=== 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: d8f36242127b9a828f127550ecd9613eefb3f1cc ################# 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/01 job id: 43399876 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/02 job id: 43399878 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/03 job id: 43399881 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/04 job id: 43399883 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/05 job id: 43399884 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/06 job id: 43399886 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/07 job id: 43399887 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/08 job id: 43399889 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/09 job id: 43399894 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/10 job id: 43399897 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/11 job id: 43399899 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/12 job id: 43399902 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/13 job id: 43399903 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/14 job id: 43399907 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/15 job id: 43399909 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/16 job id: 43399910 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/17 job id: 43399912 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/18 job id: 43399913 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/19 job id: 43399915 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/20 job id: 43399917 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/21 job id: 43399919 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/22 job id: 43399921 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/23 job id: 43399923 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/24 job id: 43399926 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/25 job id: 43399928 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/26 job id: 43399929 --- 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/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 132.623 132.624 farming_run 1 2.0 132.071 132.073 132.594 132.598 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.456583E+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 0.011 0.033 114.973 114.974 qs_energies 1 2.0 0.000 0.000 114.726 114.728 mp2_main 1 3.0 0.000 0.000 112.751 112.753 mp2_gpw_main 1 4.0 0.020 0.027 111.815 111.816 mp2_ri_gpw_compute_in 1 5.0 0.173 0.202 92.835 93.383 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.399 55.948 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.711 47.174 get_2c_integrals 1 6.0 0.000 0.000 36.737 37.264 integrate_v_rspace 273 8.0 0.434 0.452 25.094 30.276 pw_transfer 6555 10.6 0.377 0.387 27.312 27.874 grid_integrate_task_list 273 9.0 20.912 26.583 20.912 26.583 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.991 26.544 fft_wrap_pw1pw2_100 2178 12.4 1.167 1.251 23.557 24.113 compute_2c_integrals 1 7.0 0.003 0.003 19.217 19.218 rpa_ri_compute_en 1 5.0 0.011 0.015 18.871 19.077 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.827 18.923 mp2_eri_2c_integrate_gpw 1 9.0 2.403 2.442 18.824 18.920 cp_fm_cholesky_decompose 12 8.2 17.493 18.014 17.493 18.014 cholesky_decomp 1 7.0 0.000 0.000 16.358 16.870 fft3d_s 5443 13.4 16.146 16.719 16.168 16.739 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.556 10.850 15.556 calculate_wavefunction 272 8.0 5.401 5.555 12.532 13.148 rpa_num_int 1 6.0 0.001 0.014 10.557 10.566 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.523 10.542 calc_mat_Q 8 8.0 0.000 0.000 9.332 9.418 contract_S_to_Q 8 9.0 0.000 0.000 8.754 8.841 calc_potential_gpw 544 9.5 0.005 0.005 8.258 8.573 parallel_gemm_fm 14 9.1 0.000 0.000 8.345 8.428 parallel_gemm_fm_cosma 14 10.1 8.345 8.428 8.345 8.428 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.174 8.420 potential_pw2rs 545 10.0 0.107 0.109 7.642 8.327 collocate_single_gaussian 272 10.0 0.042 0.043 7.430 7.674 create_integ_mat 1 6.0 0.014 0.026 7.656 7.665 array2fm 1 7.0 0.000 0.000 6.745 7.139 pw_scatter_s 2720 13.7 4.443 4.597 4.443 4.597 pw_gather_s 2722 13.2 3.869 4.208 3.869 4.208 array2fm_buffer_send 1 8.0 3.006 3.172 3.006 3.172 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.813515, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 0.027 0.040 395.792 395.793 farming_run 1 2.0 394.612 394.690 395.752 395.755 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223700E+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 0.013 0.048 209.099 209.100 qs_energies 1 2.0 0.000 0.000 208.791 208.805 scf_env_do_scf 1 3.0 0.000 0.000 106.102 106.102 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.228 105.238 rebuild_ks_matrix 4 6.0 0.000 0.000 105.227 105.237 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.067 105.226 105.237 hfx_ks_matrix 4 8.0 0.001 0.001 104.855 104.858 integrate_four_center 4 9.0 0.144 0.456 104.854 104.857 mp2_main 1 3.0 0.000 0.000 102.402 102.416 mp2_gpw_main 1 4.0 0.031 0.052 101.508 101.526 integrate_four_center_main 4 10.0 0.122 0.564 96.828 99.365 integrate_four_center_bin 265 11.0 96.706 99.015 96.706 99.015 init_scf_loop 1 4.0 0.000 0.000 91.812 91.813 mp2_ri_gpw_compute_in 1 5.0 0.068 0.097 74.834 75.856 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.432 55.461 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 42.168 47.277 integrate_v_rspace 95 8.0 0.399 0.568 28.535 33.453 pw_transfer 2240 10.6 0.145 0.166 29.937 30.347 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.936 29.366 grid_integrate_task_list 95 9.0 23.831 28.935 23.831 28.935 ao_to_mo_and_store_B_mult_1 91 7.0 10.578 28.906 10.578 28.906 mp2_ri_gpw_compute_en 1 5.0 0.061 0.085 26.499 28.192 fft_wrap_pw1pw2_100 730 12.4 1.305 1.486 26.682 27.100 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.904 24.795 24.810 get_2c_integrals 1 6.0 0.002 0.015 20.306 20.339 compute_2c_integrals 1 7.0 0.005 0.025 19.281 19.290 compute_2c_integrals_loop_lm 1 8.0 0.006 0.044 18.970 19.149 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.874 18.964 19.147 fft3d_s 1823 13.4 18.409 18.888 18.422 18.900 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.288 14.289 calculate_wavefunction 91 8.0 2.024 2.056 9.744 9.999 potential_pw2rs 186 10.0 0.034 0.035 8.632 9.244 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.594 8.741 9.240 local_gemm 172 8.0 8.184 8.666 8.184 8.666 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.018 8.304 8.638 mp2_ri_gpw_compute_en_comm 22 7.0 0.517 0.546 7.805 8.238 calc_potential_gpw 182 9.5 0.002 0.002 7.971 8.185 collocate_single_gaussian 91 10.0 0.019 0.065 7.920 8.122 mp2_ri_gpw_compute_en_ener 172 7.0 6.363 6.455 6.363 6.455 mp_sendrecv_dm3 2068 8.0 5.793 6.229 5.793 6.229 pw_gather_s 912 13.2 4.899 5.457 4.899 5.457 mp_sync 38 10.4 2.709 5.255 2.709 5.255 pw_scatter_s 910 13.7 3.963 4.376 3.963 4.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.511350, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 451.784704E+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 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 0.027 0.056 52.189 52.190 qs_mol_dyn_low 1 2.0 0.005 0.016 51.867 51.874 qs_forces 11 3.9 0.002 0.006 51.803 51.806 qs_energies 11 4.9 0.010 0.017 50.337 50.348 scf_env_do_scf 11 5.9 0.002 0.006 44.127 44.127 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.983 41.984 dbcsr_multiply_generic 2286 12.5 0.095 0.112 33.448 33.912 qs_scf_new_mos 108 7.5 0.000 0.001 31.930 32.223 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.929 32.223 ot_scf_mini 108 9.5 0.002 0.002 30.259 30.480 multiply_cannon 2286 13.5 0.181 0.189 25.957 27.702 multiply_cannon_loop 2286 14.5 1.507 1.581 25.221 26.980 velocity_verlet 10 3.0 0.001 0.001 26.051 26.051 ot_mini 108 10.5 0.001 0.001 19.177 19.422 qs_ot_get_derivative 108 11.5 0.001 0.001 16.172 16.419 mp_waitall_1 267858 16.1 8.807 14.655 8.807 14.655 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 6.005 12.871 multiply_cannon_multrec 54864 15.5 4.168 6.481 7.483 11.037 rebuild_ks_matrix 119 8.3 0.000 0.000 7.896 8.039 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.896 8.038 multiply_cannon_sync_h2d 54864 15.5 5.864 7.234 5.864 7.234 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.965 7.099 mp_sum_l 7207 12.9 5.436 7.033 5.436 7.033 qs_ot_get_p 119 10.4 0.001 0.001 6.395 6.690 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.608 6.047 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.229 5.357 init_scf_run 11 5.9 0.000 0.001 4.903 4.903 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.903 4.903 dbcsr_mm_accdrv_process 76910 16.1 1.197 1.859 3.237 4.664 sum_up_and_integrate 119 10.3 0.012 0.015 4.523 4.532 integrate_v_rspace 119 11.3 0.002 0.002 4.511 4.522 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.116 4.226 calculate_rho_elec 119 8.7 0.011 0.016 4.116 4.226 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.488 3.561 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.897 3.198 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.899 3.089 apply_single 119 13.6 0.000 0.000 2.898 3.089 calculate_dm_sparse 119 9.5 0.000 0.000 2.884 3.004 rs_pw_transfer 974 11.9 0.011 0.012 2.834 2.953 jit_kernel_multiply 13 15.8 1.977 2.745 1.977 2.745 ot_diis_step 108 11.5 0.006 0.006 2.709 2.710 calculate_first_density_matrix 1 7.0 0.000 0.001 2.707 2.709 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.669 2.670 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.499 2.565 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.496 2.498 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.441 2.442 cp_fm_redistribute_end 50 14.0 2.216 2.419 2.221 2.420 cp_fm_diag_elpa_base 50 14.0 0.198 2.379 0.199 2.389 density_rs2pw 119 9.7 0.004 0.004 2.192 2.275 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.197 2.252 acc_transpose_blocks 54864 15.5 0.231 0.247 1.751 2.171 wfi_extrapolate 11 7.9 0.001 0.001 2.142 2.142 init_scf_loop 11 6.9 0.000 0.000 2.119 2.121 grid_integrate_task_list 119 12.3 2.032 2.117 2.032 2.117 mp_sum_d 4125 12.0 1.376 2.008 1.376 2.008 potential_pw2rs 119 12.3 0.004 0.004 1.884 1.900 make_m2s 4572 13.5 0.053 0.055 1.747 1.788 pw_transfer 1439 11.6 0.051 0.055 1.646 1.723 make_images 4572 14.5 0.132 0.138 1.666 1.706 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.571 1.649 mp_alltoall_d11v 2130 13.8 1.352 1.522 1.352 1.522 fft3d_ps 1201 14.6 0.355 0.458 1.349 1.424 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.373 1.394 mp_waitany 12084 13.8 1.252 1.387 1.252 1.387 grid_collocate_task_list 119 9.7 1.288 1.350 1.288 1.350 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.185 1.261 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.774 1.153 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.190000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.567360E+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 1157366. 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 0.032 0.075 39.835 39.835 qs_mol_dyn_low 1 2.0 0.003 0.003 39.520 39.531 qs_forces 11 3.9 0.002 0.002 39.460 39.460 qs_energies 11 4.9 0.037 0.118 37.765 37.769 scf_env_do_scf 11 5.9 0.001 0.002 32.314 32.314 scf_env_do_scf_inner_loop 108 6.5 0.003 0.010 29.693 29.693 dbcsr_multiply_generic 2286 12.5 0.099 0.103 21.718 22.066 qs_scf_new_mos 108 7.5 0.001 0.001 20.353 20.603 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.352 20.602 ot_scf_mini 108 9.5 0.003 0.003 19.428 19.602 velocity_verlet 10 3.0 0.001 0.001 18.693 18.693 multiply_cannon 2286 13.5 0.209 0.219 16.424 18.280 multiply_cannon_loop 2286 14.5 0.898 0.969 15.251 16.593 ot_mini 108 10.5 0.001 0.001 12.009 12.248 mp_waitall_1 217478 16.2 6.356 11.530 6.356 11.530 qs_ot_get_derivative 108 11.5 0.001 0.001 9.516 9.696 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.208 9.544 multiply_cannon_multrec 27432 15.5 1.965 4.458 5.875 8.702 rebuild_ks_matrix 119 8.3 0.000 0.000 7.336 7.473 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 7.335 7.472 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.496 6.619 dbcsr_mm_accdrv_process 47894 16.0 3.044 4.907 3.841 5.709 qs_ot_get_p 119 10.4 0.001 0.001 4.453 4.691 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.632 4.484 sum_up_and_integrate 119 10.3 0.030 0.036 4.329 4.336 integrate_v_rspace 119 11.3 0.002 0.002 4.299 4.312 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.064 4.140 apply_single 119 13.6 0.000 0.000 3.064 4.140 init_scf_run 11 5.9 0.000 0.001 4.038 4.038 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.037 4.038 mp_sum_l 7207 12.9 2.175 4.036 2.175 4.036 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.822 3.855 calculate_rho_elec 119 8.7 0.021 0.024 3.821 3.855 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.923 2.942 rs_pw_transfer 974 11.9 0.010 0.011 2.800 2.919 make_m2s 4572 13.5 0.052 0.054 2.694 2.918 multiply_cannon_sync_h2d 27432 15.5 2.185 2.893 2.185 2.893 make_images 4572 14.5 0.199 0.236 2.607 2.830 init_scf_loop 11 6.9 0.001 0.005 2.600 2.601 calculate_first_density_matrix 1 7.0 0.007 0.035 2.463 2.465 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.461 2.461 ot_diis_step 108 11.5 0.010 0.011 2.435 2.436 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.235 2.327 density_rs2pw 119 9.7 0.004 0.004 2.175 2.260 calculate_dm_sparse 119 9.5 0.000 0.000 2.066 2.149 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.077 2.078 potential_pw2rs 119 12.3 0.006 0.006 2.055 2.070 cp_fm_redistribute_end 50 14.0 1.711 2.052 1.716 2.054 cp_fm_diag_elpa_base 50 14.0 0.328 1.988 0.337 2.021 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.971 1.974 grid_integrate_task_list 119 12.3 1.835 1.950 1.835 1.950 pw_transfer 1439 11.6 0.063 0.067 1.919 1.946 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.829 1.858 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.733 1.772 prepare_preconditioner 11 7.9 0.000 0.000 1.670 1.699 make_preconditioner 11 8.9 0.000 0.002 1.670 1.699 make_images_data 4572 15.5 0.045 0.052 1.216 1.