=== 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: 48d3695c3ced91faad61a1fd2c67ef2ebe356476 ################# 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 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231215 # # 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. # # Last update: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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.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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/01 job id: 50956284 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/02 job id: 50956286 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/03 job id: 50956288 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/04 job id: 50956290 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/05 job id: 50956291 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/06 job id: 50956292 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/07 job id: 50956293 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/08 job id: 50956294 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/09 job id: 50956295 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/10 job id: 50956297 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/11 job id: 50956299 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/12 job id: 50956301 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/13 job id: 50956302 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/14 job id: 50956305 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/15 job id: 50956308 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/16 job id: 50956310 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/17 job id: 50956312 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/18 job id: 50956316 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/19 job id: 50956317 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/20 job id: 50956318 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/21 job id: 50956320 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/22 job id: 50956321 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/23 job id: 50956322 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/24 job id: 50956323 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/25 job id: 50956324 --- 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/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/26 job id: 50956325 --- 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 ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/27 job id: 50956327 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ 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]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 172669. MP_Allreduce 424 8. 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 0.018 0.033 133.994 133.995 farming_run 1 2.0 133.441 133.443 133.966 133.970 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.492103E+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 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 115.047 115.048 qs_energies 1 2.0 0.000 0.000 114.852 114.854 mp2_main 1 3.0 0.000 0.000 112.758 112.760 mp2_gpw_main 1 4.0 0.028 0.034 111.707 111.709 mp2_ri_gpw_compute_in 1 5.0 0.174 0.201 92.552 93.072 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.611 55.131 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.163 41.028 46.247 get_2c_integrals 1 6.0 0.008 0.009 37.170 37.768 integrate_v_rspace 273 8.0 0.439 0.456 24.615 29.519 pw_transfer 6555 10.6 0.384 0.410 26.902 27.586 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.521 26.285 grid_integrate_task_list 273 9.0 20.507 25.878 20.507 25.878 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.030 23.057 23.787 compute_2c_integrals 1 7.0 0.002 0.002 19.462 19.463 rpa_ri_compute_en 1 5.0 0.019 0.021 19.051 19.243 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.636 19.195 mp2_eri_2c_integrate_gpw 1 9.0 2.366 2.418 18.635 19.194 cp_fm_cholesky_decompose 12 8.2 17.644 18.254 17.644 18.254 cholesky_decomp 1 7.0 0.000 0.000 16.549 17.157 fft3d_s 5443 13.4 16.146 16.617 16.168 16.638 ao_to_mo_and_store_B_mult_1 272 7.0 10.763 15.319 10.763 15.319 calculate_wavefunction 272 8.0 5.376 5.516 12.318 12.959 rpa_num_int 1 6.0 0.001 0.014 10.668 10.668 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.636 10.666 calc_mat_Q 8 8.0 0.000 0.000 9.496 9.608 contract_S_to_Q 8 9.0 0.000 0.000 8.922 9.032 calc_potential_gpw 544 9.5 0.004 0.005 8.246 8.749 parallel_gemm_fm 14 9.1 0.000 0.000 8.493 8.632 parallel_gemm_fm_cosma 14 10.1 8.493 8.632 8.493 8.632 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.106 8.466 potential_pw2rs 545 10.0 0.106 0.107 7.521 8.168 create_integ_mat 1 6.0 0.024 0.028 7.772 7.772 collocate_single_gaussian 272 10.0 0.038 0.041 7.358 7.654 array2fm 1 7.0 0.000 0.000 6.720 7.207 pw_scatter_s 2720 13.7 4.362 4.498 4.362 4.498 pw_gather_s 2722 13.2 3.449 3.762 3.449 3.762 array2fm_buffer_send 1 8.0 2.984 3.129 2.984 3.129 pw_poisson_solve 545 10.5 1.120 1.171 2.203 2.445 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.708667, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 200775. MP_Allreduce 424 9. 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 0.025 0.035 421.872 421.874 farming_run 1 2.0 421.156 421.163 421.823 421.826 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.244738E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 217.448 217.449 qs_energies 1 2.0 0.000 0.000 217.228 217.240 scf_env_do_scf 1 3.0 0.000 0.000 115.438 115.439 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.553 114.560 rebuild_ks_matrix 4 6.0 0.000 0.000 114.552 114.559 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.552 114.559 hfx_ks_matrix 4 8.0 0.001 0.001 114.213 114.217 integrate_four_center 4 9.0 0.154 0.469 114.212 114.216 integrate_four_center_main 4 10.0 0.160 0.618 101.891 105.832 integrate_four_center_bin 265 11.0 101.732 105.772 101.732 105.772 mp2_main 1 3.0 0.002 0.012 101.501 101.514 mp2_gpw_main 1 4.0 0.052 0.081 100.647 100.656 init_scf_loop 1 4.0 0.000 0.000 96.831 96.831 mp2_ri_gpw_compute_in 1 5.0 0.065 0.072 73.899 74.965 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.782 54.845 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.163 41.341 46.540 integrate_v_rspace 95 8.0 0.397 0.572 27.795 32.815 pw_transfer 2240 10.6 0.147 0.167 29.476 29.874 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.494 28.899 mp2_ri_gpw_compute_en 1 5.0 0.055 0.066 26.599 28.701 grid_integrate_task_list 95 9.0 23.221 28.430 23.221 28.430 ao_to_mo_and_store_B_mult_1 91 7.0 10.765 28.039 10.765 28.039 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.261 26.632 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.891 24.887 24.899 get_2c_integrals 1 6.0 0.000 0.000 20.019 20.039 compute_2c_integrals 1 7.0 0.002 0.003 19.001 19.009 fft3d_s 1823 13.4 18.494 18.858 18.507 18.871 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.569 18.861 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.843 18.568 18.859 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.606 18.606 calculate_wavefunction 91 8.0 2.017 2.043 9.630 9.847 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.581 8.761 9.436 potential_pw2rs 186 10.0 0.033 0.034 8.416 9.002 local_gemm 172 8.0 8.205 8.867 8.205 8.867 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.120 8.539 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.517 7.905 8.257 calc_potential_gpw 182 9.5 0.002 0.002 7.869 8.059 collocate_single_gaussian 91 10.0 0.016 0.020 7.727 7.973 mp_sync 37 10.5 4.276 7.568 4.276 7.568 integrate_four_center_load 4 10.0 0.000 0.000 6.784 6.788 hfx_load_balance 1 11.0 0.000 0.000 6.784 6.788 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.390 6.339 6.390 mp_sendrecv_dm3 2068 8.0 5.942 6.270 5.942 6.270 pw_gather_s 912 13.2 4.499 5.035 4.499 5.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.650057, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 453.435392E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 53.724 53.725 qs_mol_dyn_low 1 2.0 0.004 0.010 53.441 53.449 qs_forces 11 3.9 0.002 0.003 53.365 53.366 qs_energies 11 4.9 0.002 0.003 51.868 51.897 scf_env_do_scf 11 5.9 0.000 0.001 45.436 45.436 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 43.365 43.366 dbcsr_multiply_generic 2286 12.5 0.093 0.099 33.546 33.948 qs_scf_new_mos 108 7.5 0.000 0.001 33.268 33.589 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.268 33.589 ot_scf_mini 108 9.5 0.002 0.002 31.564 31.796 multiply_cannon 2286 13.5 0.189 0.195 26.411 27.731 multiply_cannon_loop 2286 14.5 1.805 1.911 25.753 27.104 velocity_verlet 10 3.0 0.001 0.002 26.529 26.530 ot_mini 108 10.5 0.001 0.001 18.862 19.112 qs_ot_get_derivative 108 11.5 0.001 0.001 15.938 16.128 mp_waitall_1 245248 16.5 8.258 14.392 8.258 14.392 multiply_cannon_metrocomm3 54864 15.5 0.072 0.076 5.980 12.864 multiply_cannon_multrec 54864 15.5 3.666 5.797 7.842 11.338 qs_ot_get_p 119 10.4 0.001 0.001 7.954 8.226 rebuild_ks_matrix 119 8.3 0.000 0.000 7.825 7.963 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.825 7.962 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.899 7.024 mp_sum_l 7287 12.8 5.207 6.823 5.207 6.823 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.824 6.267 multiply_cannon_sync_h2d 54864 15.5 5.190 6.138 5.190 6.138 dbcsr_mm_accdrv_process 76910 16.1 1.819 2.851 4.089 5.827 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.220 5.291 init_scf_run 11 5.9 0.000 0.001 5.082 5.082 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.082 5.082 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.904 5.012 sum_up_and_integrate 119 10.3 0.001 0.002 4.434 4.441 integrate_v_rspace 119 11.3 0.002 0.002 4.423 4.431 cp_dbcsr_syevd 50 12.0 0.003 0.004 4.411 4.412 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.193 4.363 calculate_rho_elec 119 8.7 0.011 0.016 4.192 4.363 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.219 4.220 cp_fm_redistribute_end 50 14.0 2.153 4.195 2.158 4.197 cp_fm_diag_elpa_base 50 14.0 2.031 4.085 2.036 4.093 calculate_dm_sparse 119 9.5 0.000 0.000 3.041 3.186 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.917 3.101 apply_single 119 13.6 0.000 0.000 2.917 3.101 acc_transpose_blocks 54864 15.5 0.225 0.249 2.396 2.993 jit_kernel_multiply 13 15.8 2.208 2.881 2.208 2.881 calculate_first_density_matrix 1 7.0 0.000 0.000 2.865 2.869 ot_diis_step 108 11.5 0.006 0.006 2.686 2.686 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.547 2.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.557 2.559 multiply_cannon_metrocomm1 54864 15.5 0.055 0.062 1.629 2.531 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.252 2.318 density_rs2pw 119 9.7 0.004 0.004 2.195 2.304 wfi_extrapolate 11 7.9 0.001 0.001 2.149 2.149 grid_integrate_task_list 119 12.3 2.018 2.123 2.018 2.123 init_scf_loop 11 6.9 0.000 0.000 2.053 2.053 mp_sum_d 4137 12.0 1.303 1.924 1.303 1.924 potential_pw2rs 119 12.3 0.004 0.004 1.818 1.835 pw_transfer 1439 11.6 0.052 0.057 1.713 1.788 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.637 1.716 make_m2s 4572 13.5 0.053 0.055 1.620 1.669 make_images 4572 14.5 0.133 0.139 1.538 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.505 1.540 transfer_rs2pw 487 10.6 0.005 0.006 1.396 1.484 mp_alltoall_d11v 2130 13.8 1.306 1.474 1.306 1.474 mp_waitany 12084 13.8 1.282 1.450 1.282 1.450 grid_collocate_task_list 119 9.7 1.351 1.438 1.351 1.438 acc_transpose_blocks_sync 164592 16.5 1.194 1.426 1.194 1.426 fft3d_ps 1201 14.6 0.369 0.477 1.350 1.423 fft_wrap_pw1pw2_140 487 13.2 0.108 0.114 1.271 1.347 transfer_pw2rs 487 13.2 0.006 0.006 1.327 1.340 acc_transpose_blocks_kernels 54864 16.5 0.247 0.383 0.933 1.270 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.711 1.086 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.725000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.181818, yerr=1.266217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 490.106880E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1187056. