053 5.366 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.985 5.240 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.769 4.858 dbcsr_mm_accdrv_process 1814 10.4 0.238 0.335 4.507 4.645 dbcsr_mm_accdrv_process_sort 1814 11.4 4.197 4.350 4.197 4.350 ls_scf_post 1 4.0 0.000 0.000 3.693 3.704 make_images_sizes 222 9.7 0.000 0.000 1.424 3.551 mp_alltoall_i44 222 10.7 1.424 3.550 1.424 3.550 mp_allgather_i34 111 8.7 1.143 3.535 1.143 3.535 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.421 calculate_norms 792 9.8 3.241 3.274 3.241 3.274 dbcsr_finalize 304 7.8 0.082 0.090 3.092 3.133 dbcsr_merge_all 275 8.9 0.892 0.921 2.877 2.923 dbcsr_complete_redistribute 5 7.6 1.432 1.466 2.766 2.871 dbcsr_data_release 12724 10.6 2.338 2.862 2.338 2.862 qs_energies_init_hamiltonians 1 3.0 0.008 0.009 2.852 2.853 matrix_ls_to_qs 2 6.0 0.000 0.000 2.415 2.544 dbcsr_sort_data 325 11.1 2.446 2.504 2.446 2.504 dbcsr_new_transposed 4 7.5 0.243 0.252 2.269 2.285 dbcsr_frobenius_norm 74 6.6 2.057 2.134 2.185 2.214 dbcsr_add_d 103 6.2 0.000 0.000 2.135 2.207 dbcsr_add_anytype 103 7.2 0.860 0.894 2.134 2.207 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.203 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.161000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6856.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: bbe71a2287d23928c5ebe2640c2d74e0055badb1 Summary: empty Status: OK