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.568 1.630 jit_kernel_multiply 8 16.1 0.745 1.575 0.745 1.575 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.055 1.563 fft3d_ps 1201 14.6 0.501 0.552 1.537 1.562 acc_transpose_blocks 27432 15.5 0.108 0.113 1.199 1.506 wfi_extrapolate 11 7.9 0.001 0.001 1.502 1.503 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.377 1.406 mp_allgather_i34 2286 14.5 0.609 1.390 0.609 1.390 mp_alltoall_d11v 2130 13.8 1.277 1.374 1.277 1.374 grid_collocate_task_list 119 9.7 1.237 1.341 1.237 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.256 1.266 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.218 1.263 mp_sum_d 4125 12.0 0.684 1.176 0.684 1.176 qs_energies_init_hamiltonians 11 5.9 0.018 0.047 1.017 1.019 make_images_sizes 4572 15.5 0.005 0.005 0.667 0.951 mp_alltoall_i44 4572 16.5 0.662 0.947 0.662 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.914 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.671 0.895 mp_alltoall_z22v 1201 16.6 0.786 0.860 0.786 0.860 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.609 0.798 0.854 dbcsr_dot_sd 1205 11.9 0.071 0.086 0.442 0.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.835000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=1.431638 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 522.027008E+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 0.071 0.132 32.673 32.693 qs_mol_dyn_low 1 2.0 0.003 0.003 32.193 32.211 qs_forces 11 3.9 0.002 0.002 32.106 32.107 qs_energies 11 4.9 0.004 0.022 30.534 30.535 scf_env_do_scf 11 5.9 0.001 0.013 25.469 25.470 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.880 22.883 dbcsr_multiply_generic 2286 12.5 0.094 0.101 16.542 16.622 velocity_verlet 10 3.0 0.001 0.001 15.276 15.277 qs_scf_new_mos 108 7.5 0.001 0.001 14.835 14.856 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.834 14.855 ot_scf_mini 108 9.5 0.002 0.004 14.113 14.125 multiply_cannon 2286 13.5 0.196 0.200 13.162 13.934 multiply_cannon_loop 2286 14.5 0.634 0.667 12.362 13.153 ot_mini 108 10.5 0.001 0.001 8.794 8.812 qs_ot_get_derivative 108 11.5 0.001 0.001 7.291 7.302 multiply_cannon_multrec 18288 15.5 1.917 2.809 6.825 7.116 rebuild_ks_matrix 119 8.3 0.000 0.000 6.349 6.364 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.349 6.364 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.607 5.621 dbcsr_mm_accdrv_process 38222 16.0 4.158 5.448 4.824 5.594 sum_up_and_integrate 119 10.3 0.031 0.034 3.979 3.984 integrate_v_rspace 119 11.3 0.002 0.003 3.948 3.952 mp_waitall_1 169478 16.3 3.018 3.829 3.018 3.829 init_scf_run 11 5.9 0.000 0.001 3.665 3.666 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.665 3.665 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.744 3.361 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.295 3.298 calculate_rho_elec 119 8.7 0.030 0.031 3.294 3.298 qs_ot_get_p 119 10.4 0.001 0.002 3.143 3.160 init_scf_loop 11 6.9 0.001 0.005 2.560 2.566 rs_pw_transfer 974 11.9 0.009 0.010 2.350 2.442 calculate_first_density_matrix 1 7.0 0.000 0.003 2.434 2.435 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.037 2.361 apply_single 119 13.6 0.000 0.000 2.037 2.360 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.053 2.058 make_m2s 4572 13.5 0.045 0.048 1.889 2.051 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.424 2.033 make_images 4572 14.5 0.191 0.204 1.805 1.966 density_rs2pw 119 9.7 0.004 0.004 1.879 1.964 grid_integrate_task_list 119 12.3 1.806 1.884 1.806 1.884 calculate_dm_sparse 119 9.5 0.000 0.000 1.857 1.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.821 1.823 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.780 1.780 prepare_preconditioner 11 7.9 0.000 0.000 1.772 1.774 make_preconditioner 11 8.9 0.000 0.002 1.772 1.774 potential_pw2rs 119 12.3 0.007 0.008 1.763 1.771 pw_transfer 1439 11.6 0.064 0.066 1.756 1.768 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.627 1.708 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.664 1.677 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.623 1.630 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.601 1.609 jit_kernel_multiply 8 16.5 0.615 1.606 0.615 1.606 multiply_cannon_sync_h2d 18288 15.5 1.426 1.547 1.426 1.547 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.540 1.541 cp_fm_redistribute_end 50 14.0 1.143 1.519 1.144 1.520 cp_fm_diag_elpa_base 50 14.0 0.359 1.456 0.374 1.495 ot_diis_step 108 11.5 0.011 0.011 1.482 1.482 mp_sum_l 7207 12.9 1.134 1.420 1.134 1.420 fft3d_ps 1201 14.6 0.507 0.523 1.349 1.364 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.314 1.326 grid_collocate_task_list 119 9.7 1.206 1.284 1.206 1.284 acc_transpose_blocks 18288 15.5 0.074 0.076 1.255 1.279 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.229 1.233 wfi_extrapolate 11 7.9 0.001 0.001 1.183 1.184 make_images_data 4572 15.5 0.045 0.050 0.841 1.022 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.999 0.999 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.707 0.915 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.887 0.906 acc_transpose_blocks_kernels 18288 16.5 0.208 0.217 0.820 0.833 mp_alltoall_d11v 2130 13.8 0.707 0.821 0.707 0.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.815 0.817 mp_alltoall_z22v 1201 16.6 0.684 0.753 0.684 0.753 cp_fm_cholesky_invert 11 10.9 0.748 0.752 0.748 0.752 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.733 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.693000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.090909, yerr=1.781447 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.224512E+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 0.020 0.064 35.325 35.326 qs_mol_dyn_low 1 2.0 0.003 0.003 35.033 35.040 qs_forces 11 3.9 0.002 0.003 34.975 34.976 qs_energies 11 4.9 0.002 0.008 33.268 33.275 scf_env_do_scf 11 5.9 0.001 0.001 27.930 27.931 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.440 24.440 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.173 18.304 velocity_verlet 10 3.0 0.001 0.001 17.865 17.867 qs_scf_new_mos 108 7.5 0.001 0.001 16.149 16.205 qs_scf_loop_do_ot 108 8.5 0.001 0.002 16.148 16.204 ot_scf_mini 108 9.5 0.002 0.004 15.209 15.267 multiply_cannon 2286 13.5 0.229 0.269 14.377 14.828 multiply_cannon_loop 2286 14.5 0.937 0.967 13.389 13.706 ot_mini 108 10.5 0.001 0.001 9.313 9.376 multiply_cannon_multrec 27432 15.5 2.324 2.983 8.589 8.954 qs_ot_get_derivative 108 11.5 0.001 0.001 7.498 7.556 dbcsr_mm_accdrv_process 47916 15.9 5.394 7.066 6.170 7.365 rebuild_ks_matrix 119 8.3 0.000 0.000 6.544 6.588 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.544 6.587 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.814 5.852 init_scf_run 11 5.9 0.000 0.001 3.874 3.875 scf_env_initial_rho_setup 11 6.9 0.002 0.003 3.874 3.875 sum_up_and_integrate 119 10.3 0.035 0.038 3.801 3.812 integrate_v_rspace 119 11.3 0.002 0.003 3.766 3.777 init_scf_loop 11 6.9 0.000 0.001 3.467 3.468 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.382 3.419 calculate_rho_elec 119 8.7 0.040 0.046 3.382 3.418 qs_ot_get_p 119 10.4 0.001 0.001 3.260 3.342 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.726 3.192 prepare_preconditioner 11 7.9 0.000 0.000 2.608 2.615 make_preconditioner 11 8.9 0.000 0.001 2.608 2.615 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.111 2.531 apply_single 119 13.6 0.000 0.000 2.110 2.531 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.200 2.528 mp_waitall_1 145218 16.4 2.162 2.522 2.162 2.522 calculate_first_density_matrix 1 7.0 0.000 0.003 2.478 2.480 make_m2s 4572 13.5 0.054 0.056 2.291 2.390 make_images 4572 14.5 0.272 0.333 2.184 2.281 rs_pw_transfer 974 11.9 0.009 0.009 2.135 2.276 calculate_dm_sparse 119 9.5 0.000 0.000 2.028 2.088 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.000 2.027 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.964 1.975 density_rs2pw 119 9.7 0.003 0.004 1.855 1.974 grid_integrate_task_list 119 12.3 1.835 1.920 1.835 1.920 pw_transfer 1439 11.6 0.063 0.067 1.826 1.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.854 1.855 ot_diis_step 108 11.5 0.012 0.012 1.771 1.771 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.736 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.659 1.671 mp_sum_l 7207 12.9 1.090 1.659 1.090 1.659 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.641 1.641 potential_pw2rs 119 12.3 0.008 0.009 1.585 1.590 acc_transpose_blocks 27432 15.5 0.111 0.112 1.453 1.471 jit_kernel_multiply 8 15.8 0.715 1.464 0.715 1.464 fft3d_ps 1201 14.6 0.535 0.589 1.418 1.441 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.392 1.426 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.389 1.390 cp_fm_redistribute_end 50 14.0 0.912 1.362 0.914 1.363 wfi_extrapolate 11 7.9 0.001 0.001 1.348 1.348 cp_fm_diag_elpa_base 50 14.0 0.427 1.301 0.447 1.342 grid_collocate_task_list 119 9.7 1.222 1.323 1.222 1.323 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.211 1.223 qs_energies_init_hamiltonians 11 5.9 0.004 0.023 1.197 1.198 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.127 1.143 cp_fm_upper_to_full 72 13.5 0.792 1.104 0.792 1.104 dbcsr_complete_redistribute 329 12.2 0.124 0.148 0.824 1.097 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.690 1.094 multiply_cannon_sync_h2d 27432 15.5 0.994 1.085 0.994 1.085 make_images_data 4572 15.5 0.045 0.048 0.910 1.040 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.777 0.961 mp_alltoall_d11v 2130 13.8 0.781 0.903 0.781 0.903 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.601 0.868 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 acc_transpose_blocks_kernels 27432 16.5 0.267 0.276 0.833 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.820 0.824 cp_fm_cholesky_invert 11 10.9 0.793 0.796 0.793 0.796 mp_alltoall_z22v 1201 16.6 0.765 0.794 0.765 0.794 mp_alltoall_i22 627 13.8 0.440 0.735 0.440 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.326000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.090909, yerr=3.502066 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 599.769088E+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 0.012 0.029 27.729 27.729 qs_mol_dyn_low 1 2.0 0.003 0.003 27.559 27.566 qs_forces 11 3.9 0.002 0.002 27.501 27.502 qs_energies 11 4.9 0.001 0.001 25.806 25.809 scf_env_do_scf 11 5.9 0.000 0.001 21.160 21.160 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.643 18.643 velocity_verlet 10 3.0 0.001 0.001 14.159 14.162 dbcsr_multiply_generic 2286 12.5 0.091 0.094 12.208 12.308 qs_scf_new_mos 108 7.5 0.001 0.001 11.147 11.167 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.147 11.166 ot_scf_mini 108 9.5 0.002 0.002 10.475 10.499 multiply_cannon 2286 13.5 0.229 0.233 9.660 10.137 multiply_cannon_loop 2286 14.5 0.329 0.340 8.750 8.961 multiply_cannon_multrec 9144 15.5 1.613 1.870 5.763 5.957 ot_mini 108 10.5 0.001 0.001 5.867 5.898 rebuild_ks_matrix 119 8.3 0.000 0.000 5.763 5.787 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.763 5.787 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.137 5.159 qs_ot_get_derivative 108 11.5 0.001 0.001 4.577 4.601 dbcsr_mm_accdrv_process 12550 15.8 3.081 4.061 4.049 4.119 sum_up_and_integrate 119 10.3 0.038 0.042 3.542 3.548 integrate_v_rspace 119 11.3 0.003 0.003 3.504 3.510 init_scf_run 11 5.9 0.000 0.001 3.225 3.225 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.225 3.225 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.174 3.179 calculate_rho_elec 119 8.7 0.059 0.061 3.174 3.178 qs_ot_get_p 119 10.4 0.001 0.001 2.834 2.873 init_scf_loop 11 6.9 0.000 0.000 2.495 2.497 mp_waitall_1 121218 16.5 1.798 2.352 1.798 2.352 calculate_first_density_matrix 1 7.0 0.000 0.000 2.095 2.095 make_m2s 4572 13.5 0.034 0.035 1.756 1.931 grid_integrate_task_list 119 12.3 1.853 1.922 1.853 1.922 qs_ot_p2m_diag 50 11.0 0.023 0.023 1.877 1.880 make_images 4572 14.5 0.268 0.302 1.668 1.841 prepare_preconditioner 11 7.9 0.000 0.000 1.768 1.773 make_preconditioner 11 8.9 0.000 0.000 1.768 1.773 jit_kernel_multiply 10 15.5 0.929 1.742 0.929 1.742 rs_pw_transfer 974 11.9 0.008 0.008 1.669 1.738 calculate_dm_sparse 119 9.5 0.000 0.000 1.703 1.