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 39.786 39.787 qs_mol_dyn_low 1 2.0 0.003 0.003 39.563 39.570 qs_forces 11 3.9 0.003 0.008 39.508 39.509 qs_energies 11 4.9 0.003 0.008 37.812 37.817 scf_env_do_scf 11 5.9 0.001 0.002 32.272 32.273 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 29.807 29.808 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.698 23.054 qs_scf_new_mos 108 7.5 0.001 0.001 20.934 21.163 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.934 21.162 ot_scf_mini 108 9.5 0.003 0.003 20.007 20.175 multiply_cannon 2286 13.5 0.207 0.214 17.421 18.930 velocity_verlet 10 3.0 0.001 0.002 18.273 18.274 multiply_cannon_loop 2286 14.5 1.180 1.242 16.236 17.642 ot_mini 108 10.5 0.001 0.001 12.319 12.551 mp_waitall_1 200699 16.5 5.840 11.308 5.840 11.308 qs_ot_get_derivative 108 11.5 0.001 0.001 9.870 10.040 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.224 9.994 multiply_cannon_multrec 27432 15.5 1.828 4.123 6.435 9.353 rebuild_ks_matrix 119 8.3 0.000 0.000 6.968 7.103 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 6.967 7.103 dbcsr_mm_accdrv_process 47894 16.0 3.843 6.253 4.526 6.729 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.133 6.256 qs_ot_get_p 119 10.4 0.001 0.002 4.764 4.987 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.687 4.532 init_scf_run 11 5.9 0.000 0.001 4.298 4.298 scf_env_initial_rho_setup 11 6.9 0.003 0.014 4.297 4.298 mp_sum_l 7287 12.8 2.086 4.178 2.086 4.178 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.072 4.134 apply_single 119 13.6 0.000 0.000 3.072 4.134 sum_up_and_integrate 119 10.3 0.002 0.003 3.863 3.876 integrate_v_rspace 119 11.3 0.002 0.006 3.848 3.863 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.696 3.725 calculate_rho_elec 119 8.7 0.021 0.024 3.695 3.724 qs_ot_p2m_diag 50 11.0 0.009 0.014 3.056 3.074 make_m2s 4572 13.5 0.052 0.053 2.740 3.065 make_images 4572 14.5 0.206 0.245 2.651 2.979 calculate_first_density_matrix 1 7.0 0.001 0.003 2.787 2.791 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.662 2.662 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.460 2.548 init_scf_loop 11 6.9 0.001 0.004 2.441 2.442 ot_diis_step 108 11.5 0.011 0.011 2.396 2.399 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.340 2.343 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.328 2.329 cp_fm_redistribute_end 50 14.0 1.182 2.300 1.189 2.304 multiply_cannon_sync_h2d 27432 15.5 1.700 2.253 1.700 2.253 cp_fm_diag_elpa_base 50 14.0 1.083 2.203 1.111 2.236 jit_kernel_multiply 9 16.3 0.625 2.217 0.625 2.217 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.115 2.156 calculate_dm_sparse 119 9.5 0.000 0.001 2.065 2.144 density_rs2pw 119 9.7 0.004 0.004 1.993 2.067 pw_transfer 1439 11.6 0.065 0.069 1.925 1.957 acc_transpose_blocks 27432 15.5 0.115 0.121 1.609 1.955 grid_integrate_task_list 119 12.3 1.836 1.922 1.836 1.922 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.833 1.867 make_images_data 4572 15.5 0.047 0.054 1.265 1.618 prepare_preconditioner 11 7.9 0.000 0.000 1.547 1.573 make_preconditioner 11 8.9 0.000 0.001 1.547 1.573 potential_pw2rs 119 12.3 0.006 0.006 1.528 1.534 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.453 1.510 fft3d_ps 1201 14.6 0.522 0.580 1.464 1.494 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.122 1.474 fft_wrap_pw1pw2_140 487 13.2 0.127 0.132 1.438 1.473 wfi_extrapolate 11 7.9 0.001 0.001 1.443 1.443 mp_alltoall_d11v 2130 13.8 1.274 1.409 1.274 1.409 grid_collocate_task_list 119 9.7 1.289 1.360 1.289 1.360 transfer_rs2pw 487 10.6 0.005 0.005 1.173 1.260 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.241 mp_sum_d 4137 12.0 0.586 1.043 0.586 1.043 mp_allgather_i34 2286 14.5 0.610 0.997 0.610 0.997 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 0.971 0.973 acc_transpose_blocks_sync 82296 16.5 0.812 0.939 0.812 0.939 transfer_pw2rs 487 13.2 0.004 0.005 0.922 0.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.906 make_images_sizes 4572 15.5 0.005 0.005 0.612 0.881 mp_alltoall_i44 4572 16.5 0.608 0.876 0.608 0.876 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.656 0.869 mp_alltoall_z22v 1201 16.6 0.728 0.798 0.728 0.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.787000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.818182, yerr=1.748671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.056256E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.034 32.713 32.715 qs_mol_dyn_low 1 2.0 0.003 0.003 32.486 32.494 qs_forces 11 3.9 0.002 0.003 32.177 32.178 qs_energies 11 4.9 0.002 0.008 30.631 30.633 scf_env_do_scf 11 5.9 0.001 0.002 25.358 25.359 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 22.930 22.930 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.405 17.483 qs_scf_new_mos 108 7.5 0.001 0.001 15.288 15.299 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.287 15.298 velocity_verlet 10 3.0 0.001 0.002 15.259 15.268 multiply_cannon 2286 13.5 0.194 0.200 14.162 14.981 ot_scf_mini 108 9.5 0.002 0.004 14.535 14.543 multiply_cannon_loop 2286 14.5 0.857 0.881 13.396 14.233 ot_mini 108 10.5 0.001 0.001 8.995 9.005 multiply_cannon_multrec 18288 15.5 1.877 2.838 7.543 7.788 qs_ot_get_derivative 108 11.5 0.001 0.001 7.513 7.521 dbcsr_mm_accdrv_process 38222 16.0 5.512 6.278 5.570 6.344 rebuild_ks_matrix 119 8.3 0.000 0.000 5.963 5.978 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 5.962 5.978 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.258 5.275 init_scf_run 11 5.9 0.000 0.001 4.012 4.012 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.011 4.012 sum_up_and_integrate 119 10.3 0.001 0.003 3.527 3.532 integrate_v_rspace 119 11.3 0.003 0.008 3.513 3.521 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.798 3.437 qs_ot_get_p 119 10.4 0.001 0.001 3.329 3.354 mp_waitall_1 158411 16.6 2.535 3.349 2.535 3.349 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.243 3.247 calculate_rho_elec 119 8.7 0.031 0.031 3.243 3.247 calculate_first_density_matrix 1 7.0 0.000 0.002 2.793 2.794 calculate_dm_sparse 119 9.5 0.000 0.001 2.447 2.455 init_scf_loop 11 6.9 0.002 0.013 2.411 2.412 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.038 2.339 apply_single 119 13.6 0.000 0.000 2.037 2.339 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.204 2.211 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.375 2.005 make_m2s 4572 13.5 0.044 0.045 1.790 1.931 cp_dbcsr_syevd 50 12.0 0.005 0.014 1.913 1.913 grid_integrate_task_list 119 12.3 1.807 1.889 1.807 1.889 density_rs2pw 119 9.7 0.004 0.004 1.807 1.869 pw_transfer 1439 11.6 0.065 0.069 1.847 1.857 make_images 4572 14.5 0.192 0.204 1.704 1.843 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.791 1.795 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.754 1.766 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.689 1.695 acc_transpose_blocks 18288 15.5 0.079 0.082 1.624 1.693 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.668 1.678 cp_fm_diag_elpa_base 50 14.0 1.636 1.650 1.658 1.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.653 1.655 prepare_preconditioner 11 7.9 0.000 0.000 1.626 1.629 make_preconditioner 11 8.9 0.000 0.001 1.626 1.629 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.482 1.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.501 1.508 ot_diis_step 108 11.5 0.011 0.011 1.467 1.467 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.420 1.431 mp_sum_l 7287 12.8 1.063 1.415 1.063 1.415 potential_pw2rs 119 12.3 0.007 0.008 1.330 1.334 fft3d_ps 1201 14.6 0.546 0.567 1.309 1.322 grid_collocate_task_list 119 9.7 1.235 1.315 1.235 1.315 multiply_cannon_sync_h2d 18288 15.5 1.055 1.200 1.055 1.200 wfi_extrapolate 11 7.9 0.001 0.001 1.165 1.165 transfer_rs2pw 487 10.6 0.005 0.005 1.027 1.097 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.025 1.026 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.937 0.955 make_images_data 4572 15.5 0.047 0.050 0.775 0.934 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.665 0.868 transfer_pw2rs 487 13.2 0.004 0.004 0.804 0.808 acc_transpose_blocks_sync 54864 16.5 0.735 0.803 0.735 0.803 acc_transpose_blocks_kernels 18288 16.5 0.218 0.223 0.790 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.785 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.671 0.741 mp_alltoall_d11v 2130 13.8 0.647 0.710 0.647 0.710 cp_fm_cholesky_invert 11 10.9 0.691 0.695 0.691 0.695 mp_alltoall_z22v 1201 16.6 0.614 0.690 0.614 0.690 mp_waitany 9880 13.7 0.585 0.658 0.585 0.658 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.715000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 558.993408E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.031 36.665 36.670 qs_mol_dyn_low 1 2.0 0.003 0.003 36.464 36.473 qs_forces 11 3.9 0.002 0.002 36.409 36.409 qs_energies 11 4.9 0.001 0.001 34.700 34.706 scf_env_do_scf 11 5.9 0.000 0.001 29.137 29.139 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.777 25.778 dbcsr_multiply_generic 2286 12.5 0.102 0.104 20.380 20.486 velocity_verlet 10 3.0 0.001 0.002 18.270 18.272 qs_scf_new_mos 108 7.5 0.001 0.001 17.743 17.799 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.742 17.798 multiply_cannon 2286 13.5 0.218 0.226 16.778 17.120 ot_scf_mini 108 9.5 0.002 0.003 16.718 16.772 multiply_cannon_loop 2286 14.5 1.534 1.633 15.838 16.303 ot_mini 108 10.5 0.001 0.001 10.301 10.363 multiply_cannon_multrec 27432 15.5 2.448 3.096 9.634 9.893 qs_ot_get_derivative 108 11.5 0.001 0.001 8.443 8.496 dbcsr_mm_accdrv_process 47916 15.9 6.429 8.364 7.084 8.425 rebuild_ks_matrix 119 8.3 0.000 0.000 6.302 6.353 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.302 6.353 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.581 5.625 init_scf_run 11 5.9 0.000 0.001 4.238 4.238 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.238 4.238 qs_ot_get_p 119 10.4 0.001 0.001 3.588 3.666 sum_up_and_integrate 119 10.3 0.001 0.001 3.478 3.486 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.088 3.483 integrate_v_rspace 119 11.3 0.002 0.003 3.467 3.475 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.336 3.374 calculate_rho_elec 119 8.7 0.040 0.046 3.335 3.374 init_scf_loop 11 6.9 0.000 0.000 3.341 3.341 calculate_first_density_matrix 1 7.0 0.000 0.000 2.865 2.869 prepare_preconditioner 11 7.9 0.000 0.000 2.518 2.527 make_preconditioner 11 8.9 0.000 0.000 2.518 2.527 acc_transpose_blocks 27432 15.5 0.121 0.124 2.413 2.509 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.128 2.456 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.426 2.427 make_m2s 4572 13.5 0.054 0.056 2.221 2.338 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.298 2.325 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.264 2.278 make_images 4572 14.5 0.273 0.337 2.113 2.228 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.090 2.221 apply_single 119 13.6 0.000 0.000 2.090 2.220 mp_waitall_1 137007 16.6 1.788 2.218 1.788 2.218 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.183 2.191 calculate_dm_sparse 119 9.5 0.000 0.000 2.129 2.183 pw_transfer 1439 11.6 0.065 0.069 1.976 2.011 jit_kernel_multiply 7 15.7 0.589 1.962 0.589 1.962 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.884 1.922 density_rs2pw 119 9.7 0.003 0.004 1.803 1.915 grid_integrate_task_list 119 12.3 1.823 1.897 1.823 1.897 ot_diis_step 108 11.5 0.012 0.012 1.817 1.817 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.784 1.784 fft_wrap_pw1pw2_140 487 13.2 0.227 0.235 1.564 1.604 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.544 1.553 cp_fm_diag_elpa_base 50 14.0 1.510 1.528 1.542 1.552 acc_transpose_blocks_sync 82296 16.5 1.447 1.544 1.447 1.544 fft3d_ps 1201 14.6 0.591 0.655 1.373 1.397 grid_collocate_task_list 119 9.7 1.248 1.338 1.248 1.338 wfi_extrapolate 11 7.9 0.001 0.001 1.318 1.318 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.275 1.289 potential_pw2rs 119 12.3 0.009 0.009 1.284 1.288 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.821 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.230 1.241 mp_sum_l 7287 12.8 0.949 1.217 0.949 1.217 cp_fm_upper_to_full 72 14.2 0.822 1.167 0.822 1.167 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.069 1.070 dbcsr_complete_redistribute 329 12.2 0.119 0.142 0.754 1.029 transfer_rs2pw 487 10.6 0.004 0.005 0.884 0.996 make_images_data 4572 15.5 0.048 0.052 0.820 0.932 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.802 0.879 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.702 0.876 mp_alltoall_d11v 2130 13.8 0.710 0.840 0.710 0.840 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 0.815 0.834 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.554 0.824 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.814 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.670000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.909091, yerr=2.745394 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 627.638272E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.029 28.699 28.700 qs_mol_dyn_low 1 2.0 0.003 0.003 28.522 28.538 qs_forces 11 3.9 0.002 0.003 28.432 28.432 qs_energies 11 4.9 0.001 0.002 26.712 26.715 scf_env_do_scf 11 5.9 0.000 0.001 21.744 21.744 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.305 19.306 velocity_verlet 10 3.0 0.018 0.042 14.723 14.726 dbcsr_multiply_generic 2286 12.5 0.095 0.099 13.211 13.281 qs_scf_new_mos 108 7.5 0.001 0.001 11.629 11.655 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.628 11.654 multiply_cannon 2286 13.5 0.225 0.233 10.646 11.099 ot_scf_mini 108 9.5 0.002 0.002 10.924 10.951 multiply_cannon_loop 2286 14.5 0.639 0.658 9.720 9.896 multiply_cannon_multrec 9144 15.5 1.655 1.888 6.178 6.391 ot_mini 108 10.5 0.001 0.001 6.232 6.261 rebuild_ks_matrix 119 8.3 0.000 0.000 5.782 5.809 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.782 5.808 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.144 5.168 qs_ot_get_derivative 108 11.5 0.001 0.001 4.922 4.949 dbcsr_mm_accdrv_process 12550 15.8 3.744 4.507 4.414 4.549 init_scf_run 11 5.9 0.000 0.001 3.532 3.532 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.531 3.531 sum_up_and_integrate 119 10.3 0.001 0.001 3.425 3.429 integrate_v_rspace 119 11.3 0.003 0.003 3.415 3.419 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.373 3.380 calculate_rho_elec 119 8.7 0.060 0.061 3.373 3.379 qs_ot_get_p 119 10.4 0.001 0.001 2.794 2.829 init_scf_loop 11 6.9 0.000 0.000 2.419 2.421 calculate_first_density_matrix 1 7.0 0.000 0.000 2.378 2.379 pw_transfer 1439 11.6 0.065 0.068 2.095 2.102 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.002 2.010 make_m2s 4572 13.5 0.033 0.035 1.792 1.953 grid_integrate_task_list 119 12.3 1.868 1.923 1.868 1.923 make_images 4572 14.5 0.269 0.298 1.702 1.862 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.825 1.826 mp_waitall_1 115863 16.7 1.364 1.825 1.364 1.825 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.820 1.824 density_rs2pw 119 9.7 0.003 0.003 1.751 1.822 calculate_dm_sparse 119 9.5 0.000 0.000 1.741 1.759 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.729 1.750 prepare_preconditioner 11 7.9 0.000 0.000 1.693 1.698 make_preconditioner 11 8.9 0.000 0.000 1.693 1.698 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.666 1.676 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.661 1.