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.660 1.689 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.681 1.682 pw_transfer 1439 11.6 0.064 0.065 1.661 1.670 density_rs2pw 119 9.7 0.003 0.003 1.594 1.658 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.570 1.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.556 1.557 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.529 1.545 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.385 1.386 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.367 1.375 cp_fm_redistribute_end 50 14.0 0.687 1.362 0.688 1.363 grid_collocate_task_list 119 9.7 1.273 1.350 1.273 1.350 cp_fm_diag_elpa_base 50 14.0 0.628 1.288 0.673 1.346 potential_pw2rs 119 12.3 0.010 0.010 1.336 1.338 ot_diis_step 108 11.5 0.012 0.013 1.277 1.278 fft_wrap_pw1pw2_140 487 13.2 0.083 0.089 1.269 1.276 fft3d_ps 1201 14.6 0.540 0.558 1.236 1.244 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.214 1.215 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.165 1.179 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.152 1.177 apply_single 119 13.6 0.000 0.000 1.152 1.177 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.123 1.129 wfi_extrapolate 11 7.9 0.001 0.001 1.086 1.086 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.793 1.046 make_images_data 4572 15.5 0.039 0.043 0.817 1.022 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.926 acc_transpose_blocks 9144 15.5 0.038 0.039 0.901 0.908 cp_fm_cholesky_invert 11 10.9 0.896 0.899 0.896 0.899 mp_alltoall_d11v 2130 13.8 0.775 0.877 0.775 0.877 multiply_cannon_sync_h2d 9144 15.5 0.707 0.800 0.707 0.800 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.748 0.750 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.729 0.737 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.668 0.715 acc_transpose_blocks_kernels 9144 16.5 0.116 0.118 0.662 0.666 mp_allgather_i34 2286 14.5 0.244 0.643 0.244 0.643 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.369 0.619 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.390 0.613 mp_alltoall_z22v 1201 16.6 0.571 0.607 0.571 0.607 dbcsr_complete_redistribute 329 12.2 0.183 0.198 0.554 0.584 qs_create_task_list 11 7.9 0.000 0.000 0.545 0.569 generate_qs_task_list 11 8.9 0.190 0.213 0.544 0.569 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.729000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.818182, yerr=4.647936 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 752.771072E+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 0.040 0.063 41.667 41.667 qs_mol_dyn_low 1 2.0 0.003 0.003 41.338 41.346 qs_forces 11 3.9 0.002 0.002 41.268 41.269 qs_energies 11 4.9 0.001 0.001 39.301 39.304 scf_env_do_scf 11 5.9 0.001 0.002 33.331 33.331 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.473 25.475 velocity_verlet 10 3.0 0.001 0.001 23.259 23.265 dbcsr_multiply_generic 2286 12.5 0.099 0.100 18.246 18.394 qs_scf_new_mos 108 7.5 0.001 0.001 16.400 16.501 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.399 16.500 ot_scf_mini 108 9.5 0.002 0.002 15.290 15.396 multiply_cannon 2286 13.5 0.303 0.309 14.197 15.109 multiply_cannon_loop 2286 14.5 0.342 0.348 12.903 13.840 ot_mini 108 10.5 0.001 0.001 9.290 9.413 multiply_cannon_multrec 9144 15.5 3.598 4.955 8.878 8.929 init_scf_loop 11 6.9 0.000 0.000 7.828 7.831 qs_ot_get_derivative 108 11.5 0.001 0.001 7.249 7.356 rebuild_ks_matrix 119 8.3 0.000 0.000 6.934 7.082 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.933 7.081 prepare_preconditioner 11 7.9 0.000 0.000 6.906 6.920 make_preconditioner 11 8.9 0.000 0.000 6.906 6.920 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.496 6.795 dbcsr_mm_accdrv_process 12550 15.8 4.114 5.673 5.156 6.541 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.284 6.418 cp_fm_upper_to_full 72 14.2 3.159 4.507 3.159 4.507 init_scf_run 11 5.9 0.000 0.001 3.943 3.943 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.943 3.943 mp_waitall_1 97218 16.6 2.793 3.860 2.793 3.860 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.755 3.761 calculate_rho_elec 119 8.7 0.118 0.121 3.754 3.761 sum_up_and_integrate 119 10.3 0.064 0.065 3.733 3.743 integrate_v_rspace 119 11.3 0.003 0.003 3.668 3.678 qs_ot_get_p 119 10.4 0.001 0.001 3.234 3.386 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.573 3.005 dbcsr_complete_redistribute 329 12.2 0.290 0.296 1.971 2.814 make_m2s 4572 13.5 0.038 0.038 2.402 2.592 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.641 2.482 make_images 4572 14.5 0.355 0.385 2.281 2.471 calculate_first_density_matrix 1 7.0 0.000 0.000 2.465 2.465 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.215 2.435 apply_single 119 13.6 0.000 0.000 2.215 2.435 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.439 2.401 calculate_dm_sparse 119 9.5 0.000 0.000 2.271 2.291 mp_alltoall_i22 627 13.8 1.437 2.270 1.437 2.270 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.405 2.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.136 2.136 pw_transfer 1439 11.6 0.066 0.067 2.084 2.090 grid_integrate_task_list 119 12.3 2.019 2.035 2.019 2.035 ot_diis_step 108 11.5 0.014 0.014 2.016 2.016 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.989 1.993 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.944 1.946 jit_kernel_multiply 10 15.6 1.015 1.908 1.015 1.908 density_rs2pw 119 9.7 0.003 0.004 1.836 1.866 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.789 1.843 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.813 1.824 mp_sum_l 7207 12.9 1.108 1.794 1.108 1.794 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.748 1.749 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.666 1.666 fft_wrap_pw1pw2_140 487 13.2 0.088 0.092 1.655 1.661 fft3d_ps 1201 14.6 0.568 0.581 1.622 1.627 cp_fm_cholesky_invert 11 10.9 1.552 1.556 1.552 1.556 rs_pw_transfer 974 11.9 0.009 0.009 1.456 1.485 grid_collocate_task_list 119 9.7 1.458 1.477 1.458 1.477 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.208 1.446 wfi_extrapolate 11 7.9 0.001 0.001 1.423 1.423 make_images_data 4572 15.5 0.043 0.046 1.153 1.412 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.367 1.367 cp_fm_diag_elpa_base 50 14.0 1.221 1.271 1.365 1.365 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.317 1.326 potential_pw2rs 119 12.3 0.014 0.014 1.259 1.263 mp_alltoall_d11v 2130 13.8 1.197 1.208 1.197 1.208 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.156 1.179 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.090 1.110 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.990 1.070 acc_transpose_blocks 9144 15.5 0.038 0.038 1.045 1.068 multiply_cannon_sync_h2d 9144 15.5 1.043 1.046 1.043 1.046 qs_create_task_list 11 7.9 0.000 0.000 0.945 0.957 generate_qs_task_list 11 8.9 0.373 0.392 0.945 0.957 mp_alltoall_z22v 1201 16.6 0.918 0.942 0.918 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.667000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=703.363636, yerr=16.176710 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.619584E+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 313056. 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 0.024 0.050 83.773 83.774 qs_mol_dyn_low 1 2.0 0.003 0.003 83.486 83.496 qs_forces 11 3.9 0.002 0.003 83.401 83.402 qs_energies 11 4.9 0.001 0.002 80.576 80.588 scf_env_do_scf 11 5.9 0.001 0.001 71.723 71.725 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.934 65.934 dbcsr_multiply_generic 2055 12.4 0.105 0.109 52.700 53.007 qs_scf_new_mos 99 7.5 0.000 0.001 48.794 48.903 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.794 48.903 ot_scf_mini 99 9.5 0.002 0.002 46.354 46.472 multiply_cannon 2055 13.4 0.176 0.184 42.684 43.519 multiply_cannon_loop 2055 14.4 1.562 1.613 41.638 42.478 velocity_verlet 10 3.0 0.001 0.001 42.471 42.472 ot_mini 99 10.5 0.001 0.001 28.250 28.367 qs_ot_get_derivative 99 11.5 0.001 0.001 21.357 21.476 multiply_cannon_multrec 49320 15.4 12.056 12.820 17.103 17.811 rebuild_ks_matrix 110 8.3 0.000 0.001 14.512 14.633 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.512 14.632 mp_waitall_1 241148 16.1 12.034 13.000 12.034 13.000 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.772 12.872 multiply_cannon_sync_h2d 49320 15.4 9.860 10.499 9.860 10.499 qs_ot_get_p 110 10.4 0.001 0.001 9.501 9.642 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.046 8.075 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.158 7.716 apply_single 110 13.6 0.000 0.001 7.158 7.716 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.099 7.607 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.066 7.124 sum_up_and_integrate 110 10.3 0.037 0.044 7.090 7.105 integrate_v_rspace 110 11.3 0.002 0.003 7.053 7.077 init_scf_run 11 5.9 0.000 0.001 6.784 6.784 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.784 6.784 ot_diis_step 99 11.5 0.005 0.005 6.598 6.598 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.364 6.512 calculate_rho_elec 110 8.6 0.020 0.024 6.363 6.511 mp_sum_l 6514 12.8 5.500 6.317 5.500 6.317 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.211 6.247 init_scf_loop 11 6.9 0.000 0.000 5.755 5.756 dbcsr_mm_accdrv_process 87628 16.1 2.091 2.206 4.926 5.271 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.183 5.184 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.664 4.667 cp_fm_redistribute_end 48 14.0 4.056 4.637 4.061 4.639 cp_fm_diag_elpa_base 48 14.0 0.570 4.564 0.574 4.591 make_m2s 4110 13.4 0.060 0.064 4.061 4.228 make_images 4110 14.4 0.176 0.191 3.967 4.136 wfi_extrapolate 11 7.9 0.001 0.001 4.042 4.042 rs_pw_transfer 902 11.9 0.012 0.013 3.843 4.002 calculate_dm_sparse 110 9.5 0.000 0.001 3.834 3.943 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.762 3.885 prepare_preconditioner 11 7.9 0.000 0.000 3.581 3.596 make_preconditioner 11 8.9 0.000 0.000 3.581 3.596 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.551 3.556 density_rs2pw 110 9.6 0.004 0.004 3.410 3.538 grid_integrate_task_list 110 12.3 3.261 3.420 3.261 3.420 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.348 3.400 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.356 3.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.233 3.262 pw_transfer 1331 11.6 0.054 0.066 2.985 3.043 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.897 2.957 potential_pw2rs 110 12.3 0.006 0.007 2.659 2.683 calculate_first_density_matrix 1 7.0 0.000 0.001 2.634 2.642 jit_kernel_multiply 13 15.9 2.551 2.597 2.551 2.597 mp_alltoall_d11v 2046 13.8 2.187 2.518 2.187 2.518 fft3d_ps 1111 14.6 0.751 0.839 2.434 2.486 fft_wrap_pw1pw2_140 451 13.1 0.170 0.191 2.406 2.471 acc_transpose_blocks 49320 15.4 0.222 0.235 2.210 2.295 grid_collocate_task_list 110 9.6 2.088 2.193 2.088 2.193 mp_waitany 14300 13.8 1.846 2.142 1.846 2.142 mp_sum_d 3879 11.9 1.543 2.033 1.543 2.033 make_images_data 4110 15.4 0.044 0.046 1.847 1.999 cp_fm_cholesky_invert 11 10.9 1.989 1.993 1.989 1.993 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.881 1.902 hybrid_alltoall_any 4261 16.3 0.083 0.477 1.581 1.890 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.673 1.693 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.774000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.181818, yerr=2.886274 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 586.747904E+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 2256357. 