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.583 1.610 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.575 1.576 acc_transpose_blocks 9144 15.5 0.043 0.044 1.435 1.461 fft3d_ps 1201 14.6 0.648 0.660 1.354 1.365 grid_collocate_task_list 119 9.7 1.296 1.343 1.296 1.343 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.328 1.341 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.289 1.298 cp_fm_diag_elpa_base 50 14.0 1.262 1.279 1.287 1.295 ot_diis_step 108 11.5 0.013 0.013 1.295 1.295 jit_kernel_multiply 6 15.4 0.631 1.266 0.631 1.266 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.234 1.262 apply_single 119 13.6 0.000 0.000 1.234 1.262 potential_pw2rs 119 12.3 0.010 0.010 1.235 1.237 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.222 1.222 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.133 1.138 wfi_extrapolate 11 7.9 0.001 0.001 1.100 1.100 hybrid_alltoall_any 4725 16.4 0.064 0.176 0.777 0.998 make_images_data 4572 15.5 0.042 0.045 0.787 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.930 mp_alltoall_d11v 2130 13.8 0.796 0.882 0.796 0.882 cp_fm_cholesky_invert 11 10.9 0.854 0.857 0.854 0.857 transfer_rs2pw 487 10.6 0.004 0.004 0.767 0.836 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.820 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.758 0.760 acc_transpose_blocks_sync 27432 16.5 0.714 0.738 0.714 0.738 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.668 0.718 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.348 0.689 mp_allgather_i34 2286 14.5 0.242 0.684 0.242 0.684 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.662 0.665 transfer_pw2rs 487 13.2 0.003 0.004 0.640 0.642 mp_alltoall_z22v 1201 16.6 0.580 0.616 0.580 0.616 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.700000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=593.272727, yerr=7.782158 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 758.747136E+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 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 44.683 44.684 qs_mol_dyn_low 1 2.0 0.003 0.003 44.469 44.476 qs_forces 11 3.9 0.002 0.003 44.408 44.409 qs_energies 11 4.9 0.001 0.001 42.414 42.417 scf_env_do_scf 11 5.9 0.001 0.001 36.508 36.508 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.541 28.542 velocity_verlet 10 3.0 0.003 0.006 24.976 24.982 dbcsr_multiply_generic 2286 12.5 0.103 0.105 20.030 20.281 qs_scf_new_mos 108 7.5 0.001 0.001 18.361 18.465 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.360 18.465 ot_scf_mini 108 9.5 0.002 0.002 17.188 17.292 multiply_cannon 2286 13.5 0.299 0.307 15.383 16.738 multiply_cannon_loop 2286 14.5 0.867 0.884 14.083 15.484 ot_mini 108 10.5 0.001 0.001 10.574 10.693 multiply_cannon_multrec 9144 15.5 3.391 4.713 8.924 9.437 qs_ot_get_derivative 108 11.5 0.001 0.001 8.456 8.560 init_scf_loop 11 6.9 0.000 0.000 7.941 7.944 rebuild_ks_matrix 119 8.3 0.000 0.000 7.432 7.577 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.431 7.577 dbcsr_mm_accdrv_process 12550 15.8 4.566 6.320 5.397 7.136 prepare_preconditioner 11 7.9 0.000 0.000 6.978 6.993 make_preconditioner 11 8.9 0.000 0.000 6.978 6.993 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.534 6.870 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.710 6.842 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.684 4.689 calculate_rho_elec 119 8.7 0.117 0.120 4.683 4.688 cp_fm_upper_to_full 72 14.2 3.148 4.535 3.148 4.535 sum_up_and_integrate 119 10.3 0.001 0.001 3.990 3.999 integrate_v_rspace 119 11.3 0.003 0.003 3.980 3.988 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.003 3.855 init_scf_run 11 5.9 0.000 0.001 3.839 3.839 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.838 3.839 qs_ot_get_p 119 10.4 0.001 0.001 3.585 3.731 mp_waitall_1 94719 16.7 2.509 3.514 2.509 3.514 pw_transfer 1439 11.6 0.068 0.069 3.449 3.460 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.351 3.362 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.698 3.064 apply_single 119 13.6 0.000 0.000 2.698 3.064 fft_wrap_pw1pw2_140 487 13.2 0.649 0.651 2.798 2.812 dbcsr_complete_redistribute 329 12.2 0.285 0.291 1.971 2.795 density_rs2pw 119 9.7 0.003 0.003 2.706 2.725 make_m2s 4572 13.5 0.037 0.038 2.504 2.684 make_images 4572 14.5 0.356 0.390 2.384 2.563 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.673 2.485 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.448 2.392 mp_sum_l 7287 12.8 1.595 2.355 1.595 2.355 calculate_dm_sparse 119 9.5 0.000 0.001 2.303 2.317 fft3d_ps 1201 14.6 0.870 0.886 2.272 2.283 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.438 2.246 mp_alltoall_i22 627 13.8 1.419 2.239 1.419 2.239 calculate_first_density_matrix 1 7.0 0.000 0.000 2.217 2.218 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.056 2.110 grid_integrate_task_list 119 12.3 2.081 2.109 2.081 2.109 ot_diis_step 108 11.5 0.014 0.015 2.090 2.090 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.972 1.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.875 1.876 acc_transpose_blocks 9144 15.5 0.045 0.046 1.839 1.866 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.779 1.780 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.630 1.674 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.615 1.615 cp_fm_cholesky_invert 11 10.9 1.604 1.608 1.604 1.608 wfi_extrapolate 11 7.9 0.001 0.001 1.558 1.558 grid_collocate_task_list 119 9.7 1.522 1.539 1.522 1.539 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.361 1.382 potential_pw2rs 119 12.3 0.013 0.014 1.377 1.378 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.125 1.334 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.331 1.332 cp_fm_diag_elpa_base 50 14.0 1.189 1.241 1.330 1.330 mp_alltoall_d11v 2130 13.8 1.292 1.319 1.292 1.319 make_images_data 4572 15.5 0.046 0.049 1.085 1.303 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.296 1.302 mp_alltoall_z22v 1201 16.6 1.265 1.299 1.265 1.299 acc_transpose_blocks_sync 27432 16.5 1.118 1.145 1.118 1.145 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.104 1.117 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.987 1.026 yz_to_x 606 15.1 0.058 0.062 0.953 0.972 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.930 0.944 qs_create_task_list 11 7.9 0.001 0.001 0.927 0.937 generate_qs_task_list 11 8.9 0.367 0.386 0.926 0.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.684000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=715.090909, yerr=14.853555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 502.513664E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 631069. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.049 83.716 83.718 qs_mol_dyn_low 1 2.0 0.003 0.003 83.300 83.371 qs_forces 11 3.9 0.005 0.020 83.215 83.216 qs_energies 11 4.9 0.003 0.012 80.340 80.359 scf_env_do_scf 11 5.9 0.000 0.001 70.937 70.939 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.510 65.511 dbcsr_multiply_generic 2055 12.4 0.105 0.108 51.943 52.324 qs_scf_new_mos 99 7.5 0.000 0.001 48.098 48.229 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.097 48.229 ot_scf_mini 99 9.5 0.002 0.003 45.694 45.812 multiply_cannon 2055 13.4 0.184 0.190 43.217 44.111 velocity_verlet 10 3.0 0.001 0.002 43.450 43.450 multiply_cannon_loop 2055 14.4 1.762 1.802 42.191 43.138 ot_mini 99 10.5 0.001 0.001 26.920 27.058 qs_ot_get_derivative 99 11.5 0.001 0.001 20.029 20.150 multiply_cannon_multrec 49320 15.4 11.429 12.049 17.829 18.430 rebuild_ks_matrix 110 8.3 0.000 0.000 14.543 14.662 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.017 14.542 14.661 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.771 12.878 mp_waitall_1 220248 16.4 11.024 11.805 11.024 11.805 multiply_cannon_sync_h2d 49320 15.4 9.647 10.273 9.647 10.273 qs_ot_get_p 110 10.4 0.001 0.003 10.157 10.266 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.597 8.136 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.628 7.955 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.296 7.811 apply_single 110 13.6 0.000 0.001 7.296 7.810 init_scf_run 11 5.9 0.000 0.001 7.245 7.245 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.245 7.245 sum_up_and_integrate 110 10.3 0.002 0.005 7.012 7.023 integrate_v_rspace 110 11.3 0.003 0.003 6.986 7.004 qs_ot_p2m_diag 48 11.0 0.016 0.044 6.894 6.924 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.633 6.791 calculate_rho_elec 110 8.6 0.020 0.024 6.632 6.791 ot_diis_step 99 11.5 0.005 0.006 6.692 6.692 dbcsr_mm_accdrv_process 87628 16.1 2.990 3.101 6.269 6.589 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.047 6.048 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.448 5.510 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.481 5.502 cp_fm_diag_elpa_base 48 14.0 5.468 5.488 5.479 5.500 init_scf_loop 11 6.9 0.001 0.004 5.399 5.401 mp_sum_l 6594 12.7 4.264 5.361 4.264 5.361 make_m2s 4110 13.4 0.060 0.065 4.006 4.136 make_images 4110 14.4 0.177 0.190 3.910 4.042 wfi_extrapolate 11 7.9 0.001 0.001 4.042 4.042 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 4.027 4.032 density_rs2pw 110 9.6 0.004 0.004 3.646 3.882 calculate_dm_sparse 110 9.5 0.001 0.001 3.769 3.862 multiply_cannon_metrocomm1 49320 15.4 0.066 0.067 2.486 3.746 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.668 3.726 grid_integrate_task_list 110 12.3 3.248 3.410 3.248 3.410 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.263 3.330 prepare_preconditioner 11 7.9 0.000 0.000 3.228 3.251 make_preconditioner 11 8.9 0.000 0.001 3.228 3.251 pw_transfer 1331 11.6 0.055 0.067 3.086 3.165 calculate_first_density_matrix 1 7.0 0.000 0.002 3.093 3.106 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.997 3.079 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.014 3.055 jit_kernel_multiply 13 15.9 3.001 3.036 3.001 3.036 fft_wrap_pw1pw2_140 451 13.1 0.309 0.334 2.563 2.656 acc_transpose_blocks 49320 15.4 0.208 0.214 2.534 2.602 potential_pw2rs 110 12.3 0.005 0.006 2.556 2.577 mp_alltoall_d11v 2046 13.8 2.148 2.561 2.148 2.561 transfer_rs2pw 451 10.6 0.005 0.006 2.200 2.424 fft3d_ps 1111 14.6 0.797 0.887 2.290 2.351 grid_collocate_task_list 110 9.6 2.160 2.276 2.160 2.276 mp_waitany 14300 13.8 1.921 2.142 1.921 2.142 mp_sum_d 3891 11.9 1.409 1.974 1.409 1.974 make_images_data 4110 15.4 0.043 0.047 1.813 1.955 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.903 1.918 hybrid_alltoall_any 4261 16.3 0.083 0.485 1.568 1.829 cp_fm_cholesky_invert 11 10.9 1.776 1.780 1.776 1.780 transfer_pw2rs 451 13.1 0.006 0.007 1.723 1.732 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.653 1.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.718000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.181818, yerr=2.948595 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 595.144704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3520197. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 69.347 69.349 qs_mol_dyn_low 1 2.0 0.003 0.003 69.109 69.119 qs_forces 11 3.9 0.003 0.003 68.992 68.993 qs_energies 11 4.9 0.004 0.022 65.680 65.683 scf_env_do_scf 11 5.9 0.000 0.001 56.605 56.609 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.064 49.064 dbcsr_multiply_generic 2055 12.4 0.114 0.117 38.617 38.805 velocity_verlet 10 3.0 0.002 0.002 35.895 35.897 qs_scf_new_mos 99 7.5 0.001 0.001 33.281 33.386 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.280 33.386 multiply_cannon 2055 13.4 0.223 0.244 32.067 33.173 ot_scf_mini 99 9.5 0.003 0.004 31.617 31.710 multiply_cannon_loop 2055 14.4 1.156 1.185 30.836 31.509 ot_mini 99 10.5 0.001 0.001 18.471 18.581 multiply_cannon_multrec 24660 15.4 6.994 8.790 14.723 16.509 rebuild_ks_matrix 110 8.3 0.000 0.000 13.486 13.575 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.486 13.574 qs_ot_get_derivative 99 11.5 0.001 0.001 12.690 12.787 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.904 11.976 mp_waitall_1 176588 16.5 7.548 9.821 7.548 9.821 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.159 7.907 dbcsr_mm_accdrv_process 52282 16.1 5.661 7.054 7.559 7.883 multiply_cannon_sync_h2d 24660 15.4 6.466 7.824 6.466 7.824 init_scf_loop 11 6.9 0.006 0.050 7.506 7.507 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.448 7.074 apply_single 110 13.6 0.000 0.001 6.447 7.074 init_scf_run 11 5.9 0.000 0.001 6.682 6.683 scf_env_initial_rho_setup 11 6.9 0.002 0.009 6.682 6.683 qs_ot_get_p 110 10.4 0.001 0.001 6.472 6.615 sum_up_and_integrate 110 10.3 0.001 0.003 6.321 6.332 integrate_v_rspace 110 11.3 0.002 0.003 6.294 6.306 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.849 5.856 calculate_rho_elec 110 8.6 0.039 0.047 5.849 5.856 ot_diis_step 99 11.5 0.010 0.010 5.734 5.734 prepare_preconditioner 11 7.9 0.000 0.000 5.460 5.482 make_preconditioner 11 8.9 0.000 0.001 5.460 5.482 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.731 5.438 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.052 5.208 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.585 4.606 make_m2s 4110 13.4 0.057 0.059 4.178 4.599 make_images 4110 14.4 0.407 0.470 4.067 4.483 cp_dbcsr_syevd 48 12.0 0.004 0.004 4.119 4.119 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.850 3.852 pw_transfer 1331 11.6 0.066 0.073 3.512 3.645 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.550 3.564 cp_fm_diag_elpa_base 48 14.0 3.503 3.522 3.547 3.560 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.405 3.543 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.512 3.538 wfi_extrapolate 11 7.9 0.001 0.001 3.509 3.509 grid_integrate_task_list 110 12.3 3.173 3.325 3.173 3.325 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.267 3.318 density_rs2pw 110 9.6 0.004 0.004 3.103 3.309 calculate_first_density_matrix 1 7.0 0.001 0.005 3.074 3.077 fft_wrap_pw1pw2_140 451 13.1 0.356 0.373 2.924 3.063 jit_kernel_multiply 12 16.3 1.530 3.024 1.530 3.024 calculate_dm_sparse 110 9.5 0.001 0.001 2.965 2.989 hybrid_alltoall_any 4261 16.3 0.105 0.454 1.957 2.747 make_images_data 4110 15.4 0.048 0.053 2.280 2.719 fft3d_ps 1111 14.6 1.114 1.341 2.498 2.645 cp_fm_cholesky_invert 11 10.9 2.510 2.516 2.510 2.516 mp_sum_l 6594 12.7 1.742 2.387 1.742 2.387 grid_collocate_task_list 110 9.6 2.167 2.303 2.167 2.303 potential_pw2rs 110 12.3 0.008 0.009 2.215 2.227 acc_transpose_blocks 24660 15.4 0.117 0.120 1.974 2.020 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.953 1.973 mp_alltoall_d11v 2046 13.8 1.699 1.908 1.699 1.908 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.872 1.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.779 1.792 transfer_rs2pw 451 10.6 0.006 0.007 1.433 1.643 mp_allgather_i34 2055 14.4 0.539 1.550 0.539 1.550 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.544 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.492 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.744 1.443 mp_waitany 10164 13.8 1.195 1.433 1.195 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.349000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=562.636364, yerr=7.314652 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 670.162944E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.029 59.881 59.883 qs_mol_dyn_low 1 2.0 0.003 0.003 59.626 59.644 qs_forces 11 3.9 0.003 0.003 59.561 59.562 qs_energies 11 4.9 0.001 0.002 56.388 56.391 scf_env_do_scf 11 5.9 0.000 0.001 48.042 48.042 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.583 39.583 velocity_verlet 10 3.0 0.001 0.002 32.297 32.298 dbcsr_multiply_generic 2055 12.4 0.109 0.111 29.142 29.398 qs_scf_new_mos 99 7.5 0.001 0.001 25.017 25.117 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.016 25.117 multiply_cannon 2055 13.4 0.213 0.223 22.971 24.178 ot_scf_mini 99 9.5 0.002 0.003 23.772 23.891 multiply_cannon_loop 2055 14.4 0.813 0.837 21.808 22.925 ot_mini 99 10.5 0.001 0.001 13.590 13.707 rebuild_ks_matrix 110 8.3 0.000 0.000 12.108 12.254 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.108 12.254 multiply_cannon_multrec 16440 15.4 3.662 4.611 10.374 11.201 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.655 10.782 mp_waitall_1 139946 16.5 6.883 9.734 6.883 9.734 qs_ot_get_derivative 99 11.5 0.001 0.001 9.179 9.296 init_scf_loop 11 6.9 0.000 0.000 8.425 8.425 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.337 7.338 dbcsr_mm_accdrv_process 34862 16.1 5.798 6.164 6.556 6.718 prepare_preconditioner 11 7.9 0.000 0.000 6.696 6.714 make_preconditioner 11 8.9 0.000 0.000 6.696 6.714 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.059 6.414 sum_up_and_integrate 110 10.3 0.001 0.002 6.243 6.258 integrate_v_rspace 110 11.3 0.003 0.003 6.217 6.232 init_scf_run 11 5.9 0.000 0.001 5.970 5.970 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.970 5.970 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.683 5.695 calculate_rho_elec 110 8.6 0.059 0.059 5.682 5.695 qs_ot_get_p 110 10.4 0.001 0.001 5.352 5.492 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.866 5.322 apply_single 110 13.6 0.000 0.000 4.866 5.322 make_m2s 4110 13.4 0.049 0.051 4.139 4.523 make_images 4110 14.4 0.399 0.525 4.023 4.406 ot_diis_step 99 11.5 0.011 0.011 4.380 4.381 multiply_cannon_sync_h2d 16440 15.4 3.233 3.851 3.233 3.851 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.171 3.781 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.734 3.737 pw_transfer 1331 11.6 0.066 0.073 3.577 3.587 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.470 3.481 grid_integrate_task_list 110 12.3 3.193 3.400 3.193 3.400 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.370 3.370 density_rs2pw 110 9.6 0.004 0.004 2.929 3.137 calculate_dm_sparse 110 9.5 0.001 0.001 3.069 3.102 fft_wrap_pw1pw2_140 451 13.1 0.460 0.468 3.009 3.022 wfi_extrapolate 11 7.9 0.001 0.001 2.952 2.952 calculate_first_density_matrix 1 7.0 0.000 0.000 2.927 2.929 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.837 2.846 cp_fm_diag_elpa_base 48 14.0 2.771 2.803 2.835 2.844 hybrid_alltoall_any 4261 16.3 0.108 0.376 2.153 2.832 make_images_data 4110 15.4 0.046 0.050 2.348 2.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.777 2.779 cp_fm_cholesky_invert 11 10.9 2.596 2.602 2.596 2.602 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.453 2.509 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.414 2.479 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.882 2.466 fft3d_ps 1111 14.6 1.119 1.132 2.387 2.403 grid_collocate_task_list 110 9.6 2.221 2.358 2.221 2.358 mp_irecv_dv 48980 15.7 0.808 2.335 0.808 2.335 potential_pw2rs 110 12.3 0.011 0.011 2.107 2.112 mp_sum_l 6594 12.7 1.406 2.057 1.406 2.057 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.008 2.022 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.946 1.947 mp_alltoall_d11v 2046 13.8 1.720 1.928 1.720 1.928 dbcsr_complete_redistribute 325 12.2 0.323 0.351 1.387 1.830 cp_fm_upper_to_full 70 14.2 1.384 1.765 1.384 1.765 jit_kernel_multiply 6 16.7 0.350 1.697 0.350 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.644 1.663 acc_transpose_blocks 16440 15.4 0.079 0.082 1.576 1.625 cp_fm_cholesky_decompose 22 10.9 1.573 1.591 1.573 1.591 mp_allgather_i34 2055 14.4 0.477 1.541 0.477 1.541 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.358 1.440 transfer_rs2pw 451 10.6 0.005 0.006 1.230 1.418 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.919 1.357 mp_waitany 17072 13.8 1.113 1.343 1.113 1.343 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.266 1.276 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.883000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.363636, yerr=9.679150 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 741.154816E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.029 65.781 65.782 qs_mol_dyn_low 1 2.0 0.003 0.003 65.565 65.575 qs_forces 11 3.9 0.003 0.003 65.503 65.503 qs_energies 11 4.9 0.001 0.002 62.118 62.122 scf_env_do_scf 11 5.9 0.001 0.001 53.548 53.551 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.935 41.936 velocity_verlet 10 3.0 0.001 0.002 37.140 37.142 dbcsr_multiply_generic 2055 12.4 0.116 0.123 30.820 31.029 qs_scf_new_mos 99 7.5 0.001 0.001 27.325 27.427 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.325 27.426 ot_scf_mini 99 9.5 0.003 0.003 25.639 25.757 multiply_cannon 2055 13.4 0.237 0.255 23.822 24.769 multiply_cannon_loop 2055 14.4 1.427 1.504 22.480 23.056 ot_mini 99 10.5 0.001 0.001 14.769 14.887 multiply_cannon_multrec 24660 15.4 4.063 6.661 12.952 14.100 rebuild_ks_matrix 110 8.3 0.000 0.000 11.917 12.033 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.917 12.033 init_scf_loop 11 6.9 0.000 0.000 11.572 11.572 qs_ot_get_derivative 99 11.5 0.001 0.001 10.582 10.698 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.528 10.631 prepare_preconditioner 11 7.9 0.000 0.000 9.853 9.873 make_preconditioner 11 8.9 0.000 0.000 9.853 9.873 dbcsr_mm_accdrv_process 52304 16.0 7.682 9.025 8.730 9.632 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.087 9.553 sum_up_and_integrate 110 10.3 0.001 0.002 6.219 6.234 integrate_v_rspace 110 11.3 0.003 0.003 6.193 6.207 mp_waitall_1 121746 16.5 4.180 6.123 4.180 6.123 qs_ot_get_p 110 10.4 0.001 0.001 5.779 5.940 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.876 5.888 calculate_rho_elec 110 8.6 0.077 0.081 5.876 5.887 init_scf_run 11 5.9 0.000 0.001 5.824 5.824 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.823 5.824 make_m2s 4110 13.4 0.059 0.061 5.380 5.638 make_images 4110 14.4 0.575 0.696 5.239 5.492 cp_fm_upper_to_full 70 14.2 3.398 4.908 3.398 4.908 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.055 4.163 apply_single 110 13.6 0.000 0.000 4.055 4.163 ot_diis_step 99 11.5 0.011 0.011 4.147 4.147 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.902 3.917 pw_transfer 1331 11.6 0.065 0.073 3.774 3.807 dbcsr_complete_redistribute 325 12.2 0.417 0.451 2.648 3.784 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.667 3.703 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.502 3.560 grid_integrate_task_list 110 12.3 3.295 3.438 3.295 3.438 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.358 3.419 calculate_dm_sparse 110 9.5 0.001 0.001 3.328 3.364 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.339 3.340 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.103 3.222 fft_wrap_pw1pw2_140 451 13.1 0.531 0.545 3.160 3.199 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.387 3.169 density_rs2pw 110 9.6 0.004 0.004 2.991 3.129 wfi_extrapolate 11 7.9 0.001 0.001 2.967 2.967 make_images_data 4110 15.4 0.048 0.052 2.654 2.956 acc_transpose_blocks 24660 15.4 0.113 0.117 2.782 2.924 hybrid_alltoall_any 4261 16.3 0.122 0.460 2.304 2.870 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.757 2.853 mp_alltoall_i22 605 13.7 1.658 2.844 1.658 2.844 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.806 2.814 cp_fm_diag_elpa_base 48 14.0 2.657 2.717 2.803 2.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.781 2.782 calculate_first_density_matrix 1 7.0 0.000 0.000 2.760 2.764 cp_fm_cholesky_invert 11 10.9 2.653 2.661 2.653 2.661 multiply_cannon_sync_h2d 24660 15.4 2.388 2.561 2.388 2.561 fft3d_ps 1111 14.6 1.154 1.198 2.493 2.511 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.424 2.463 grid_collocate_task_list 110 9.6 2.263 2.405 2.263 2.405 qs_energies_init_hamiltonians 11 5.9 0.001 0.008 2.257 2.257 potential_pw2rs 110 12.3 0.012 0.013 2.041 2.046 mp_alltoall_d11v 2046 13.8 1.730 1.973 1.730 1.973 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.883 1.901 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.809 1.839 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.616 1.717 cp_fm_cholesky_decompose 22 10.9 1.611 1.666 1.611 1.666 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.628 1.641 acc_transpose_blocks_sync 73980 16.4 1.432 1.560 1.432 1.560 mp_sum_l 6594 12.7 0.925 1.480 0.925 1.480 multiply_cannon_metrocomm4 20550 15.4 0.058 0.062 0.848 1.470 mp_irecv_dv 62702 16.1 0.746 1.393 0.746 1.393 transfer_rs2pw 451 10.6 0.005 0.005 1.180 1.