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 0.039 0.121 69.425 69.433 qs_mol_dyn_low 1 2.0 0.003 0.004 69.004 69.015 qs_forces 11 3.9 0.002 0.002 68.725 68.726 qs_energies 11 4.9 0.006 0.020 65.422 65.425 scf_env_do_scf 11 5.9 0.001 0.001 56.731 56.735 scf_env_do_scf_inner_loop 99 6.5 0.011 0.039 48.818 48.819 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.714 37.911 velocity_verlet 10 3.0 0.001 0.001 36.452 36.456 qs_scf_new_mos 99 7.5 0.001 0.001 32.703 32.832 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.702 32.832 multiply_cannon 2055 13.4 0.223 0.248 30.944 32.098 ot_scf_mini 99 9.5 0.003 0.004 31.039 31.168 multiply_cannon_loop 2055 14.4 0.920 0.941 29.604 30.454 ot_mini 99 10.5 0.001 0.001 18.262 18.399 multiply_cannon_multrec 24660 15.4 7.661 9.304 13.824 15.622 rebuild_ks_matrix 110 8.3 0.000 0.001 13.794 13.901 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.020 13.794 13.901 qs_ot_get_derivative 99 11.5 0.001 0.001 12.431 12.564 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.182 12.278 mp_waitall_1 186928 16.3 7.928 10.283 7.928 10.283 multiply_cannon_sync_h2d 24660 15.4 6.987 8.048 6.987 8.048 init_scf_loop 11 6.9 0.001 0.003 7.878 7.879 multiply_cannon_metrocomm3 24660 15.4 0.068 0.070 5.089 7.506 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.500 7.230 apply_single 110 13.6 0.000 0.001 6.500 7.230 sum_up_and_integrate 110 10.3 0.055 0.058 6.692 6.704 integrate_v_rspace 110 11.3 0.002 0.002 6.636 6.651 dbcsr_mm_accdrv_process 52282 16.1 4.738 5.450 6.003 6.305 qs_ot_get_p 110 10.4 0.001 0.001 6.107 6.272 init_scf_run 11 5.9 0.000 0.001 6.201 6.202 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.201 6.202 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.897 5.913 calculate_rho_elec 110 8.6 0.039 0.047 5.896 5.913 prepare_preconditioner 11 7.9 0.000 0.000 5.836 5.857 make_preconditioner 11 8.9 0.007 0.054 5.836 5.857 ot_diis_step 99 11.5 0.010 0.010 5.782 5.783 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.399 5.561 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.681 5.395 make_m2s 4110 13.4 0.057 0.059 4.348 4.819 make_images 4110 14.4 0.399 0.443 4.243 4.712 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.196 4.216 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.749 3.750 pw_transfer 1331 11.6 0.066 0.070 3.435 3.576 wfi_extrapolate 11 7.9 0.001 0.001 3.532 3.532 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.328 3.468 grid_integrate_task_list 110 12.3 3.149 3.367 3.149 3.367 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.321 3.323 density_rs2pw 110 9.6 0.004 0.004 3.138 3.292 rs_pw_transfer 902 11.9 0.012 0.014 3.099 3.291 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.213 3.283 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.192 3.194 cp_fm_redistribute_end 48 14.0 2.377 3.164 2.380 3.166 cp_fm_diag_elpa_base 48 14.0 0.750 3.043 0.782 3.132 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.959 3.001 calculate_dm_sparse 110 9.5 0.001 0.001 2.959 2.989 make_images_data 4110 15.4 0.047 0.051 2.436 2.931 hybrid_alltoall_any 4261 16.3 0.103 0.449 2.188 2.924 fft_wrap_pw1pw2_140 451 13.1 0.200 0.218 2.731 2.869 fft3d_ps 1111 14.6 1.076 1.270 2.686 2.811 cp_fm_cholesky_invert 11 10.9 2.697 2.704 2.697 2.704 calculate_first_density_matrix 1 7.0 0.001 0.004 2.581 2.584 potential_pw2rs 110 12.3 0.008 0.009 2.535 2.547 mp_sum_l 6514 12.8 1.836 2.441 1.836 2.441 grid_collocate_task_list 110 9.6 2.119 2.254 2.119 2.254 mp_alltoall_d11v 2046 13.8 1.917 2.177 1.917 2.177 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.970 1.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.904 1.905 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.778 1.789 multiply_cannon_metrocomm4 22605 15.4 0.075 0.079 0.797 1.759 mp_allgather_i34 2055 14.4 0.665 1.725 0.665 1.725 mp_irecv_dv 57340 16.2 0.672 1.641 0.672 1.641 acc_transpose_blocks 24660 15.4 0.110 0.113 1.538 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.528 1.537 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.475 mp_waitany 10164 13.8 1.255 1.469 1.255 1.469 jit_kernel_multiply 8 16.3 0.918 1.461 0.918 1.461 mp_sum_d 3879 11.9 0.974 1.458 0.974 1.458 dbcsr_complete_redistribute 325 12.2 0.245 0.328 1.173 1.446 mp_alltoall_z22v 1111 16.6 1.345 1.414 1.345 1.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.433000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=7.094405 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 659.349504E+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 0.095 0.124 59.382 59.383 qs_mol_dyn_low 1 2.0 0.003 0.003 58.985 58.995 qs_forces 11 3.9 0.002 0.002 58.918 58.920 qs_energies 11 4.9 0.002 0.003 55.745 55.750 scf_env_do_scf 11 5.9 0.001 0.001 47.907 47.908 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.244 39.244 velocity_verlet 10 3.0 0.001 0.001 32.336 32.337 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.463 28.714 qs_scf_new_mos 99 7.5 0.001 0.001 24.862 24.950 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.861 24.949 ot_scf_mini 99 9.5 0.002 0.003 23.639 23.744 multiply_cannon 2055 13.4 0.212 0.223 22.147 23.418 multiply_cannon_loop 2055 14.4 0.614 0.633 20.871 21.801 ot_mini 99 10.5 0.001 0.001 13.503 13.612 rebuild_ks_matrix 110 8.3 0.000 0.000 12.115 12.236 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.115 12.235 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.695 10.805 multiply_cannon_multrec 16440 15.4 3.899 4.951 9.666 10.646 mp_waitall_1 146766 16.3 7.132 10.228 7.132 10.228 qs_ot_get_derivative 99 11.5 0.001 0.001 9.096 9.203 init_scf_loop 11 6.9 0.000 0.002 8.627 8.628 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.289 7.175 prepare_preconditioner 11 7.9 0.000 0.000 6.866 6.889 make_preconditioner 11 8.9 0.000 0.002 6.866 6.889 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.168 6.537 sum_up_and_integrate 110 10.3 0.061 0.062 6.346 6.362 integrate_v_rspace 110 11.3 0.002 0.003 6.285 6.300 dbcsr_mm_accdrv_process 34862 16.1 4.684 5.259 5.621 5.740 init_scf_run 11 5.9 0.000 0.001 5.449 5.449 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.439 5.449 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.449 5.449 calculate_rho_elec 110 8.6 0.058 0.058 5.438 5.449 qs_ot_get_p 110 10.4 0.001 0.001 5.267 5.408 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.923 5.341 apply_single 110 13.6 0.000 0.000 4.923 5.341 make_m2s 4110 13.4 0.050 0.052 4.192 4.610 make_images 4110 14.4 0.394 0.510 4.081 4.499 ot_diis_step 99 11.5 0.011 0.011 4.374 4.374 multiply_cannon_sync_h2d 16440 15.4 3.683 4.325 3.683 4.325 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.104 3.747 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.673 3.680 grid_integrate_task_list 110 12.3 3.186 3.368 3.186 3.368 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.322 3.323 pw_transfer 1331 11.6 0.065 0.071 3.115 3.122 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.008 3.019 make_images_data 4110 15.4 0.044 0.048 2.477 2.987 wfi_extrapolate 11 7.9 0.001 0.001 2.955 2.955 hybrid_alltoall_any 4261 16.3 0.105 0.376 2.158 2.920 density_rs2pw 110 9.6 0.004 0.004 2.729 2.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.803 2.804 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.799 2.801 cp_fm_redistribute_end 48 14.0 1.739 2.777 1.741 2.778 rs_pw_transfer 902 11.9 0.011 0.012 2.613 2.756 cp_fm_diag_elpa_base 48 14.0 0.972 2.645 1.031 2.752 cp_fm_cholesky_invert 11 10.9 2.707 2.713 2.707 2.713 fft_wrap_pw1pw2_140 451 13.1 0.211 0.216 2.568 2.580 calculate_dm_sparse 110 9.5 0.001 0.001 2.494 2.525 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.480 2.523 calculate_first_density_matrix 1 7.0 0.000 0.000 2.414 2.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.324 2.389 grid_collocate_task_list 110 9.6 2.180 2.382 2.180 2.382 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.875 2.355 fft3d_ps 1111 14.6 1.065 1.090 2.322 2.335 mp_irecv_dv 48980 15.7 0.805 2.229 0.805 2.229 potential_pw2rs 110 12.3 0.011 0.011 2.197 2.203 mp_sum_l 6514 12.8 1.560 2.188 1.560 2.188 mp_alltoall_d11v 2046 13.8 1.741 2.098 1.741 2.098 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.957 1.958 dbcsr_complete_redistribute 325 12.2 0.329 0.355 1.490 1.957 cp_fm_upper_to_full 70 13.6 1.403 1.852 1.403 1.852 mp_allgather_i34 2055 14.4 0.597 1.705 0.597 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.622 1.640 cp_fm_cholesky_decompose 22 10.9 1.595 1.611 1.595 1.611 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.375 1.501 jit_kernel_multiply 8 16.8 0.552 1.501 0.552 1.501 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.007 1.472 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.452 1.465 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.271 1.278 mp_waitany 17072 13.8 1.141 1.275 1.141 1.275 acc_transpose_blocks 16440 15.4 0.073 0.076 1.228 1.254 rs_gather_matrices 110 12.3 0.138 0.151 0.852 1.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.383000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.000000, yerr=9.145292 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 724.983808E+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 0.045 0.137 65.282 65.283 qs_mol_dyn_low 1 2.0 0.003 0.005 64.843 64.878 qs_forces 11 3.9 0.003 0.008 64.767 64.769 qs_energies 11 4.9 0.002 0.005 61.406 61.411 scf_env_do_scf 11 5.9 0.001 0.005 52.794 52.797 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.925 40.926 velocity_verlet 10 3.0 0.001 0.002 36.571 36.573 dbcsr_multiply_generic 2055 12.4 0.113 0.117 29.995 30.152 qs_scf_new_mos 99 7.5 0.001 0.001 26.603 26.705 qs_scf_loop_do_ot 99 8.5 0.003 0.023 26.603 26.704 ot_scf_mini 99 9.5 0.003 0.006 24.975 25.085 multiply_cannon 2055 13.4 0.247 0.267 22.942 23.709 multiply_cannon_loop 2055 14.4 0.886 0.904 21.516 22.001 multiply_cannon_multrec 24660 15.4 4.250 6.851 13.050 14.463 ot_mini 99 10.5 0.001 0.001 14.296 14.424 rebuild_ks_matrix 110 8.3 0.000 0.000 11.933 12.060 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.023 11.933 12.059 init_scf_loop 11 6.9 0.006 0.047 11.824 11.825 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.614 10.731 qs_ot_get_derivative 99 11.5 0.001 0.002 10.121 10.236 prepare_preconditioner 11 7.9 0.000 0.000 9.965 9.980 make_preconditioner 11 8.9 0.000 0.001 9.965 9.980 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.207 9.670 dbcsr_mm_accdrv_process 52304 16.0 6.769 7.965 8.653 9.542 mp_waitall_1 126806 16.4 4.504 6.342 4.504 6.342 sum_up_and_integrate 110 10.3 0.070 0.091 6.312 6.327 integrate_v_rspace 110 11.3 0.003 0.004 6.241 6.257 init_scf_run 11 5.9 0.000 0.001 5.850 5.850 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.850 5.850 qs_ot_get_p 110 10.4 0.001 0.001 5.686 5.826 make_m2s 4110 13.4 0.059 0.061 5.387 5.763 make_images 4110 14.4 0.578 0.702 5.247 5.619 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.548 5.557 calculate_rho_elec 110 8.6 0.077 0.081 5.547 5.556 cp_fm_upper_to_full 70 13.8 3.250 4.632 3.250 4.632 ot_diis_step 99 11.5 0.011 0.013 4.136 4.136 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.039 4.127 apply_single 110 13.6 0.000 0.000 4.039 4.127 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.867 3.881 dbcsr_complete_redistribute 325 12.2 0.430 0.468 2.658 3.778 multiply_cannon_sync_h2d 24660 15.4 3.199 3.475 3.199 3.475 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.400 3.455 grid_integrate_task_list 110 12.3 3.269 3.449 3.269 3.449 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.386 3.386 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.133 3.247 pw_transfer 1331 11.6 0.065 0.072 3.204 3.228 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.442 3.189 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.414 3.149 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.098 3.127 make_images_data 4110 15.4 0.047 0.051 2.747 3.124 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.014 3.074 calculate_dm_sparse 110 9.5 0.001 0.001 3.022 3.050 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.024 3.025 wfi_extrapolate 11 7.9 0.001 0.004 2.971 2.972 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.749 2.841 density_rs2pw 110 9.6 0.004 0.004 2.721 2.841 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.831 2.834 mp_alltoall_i22 605 13.7 1.650 2.833 1.650 2.833 cp_fm_cholesky_invert 11 10.9 2.811 2.819 2.811 2.819 cp_fm_redistribute_end 48 14.0 1.410 2.805 1.413 2.807 calculate_first_density_matrix 1 7.0 0.000 0.002 2.790 2.792 cp_fm_diag_elpa_base 48 14.0 1.310 2.670 1.389 2.781 jit_kernel_multiply 12 15.8 1.551 2.688 1.551 2.688 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.640 2.680 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.622 2.651 rs_pw_transfer 902 11.9 0.010 0.011 2.433 2.583 grid_collocate_task_list 110 9.6 2.224 2.440 2.224 2.440 fft3d_ps 1111 14.6 1.061 1.098 2.399 2.419 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 2.262 2.265 mp_alltoall_d11v 2046 13.8 1.758 2.108 1.758 2.108 potential_pw2rs 110 12.3 0.012 0.013 2.099 2.107 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.705 1.743 cp_fm_cholesky_decompose 22 10.9 1.698 1.743 1.698 1.743 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.596 1.694 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.635 1.653 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.613 