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.782000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.818182, yerr=11.960679 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 850.956288E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 55.502 55.503 qs_mol_dyn_low 1 2.0 0.003 0.003 55.203 55.212 qs_forces 11 3.9 0.003 0.003 55.124 55.126 qs_energies 11 4.9 0.001 0.001 51.484 51.489 scf_env_do_scf 11 5.9 0.000 0.001 43.019 43.019 scf_env_do_scf_inner_loop 99 6.5 0.004 0.009 35.297 35.298 velocity_verlet 10 3.0 0.009 0.022 31.183 31.186 dbcsr_multiply_generic 2055 12.4 0.106 0.109 23.416 23.530 qs_scf_new_mos 99 7.5 0.001 0.001 20.671 20.737 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.670 20.736 ot_scf_mini 99 9.5 0.002 0.002 19.420 19.483 multiply_cannon 2055 13.4 0.239 0.250 17.883 19.201 multiply_cannon_loop 2055 14.4 0.600 0.621 16.641 16.910 rebuild_ks_matrix 110 8.3 0.000 0.000 11.645 11.678 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.644 11.677 ot_mini 99 10.5 0.001 0.001 10.738 10.793 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.362 10.391 multiply_cannon_multrec 8220 15.4 3.166 4.290 7.707 8.629 init_scf_loop 11 6.9 0.000 0.000 7.674 7.675 mp_waitall_1 103326 16.6 5.912 7.596 5.912 7.596 qs_ot_get_derivative 99 11.5 0.001 0.001 7.004 7.066 sum_up_and_integrate 110 10.3 0.001 0.002 6.212 6.225 integrate_v_rspace 110 11.3 0.003 0.003 6.185 6.198 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.041 6.054 calculate_rho_elec 110 8.6 0.112 0.113 6.041 6.054 prepare_preconditioner 11 7.9 0.000 0.000 6.032 6.039 make_preconditioner 11 8.9 0.000 0.000 6.032 6.039 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.614 5.689 dbcsr_mm_accdrv_process 17442 15.9 3.167 4.167 4.401 5.358 init_scf_run 11 5.9 0.000 0.001 5.341 5.341 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.340 5.341 qs_ot_get_p 110 10.4 0.001 0.001 4.723 4.804 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.147 4.480 make_m2s 4110 13.4 0.038 0.039 4.178 4.425 make_images 4110 14.4 0.648 0.706 4.048 4.291 pw_transfer 1331 11.6 0.066 0.071 4.077 4.091 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.969 3.987 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.772 3.814 apply_single 110 13.6 0.000 0.000 3.772 3.814 ot_diis_step 99 11.5 0.012 0.012 3.711 3.711 grid_integrate_task_list 110 12.3 3.383 3.491 3.383 3.491 fft_wrap_pw1pw2_140 451 13.1 0.716 0.728 3.447 3.469 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.356 3.361 density_rs2pw 110 9.6 0.004 0.004 3.020 3.171 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.017 3.018 cp_fm_cholesky_invert 11 10.9 2.887 2.891 2.887 2.891 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.738 2.739 wfi_extrapolate 11 7.9 0.001 0.001 2.692 2.692 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.672 2.673 hybrid_alltoall_any 4261 16.3 0.200 0.852 2.240 2.656 calculate_first_density_matrix 1 7.0 0.000 0.000 2.549 2.552 grid_collocate_task_list 110 9.6 2.366 2.551 2.366 2.551 make_images_data 4110 15.4 0.040 0.046 2.226 2.548 calculate_dm_sparse 110 9.5 0.001 0.001 2.492 2.533 fft3d_ps 1111 14.6 1.289 1.295 2.514 2.526 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.469 2.477 cp_fm_diag_elpa_base 48 14.0 2.415 2.438 2.468 2.476 multiply_cannon_sync_h2d 8220 15.4 2.371 2.455 2.371 2.455 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.349 2.360 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.021 2.064 potential_pw2rs 110 12.3 0.015 0.015 2.004 2.008 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.794 1.998 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.924 1.952 mp_alltoall_d11v 2046 13.8 1.731 1.919 1.731 1.919 cp_fm_cholesky_decompose 22 10.9 1.697 1.714 1.697 1.714 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.611 1.614 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.497 1.610 jit_kernel_multiply 8 15.3 0.918 1.574 0.918 1.574 dbcsr_complete_redistribute 325 12.2 0.565 0.620 1.448 1.545 acc_transpose_blocks 8220 15.4 0.041 0.042 1.412 1.474 mp_allgather_i34 2055 14.4 0.413 1.473 0.413 1.473 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.444 1.458 qs_create_task_list 11 7.9 0.000 0.001 1.209 1.307 generate_qs_task_list 11 8.9 0.374 0.446 1.209 1.307 transfer_rs2pw 451 10.6 0.005 0.005 1.104 1.267 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.194 1.233 mp_waitany 9240 13.8 1.037 1.212 1.037 1.212 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.470 1.166 mp_irecv_dv 24056 15.7 0.444 1.121 0.444 1.121 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.503000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.545455, yerr=10.611860 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 1.373000E+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 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 89.470 89.471 qs_mol_dyn_low 1 2.0 0.003 0.003 89.167 89.189 qs_forces 11 3.9 0.003 0.003 89.078 89.083 qs_energies 11 4.9 0.001 0.002 84.909 84.915 scf_env_do_scf 11 5.9 0.000 0.001 74.842 74.842 velocity_verlet 10 3.0 0.002 0.002 56.291 56.297 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 46.515 46.517 dbcsr_multiply_generic 2055 12.4 0.121 0.126 30.159 30.318 init_scf_loop 11 6.9 0.000 0.000 28.256 28.260 qs_scf_new_mos 99 7.5 0.001 0.001 27.776 27.876 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.775 27.875 prepare_preconditioner 11 7.9 0.000 0.000 26.220 26.236 make_preconditioner 11 8.9 0.000 0.000 26.220 26.236 ot_scf_mini 99 9.5 0.002 0.002 25.970 26.062 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.415 25.699 multiply_cannon 2055 13.4 0.333 0.359 22.843 23.504 multiply_cannon_loop 2055 14.4 0.826 0.838 21.009 21.423 cp_fm_upper_to_full 70 14.2 12.692 18.223 12.692 18.223 rebuild_ks_matrix 110 8.3 0.000 0.001 14.704 14.823 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.704 14.823 ot_mini 99 10.5 0.001 0.001 14.389 14.487 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.415 13.525 dbcsr_complete_redistribute 325 12.2 1.006 1.034 7.369 10.572 multiply_cannon_multrec 8220 15.4 4.047 4.207 9.732 9.835 qs_ot_get_derivative 99 11.5 0.001 0.001 9.561 9.651 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.354 9.572 mp_waitall_1 84994 16.7 7.947 8.990 7.947 8.990 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.790 8.980 mp_alltoall_i22 605 13.7 5.444 8.704 5.444 8.704 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.537 7.573 calculate_rho_elec 110 8.6 0.222 0.223 7.536 7.573 sum_up_and_integrate 110 10.3 0.002 0.002 7.204 7.219 integrate_v_rspace 110 11.3 0.003 0.004 7.176 7.191 pw_transfer 1331 11.6 0.075 0.075 6.207 6.215 qs_ot_get_p 110 10.4 0.001 0.001 6.100 6.194 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.089 6.098 make_m2s 4110 13.4 0.043 0.044 5.447 5.975 make_images 4110 14.4 0.885 0.937 5.258 5.786 init_scf_run 11 5.9 0.000 0.001 5.764 5.764 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.764 5.764 dbcsr_mm_accdrv_process 11614 15.7 3.888 4.114 5.537 5.740 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.226 5.556 fft_wrap_pw1pw2_140 451 13.1 1.368 1.373 5.397 5.407 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.863 5.387 apply_single 110 13.6 0.000 0.000 4.863 5.387 cp_fm_cholesky_invert 11 10.9 5.223 5.229 5.223 5.229 ot_diis_step 99 11.5 0.016 0.016 4.811 4.813 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.470 4.476 density_rs2pw 110 9.6 0.004 0.004 3.977 4.008 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.952 3.953 grid_integrate_task_list 110 12.3 3.708 3.805 3.708 3.805 fft3d_ps 1111 14.6 1.887 1.896 3.792 3.800 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.699 3.704 hybrid_alltoall_any 4261 16.3 0.264 0.568 2.896 3.587 make_images_data 4110 15.4 0.045 0.048 2.892 3.535 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.032 3.503 wfi_extrapolate 11 7.9 0.001 0.001 3.369 3.369 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.338 3.338 cp_fm_diag_elpa_base 48 14.0 2.806 3.004 3.335 3.336 calculate_dm_sparse 110 9.5 0.001 0.001 3.196 3.220 multiply_cannon_sync_h2d 8220 15.4 3.121 3.152 3.121 3.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.916 2.921 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.694 2.753 grid_collocate_task_list 110 9.6 2.692 2.720 2.692 2.720 potential_pw2rs 110 12.3 0.021 0.021 2.609 2.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.395 2.431 qs_env_update_s_mstruct 11 6.9 0.001 0.004 2.219 2.278 calculate_first_density_matrix 1 7.0 0.000 0.000 2.276 2.277 mp_alltoall_d11v 2046 13.8 2.140 2.227 2.140 2.227 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.208 cp_fm_cholesky_decompose 22 10.9 1.970 1.988 1.970 1.988 qs_create_task_list 11 7.9 0.000 0.001 1.880 1.923 generate_qs_task_list 11 8.9 0.731 0.785 1.879 1.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.911 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.783 1.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.471000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1250.545455, yerr=43.468617 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.771136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2990692. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.055 208.098 208.099 qs_mol_dyn_low 1 2.0 0.003 0.003 207.619 207.632 qs_forces 11 3.9 0.005 0.007 207.099 207.102 qs_energies 11 4.9 0.002 0.003 201.504 201.528 scf_env_do_scf 11 5.9 0.001 0.001 185.069 185.073 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 164.646 164.648 dbcsr_multiply_generic 2507 12.6 0.177 0.181 126.148 126.706 qs_scf_new_mos 117 7.6 0.001 0.001 125.131 125.424 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.131 125.424 velocity_verlet 10 3.0 0.001 0.002 124.029 124.035 ot_scf_mini 117 9.6 0.003 0.003 118.397 118.677 multiply_cannon 2507 13.6 0.237 0.245 101.731 103.128 multiply_cannon_loop 2507 14.6 2.378 2.445 99.538 101.083 ot_mini 117 10.6 0.001 0.001 66.932 67.247 multiply_cannon_multrec 60168 15.6 31.849 33.940 41.535 43.460 qs_ot_get_derivative 117 11.6 0.001 0.001 42.217 42.508 rebuild_ks_matrix 128 8.3 0.001 0.001 33.700 34.078 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.699 34.077 mp_waitall_1 267128 16.5 28.987 31.953 28.987 31.953 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.268 30.631 qs_ot_get_p 128 10.4 0.001 0.001 29.262 29.717 multiply_cannon_sync_h2d 60168 15.6 26.547 28.759 26.547 28.759 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.296 25.040 apply_single 128 13.6 0.001 0.001 24.296 25.040 ot_diis_step 117 11.6 0.007 0.008 24.511 24.512 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.630 22.700 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.583 20.881 init_scf_loop 11 6.9 0.000 0.000 20.346 20.347 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.991 19.992 multiply_cannon_metrocomm3 60168 15.6 0.118 0.124 16.202 18.304 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.040 17.084 cp_fm_diag_elpa_base 83 14.4 16.973 17.017 17.036 17.080 prepare_preconditioner 11 7.9 0.000 0.000 15.768 15.814 make_preconditioner 11 8.9 0.000 0.000 15.768 15.814 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.010 15.191 make_m2s 5014 13.6 0.104 0.115 14.018 14.448 make_images 5014 14.6 0.397 0.415 13.833 14.278 sum_up_and_integrate 128 10.3 0.002 0.004 14.137 14.158 integrate_v_rspace 128 11.3 0.003 0.004 14.078 14.101 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.109 13.219 calculate_rho_elec 128 8.7 0.045 0.064 13.109 13.219 init_scf_run 11 5.9 0.000 0.001 12.279 12.280 scf_env_initial_rho_setup 11 6.9 0.001 0.005 12.279 12.280 mp_sum_l 7950 12.9 9.158 10.296 9.158 10.296 dbcsr_mm_accdrv_process 124484 16.2 4.647 4.833 9.245 9.791 wfi_extrapolate 11 7.9 0.001 0.001 9.006 9.006 cp_fm_cholesky_invert 11 10.9 8.984 8.992 8.984 8.992 calculate_dm_sparse 128 9.5 0.001 0.001 8.540 8.632 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.254 8.438 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.077 8.179 multiply_cannon_metrocomm1 60168 15.6 0.095 0.099 6.205 7.929 make_images_data 5014 15.6 0.066 0.072 6.847 7.777 pw_transfer 1547 11.6 0.074 0.088 7.576 7.760 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.373 7.556 density_rs2pw 128 9.7 0.006 0.007 6.780 7.529 grid_integrate_task_list 128 12.3 7.091 7.479 7.091 7.479 hybrid_alltoall_any 5200 16.5 0.294 2.270 5.991 7.401 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.666 6.676 fft_wrap_pw1pw2_140 523 13.2 0.858 0.895 6.479 6.663 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.730 5.880 mp_alltoall_d11v 2415 14.1 4.406 5.784 4.406 5.784 fft3d_ps 1291 14.7 2.196 2.840 5.378 5.712 grid_collocate_task_list 128 9.7 4.836 5.288 4.836 5.288 cp_fm_cholesky_decompose 22 10.9 4.626 4.641 4.626 4.641 potential_pw2rs 128 12.3 0.009 0.010 4.482 4.507 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.099000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.363636, yerr=5.881755 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444712984576 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 2.166474E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 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 843.808768E+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 1453843365392 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58199. MP_Allreduce 11086 960. MP_Sync 86 MP_Alltoall 1955 4940749. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 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 0.022 0.041 187.086 187.087 qs_mol_dyn_low 1 2.0 0.003 0.004 186.582 186.638 qs_forces 11 3.9 0.004 0.005 186.464 186.469 qs_energies 11 4.9 0.005 0.032 179.725 179.739 scf_env_do_scf 11 5.9 0.001 0.001 163.451 163.461 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.889 130.892 velocity_verlet 10 3.0 0.001 0.002 117.104 117.106 dbcsr_multiply_generic 2485 12.5 0.186 0.193 95.917 97.238 qs_scf_new_mos 116 7.6 0.001 0.001 93.047 93.617 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.046 93.616 ot_scf_mini 116 9.6 0.004 0.005 88.293 88.949 multiply_cannon 2485 13.5 0.496 0.551 76.022 80.264 multiply_cannon_loop 2485 14.5 1.547 1.621 73.205 75.872 ot_mini 116 10.6 0.001 0.001 49.652 50.240 mp_waitall_1 212858 16.6 23.798 37.486 23.798 37.486 multiply_cannon_multrec 29820 15.5 20.727 24.972 31.248 35.797 init_scf_loop 11 6.9 0.002 0.016 32.474 32.475 rebuild_ks_matrix 127 8.3 0.001 0.001 31.712 32.379 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 31.711 32.379 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.484 29.087 multiply_cannon_metrocomm3 29820 15.5 0.098 0.105 15.591 28.704 qs_ot_get_derivative 116 11.6 0.001 0.002 27.863 28.510 prepare_preconditioner 11 7.9 0.000 0.000 28.194 28.273 make_preconditioner 11 8.9 0.000 0.001 28.194 28.273 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.886 27.418 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.933 23.181 apply_single 127 13.6 0.001 0.001 21.932 23.180 qs_ot_get_p 127 10.4 0.001 0.002 20.980 21.725 ot_diis_step 116 11.6 0.014 0.015 21.613 21.615 multiply_cannon_sync_h2d 29820 15.5 17.778 20.344 17.778 20.344 cp_fm_cholesky_invert 11 10.9 16.453 16.466 16.453 16.466 qs_ot_p2m_diag 82 11.4 0.186 0.215 16.132 16.167 make_m2s 4970 13.5 0.086 0.091 13.979 15.064 cp_dbcsr_syevd 82 12.4 0.006 0.006 14.939 14.940 make_images 4970 14.5 1.158 1.348 13.765 14.850 sum_up_and_integrate 127 10.3 0.002 0.004 13.811 13.843 integrate_v_rspace 127 11.3 0.003 0.003 13.752 13.787 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.181 13.218 calculate_rho_elec 127 8.7 0.086 0.102 13.180 13.217 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.791 11.819 cp_fm_diag_elpa_base 82 14.4 11.542 11.631 11.785 11.809 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 10.984 11.458 init_scf_run 11 5.9 0.000 0.001 11.423 11.424 scf_env_initial_rho_setup 11 6.9 0.001 0.003 11.422 11.424 multiply_cannon_metrocomm4 27335 15.5 0.097 0.112 3.847 10.798 dbcsr_mm_accdrv_process 61726 16.2 5.395 6.031 9.976 10.558 mp_irecv_dv 68888 16.3 3.649 10.402 3.649 10.402 make_images_data 4970 15.5 0.066 0.077 8.227 9.576 hybrid_alltoall_any 5155 16.4 0.346 1.546 7.025 9.067 pw_transfer 1535 11.6 0.084 0.099 8.463 8.517 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.242 8.299 wfi_extrapolate 11 7.9 0.001 0.001 8.265 8.265 grid_integrate_task_list 127 12.3 7.123 7.539 7.123 7.539 density_rs2pw 127 9.7 0.006 0.006 6.942 7.387 fft_wrap_pw1pw2_140 519 13.2 0.924 0.945 7.303 7.380 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.231 6.961 cp_fm_cholesky_decompose 22 10.9 6.855 6.953 6.855 6.953 calculate_dm_sparse 127 9.5 0.001 0.001 6.409 6.552 mp_sum_l 7884 12.9 4.132 6.165 4.132 6.165 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.000 6.009 fft3d_ps 1281 14.7 2.805 2.976 5.820 5.882 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.328 5.396 grid_collocate_task_list 127 9.7 4.962 5.361 4.962 5.361 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.122 5.313 mp_alltoall_d11v 2401 14.1 4.089 4.691 4.089 4.691 potential_pw2rs 127 12.3 0.015 0.018 4.393 4.411 mp_allgather_i34 2485 14.5 1.348 4.341 1.348 4.341 mp_sum_d 4454 12.1 2.617 4.041 2.617 4.041 dbcsr_complete_redistribute 393 12.7 0.768 0.854 3.132 3.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.087000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.181818, yerr=1.991718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 963.039232E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58307. MP_Allreduce 11065 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.031 173.109 173.111 qs_mol_dyn_low 1 2.0 0.003 0.003 172.720 172.733 qs_forces 11 3.9 0.004 0.004 172.475 172.477 qs_energies 11 4.9 0.001 0.002 165.958 165.966 scf_env_do_scf 11 5.9 0.001 0.001 150.551 150.553 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.847 115.848 velocity_verlet 10 3.0 0.003 0.004 110.754 110.762 dbcsr_multiply_generic 2485 12.5 0.184 0.189 80.890 81.997 qs_scf_new_mos 116 7.6 0.001 0.001 79.361 79.777 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.360 79.776 ot_scf_mini 116 9.6 0.003 0.004 75.196 75.649 multiply_cannon 2485 13.5 0.500 0.527 61.441 66.401 multiply_cannon_loop 2485 14.5 1.119 1.182 58.746 61.410 ot_mini 116 10.6 0.001 0.001 42.098 42.523 init_scf_loop 11 6.9 0.000 0.000 34.606 34.607 mp_waitall_1 169034 16.6 24.483 34.136 24.483 34.136 prepare_preconditioner 11 7.9 0.000 0.000 30.642 30.690 make_preconditioner 11 8.9 0.000 0.000 30.642 30.690 rebuild_ks_matrix 127 8.3 0.001 0.001 29.790 30.358 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.021 29.789 30.357 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.265 29.658 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.812 27.329 multiply_cannon_multrec 19880 15.5 12.924 16.401 22.214 25.700 multiply_cannon_metrocomm3 19880 15.5 0.061 0.066 15.292 24.526 qs_ot_get_derivative 116 11.6 0.001 0.002 22.535 22.998 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.711 20.754 apply_single 127 13.6 0.001 0.001 19.711 20.754 ot_diis_step 116 11.6 0.018 0.018 19.452 19.452 qs_ot_get_p 127 10.4 0.001 0.001 18.662 19.198 make_m2s 4970 13.5 0.077 0.084 14.510 15.228 multiply_cannon_sync_h2d 19880 15.5 13.472 15.024 13.472 15.024 make_images 4970 14.5 1.141 1.234 14.277 14.989 qs_ot_p2m_diag 82 11.4 0.261 0.268 14.477 14.486 cp_fm_cholesky_invert 11 10.9 14.264 14.272 14.264 14.272 sum_up_and_integrate 127 10.3 0.002 0.003 13.782 13.810 integrate_v_rspace 127 11.3 0.003 0.003 13.723 13.752 cp_dbcsr_syevd 82 12.4 0.006 0.011 13.438 13.439 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.358 13.392 calculate_rho_elec 127 8.7 0.128 0.142 13.358 13.391 init_scf_run 11 5.9 0.000 0.001 10.428 10.428 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.427 10.428 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.280 10.304 cp_fm_diag_elpa_base 82 14.4 9.869 10.017 10.272 10.296 make_images_data 4970 15.5 0.063 0.073 8.709 9.797 multiply_cannon_metrocomm4 17395 15.5 0.064 0.074 3.496 9.607 hybrid_alltoall_any 5155 16.4 0.444 2.025 7.476 9.381 mp_irecv_dv 49801 16.2 3.368 9.339 3.368 9.339 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.829 9.143 dbcsr_mm_accdrv_process 41158 16.2 5.648 5.959 8.757 8.886 pw_transfer 1535 11.6 0.085 0.103 8.669 8.770 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.447 8.554 grid_integrate_task_list 127 12.3 7.253 7.632 7.253 7.632 fft_wrap_pw1pw2_140 519 13.2 1.045 1.083 7.512 7.625 wfi_extrapolate 11 7.9 0.001 0.001 7.381 7.381 cp_fm_cholesky_decompose 22 10.9 7.220 7.243 7.220 7.243 cp_fm_upper_to_full 104 14.8 5.727 7.187 5.727 7.187 density_rs2pw 127 9.7 0.005 0.006 6.808 7.098 dbcsr_complete_redistribute 393 12.7 1.162 1.193 4.568 6.292 calculate_dm_sparse 127 9.5 0.001 0.001 5.807 5.882 fft3d_ps 1281 14.7 2.791 3.024 5.792 5.870 grid_collocate_task_list 127 9.7 5.120 5.460 5.120 5.460 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.383 5.389 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.533 5.189 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.413 5.131 mp_alltoall_d11v 2401 14.1 4.249 4.829 4.249 4.829 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.500 4.668 mp_sum_l 7884 12.9 3.119 4.601 3.119 4.601 potential_pw2rs 127 12.3 0.020 0.021 4.314 4.325 mp_allgather_i34 2485 14.5 1.203 4.234 1.203 4.234 transfer_fm_to_dbcsr 11 9.9 0.019 0.022 2.357 4.061 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.928 3.966 mp_alltoall_i22 712 14.1 1.889 3.780 1.889 3.780 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.714 3.714 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.469 3.519 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.111000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=912.000000, yerr=14.887457 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.144689E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58275. MP_Allreduce 11070 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.041 186.210 186.211 qs_mol_dyn_low 1 2.0 0.003 0.003 185.758 185.770 qs_forces 11 3.9 0.004 0.005 185.631 185.641 qs_energies 11 4.9 0.011 0.046 178.505 178.530 scf_env_do_scf 11 5.9 0.001 0.001 161.559 161.572 velocity_verlet 10 3.0 0.002 0.002 123.123 123.125 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.411 115.412 dbcsr_multiply_generic 2485 12.5 0.192 0.198 79.814 80.561 qs_scf_new_mos 116 7.6 0.001 0.001 79.428 79.736 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.427 79.735 ot_scf_mini 116 9.6 0.003 0.004 74.916 75.236 multiply_cannon 2485 13.5 0.546 0.579 55.362 57.939 multiply_cannon_loop 2485 14.5 1.825 1.888 51.734 53.491 init_scf_loop 11 6.9 0.001 0.004 46.026 46.027 ot_mini 116 10.6 0.001 0.001 42.546 42.861 prepare_preconditioner 11 7.9 0.000 0.000 41.940 41.964 make_preconditioner 11 8.9 0.000 0.001 41.940 41.964 make_full_inverse_cholesky 11 9.9 0.013 0.022 35.527 40.618 multiply_cannon_multrec 29820 15.5 13.657 18.644 26.427 31.023 rebuild_ks_matrix 127 8.3 0.001 0.001 29.072 29.317 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.029 29.071 29.316 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.185 26.390 mp_waitall_1 146592 16.7 16.862 25.333 16.862 25.333 qs_ot_get_derivative 116 11.6 0.001 0.002 23.146 23.458 make_m2s 4970 13.5 0.091 0.096 20.082 21.095 make_images 4970 14.5 1.948 2.298 19.776 20.790 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.863 19.406 apply_single 127 13.6 0.001 0.001 18.863 19.406 ot_diis_step 116 11.6 0.017 0.018 19.269 19.271 qs_ot_get_p 127 10.4 0.001 0.002 18.653 19.013 cp_fm_upper_to_full 104 14.8 11.580 17.071 11.580 17.071 cp_fm_cholesky_invert 11 10.9 15.827 15.835 15.827 15.835 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.523 14.573 multiply_cannon_metrocomm3 29820 15.5 0.048 0.051 6.172 14.526 sum_up_and_integrate 127 10.3 0.002 0.004 13.912 13.940 integrate_v_rspace 127 11.3 0.003 0.003 13.852 13.881 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.700 13.746 calculate_rho_elec 127 8.7 0.169 0.184 13.700 13.745 cp_dbcsr_syevd 82 12.4 0.006 0.007 13.144 13.146 dbcsr_complete_redistribute 393 12.7 1.515 1.623 9.174 12.947 dbcsr_mm_accdrv_process 61748 16.2 8.395 9.027 12.346 12.871 make_images_data 4970 15.5 0.064 0.071 10.764 12.469 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.835 11.604 hybrid_alltoall_any 5155 16.4 0.522 2.181 9.559 11.043 init_scf_run 11 5.9 0.000 0.001 10.939 10.940 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.938 10.940 multiply_cannon_sync_h2d 29820 15.5 10.335 10.915 10.335 10.915 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.389 10.099 cp_fm_diag_elpa 82 13.4 0.001 0.001 10.004 10.017 cp_fm_diag_elpa_base 82 14.4 9.042 9.362 9.997 10.008 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.641 9.863 mp_alltoall_i22 712 14.1 5.698 9.609 5.698 9.609 pw_transfer 1535 11.6 0.085 0.102 9.143 9.224 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.920 9.005 fft_wrap_pw1pw2_140 519 13.2 1.239 1.283 7.953 8.054 grid_integrate_task_list 127 12.3 7.491 7.838 7.491 7.838 wfi_extrapolate 11 7.9 0.009 0.049 7.784 7.784 cp_fm_cholesky_decompose 22 10.9 7.447 7.557 7.447 7.557 density_rs2pw 127 9.7 0.005 0.006 6.775 7.154 multiply_cannon_metrocomm4 24850 15.5 0.078 0.086 2.751 6.432 calculate_dm_sparse 127 9.5 0.001 0.001 6.203 6.294 mp_irecv_dv 75445 16.2 2.600 6.166 2.600 6.166 fft3d_ps 1281 14.7 2.985 3.076 5.992 6.046 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.620 5.713 grid_collocate_task_list 127 9.7 5.245 5.661 5.245 5.661 mp_alltoall_d11v 2401 14.1 4.844 5.066 4.844 5.066 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.482 4.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.466 4.555 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.443 4.449 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.188 4.256 potential_pw2rs 127 12.3 0.022 0.023 4.231 4.242 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.211000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.727273, yerr=17.560756 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/19/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 5.820059E+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 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.573974E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828189. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.037 167.417 167.418 qs_mol_dyn_low 1 2.0 0.003 0.003 167.013 167.025 qs_forces 11 3.9 0.004 0.004 166.704 166.710 qs_energies 11 4.9 0.002 0.002 159.304 159.313 scf_env_do_scf 11 5.9 0.001 0.001 142.095 142.104 velocity_verlet 10 3.0 0.012 0.014 109.731 109.746 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 106.665 106.667 dbcsr_multiply_generic 2485 12.5 0.183 0.190 70.919 71.197 qs_scf_new_mos 116 7.6 0.001 0.001 69.952 70.060 qs_scf_loop_do_ot 116 8.6 0.001 0.001 69.951 70.059 ot_scf_mini 116 9.6 0.003 0.004 65.597 65.678 multiply_cannon 2485 13.5 0.557 0.582 52.687 55.970 multiply_cannon_loop 2485 14.5 0.795 0.819 49.811 50.383 ot_mini 116 10.6 0.001 0.001 36.860 36.939 init_scf_loop 11 6.9 0.000 0.000 35.283 35.285 prepare_preconditioner 11 7.9 0.000 0.000 31.429 31.454 make_preconditioner 11 8.9 0.000 0.000 31.429 31.454 mp_waitall_1 124680 16.7 24.274 30.642 24.274 30.642 make_full_inverse_cholesky 11 9.9 0.014 0.025 29.356 29.644 rebuild_ks_matrix 127 8.3 0.001 0.001 28.926 28.995 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.925 28.994 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.357 26.418 multiply_cannon_multrec 9940 15.5 10.322 14.460 17.820 20.604 qs_ot_get_derivative 116 11.6 0.002 0.002 20.078 20.157 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.437 19.319 cp_fm_cholesky_invert 11 10.9 17.954 17.960 17.954 17.960 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.819 16.933 apply_single 127 13.6 0.001 0.001 16.819 16.932 ot_diis_step 116 11.6 0.020 0.020 16.717 16.718 qs_ot_get_p 127 10.4 0.001 0.001 16.192 16.263 make_m2s 4970 13.5 0.064 0.069 14.487 15.291 make_images 4970 14.5 2.146 2.555 14.181 14.978 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.144 14.169 calculate_rho_elec 127 8.7 0.251 0.262 14.144 14.168 sum_up_and_integrate 127 10.3 0.002 0.002 14.046 14.091 integrate_v_rspace 127 11.3 0.004 0.004 13.986 14.033 qs_ot_p2m_diag 82 11.4 0.489 0.496 12.507 12.523 cp_dbcsr_syevd 82 12.4 0.006 0.008 11.367 11.368 multiply_cannon_sync_h2d 9940 15.5 10.826 11.076 10.826 11.076 init_scf_run 11 5.9 0.000 0.001 10.260 10.261 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.260 10.260 pw_transfer 1535 11.6 0.084 0.092 9.793 9.836 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.572 9.621 hybrid_alltoall_any 5155 16.4 0.835 3.746 8.237 9.585 make_images_data 4970 15.5 0.054 0.065 8.349 9.581 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.574 8.585 cp_fm_diag_elpa_base 82 14.4 8.338 8.419 8.569 8.579 fft_wrap_pw1pw2_140 519 13.2 1.664 1.711 8.494 8.535 grid_integrate_task_list 127 12.3 7.719 8.029 7.719 8.029 cp_fm_cholesky_decompose 22 10.9 7.847 7.967 7.847 7.967 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.835 7.897 dbcsr_mm_accdrv_process 20590 16.1 3.218 4.470 7.146 7.828 wfi_extrapolate 11 7.9 0.001 0.001 7.265 7.265 density_rs2pw 127 9.7 0.005 0.005 6.920 7.110 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.275 7.090 fft3d_ps 1281 14.7 3.119 3.220 6.050 6.092 calculate_dm_sparse 127 9.5 0.001 0.001 5.995 6.061 grid_collocate_task_list 127 9.7 5.531 5.760 5.531 5.760 dbcsr_complete_redistribute 393 12.7 2.094 2.138 5.129 5.508 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.367 5.372 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.088 5.102 mp_alltoall_d11v 2401 14.1 4.475 4.992 4.475 4.992 mp_allgather_i34 2485 14.5 1.078 4.509 1.078 4.509 potential_pw2rs 127 12.3 0.026 0.026 4.298 4.304 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.030 4.084 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.597 3.897 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.314 3.620 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.549 3.607 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.518 3.550 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.706 3.539 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.462 3.472 mp_irecv_dv 28618 15.9 1.667 3.454 1.667 3.454 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.377 3.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=167.418000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1491.818182, yerr=13.469279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 3.213971E+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 3955 59814. MP_Allreduce 10965 1506. MP_Sync 86 MP_Alltoall 1700 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.052 289.311 289.312 qs_mol_dyn_low 1 2.0 0.003 0.003 288.593 288.625 qs_forces 11 3.9 0.006 0.007 288.409 288.410 qs_energies 11 4.9 0.002 0.002 279.494 279.499 scf_env_do_scf 11 5.9 0.001 0.001 257.030 257.035 velocity_verlet 10 3.0 0.008 0.009 208.776 208.795 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 132.797 132.800 init_scf_loop 11 6.9 0.000 0.000 123.961 123.964 prepare_preconditioner 11 7.9 0.000 0.000 119.149 119.166 make_preconditioner 11 8.9 0.000 0.000 119.149 119.166 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.182 116.299 qs_scf_new_mos 116 7.6 0.001 0.001 88.521 88.589 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.520 88.588 ot_scf_mini 116 9.6 0.004 0.004 83.735 83.759 dbcsr_multiply_generic 2485 12.5 0.215 0.228 81.305 81.833 cp_fm_upper_to_full 104 14.8 53.044 75.935 53.044 75.935 multiply_cannon 2485 13.5 0.678 0.720 58.349 59.514 multiply_cannon_loop 2485 14.5 1.035 1.055 54.340 55.772 ot_mini 116 10.6 0.001 0.001 44.083 44.126 dbcsr_complete_redistribute 393 12.7 3.976 4.027 29.841 42.806 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.368 39.328 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.925 36.753 mp_alltoall_i22 712 14.1 21.666 34.762 21.666 34.762 rebuild_ks_matrix 127 8.3 0.001 0.001 34.327 34.381 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 34.326 34.381 cp_fm_cholesky_invert 11 10.9 33.077 33.083 33.077 33.083 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.809 31.871 mp_waitall_1 102768 16.8 27.607 31.812 27.607 31.812 qs_ot_get_p 127 10.4 0.001 0.001 24.941 25.023 qs_ot_get_derivative 116 11.6 0.002 0.002 24.292 24.320 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.932 20.966 make_m2s 4970 13.5 0.073 0.077 18.452 19.939 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 18.451 19.838 ot_diis_step 116 11.6 0.022 0.022 19.756 19.757 make_images 4970 14.5 3.053 3.232 17.975 19.461 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.117 19.256 apply_single 127 13.6 0.001 0.001 19.117 19.255 cp_dbcsr_syevd 82 12.4 0.007 0.007 19.164 19.165 multiply_cannon_multrec 9940 15.5 10.176 11.960 18.073 18.160 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.139 17.160 calculate_rho_elec 127 8.7 0.475 0.476 17.138 17.159 sum_up_and_integrate 127 10.3 0.002 0.002 16.241 16.327 integrate_v_rspace 127 11.3 0.004 0.004 16.179 16.265 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.039 16.040 cp_fm_diag_elpa_base 82 14.4 11.733 13.271 16.036 16.036 multiply_cannon_sync_h2d 9940 15.5 14.199 14.257 14.199 14.257 pw_transfer 1535 11.6 0.095 0.096 13.038 13.054 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 12.800 12.816 hybrid_alltoall_any 5155 16.4 1.304 3.047 10.650 12.758 make_images_data 4970 15.5 0.063 0.067 10.461 12.627 init_scf_run 11 5.9 0.000 0.001 12.161 12.162 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.160 12.161 fft_wrap_pw1pw2_140 519 13.2 3.199 3.238 11.400 11.427 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.532 9.554 dbcsr_mm_accdrv_process 20590 16.0 4.233 5.972 7.649 9.529 wfi_extrapolate 11 7.9 0.001 0.001 8.841 8.841 cp_fm_cholesky_decompose 22 10.9 8.789 8.803 8.789 8.803 grid_integrate_task_list 127 12.3 8.481 8.685 8.481 8.685 density_rs2pw 127 9.7 0.005 0.005 8.264 8.318 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.182 8.183 fft3d_ps 1281 14.7 3.958 3.974 7.291 7.349 mp_alltoall_d11v 2401 14.1 6.733 6.881 6.733 6.881 calculate_dm_sparse 127 9.5 0.001 0.001 6.777 6.820 grid_collocate_task_list 127 9.7 6.362 6.435 6.362 6.435 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.317 6.364 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.104 6.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.312000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2907.090909, yerr=158.905034 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 1.261113E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 17891635. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.079 85.083 85.085 qs_energies 1 2.0 0.001 0.009 84.530 84.566 ls_scf 1 3.0 0.003 0.020 83.627 83.640 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.563 72.714 multiply_cannon 111 7.7 0.017 0.020 55.962 57.259 multiply_cannon_loop 111 8.7 0.228 0.245 52.616 54.032 ls_scf_main 1 4.0 0.000 0.003 52.245 52.246 density_matrix_trs4 2 5.0 0.002 0.004 46.657 46.735 ls_scf_init_scf 1 4.0 0.001 0.005 28.288 28.290 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.167 27.219 mp_waitall_1 11031 10.9 22.375 26.026 22.375 26.026 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.072 25.087 multiply_cannon_multrec 2664 9.7 8.170 8.908 15.558 17.331 multiply_cannon_sync_h2d 2664 9.7 13.528 15.836 13.528 15.836 make_m2s 222 7.7 0.009 0.011 12.923 13.319 make_images 222 8.7 0.098 0.109 12.901 13.296 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.728 12.456 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.462 8.024 make_images_data 222 9.7 0.004 0.005 7.518 8.021 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.695 7.008 7.982 hybrid_alltoall_any 227 10.6 0.215 1.848 6.508 7.553 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.088 6.208 7.088 calculate_norms 4752 9.8 5.502 6.154 5.502 6.154 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.216 mp_sum_l 887 5.1 3.152 4.865 3.152 4.865 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.775 mp_irecv_dv 6231 10.9 2.039 3.760 2.039 3.760 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.310 3.655 make_images_sizes 222 9.7 0.000 0.000 0.654 3.533 mp_alltoall_i44 222 10.7 0.654 3.532 0.654 3.532 arnoldi_extremal 4 6.8 0.000 0.000 3.228 3.250 arnoldi_normal_ev 4 7.8 0.001 0.003 3.228 3.250 build_subspace 16 8.4 0.009 0.012 3.138 3.140 ls_scf_post 1 4.0 0.001 0.007 3.092 3.106 ls_scf_store_result 1 5.0 0.000 0.000 2.892 2.937 dbcsr_special_finalize 555 9.7 0.005 0.006 2.399 2.735 dbcsr_merge_single_wm 555 10.7 0.461 0.593 2.390 2.726 make_images_pack 222 9.7 2.206 2.626 2.208 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.322 2.586 dbcsr_matrix_vector_mult_local 304 10.0 2.074 2.490 2.076 2.492 dbcsr_sort_data 658 11.4 2.187 2.488 2.187 2.488 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.339 2.416 buffer_matrices_ensure_size 222 8.7 1.749 2.101 1.749 2.101 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.795 1.799 rebuild_ks_matrix 3 7.3 0.000 0.000 1.786 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.786 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.085000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 2.139279E+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 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.085 90.777 90.779 qs_energies 1 2.0 0.000 0.004 90.155 90.174 ls_scf 1 3.0 0.009 0.036 88.604 88.619 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.464 74.789 multiply_cannon 111 7.7 0.026 0.038 52.486 57.513 ls_scf_main 1 4.0 0.013 0.079 54.670 54.682 multiply_cannon_loop 111 8.7 0.135 0.147 49.677 53.510 density_matrix_trs4 2 5.0 0.002 0.016 48.905 49.184 ls_scf_init_scf 1 4.0 0.000 0.003 30.275 30.287 mp_waitall_1 9105 10.9 20.546 29.141 20.546 29.141 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.888 28.976 multiply_cannon_multrec 1332 9.7 13.340 16.543 22.630 26.906 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.349 26.360 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.231 20.744 make_m2s 222 7.7 0.006 0.007 15.189 15.744 make_images 222 8.7 1.372 1.692 15.159 15.715 dbcsr_mm_accdrv_process 4041 10.4 0.370 0.595 8.888 10.456 dbcsr_mm_accdrv_process_sort 4041 11.4 8.378 9.861 8.378 9.861 hybrid_alltoall_any 227 10.6 0.542 2.559 8.147 9.834 make_images_data 222 9.7 0.004 0.005 8.855 9.748 mp_sum_l 887 5.1 5.348 8.466 5.348 8.466 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.217 7.878 mp_irecv_dv 3311 11.0 3.197 7.825 3.197 7.825 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.991 6.695 calculate_norms 2376 9.8 6.070 6.694 6.070 6.694 multiply_cannon_sync_h2d 1332 9.7 4.895 6.414 4.895 6.414 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.078 5.300 arnoldi_extremal 4 6.8 0.000 0.000 4.648 4.663 arnoldi_normal_ev 4 7.8 0.004 0.016 4.648 4.663 build_subspace 16 8.4 0.014 0.021 4.391 4.394 ls_scf_post 1 4.0 0.001 0.008 3.650 3.656 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.473 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.132 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.229 2.747 3.230 mp_allgather_i34 111 8.7 0.799 3.144 0.799 3.144 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.605 2.683 dbcsr_data_new 4174 10.1 2.119 2.402 2.119 2.402 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.108 2.362 make_images_pack 222 9.7 1.836 2.138 1.838 2.140 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.006 2.017 rebuild_ks_matrix 3 7.3 0.000 0.000 1.993 2.005 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.028 1.993 2.005 dbcsr_sort_data 436 11.2 1.808 1.987 1.808 1.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.779000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1790.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 2.916221E+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 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.070 93.307 93.309 qs_energies 1 2.0 0.000 0.002 92.581 92.614 ls_scf 1 3.0 0.000 0.002 91.132 91.167 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.875 76.145 ls_scf_main 1 4.0 0.000 0.002 57.059 57.064 multiply_cannon 111 7.7 0.035 0.138 52.333 56.334 multiply_cannon_loop 111 8.7 0.116 0.131 49.562 53.667 density_matrix_trs4 2 5.0 0.002 0.004 51.141 51.290 mp_waitall_1 7281 11.0 23.936 33.264 23.936 33.264 ls_scf_init_scf 1 4.0 0.000 0.002 30.378 30.381 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.137 29.220 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.838 26.851 multiply_cannon_multrec 888 9.7 12.596 15.035 21.134 24.031 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.652 23.040 make_m2s 222 7.7 0.006 0.007 16.452 17.209 make_images 222 8.7 1.586 1.852 16.414 17.168 make_images_data 222 9.7 0.004 0.004 9.585 10.667 hybrid_alltoall_any 227 10.6 0.641 2.912 9.096 10.351 dbcsr_mm_accdrv_process 3754 10.4 0.314 0.485 8.064 9.306 mp_sum_l 887 5.1 5.393 9.153 5.393 9.153 dbcsr_mm_accdrv_process_sort 3754 11.4 7.633 8.821 7.633 8.821 multiply_cannon_sync_h2d 888 9.7 6.128 7.271 6.128 7.271 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.205 7.193 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.493 6.967 mp_irecv_dv 2335 11.1 2.478 6.920 2.478 6.920 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.346 6.563 arnoldi_extremal 4 6.8 0.000 0.000 5.050 5.067 arnoldi_normal_ev 4 7.8 0.001 0.005 5.050 5.067 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.794 4.972 build_subspace 16 8.4 0.014 0.020 4.748 4.753 calculate_norms 1584 9.8 4.347 4.732 4.347 4.732 mp_allgather_i34 111 8.7 0.889 3.788 0.889 3.788 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.429 3.761 ls_scf_post 1 4.0 0.000 0.002 3.695 3.728 dbcsr_matrix_vector_mult_local 304 10.0 3.023 3.597 3.025 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.420 3.516 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.835 2.931 dbcsr_data_new 4116 9.9 2.099 2.449 2.099 2.449 dbcsr_sort_data 325 11.1 1.886 2.158 1.886 2.158 make_images_sizes 222 9.7 0.000 0.000 1.030 2.099 mp_alltoall_i44 222 10.7 1.029 2.098 1.029 2.098 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.926 1.929 dbcsr_finalize 304 7.8 0.026 0.032 1.612 1.889 rebuild_ks_matrix 3 7.3 0.000 0.000 1.885 1.887 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.016 1.885 1.887 make_images_pack 222 9.7 1.622 1.880 1.625 1.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.309000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2283.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 3.311550E+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 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.043 97.235 97.236 qs_energies 1 2.0 0.000 0.000 96.572 96.580 ls_scf 1 3.0 0.000 0.000 94.815 94.822 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.645 78.870 ls_scf_main 1 4.0 0.000 0.000 58.844 58.846 multiply_cannon 111 7.7 0.041 0.084 51.895 56.249 density_matrix_trs4 2 5.0 0.002 0.003 52.727 52.842 multiply_cannon_loop 111 8.7 0.153 0.166 46.870 50.020 ls_scf_init_scf 1 4.0 0.000 0.001 32.682 32.685 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.475 31.545 mp_waitall_1 6369 11.0 22.886 29.762 22.886 29.762 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.926 28.936 multiply_cannon_multrec 1332 9.7 14.145 16.981 22.087 25.562 make_m2s 222 7.7 0.007 0.007 21.199 22.604 make_images 222 8.7 3.142 3.602 21.149 22.556 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.389 16.814 make_images_data 222 9.7 0.004 0.004 11.832 13.462 hybrid_alltoall_any 227 10.6 0.796 3.780 11.215 12.858 dbcsr_mm_accdrv_process 3641 10.4 0.303 0.481 7.555 9.050 dbcsr_mm_accdrv_process_sort 3641 11.4 7.084 8.532 7.084 8.532 mp_sum_l 887 5.1 3.958 6.919 3.958 6.919 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.085 6.150 mp_irecv_dv 3229 10.9 2.059 6.068 2.059 6.068 multiply_cannon_sync_h2d 1332 9.7 5.513 6.047 5.513 6.047 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.019 5.328 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.497 5.312 arnoldi_extremal 4 6.8 0.000 0.000 5.214 5.226 arnoldi_normal_ev 4 7.8 0.008 0.020 5.214 5.226 build_subspace 16 8.4 0.014 0.021 4.866 4.877 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.645 4.826 calculate_norms 2376 9.8 4.178 4.504 4.178 4.504 mp_allgather_i34 111 8.7 2.149 4.463 2.149 4.463 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.564 3.866 dbcsr_matrix_vector_mult_local 304 10.0 3.192 3.689 3.194 3.691 dbcsr_sort_data 658 11.4 3.079 3.511 3.079 3.511 ls_scf_post 1 4.0 0.000 0.000 3.289 3.297 dbcsr_special_finalize 555 9.7 0.006 0.007 2.819 3.220 dbcsr_merge_single_wm 555 10.7 0.537 0.665 2.810 3.212 ls_scf_store_result 1 5.0 0.000 0.000 3.025 3.096 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.008 3.041 dbcsr_data_release 10477 10.7 1.609 2.439 1.609 2.439 dbcsr_finalize 304 7.8 0.049 0.061 1.810 2.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.236000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 4.773364E+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 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.053 93.437 93.438 qs_energies 1 2.0 0.000 0.000 92.705 92.710 ls_scf 1 3.0 0.000 0.000 90.744 90.748 dbcsr_multiply_generic 111 6.7 0.017 0.019 72.367 72.571 ls_scf_main 1 4.0 0.000 0.000 57.029 57.030 multiply_cannon 111 7.7 0.086 0.118 53.799 56.713 multiply_cannon_loop 111 8.7 0.087 0.094 51.202 53.008 density_matrix_trs4 2 5.0 0.002 0.003 50.099 50.169 ls_scf_init_scf 1 4.0 0.000 0.001 30.294 30.295 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.044 29.076 mp_waitall_1 5436 11.0 24.965 28.922 24.965 28.922 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.944 26.960 multiply_cannon_multrec 444 9.7 13.690 16.416 20.749 22.023 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.087 15.750 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.208 14.861 make_m2s 222 7.7 0.005 0.005 13.717 14.682 make_images 222 8.7 2.042 2.482 13.650 14.613 hybrid_alltoall_any 227 10.6 0.802 3.822 8.257 9.977 make_images_data 222 9.7 0.003 0.004 8.487 9.807 multiply_cannon_sync_h2d 444 9.7 6.731 8.024 6.731 8.024 dbcsr_mm_accdrv_process 3003 10.4 0.362 0.404 6.753 7.880 dbcsr_mm_accdrv_process_sort 3003 11.4 6.391 7.481 6.391 7.481 arnoldi_extremal 4 6.8 0.000 0.000 5.830 5.847 arnoldi_normal_ev 4 7.8 0.011 0.024 5.830 5.847 build_subspace 16 8.4 0.015 0.020 5.406 5.417 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.412 4.666 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.170 4.373 mp_sum_l 887 5.1 2.848 4.346 2.848 4.346 dbcsr_matrix_vector_mult_local 304 10.0 3.702 4.177 3.705 4.179 mp_allgather_i34 111 8.7 1.170 3.790 1.170 3.790 calculate_norms 792 9.8 3.609 3.704 3.609 3.704 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.528 3.702 mp_irecv_dv 1241 11.2 1.515 3.663 1.515 3.663 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.514 3.554 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.018 3.449 ls_scf_post 1 4.0 0.000 0.000 3.420 3.426 make_images_sizes 222 9.7 0.000 0.000 0.843 3.353 mp_alltoall_i44 222 10.7 0.843 3.352 0.843 3.352 ls_scf_store_result 1 5.0 0.000 0.000 3.210 3.246 dbcsr_finalize 304 7.8 0.062 0.076 2.194 2.298 dbcsr_data_new 4608 9.7 1.797 2.221 1.797 2.221 dbcsr_merge_all 275 8.9 0.478 0.526 2.053 2.137 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.025 2.025 rebuild_ks_matrix 3 7.3 0.000 0.000 1.992 1.993 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.992 1.993 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.946 1.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.438000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_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 8.823255E+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 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.089 105.125 105.127 qs_energies 1 2.0 0.000 0.000 103.689 103.696 ls_scf 1 3.0 0.000 0.000 100.681 100.688 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.285 74.439 ls_scf_main 1 4.0 0.000 0.000 63.308 63.309 density_matrix_trs4 2 5.0 0.002 0.003 54.243 54.303 multiply_cannon 111 7.7 0.154 0.280 48.294 50.291 multiply_cannon_loop 111 8.7 0.109 0.115 45.255 45.877 ls_scf_init_scf 1 4.0 0.000 0.000 33.595 33.596 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.038 32.062 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.336 29.345 mp_waitall_1 4527 11.1 21.456 25.090 21.456 25.090 make_m2s 222 7.7 0.005 0.005 22.378 23.501 make_images 222 8.7 3.571 3.853 22.270 23.392 multiply_cannon_multrec 444 9.7 17.803 18.412 22.424 22.979 hybrid_alltoall_any 227 10.6 1.653 3.629 12.549 15.321 make_images_data 222 9.7 0.003 0.004 12.752 14.878 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.112 10.506 multiply_cannon_sync_h2d 444 9.7 8.772 8.825 8.772 8.825 arnoldi_extremal 4 6.8 0.000 0.000 7.397 7.408 arnoldi_normal_ev 4 7.8 0.003 0.009 7.397 7.408 build_subspace 16 8.4 0.026 0.036 6.846 6.858 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.468 5.628 dbcsr_matrix_vector_mult_local 304 10.0 5.044 5.363 5.047 5.365 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.179 5.268 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.895 5.140 dbcsr_mm_accdrv_process 1814 10.4 0.244 0.351 4.444 4.570 dbcsr_mm_accdrv_process_sort 1814 11.4 4.109 4.240 4.109 4.240 ls_scf_post 1 4.0 0.000 0.000 3.778 3.785 make_images_sizes 222 9.7 0.000 0.000 1.482 3.608 mp_alltoall_i44 222 10.7 1.481 3.608 1.481 3.608 mp_allgather_i34 111 8.7 1.138 3.546 1.138 3.546 ls_scf_store_result 1 5.0 0.000 0.000 3.501 3.535 calculate_norms 792 9.8 3.238 3.277 3.238 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.077 3.160 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.977 2.977 dbcsr_merge_all 275 8.9 0.885 0.914 2.863 2.939 dbcsr_complete_redistribute 5 7.6 1.438 1.475 2.749 2.864 matrix_ls_to_qs 2 6.0 0.000 0.000 2.405 2.525 dbcsr_sort_data 325 11.1 2.439 2.498 2.439 2.498 dbcsr_data_new 6591 9.6 1.874 2.482 1.874 2.482 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.457 2.458 rebuild_ks_matrix 3 7.3 0.000 0.000 2.391 2.393 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.391 2.392 dbcsr_new_transposed 4 7.5 0.241 0.251 2.284 2.297 dbcsr_frobenius_norm 74 6.6 2.056 2.138 2.204 2.239 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.201 dbcsr_add_anytype 103 7.2 0.858 0.891 2.124 2.201 dbcsr_data_release 12724 10.6 1.972 2.175 1.972 2.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.127000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6975.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48d3695c3ced91faad61a1fd2c67ef2ebe356476_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 589.905920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 237.656 237.657 qs_mol_dyn_low 1 2.0 0.005 0.042 236.540 236.631 qs_forces 5 3.8 0.005 0.011 236.394 236.396 qs_energies 5 4.8 0.004 0.042 233.339 233.363 scf_env_do_scf 5 5.8 0.001 0.010 218.767 218.769 scf_env_do_scf_inner_loop 105 6.6 0.005 0.059 191.332 191.345 qs_scf_new_mos 105 7.6 0.000 0.001 149.999 150.151 qs_scf_loop_do_ot 105 8.6 0.001 0.001 149.999 150.150 ot_scf_mini 105 9.6 0.003 0.007 139.760 139.958 dbcsr_multiply_generic 1445 12.2 0.129 0.139 136.234 136.840 multiply_cannon 1445 13.2 0.275 0.285 115.704 118.256 multiply_cannon_loop 1445 14.2 2.799 2.975 113.836 115.511 velocity_verlet 4 3.0 0.003 0.015 110.434 110.435 ot_mini 105 10.6 0.002 0.028 62.365 62.501 qs_ot_get_p 112 10.4 0.001 0.002 46.906 47.261 mp_waitall_1 488190 16.1 37.149 44.979 37.149 44.979 multiply_cannon_multrec 69360 15.2 29.621 34.639 39.761 44.549 qs_ot_get_derivative 55 11.6 0.001 0.001 40.332 40.541 multiply_cannon_metrocomm3 69360 15.2 0.200 0.215 27.327 36.285 qs_ot_p2m_diag 40 11.0 0.020 0.031 35.202 35.271 multiply_cannon_sync_h2d 69360 15.2 28.974 32.867 28.974 32.867 cp_dbcsr_syevd 40 12.0 0.002 0.002 31.873 31.874 rebuild_ks_matrix 110 8.4 0.000 0.000 29.471 29.703 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.033 29.471 29.702 init_scf_loop 7 6.6 0.000 0.006 27.403 27.406 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.073 27.276 cp_fm_syevd 40 13.0 0.000 0.000 25.873 26.127 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.145 23.407 apply_single 62 13.6 0.000 0.000 23.145 23.407 prepare_preconditioner 7 7.6 0.000 0.000 22.447 22.490 make_preconditioner 7 8.6 0.000 0.003 22.447 22.490 ot_new_cg_direction 55 11.6 0.002 0.026 21.215 21.216 cp_fm_redistribute_end 40 14.0 10.522 20.997 10.527 20.999 cp_fm_syevd_base 40 14.0 10.462 20.938 10.462 20.938 qs_rho_update_rho_low 110 7.6 0.001 0.014 17.619 18.057 calculate_rho_elec 110 8.6 0.030 0.040 17.619 18.056 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.270 15.441 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.991 15.052 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.540 14.642 mp_sum_l 4764 12.2 13.510 14.328 13.510 14.328 calculate_dm_sparse 110 9.5 0.000 0.001 12.185 12.355 pw_transfer 1645 12.4 0.079 0.103 12.063 12.328 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.924 12.195 density_rs2pw 110 9.6 0.005 0.006 11.764 12.179 init_scf_run 5 5.8 0.000 0.001 11.707 11.708 scf_env_initial_rho_setup 5 6.8 0.002 0.017 11.707 11.708 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.767 10.873 dbcsr_mm_accdrv_process 154766 15.8 6.107 6.362 10.008 10.862 qs_vxc_create 110 10.4 0.002 0.006 10.743 10.793 fft_wrap_pw1pw2_240 915 15.0 0.827 0.928 10.432 10.683 cp_fm_cholesky_invert 7 10.6 10.369 10.376 10.369 10.376 check_diag 80 13.5 8.685 9.180 10.052 10.305 fft3d_pb 915 16.0 2.394 2.702 8.632 8.939 transfer_rs2pw 445 10.6 0.007 0.008 8.332 8.774 sum_up_and_integrate 60 10.3 0.001 0.005 8.603 8.678 integrate_v_rspace 60 11.3 0.002 0.002 8.585 8.662 multiply_cannon_metrocomm1 69360 15.2 0.098 0.106 5.183 8.181 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.330 7.596 xc_vxc_pw_create 60 11.3 0.039 0.049 7.184 7.234 acc_transpose_blocks 69360 15.2 0.356 0.377 6.659 7.220 make_m2s 2890 13.2 0.079 0.087 6.541 7.161 make_full_single_inverse 7 9.6 0.001 0.004 7.109 7.152 make_images 2890 14.2 0.238 0.260 6.433 7.053 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.729 6.746 calculate_first_density_matrix 1 7.0 0.000 0.004 6.697 6.717 xc_pw_derive 510 13.4 0.005 0.007 6.389 6.485 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.217 6.277 mp_alltoall_z22v 2340 17.7 5.735 6.008 5.735 6.008 mp_waitany 7680 13.5 4.576 5.157 4.576 5.157 potential_pw2rs 60 12.3 0.002 0.003 4.815 4.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=237.657000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 48d3695c3ced91faad61a1fd2c67ef2ebe356476 Summary: empty Status: OK