1.627 acc_transpose_blocks 24660 15.4 0.103 0.107 1.579 1.621 mp_sum_l 6514 12.8 1.017 1.606 1.017 1.606 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.846 1.591 mp_irecv_dv 62702 16.1 0.746 1.515 0.746 1.515 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.283000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.636364, yerr=7.022985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 843.907072E+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 0.014 0.043 55.596 55.597 qs_mol_dyn_low 1 2.0 0.003 0.003 55.342 55.352 qs_forces 11 3.9 0.002 0.002 55.274 55.275 qs_energies 11 4.9 0.001 0.002 51.667 51.671 scf_env_do_scf 11 5.9 0.000 0.001 43.305 43.306 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.272 35.273 velocity_verlet 10 3.0 0.001 0.001 31.084 31.087 dbcsr_multiply_generic 2055 12.4 0.104 0.105 23.163 23.307 qs_scf_new_mos 99 7.5 0.001 0.001 21.105 21.165 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.104 21.164 ot_scf_mini 99 9.5 0.002 0.002 19.852 19.890 multiply_cannon 2055 13.4 0.246 0.259 17.533 18.858 multiply_cannon_loop 2055 14.4 0.320 0.333 16.160 16.561 rebuild_ks_matrix 110 8.3 0.000 0.000 11.488 11.540 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.487 11.540 ot_mini 99 10.5 0.001 0.001 10.634 10.666 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.238 10.286 multiply_cannon_multrec 8220 15.4 3.241 4.431 7.450 8.452 mp_waitall_1 106626 16.5 6.380 8.031 6.380 8.031 init_scf_loop 11 6.9 0.000 0.000 7.976 7.978 qs_ot_get_derivative 99 11.5 0.001 0.001 6.780 6.815 prepare_preconditioner 11 7.9 0.000 0.000 6.340 6.347 make_preconditioner 11 8.9 0.000 0.000 6.340 6.347 sum_up_and_integrate 110 10.3 0.080 0.082 6.200 6.212 integrate_v_rspace 110 11.3 0.003 0.003 6.119 6.133 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.922 6.006 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.649 5.664 calculate_rho_elec 110 8.6 0.115 0.115 5.648 5.664 qs_ot_get_p 110 10.4 0.001 0.001 5.257 5.283 init_scf_run 11 5.9 0.000 0.001 5.223 5.223 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.223 5.223 dbcsr_mm_accdrv_process 17442 15.9 2.781 3.941 4.079 4.976 make_m2s 4110 13.4 0.039 0.040 4.265 4.539 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.169 4.439 make_images 4110 14.4 0.637 0.695 4.136 4.409 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.850 3.854 ot_diis_step 99 11.5 0.012 0.012 3.828 3.829 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.740 3.785 apply_single 110 13.6 0.000 0.000 3.740 3.785 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.543 3.544 grid_integrate_task_list 110 12.3 3.372 3.535 3.372 3.535 pw_transfer 1331 11.6 0.065 0.070 3.271 3.280 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.164 3.176 cp_fm_cholesky_invert 11 10.9 3.149 3.154 3.149 3.154 multiply_cannon_sync_h2d 8220 15.4 2.922 3.033 2.922 3.033 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.990 2.991 cp_fm_redistribute_end 48 14.0 0.755 2.953 0.761 2.954 cp_fm_diag_elpa_base 48 14.0 2.004 2.749 2.184 2.916 make_images_data 4110 15.4 0.039 0.044 2.414 2.840 density_rs2pw 110 9.6 0.004 0.004 2.696 2.794 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.349 2.749 wfi_extrapolate 11 7.9 0.001 0.001 2.746 2.746 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.692 2.705 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.652 2.652 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.620 2.621 calculate_dm_sparse 110 9.5 0.001 0.001 2.478 2.514 grid_collocate_task_list 110 9.6 2.330 2.507 2.330 2.507 fft3d_ps 1111 14.6 1.115 1.141 2.423 2.433 calculate_first_density_matrix 1 7.0 0.000 0.000 2.387 2.388 rs_pw_transfer 902 11.9 0.010 0.010 2.218 2.344 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.213 2.232 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.767 1.978 potential_pw2rs 110 12.3 0.015 0.015 1.949 1.953 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.925 1.949 mp_alltoall_d11v 2046 13.8 1.666 1.910 1.666 1.910 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.865 1.882 jit_kernel_multiply 11 15.6 0.991 1.861 0.991 1.861 cp_fm_cholesky_decompose 22 10.9 1.735 1.759 1.735 1.759 mp_allgather_i34 2055 14.4 0.557 1.627 0.557 1.627 dbcsr_complete_redistribute 325 12.2 0.575 0.609 1.523 1.620 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.498 1.616 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.584 1.589 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.433 1.444 qs_create_task_list 11 7.9 0.000 0.000 1.227 1.324 generate_qs_task_list 11 8.9 0.379 0.448 1.227 1.324 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.183 1.209 mp_waitany 9240 13.8 1.061 1.190 1.061 1.190 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.478 1.133 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.597000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=788.363636, yerr=14.131028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 1.354469E+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 0.019 0.049 85.221 85.222 qs_mol_dyn_low 1 2.0 0.003 0.003 84.902 84.910 qs_forces 11 3.9 0.002 0.002 84.831 84.831 qs_energies 11 4.9 0.013 0.013 80.833 80.836 scf_env_do_scf 11 5.9 0.000 0.001 70.783 70.784 velocity_verlet 10 3.0 0.001 0.001 54.608 54.615 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.369 42.370 dbcsr_multiply_generic 2055 12.4 0.118 0.120 28.298 28.409 init_scf_loop 11 6.9 0.000 0.000 28.341 28.344 prepare_preconditioner 11 7.9 0.000 0.000 26.418 26.425 make_preconditioner 11 8.9 0.000 0.000 26.417 26.425 qs_scf_new_mos 99 7.5 0.001 0.001 26.116 26.164 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.116 26.164 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.581 25.894 ot_scf_mini 99 9.5 0.002 0.002 24.354 24.390 multiply_cannon 2055 13.4 0.346 0.375 21.363 22.269 multiply_cannon_loop 2055 14.4 0.340 0.345 19.575 19.871 cp_fm_upper_to_full 70 14.2 12.645 18.210 12.645 18.210 ot_mini 99 10.5 0.001 0.001 13.500 13.543 rebuild_ks_matrix 110 8.3 0.001 0.001 13.024 13.069 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.024 13.069 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.862 11.902 dbcsr_complete_redistribute 325 12.2 1.028 1.055 7.446 10.595 multiply_cannon_multrec 8220 15.4 4.342 4.519 9.517 9.609 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.401 9.548 qs_ot_get_derivative 99 11.5 0.001 0.001 8.953 8.987 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.821 8.943 mp_alltoall_i22 605 13.7 5.448 8.616 5.448 8.616 mp_waitall_1 87304 16.6 7.637 8.473 7.637 8.473 sum_up_and_integrate 110 10.3 0.150 0.151 6.574 6.587 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.402 6.440 calculate_rho_elec 110 8.6 0.227 0.227 6.402 6.439 integrate_v_rspace 110 11.3 0.003 0.004 6.423 6.438 init_scf_run 11 5.9 0.000 0.001 5.812 5.812 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.812 5.812 qs_ot_get_p 110 10.4 0.001 0.001 5.689 5.727 make_m2s 4110 13.4 0.043 0.043 5.178 5.681 make_images 4110 14.4 0.881 0.932 4.992 5.495 cp_fm_cholesky_invert 11 10.9 5.336 5.341 5.336 5.341 dbcsr_mm_accdrv_process 11614 15.7 3.227 3.609 5.032 5.276 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.650 5.094 apply_single 110 13.6 0.000 0.000 4.650 5.094 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.704 4.958 ot_diis_step 99 11.5 0.015 0.016 4.524 4.524 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.168 4.173 multiply_cannon_sync_h2d 8220 15.4 3.953 3.955 3.953 3.955 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.727 3.728 grid_integrate_task_list 110 12.3 3.659 3.708 3.659 3.708 pw_transfer 1331 11.6 0.074 0.074 3.641 3.648 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.629 3.629 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.812 3.538 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.525 3.531 make_images_data 4110 15.4 0.042 0.045 2.784 3.427 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.852 3.320 wfi_extrapolate 11 7.9 0.001 0.001 3.203 3.203 calculate_dm_sparse 110 9.5 0.001 0.001 3.122 3.138 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.119 3.119 cp_fm_diag_elpa_base 48 14.0 2.582 2.778 3.117 3.117 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.047 3.052 density_rs2pw 110 9.6 0.004 0.004 2.913 2.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.848 2.850 fft3d_ps 1111 14.6 1.262 1.273 2.734 2.741 grid_collocate_task_list 110 9.6 2.628 2.652 2.628 2.652 calculate_first_density_matrix 1 7.0 0.000 0.000 2.502 2.503 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.384 2.403 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.342 2.355 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.168 2.220 rs_pw_transfer 902 11.9 0.010 0.011 2.120 2.167 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.086 2.151 cp_fm_cholesky_decompose 22 10.9 2.054 2.070 2.054 2.070 mp_alltoall_d11v 2046 13.8 2.032 2.069 2.032 2.069 potential_pw2rs 110 12.3 0.021 0.021 2.017 2.022 qs_create_task_list 11 7.9 0.000 0.000 1.897 1.943 generate_qs_task_list 11 8.9 0.739 0.793 1.897 1.943 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.788 1.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.772 1.777 jit_kernel_multiply 10 15.4 1.603 1.728 1.603 1.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.222000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1213.363636, yerr=58.723192 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 1.103378E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 12049536 0.0% 0.0% 100.0% average stack size 0.0 0.0 564.9 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 626.593792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10439712 MPI messages size (bytes): total size 4.526311E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.566688E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3611304 96569262080 32768 < size <= 131072 1307392 74803314688 131072 < size <= 4194304 5190114 3200149557592 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56191. MP_Allreduce 11358 781. MP_Sync 172 MP_Alltoall 2242 3074876. MP_ISendRecv 49020 18752. MP_Wait 67312 MP_comm_split 84 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.029 206.066 206.071 qs_mol_dyn_low 1 2.0 0.003 0.003 205.675 205.691 qs_forces 11 3.9 0.003 0.004 205.584 205.586 qs_energies 11 4.9 0.004 0.009 200.111 200.124 scf_env_do_scf 11 5.9 0.001 0.002 183.301 183.307 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 162.309 162.311 dbcsr_multiply_generic 2529 12.6 0.173 0.179 125.168 126.040 velocity_verlet 10 3.0 0.001 0.001 124.545 124.547 qs_scf_new_mos 118 7.6 0.001 0.001 123.368 123.694 qs_scf_loop_do_ot 118 8.6 0.001 0.002 123.368 123.694 ot_scf_mini 118 9.6 0.003 0.004 116.714 116.981 multiply_cannon 2529 13.6 0.238 0.246 101.555 103.907 multiply_cannon_loop 2529 14.6 2.141 2.206 99.346 101.294 ot_mini 118 10.6 0.001 0.001 66.241 66.517 multiply_cannon_multrec 60696 15.6 33.443 35.466 41.817 44.054 qs_ot_get_derivative 118 11.6 0.001 0.001 41.095 41.385 rebuild_ks_matrix 129 8.3 0.001 0.001 33.419 33.763 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 33.419 33.762 mp_waitall_1 293956 16.2 28.871 31.327 28.871 31.327 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.116 30.433 multiply_cannon_sync_h2d 60696 15.6 27.601 29.451 27.601 29.451 qs_ot_get_p 129 10.4 0.001 0.001 28.439 28.902 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.565 25.290 apply_single 129 13.6 0.001 0.001 24.565 25.290 ot_diis_step 118 11.6 0.008 0.008 24.777 24.779 qs_ot_p2m_diag 84 11.4 0.078 0.092 21.616 21.725 init_scf_loop 11 6.9 0.001 0.004 20.917 20.919 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 19.266 19.467 cp_dbcsr_syevd 84 12.4 0.005 0.005 18.960 18.961 multiply_cannon_metrocomm3 60696 15.6 0.118 0.123 15.829 18.544 prepare_preconditioner 11 7.9 0.000 0.000 16.316 16.375 make_preconditioner 11 8.9 0.000 0.002 16.316 16.375 cp_fm_diag_elpa 84 13.4 0.000 0.001 15.999 16.005 cp_fm_redistribute_end 84 14.4 12.606 15.918 12.619 15.921 cp_fm_diag_elpa_base 84 14.4 3.256 15.679 3.290 15.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.561 15.745 make_m2s 5058 13.6 0.103 0.113 13.994 14.385 make_images 5058 14.6 0.409 0.430 13.820 14.219 sum_up_and_integrate 129 10.3 0.090 0.109 13.999 14.015 integrate_v_rspace 129 11.3 0.003 0.004 13.909 13.926 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.647 12.771 calculate_rho_elec 129 8.7 0.045 0.064 12.646 12.770 init_scf_run 11 5.9 0.000 0.001 12.671 12.672 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.671 12.672 mp_sum_l 7936 13.1 8.352 10.190 8.352 10.190 cp_fm_cholesky_invert 11 10.9 9.332 9.340 9.332 9.340 wfi_extrapolate 11 7.9 0.001 0.001 9.062 9.062 calculate_dm_sparse 129 9.5 0.001 0.001 8.446 8.559 dbcsr_mm_accdrv_process 125516 16.2 3.322 3.481 7.936 8.480 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.972 8.115 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.873 8.009 make_images_data 5058 15.6 0.071 0.077 6.889 7.744 grid_integrate_task_list 129 12.3 7.144 7.574 7.144 7.574 multiply_cannon_metrocomm1 60696 15.6 0.087 0.091 5.948 7.334 density_rs2pw 129 9.7 0.006 0.007 6.451 7.098 hybrid_alltoall_any 5245 16.5 0.298 2.291 6.041 7.036 pw_transfer 1559 11.6 0.076 0.106 6.668 6.949 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 6.462 6.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.706 6.714 rs_pw_transfer 1054 12.0 0.017 0.018 5.776 6.425 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.839 5.936 fft_wrap_pw1pw2_140 527 13.2 0.450 0.512 5.489 5.678 fft3d_ps 1301 14.7 2.110 2.589 5.243 5.442 mp_alltoall_d11v 2429 14.1 4.066 5.398 4.066 5.398 grid_collocate_task_list 129 9.7 4.764 5.092 4.764 5.092 cp_fm_cholesky_decompose 22 10.9 4.840 4.854 4.840 4.854 potential_pw2rs 129 12.3 0.009 0.010 4.477 4.502 mp_sum_d 4499 12.2 3.642 4.281 3.642 4.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.071000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.818182, yerr=7.505370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 831.193088E+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 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57173. MP_Allreduce 11257 946. MP_Sync 168 MP_Alltoall 1955 5903454. 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 0.023 0.044 196.728 196.729 qs_mol_dyn_low 1 2.0 0.003 0.003 196.276 196.290 qs_forces 11 3.9 0.003 0.005 196.189 196.191 qs_energies 11 4.9 0.004 0.018 189.330 189.343 scf_env_do_scf 11 5.9 0.001 0.002 172.176 172.185 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 137.124 137.127 velocity_verlet 10 3.0 0.001 0.001 123.525 123.526 dbcsr_multiply_generic 2485 12.5 0.182 0.189 98.948 100.301 qs_scf_new_mos 116 7.6 0.001 0.001 96.512 96.812 qs_scf_loop_do_ot 116 8.6 0.001 0.002 96.511 96.812 ot_scf_mini 116 9.6 0.004 0.005 91.620 91.983 multiply_cannon 2485 13.5 0.477 0.531 77.267 81.683 multiply_cannon_loop 2485 14.5 1.235 1.274 73.706 76.511 ot_mini 116 10.6 0.001 0.001 51.224 51.619 mp_waitall_1 224796 16.4 26.637 40.891 26.637 40.891 multiply_cannon_multrec 29820 15.5 21.590 25.601 31.154 35.401 init_scf_loop 11 6.9 0.001 0.005 34.925 34.927 rebuild_ks_matrix 127 8.3 0.001 0.001 33.893 34.345 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 33.893 34.345 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.550 30.971 prepare_preconditioner 11 7.9 0.000 0.000 30.373 30.439 make_preconditioner 11 8.9 0.000 0.003 30.373 30.439 qs_ot_get_derivative 116 11.6 0.001 0.002 29.227 29.577 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.868 29.418 multiply_cannon_metrocomm3 29820 15.5 0.091 0.096 16.095 28.819 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.983 23.120 apply_single 127 13.6 0.001 0.001 21.983 23.119 qs_ot_get_p 127 10.4 0.001 0.001 22.073 22.552 ot_diis_step 116 11.6 0.014 0.015 21.806 21.809 multiply_cannon_sync_h2d 29820 15.5 18.864 20.807 18.864 20.807 cp_fm_cholesky_invert 11 10.9 17.618 17.631 17.618 17.631 make_m2s 4970 13.5 0.091 0.098 15.479 17.169 make_images 4970 14.5 1.155 1.344 15.279 16.971 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.845 16.900 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.602 15.604 sum_up_and_integrate 127 10.3 0.116 0.134 15.354 15.382 integrate_v_rspace 127 11.3 0.003 0.004 15.238 15.271 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.196 14.259 calculate_rho_elec 127 8.7 0.087 0.104 14.195 14.258 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.325 12.338 cp_fm_redistribute_end 82 14.4 7.165 12.241 7.181 12.247 cp_fm_diag_elpa_base 82 14.4 4.829 11.749 5.039 12.102 init_scf_run 11 5.9 0.000 0.001 12.081 12.083 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.080 12.083 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.639 11.905 make_images_data 4970 15.5 0.066 0.072 9.380 11.393 multiply_cannon_metrocomm4 27335 15.5 0.099 0.113 3.799 10.435 hybrid_alltoall_any 5155 16.4 0.341 1.490 8.042 10.203 mp_irecv_dv 68888 16.3 3.604 10.052 3.604 10.052 dbcsr_mm_accdrv_process 61726 16.2 4.509 5.337 9.039 9.640 pw_transfer 1535 11.6 0.086 0.101 9.018 9.073 wfi_extrapolate 11 7.9 0.001 0.001 8.863 8.864 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.795 8.845 density_rs2pw 127 9.7 0.006 0.006 7.628 8.100 cp_fm_cholesky_decompose 22 10.9 7.543 7.623 7.543 7.623 fft_wrap_pw1pw2_140 519 13.2 0.465 0.520 7.486 7.532 grid_integrate_task_list 127 12.3 7.104 7.523 7.104 7.523 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.475 7.225 fft3d_ps 1281 14.7 2.756 2.899 7.174 7.209 rs_pw_transfer 1038 11.9 0.014 0.016 6.304 6.771 calculate_dm_sparse 127 9.5 0.001 0.001 6.580 6.736 mp_alltoall_d11v 2401 14.1 5.514 6.706 5.514 6.706 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.332 6.343 mp_sum_l 7804 13.0 4.438 6.334 4.438 6.334 potential_pw2rs 127 12.3 0.015 0.017 5.588 5.609 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.469 5.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.343 5.483 grid_collocate_task_list 127 9.7 4.873 5.238 4.873 5.238 mp_allgather_i34 2485 14.5 2.134 5.208 2.134 5.208 dbcsr_complete_redistribute 393 12.7 0.773 0.851 3.793 4.688 mp_sum_d 4443 12.2 2.802 4.201 2.802 4.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=196.729000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.545455, yerr=3.201239 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 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 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 2.906046E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 941.187072E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172869264 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.082 187.620 187.622 qs_mol_dyn_low 1 2.0 0.003 0.003 187.078 187.093 qs_forces 11 3.9 0.003 0.004 186.957 186.961 qs_energies 11 4.9 0.004 0.017 180.268 180.282 scf_env_do_scf 11 5.9 0.002 0.031 163.959 163.961 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 125.685 125.686 velocity_verlet 10 3.0 0.001 0.001 119.373 119.375 qs_scf_new_mos 116 7.6 0.001 0.001 86.816 87.131 qs_scf_loop_do_ot 116 8.6 0.001 0.002 86.815 87.130 dbcsr_multiply_generic 2485 12.5 0.176 0.180 83.609 84.858 ot_scf_mini 116 9.6 0.004 0.005 82.617 83.020 multiply_cannon 2485 13.5 0.491 0.517 62.183 68.034 multiply_cannon_loop 2485 14.5 0.846 0.877 58.742 61.712 ot_mini 116 10.6 0.001 0.001 43.851 44.238 init_scf_loop 11 6.9 0.001 0.003 38.122 38.122 mp_waitall_1 176908 16.5 27.050 35.290 27.050 35.290 prepare_preconditioner 11 7.9 0.000 0.000 33.775 33.828 make_preconditioner 11 8.9 0.000 0.002 33.775 33.828 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.270 32.688 rebuild_ks_matrix 127 8.3 0.001 0.001 31.696 32.222 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 31.696 32.222 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.628 29.095 multiply_cannon_metrocomm3 19880 15.5 0.060 0.063 15.587 25.222 qs_ot_get_derivative 116 11.6 0.001 0.002 24.022 24.425 multiply_cannon_multrec 19880 15.5 13.279 15.916 21.891 24.375 qs_ot_get_p 127 10.4 0.001 0.001 23.813 24.288 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.872 21.011 apply_single 127 13.6 0.001 0.001 19.871 21.011 ot_diis_step 116 11.6 0.018 0.018 19.717 19.719 qs_ot_p2m_diag 82 11.4 0.263 0.270 19.131 19.161 cp_dbcsr_syevd 82 12.4 0.005 0.005 18.055 18.056 make_m2s 4970 13.5 0.082 0.086 15.845 16.860 make_images 4970 14.5 1.177 1.291 15.620 16.632 cp_fm_cholesky_invert 11 10.9 16.123 16.132 16.123 16.132 multiply_cannon_sync_h2d 19880 15.5 14.118 15.923 14.118 15.923 sum_up_and_integrate 127 10.3 0.131 0.145 15.191 15.215 integrate_v_rspace 127 11.3 0.003 0.004 15.060 15.088 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.749 14.752 cp_fm_redistribute_end 82 14.4 5.537 14.630 5.555 14.635 cp_fm_diag_elpa_base 82 14.4 8.625 14.001 9.051 14.502 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.184 14.210 calculate_rho_elec 127 8.7 0.130 0.146 14.184 14.209 make_images_data 4970 15.5 0.061 0.068 9.858 11.322 init_scf_run 11 5.9 0.000 0.001 11.067 11.067 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.066 11.067 hybrid_alltoall_any 5155 16.4 0.428 1.957 8.568 10.540 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.508 9.797 pw_transfer 1535 11.6 0.084 0.103 9.135 9.239 multiply_cannon_metrocomm4 17395 15.5 0.059 0.070 3.422 9.102 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 8.913 9.024 mp_irecv_dv 49801 16.2 3.303 8.856 3.303 8.856 dbcsr_mm_accdrv_process 41158 16.2 4.374 5.202 8.080 8.209 cp_fm_cholesky_decompose 22 10.9 8.164 8.198 8.164 8.198 grid_integrate_task_list 127 12.3 7.202 7.867 7.202 7.867 wfi_extrapolate 11 7.9 0.001 0.001 7.853 7.853 fft_wrap_pw1pw2_140 519 13.2 0.472 0.524 7.717 7.830 density_rs2pw 127 9.7 0.005 0.006 7.364 7.656 cp_fm_upper_to_full 104 14.5 5.829 7.541 5.829 7.541 fft3d_ps 1281 14.7 2.677 2.881 7.193 7.255 dbcsr_complete_redistribute 393 12.7 1.174 1.204 5.168 6.977 mp_alltoall_d11v 2401 14.1 5.551 6.786 5.551 6.786 rs_pw_transfer 1038 11.9 0.013 0.014 5.641 5.952 calculate_dm_sparse 127 9.5 0.001 0.001 5.836 5.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.745 5.752 grid_collocate_task_list 127 9.7 5.037 5.565 5.037 5.565 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.753 5.530 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.695 5.501 potential_pw2rs 127 12.3 0.020 0.022 5.282 5.307 mp_sum_l 7804 13.0 3.799 5.291 3.799 5.291 mp_allgather_i34 2485 14.5 1.968 5.258 1.968 5.258 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.670 4.791 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.486 4.269 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.148 4.176 mp_alltoall_i22 712 14.1 2.085 4.102 2.085 4.102 mp_alltoall_z22v 1281 16.7 3.925 4.082 3.925 4.082 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 3.805 3.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=187.622000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.818182, yerr=10.115858 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.163108E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769793E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921130560 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56813. MP_Allreduce 11321 1065. MP_Sync 170 MP_Alltoall 1724 12509439. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 83 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.046 194.685 194.686 qs_mol_dyn_low 1 2.0 0.003 0.003 194.275 194.294 qs_forces 11 3.9 0.003 0.003 194.141 194.148 qs_energies 11 4.9 0.003 0.012 187.047 187.059 scf_env_do_scf 11 5.9 0.001 0.002 169.446 169.457 velocity_verlet 10 3.0 0.001 0.001 128.347 128.350 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 121.677 121.678 qs_scf_new_mos 118 7.6 0.001 0.001 85.105 85.411 qs_scf_loop_do_ot 118 8.6 0.001 0.002 85.104 85.410 dbcsr_multiply_generic 2527 12.6 0.188 0.193 80.782 81.522 ot_scf_mini 118 9.6 0.003 0.004 80.537 80.832 multiply_cannon 2527 13.6 0.561 0.603 55.407 58.484 multiply_cannon_loop 2527 14.6 1.191 1.219 51.518 53.506 init_scf_loop 11 6.9 0.001 0.006 47.637 47.638 prepare_preconditioner 11 7.9 0.000 0.000 43.513 43.541 make_preconditioner 11 8.9 0.000 0.002 43.513 43.541 ot_mini 118 10.6 0.001 0.001 43.167 43.462 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.069 42.018 multiply_cannon_multrec 30324 15.6 14.252 19.548 26.448 31.161 rebuild_ks_matrix 129 8.3 0.001 0.001 29.586 29.890 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.585 29.889 mp_waitall_1 154994 16.5 18.143 27.232 18.143 27.232 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.761 27.032 qs_ot_get_p 129 10.4 0.001 0.001 23.398 23.716 qs_ot_get_derivative 118 11.6 0.001 0.002 23.378 23.674 make_m2s 5054 13.6 0.096 0.099 20.914 22.001 make_images 5054 14.6 1.986 2.258 20.610 21.698 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.161 19.686 apply_single 129 13.6 0.001 0.001 19.161 19.686 ot_diis_step 118 11.6 0.017 0.018 19.667 19.669 qs_ot_p2m_diag 83 11.4 0.343 0.391 19.025 19.078 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.711 17.713 cp_fm_cholesky_invert 11 10.9 17.042 17.052 17.042 17.052 cp_fm_upper_to_full 105 14.7 11.037 16.332 11.037 16.332 multiply_cannon_metrocomm3 30324 15.6 0.048 0.050 6.308 14.925 sum_up_and_integrate 129 10.3 0.141 0.154 14.467 14.495 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.460 14.464 cp_fm_redistribute_end 83 14.4 2.479 14.352 2.505 14.364 integrate_v_rspace 129 11.3 0.003 0.004 14.325 14.357 cp_fm_diag_elpa_base 83 14.4 11.199 13.681 11.809 14.184 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.728 13.761 calculate_rho_elec 129 8.7 0.177 0.192 13.727 13.761 make_images_data 5054 15.6 0.066 0.070 11.328 12.969 dbcsr_complete_redistribute 395 12.7 1.513 1.627 9.207 12.930 multiply_cannon_sync_h2d 30324 15.6 11.818 12.787 11.818 12.787 dbcsr_mm_accdrv_process 62756 16.2 7.412 8.667 11.766 12.298 hybrid_alltoall_any 5240 16.5 0.531 2.241 10.144 11.701 init_scf_run 11 5.9 0.000 0.001 11.590 11.591 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.590 11.591 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.737 11.445 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.421 10.054 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.595 9.804 mp_alltoall_i22 718 14.1 5.562 9.363 5.562 9.363 pw_transfer 1559 11.6 0.086 0.100 8.303 8.375 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 8.075 8.153 grid_integrate_task_list 129 12.3 7.608 8.013 7.608 8.013 cp_fm_cholesky_decompose 22 10.9 7.873 7.971 7.873 7.971 wfi_extrapolate 11 7.9 0.001 0.001 7.704 7.704 fft_wrap_pw1pw2_140 527 13.2 0.484 0.491 7.093 7.185 density_rs2pw 129 9.7 0.005 0.006 6.659 6.933 multiply_cannon_metrocomm4 25270 15.6 0.077 0.090 2.804 6.906 mp_irecv_dv 76685 16.2 2.655 6.644 2.655 6.644 calculate_dm_sparse 129 9.5 0.001 0.001 6.390 6.461 fft3d_ps 1301 14.7 2.825 2.929 6.292 6.374 mp_alltoall_d11v 2423 14.1 5.266 5.640 5.266 5.640 grid_collocate_task_list 129 9.7 5.250 5.611 5.250 5.611 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.497 5.553 rs_pw_transfer 1054 12.0 0.013 0.014 4.927 5.245 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.600 4.678 potential_pw2rs 129 12.3 0.023 0.023 4.534 4.556 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 4.468 4.470 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.383 4.435 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.231 4.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=194.686000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1091.636364, yerr=26.799840 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.525858E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.065 177.262 177.263 qs_mol_dyn_low 1 2.0 0.003 0.003 176.854 176.867 qs_forces 11 3.9 0.003 0.004 176.757 176.763 qs_energies 11 4.9 0.001 0.002 169.421 169.432 scf_env_do_scf 11 5.9 0.001 0.002 152.189 152.205 velocity_verlet 10 3.0 0.001 0.001 115.604 115.608 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.989 114.990 qs_scf_new_mos 117 7.6 0.001 0.001 79.002 79.098 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.002 79.098 dbcsr_multiply_generic 2507 12.6 0.180 0.185 74.792 75.234 ot_scf_mini 117 9.6 0.003 0.004 74.554 74.637 multiply_cannon 2507 13.6 0.583 0.623 54.576 59.411 multiply_cannon_loop 2507 14.6 0.445 0.458 49.839 50.724 ot_mini 117 10.6 0.001 0.001 40.187 40.258 init_scf_loop 11 6.9 0.000 0.000 37.027 37.028 mp_waitall_1 129618 16.6 26.534 33.876 26.534 33.876 prepare_preconditioner 11 7.9 0.000 0.000 33.208 33.233 make_preconditioner 11 8.9 0.000 0.000 33.208 33.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.009 31.289 rebuild_ks_matrix 128 8.3 0.001 0.001 28.768 28.869 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.768 28.868 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.266 26.358 qs_ot_get_p 128 10.4 0.001 0.001 21.363 21.440 multiply_cannon_multrec 10028 15.6 10.448 14.173 17.735 20.783 qs_ot_get_derivative 117 11.6 0.002 0.002 20.181 20.269 ot_diis_step 117 11.6 0.020 0.020 19.926 19.927 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.482 19.695 apply_single 128 13.6 0.001 0.001 19.482 19.695 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.179 19.199 cp_fm_cholesky_invert 11 10.9 19.071 19.078 19.071 19.078 make_m2s 5014 13.6 0.068 0.072 16.425 18.792 make_images 5014 14.6 2.306 2.619 16.122 18.489 qs_ot_p2m_diag 83 11.4 0.496 0.501 17.402 17.419 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.215 16.216 sum_up_and_integrate 128 10.3 0.181 0.191 14.272 14.319 integrate_v_rspace 128 11.3 0.004 0.004 14.090 14.146 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.338 13.371 calculate_rho_elec 128 8.7 0.258 0.269 13.337 13.370 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.022 13.023 cp_fm_diag_elpa_base 83 14.4 12.756 12.848 13.011 13.011 make_images_data 5014 15.6 0.055 0.061 10.009 12.453 hybrid_alltoall_any 5200 16.5 0.842 3.821 9.685 12.172 multiply_cannon_sync_h2d 10028 15.6 11.642 12.125 11.642 12.125 init_scf_run 11 5.9 0.000 0.001 10.459 10.460 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.459 10.459 cp_fm_cholesky_decompose 22 10.9 8.404 8.491 8.404 8.491 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.214 8.281 grid_integrate_task_list 128 12.3 7.754 8.123 7.754 8.123 pw_transfer 1547 11.6 0.084 0.090 7.895 7.910 multiply_cannon_metrocomm1 10028 15.6 0.028 0.028 4.660 7.796 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.674 7.695 mp_allgather_i34 2507 14.6 2.933 7.631 2.933 7.631 dbcsr_mm_accdrv_process 20762 16.1 2.594 3.437 6.923 7.559 wfi_extrapolate 11 7.9 0.001 0.001 7.467 7.467 fft_wrap_pw1pw2_140 523 13.2 0.502 0.524 6.678 6.698 density_rs2pw 128 9.7 0.005 0.005 6.156 6.467 calculate_dm_sparse 128 9.5 0.001 0.001 6.115 6.218 grid_collocate_task_list 128 9.7 5.479 5.877 5.479 5.877 dbcsr_complete_redistribute 395 12.7 2.125 2.196 5.439 5.867 fft3d_ps 1291 14.7 2.729 2.835 5.833 5.861 mp_alltoall_d11v 2415 14.1 4.896 5.571 4.896 5.571 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.218 5.222 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.215 5.217 rs_pw_transfer 1046 11.9 0.013 0.013 4.385 4.598 potential_pw2rs 128 12.3 0.026 0.026 4.285 4.304 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.090 4.119 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.551 3.897 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.553 3.869 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.760 3.798 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.843 3.758 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.642 3.714 mp_irecv_dv 28860 15.9 1.806 3.692 1.806 3.692 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.530 3.555 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.263000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1431.545455, yerr=55.385203 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.943836E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59908. MP_Allreduce 10870 1518. MP_Sync 86 MP_Alltoall 1700 36954374. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.049 294.494 294.495 qs_mol_dyn_low 1 2.0 0.003 0.003 293.882 293.894 qs_forces 11 3.9 0.003 0.003 293.787 293.789 qs_energies 11 4.9 0.002 0.002 285.071 285.092 scf_env_do_scf 11 5.9 0.001 0.002 263.213 263.225 velocity_verlet 10 3.0 0.001 0.001 211.412 211.420 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 137.098 137.100 init_scf_loop 11 6.9 0.000 0.000 125.839 125.842 prepare_preconditioner 11 7.9 0.000 0.000 121.008 121.046 make_preconditioner 11 8.9 0.000 0.000 121.008 121.046 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.373 118.222 qs_scf_new_mos 116 7.6 0.001 0.001 94.841 95.055 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.840 95.054 ot_scf_mini 116 9.6 0.004 0.004 89.966 90.121 dbcsr_multiply_generic 2485 12.5 0.209 0.216 84.557 85.525 cp_fm_upper_to_full 104 14.8 52.493 75.120 52.493 75.120 multiply_cannon 2485 13.5 0.702 0.773 59.127 60.136 multiply_cannon_loop 2485 14.5 0.466 0.478 55.183 57.432 ot_mini 116 10.6 0.001 0.001 45.713 45.908 dbcsr_complete_redistribute 393 12.7 3.987 4.030 29.409 42.141 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.948 38.690 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.588 36.162 cp_fm_cholesky_invert 11 10.9 35.073 35.079 35.073 35.079 mp_alltoall_i22 712 14.1 21.446 34.368 21.446 34.368 mp_waitall_1 104546 16.7 29.238 34.231 29.238 34.231 rebuild_ks_matrix 127 8.3 0.001 0.001 33.771 33.973 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.771 33.973 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.437 31.618 qs_ot_get_p 127 10.4 0.001 0.001 28.563 28.742 qs_ot_get_derivative 116 11.6 0.002 0.002 24.958 25.118 qs_ot_p2m_diag 82 11.4 0.868 0.874 24.268 24.301 cp_dbcsr_syevd 82 12.4 0.005 0.006 22.496 22.501 make_m2s 4970 13.5 0.075 0.078 20.256 21.495 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 19.260 21.381 make_images 4970 14.5 3.739 3.842 19.787 21.029 ot_diis_step 116 11.6 0.022 0.022 20.698 20.699 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.692 20.178 apply_single 127 13.6 0.001 0.001 19.691 20.177 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.185 19.185 cp_fm_diag_elpa_base 82 14.4 14.909 16.438 19.180 19.180 multiply_cannon_multrec 9940 15.5 10.440 12.240 17.829 17.899 sum_up_and_integrate 127 10.3 0.319 0.322 15.840 15.927 integrate_v_rspace 127 11.3 0.004 0.004 15.520 15.608 multiply_cannon_sync_h2d 9940 15.5 15.539 15.553 15.539 15.553 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.209 15.224 calculate_rho_elec 127 8.7 0.480 0.480 15.208 15.223 make_images_data 4970 15.5 0.061 0.065 10.909 12.975 hybrid_alltoall_any 5155 16.4 1.294 3.028 10.885 12.923 init_scf_run 11 5.9 0.000 0.001 11.985 11.985 scf_env_initial_rho_setup 11 6.9 0.002 0.002 11.984 11.985 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.804 9.931 cp_fm_cholesky_decompose 22 10.9 9.503 9.624 9.503 9.624 dbcsr_mm_accdrv_process 20590 16.0 3.685 5.332 7.156 8.941 wfi_extrapolate 11 7.9 0.001 0.001 8.791 8.791 grid_integrate_task_list 127 12.3 8.483 8.668 8.483 8.668 pw_transfer 1535 11.6 0.090 0.090 8.608 8.617 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.376 8.384 qs_energies_init_hamiltonians 11 5.9 0.005 0.007 7.855 7.856 fft_wrap_pw1pw2_140 519 13.2 0.535 0.537 7.366 7.377 mp_alltoall_d11v 2401 14.1 7.142 7.306 7.142 7.306 calculate_dm_sparse 127 9.5 0.001 0.001 6.759 6.888 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.417 6.519 fft3d_ps 1281 14.7 2.733 2.743 6.436 6.448 grid_collocate_task_list 127 9.7 6.292 6.322 6.292 6.322 density_rs2pw 127 9.7 0.005 0.005 6.265 6.294 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.147 6.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=294.495000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2608.272727, yerr=168.663242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_performance_tests/21/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 2.766000E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 419739 0.0% 0.0% 100.0% average stack size 0.0 0.0 22952.9 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 1.261486E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 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 0.059 0.185 85.394 85.395 qs_energies 1 2.0 0.001 0.007 84.774 84.799 ls_scf 1 3.0 0.002 0.014 83.874 83.897 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.721 72.893 multiply_cannon 111 7.7 0.018 0.020 55.885 57.080 multiply_cannon_loop 111 8.7 0.213 0.228 52.500 53.798 ls_scf_main 1 4.0 0.001 0.010 52.309 52.312 density_matrix_trs4 2 5.0 0.003 0.019 46.818 46.900 ls_scf_init_scf 1 4.0 0.000 0.002 28.463 28.482 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.152 27.204 mp_waitall_1 11316 10.9 22.554 25.146 22.554 25.146 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.061 25.078 multiply_cannon_multrec 2664 9.7 8.170 8.978 15.569 17.374 multiply_cannon_sync_h2d 2664 9.7 13.637 15.383 13.637 15.383 make_m2s 222 7.7 0.009 0.011 13.278 13.702 make_images 222 8.7 0.101 0.115 13.256 13.682 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.656 12.491 make_images_data 222 9.7 0.005 0.006 7.813 8.354 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.415 8.211 hybrid_alltoall_any 227 10.6 0.218 1.849 6.770 8.155 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.652 7.015 8.003 dbcsr_mm_accdrv_process_sort 4760 11.4 6.303 7.213 6.303 7.213 calculate_norms 4752 9.8 5.527 6.243 5.527 6.243 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.047 5.287 mp_sum_l 807 5.4 3.084 4.812 3.084 4.812 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.058 3.798 make_images_sizes 222 9.7 0.000 0.000 0.731 3.795 mp_alltoall_i44 222 10.7 0.730 3.795 0.730 3.795 mp_irecv_dv 6231 10.9 2.040 3.767 2.040 3.767 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.207 3.521 arnoldi_extremal 4 6.8 0.000 0.000 3.244 3.273 arnoldi_normal_ev 4 7.8 0.001 0.004 3.244 3.273 build_subspace 16 8.4 0.009 0.012 3.147 3.149 ls_scf_post 1 4.0 0.004 0.026 3.100 3.126 ls_scf_store_result 1 5.0 0.000 0.000 2.913 2.956 dbcsr_special_finalize 555 9.7 0.005 0.006 2.369 2.755 dbcsr_merge_single_wm 555 10.7 0.454 0.575 2.360 2.745 make_images_pack 222 9.7 2.216 2.634 2.218 2.636 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.325 2.569 dbcsr_sort_data 658 11.4 2.163 2.521 2.163 2.521 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.460 2.071 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.227 2.320 buffer_matrices_ensure_size 222 8.7 1.752 2.074 1.752 2.074 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.855 1.858 rebuild_ks_matrix 3 7.3 0.000 0.000 1.845 1.848 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.015 1.845 1.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.395000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 2.119164E+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 0.030 0.045 91.018 91.019 qs_energies 1 2.0 0.000 0.000 90.480 90.486 ls_scf 1 3.0 0.000 0.001 89.169 89.172 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.372 75.676 multiply_cannon 111 7.7 0.029 0.038 53.360 57.366 ls_scf_main 1 4.0 0.000 0.001 54.869 54.870 multiply_cannon_loop 111 8.7 0.117 0.124 50.036 53.503 density_matrix_trs4 2 5.0 0.002 0.003 49.238 49.446 ls_scf_init_scf 1 4.0 0.002 0.014 30.798 30.800 mp_waitall_1 9246 10.9 21.051 30.026 21.051 30.026 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.646 29.758 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.108 27.121 multiply_cannon_multrec 1332 9.7 13.128 16.214 22.458 26.638 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.748 20.807 make_m2s 222 7.7 0.008 0.009 15.365 16.107 make_images 222 8.7 1.591 1.951 15.335 16.076 dbcsr_mm_accdrv_process 4041 10.4 0.261 0.442 8.937 10.526 dbcsr_mm_accdrv_process_sort 4041 11.4 8.532 10.084 8.532 10.084 make_images_data 222 9.7 0.004 0.005 8.748 9.826 hybrid_alltoall_any 227 10.6 0.523 2.441 8.167 9.488 mp_sum_l 807 5.4 5.342 8.740 5.342 8.740 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.234 7.797 mp_irecv_dv 3311 11.0 3.215 7.734 3.215 7.734 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.166 6.965 calculate_norms 2376 9.8 6.019 6.720 6.019 6.720 multiply_cannon_sync_h2d 1332 9.7 4.786 5.890 4.786 5.890 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.084 5.287 arnoldi_extremal 4 6.8 0.000 0.000 4.596 4.616 arnoldi_normal_ev 4 7.8 0.001 0.004 4.596 4.616 build_subspace 16 8.4 0.014 0.021 4.341 4.344 ls_scf_post 1 4.0 0.000 0.001 3.501 3.507 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.129 3.356 ls_scf_store_result 1 5.0 0.000 0.000 3.210 3.323 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.225 2.747 3.227 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.188 2.745 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.507 2.586 make_images_pack 222 9.7 2.070 2.487 2.073 2.490 mp_allgather_i34 111 8.7 1.003 2.488 1.003 2.488 dbcsr_sort_data 436 11.2 1.826 2.079 1.826 2.079 dbcsr_data_new 4174 10.1 1.625 1.832 1.625 1.832 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.819 1.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.019000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 2.765984E+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 0.065 0.080 95.001 95.003 qs_energies 1 2.0 0.000 0.000 94.267 94.269 ls_scf 1 3.0 0.000 0.001 92.821 92.824 dbcsr_multiply_generic 111 6.7 0.016 0.027 77.569 77.850 ls_scf_main 1 4.0 0.000 0.001 58.132 58.136 multiply_cannon 111 7.7 0.038 0.071 53.210 57.179 multiply_cannon_loop 111 8.7 0.100 0.115 49.640 53.902 density_matrix_trs4 2 5.0 0.002 0.003 52.219 52.349 mp_waitall_1 7374 11.0 24.195 34.608 24.195 34.608 ls_scf_init_scf 1 4.0 0.002 0.008 31.026 31.028 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.867 29.937 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.404 27.414 multiply_cannon_multrec 888 9.7 12.753 15.058 21.386 24.295 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.127 23.380 make_m2s 222 7.7 0.007 0.008 17.401 18.625 make_images 222 8.7 1.981 2.305 17.362 18.585 hybrid_alltoall_any 227 10.6 0.621 2.884 9.657 11.218 make_images_data 222 9.7 0.004 0.005 9.989 11.014 dbcsr_mm_accdrv_process 3754 10.4 0.263 0.457 8.165 9.427 mp_sum_l 807 5.4 5.380 9.100 5.380 9.100 dbcsr_mm_accdrv_process_sort 3754 11.4 7.784 8.970 7.784 8.970 multiply_cannon_sync_h2d 888 9.7 6.038 7.276 6.038 7.276 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.156 7.126 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.704 7.005 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.451 6.879 mp_irecv_dv 2335 11.1 2.436 6.832 2.436 6.832 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.958 5.161 arnoldi_extremal 4 6.8 0.000 0.000 5.110 5.126 arnoldi_normal_ev 4 7.8 0.001 0.006 5.110 5.125 build_subspace 16 8.4 0.014 0.020 4.796 4.807 calculate_norms 1584 9.8 4.263 4.627 4.263 4.627 mp_allgather_i34 111 8.7 1.395 3.813 1.395 3.813 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.448 3.785 ls_scf_post 1 4.0 0.000 0.001 3.663 3.666 dbcsr_matrix_vector_mult_local 304 10.0 3.028 3.608 3.030 3.610 ls_scf_store_result 1 5.0 0.000 0.000 3.390 3.493 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.752 2.868 make_images_sizes 222 9.7 0.000 0.000 1.047 2.197 mp_alltoall_i44 222 10.7 1.047 2.197 1.047 2.197 dbcsr_sort_data 325 11.1 1.912 2.176 1.912 2.176 make_images_pack 222 9.7 1.817 2.107 1.820 2.110 dbcsr_data_release 9322 10.9 1.317 1.911 1.317 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.003000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2179.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 3.321651E+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 0.081 0.131 97.669 97.670 qs_energies 1 2.0 0.000 0.000 96.738 96.743 ls_scf 1 3.0 0.000 0.001 95.101 95.106 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.927 79.175 ls_scf_main 1 4.0 0.004 0.029 59.114 59.115 multiply_cannon 111 7.7 0.058 0.133 52.047 56.193 density_matrix_trs4 2 5.0 0.002 0.003 53.036 53.203 multiply_cannon_loop 111 8.7 0.114 0.123 46.982 50.196 ls_scf_init_scf 1 4.0 0.001 0.002 32.718 32.721 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.504 31.568 mp_waitall_1 6438 11.0 22.902 29.320 22.902 29.320 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.992 29.008 multiply_cannon_multrec 1332 9.7 14.321 17.973 22.310 25.359 make_m2s 222 7.7 0.008 0.009 21.216 22.645 make_images 222 8.7 3.153 3.607 21.165 22.596 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.418 17.420 make_images_data 222 9.7 0.004 0.005 11.829 13.499 hybrid_alltoall_any 227 10.6 0.798 3.777 11.175 12.791 dbcsr_mm_accdrv_process 3641 10.4 0.236 0.406 7.611 9.128 dbcsr_mm_accdrv_process_sort 3641 11.4 7.237 8.710 7.237 8.710 mp_sum_l 807 5.4 4.122 7.700 4.122 7.700 multiply_cannon_sync_h2d 1332 9.7 5.550 6.158 5.550 6.158 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.122 6.030 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.087 5.971 mp_irecv_dv 3229 10.9 2.063 5.877 2.063 5.877 arnoldi_extremal 4 6.8 0.000 0.000 5.250 5.261 arnoldi_normal_ev 4 7.8 0.001 0.004 5.250 5.261 build_subspace 16 8.4 0.014 0.021 4.919 4.927 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.585 4.770 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.484 4.707 calculate_norms 2376 9.8 4.183 4.537 4.183 4.537 mp_allgather_i34 111 8.7 2.174 4.477 2.174 4.477 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.613 3.911 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.696 3.205 3.698 dbcsr_sort_data 658 11.4 3.066 3.328 3.066 3.328 ls_scf_post 1 4.0 0.000 0.001 3.269 3.274 dbcsr_special_finalize 555 9.7 0.006 0.007 2.826 3.218 dbcsr_merge_single_wm 555 10.7 0.534 0.661 2.817 3.210 ls_scf_store_result 1 5.0 0.000 0.000 3.012 3.071 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.893 2.948 dbcsr_data_release 10477 10.7 1.603 2.386 1.603 2.386 dbcsr_finalize 304 7.8 0.050 0.062 1.799 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.670000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2720.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 4.629643E+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 0.042 0.057 99.315 99.315 qs_energies 1 2.0 0.000 0.000 98.499 98.505 ls_scf 1 3.0 0.000 0.000 96.595 96.600 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.103 78.323 ls_scf_main 1 4.0 0.000 0.000 62.223 62.224 multiply_cannon 111 7.7 0.087 0.180 55.574 61.005 density_matrix_trs4 2 5.0 0.002 0.003 55.212 55.296 multiply_cannon_loop 111 8.7 0.069 0.079 50.919 52.575 mp_waitall_1 5481 11.0 26.343 30.728 26.343 30.728 ls_scf_init_scf 1 4.0 0.000 0.000 30.706 30.708 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.518 29.555 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.306 27.315 multiply_cannon_multrec 444 9.7 14.164 16.404 21.268 24.487 make_m2s 222 7.7 0.006 0.006 17.688 20.261 make_images 222 8.7 3.733 4.429 17.625 20.199 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.180 16.224 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.107 15.036 make_images_data 222 9.7 0.003 0.004 9.858 12.369 hybrid_alltoall_any 227 10.6 0.791 3.775 9.551 12.250 multiply_cannon_sync_h2d 444 9.7 6.576 8.523 6.576 8.523 dbcsr_mm_accdrv_process 3003 10.4 0.195 0.380 6.812 7.924 dbcsr_mm_accdrv_process_sort 3003 11.4 6.488 7.581 6.488 7.581 mp_allgather_i34 111 8.7 2.852 7.044 2.852 7.044 arnoldi_extremal 4 6.8 0.000 0.000 5.991 6.002 arnoldi_normal_ev 4 7.8 0.002 0.005 5.991 6.002 build_subspace 16 8.4 0.015 0.019 5.602 5.613 mp_sum_l 807 5.4 2.874 4.981 2.874 4.981 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.619 4.768 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.275 4.497 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.561 4.322 mp_irecv_dv 1241 11.2 1.543 4.294 1.543 4.294 dbcsr_matrix_vector_mult_local 304 10.0 3.738 4.225 3.740 4.227 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.996 3.902 calculate_norms 792 9.8 3.560 3.719 3.560 3.719 ls_scf_post 1 4.0 0.000 0.000 3.666 3.672 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.496 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.389 3.476 make_images_sizes 222 9.7 0.000 0.000 1.150 3.402 mp_alltoall_i44 222 10.7 1.149 3.402 1.149 3.402 dbcsr_finalize 304 7.8 0.062 0.078 2.209 2.270 dbcsr_merge_all 275 8.9 0.476 0.529 2.056 2.100 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.315000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3594.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d8f36242127b9a828f127550ecd9613eefb3f1cc_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 8.785744E+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 0.084 0.109 113.390 113.390 qs_energies 1 2.0 0.000 0.000 111.933 111.953 ls_scf 1 3.0 0.000 0.000 109.008 109.028 dbcsr_multiply_generic 111 6.7 0.023 0.026 82.692 82.824 ls_scf_main 1 4.0 0.000 0.000 69.438 69.439 density_matrix_trs4 2 5.0 0.002 0.003 60.266 60.335 multiply_cannon 111 7.7 0.094 0.169 53.551 56.721 multiply_cannon_loop 111 8.7 0.068 0.070 49.431 50.007 ls_scf_init_scf 1 4.0 0.000 0.000 35.679 35.680 ls_scf_init_matrix_S 1 5.0 0.000 0.001 34.325 34.340 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.442 31.452 mp_waitall_1 4569 11.1 26.468 30.867 26.468 30.867 make_m2s 222 7.7 0.007 0.007 25.385 27.027 make_images 222 8.7 4.596 4.984 25.275 26.916 multiply_cannon_multrec 444 9.7 17.916 18.649 22.575 23.142 hybrid_alltoall_any 227 10.6 1.666 3.634 14.300 18.007 make_images_data 222 9.7 0.003 0.004 14.494 17.682 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 13.010 13.812 multiply_cannon_sync_h2d 444 9.7 8.858 9.030 8.858 9.030 arnoldi_extremal 4 6.8 0.000 0.000 7.650 7.665 arnoldi_normal_ev 4 7.8 0.004 0.010 7.650 7.664 build_subspace 16 8.4 0.026 0.036 7.095 7.108 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.618 5.815 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.324 5.595 dbcsr_matrix_vector_mult_local 304 10.0 5.120 5.435 5.122 5.437 mp_allgather_i34 111 8.7 1.725 5.028 1.725 5.028 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.893 4.983 dbcsr_mm_accdrv_process 1814 10.4 0.256 0.320 4.488 4.615 dbcsr_mm_accdrv_process_sort 1814 11.4 4.186 4.319 4.186 4.319 ls_scf_post 1 4.0 0.000 0.000 3.891 3.911 make_images_sizes 222 9.7 0.000 0.000 1.532 3.861 mp_alltoall_i44 222 10.7 1.532 3.861 1.532 3.861 ls_scf_store_result 1 5.0 0.000 0.000 3.642 3.649 calculate_norms 792 9.8 3.229 3.304 3.229 3.304 dbcsr_finalize 304 7.8 0.082 0.090 3.096 3.143 dbcsr_complete_redistribute 5 7.6 1.457 1.490 2.854 2.982 dbcsr_merge_all 275 8.9 0.894 0.928 2.880 2.921 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.894 2.894 dbcsr_data_release 12724 10.6 2.339 2.863 2.339 2.863 matrix_ls_to_qs 2 6.0 0.000 0.000 2.470 2.613 dbcsr_sort_data 325 11.1 2.444 2.500 2.444 2.500 dbcsr_new_transposed 4 7.5 0.243 0.251 2.341 2.348 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=113.390000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d8f36242127b9a828f127550ecd9613eefb3f1cc Summary: empty Status: OK