=== 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: 3bcd6e6d97c8df411d49f3e0e700dc328631558d ################# 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 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.8.2, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 28.08.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/01 job id: 48762766 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/02 job id: 48762767 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/03 job id: 48762768 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/04 job id: 48762769 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/05 job id: 48762770 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/06 job id: 48762771 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/07 job id: 48762772 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/08 job id: 48762773 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/09 job id: 48762774 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/10 job id: 48762775 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/11 job id: 48762776 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/12 job id: 48762777 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/13 job id: 48762778 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/14 job id: 48762779 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/15 job id: 48762780 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/16 job id: 48762781 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/17 job id: 48762782 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/18 job id: 48762783 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/19 job id: 48762784 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/20 job id: 48762785 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/21 job id: 48762786 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/22 job id: 48762787 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/23 job id: 48762788 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/24 job id: 48762789 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/25 job id: 48762790 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/26 job id: 48762791 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/27 job id: 48762792 --- 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/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 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 0.019 0.033 136.767 136.768 farming_run 1 2.0 135.988 135.993 136.733 136.737 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.495273E+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 0.010 0.029 115.721 115.721 qs_energies 1 2.0 0.000 0.000 115.241 115.296 mp2_main 1 3.0 0.000 0.000 112.974 113.029 mp2_gpw_main 1 4.0 0.028 0.034 111.813 111.868 mp2_ri_gpw_compute_in 1 5.0 0.205 0.706 92.832 93.167 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.645 54.980 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.027 46.103 get_2c_integrals 1 6.0 0.008 0.009 37.277 38.013 integrate_v_rspace 273 8.0 0.438 0.455 24.602 29.508 pw_transfer 6555 10.6 0.377 0.413 26.835 27.261 fft_wrap_pw1pw2 5465 11.4 0.045 0.050 25.433 25.935 grid_integrate_task_list 273 9.0 20.516 25.885 20.516 25.885 fft_wrap_pw1pw2_100 2178 12.4 1.140 1.318 22.986 23.471 compute_2c_integrals 1 7.0 0.003 0.003 19.520 19.523 compute_2c_integrals_loop_lm 1 8.0 0.003 0.005 18.663 19.205 mp2_eri_2c_integrate_gpw 1 9.0 2.360 2.419 18.660 19.203 rpa_ri_compute_en 1 5.0 0.033 0.039 18.863 18.970 cp_fm_cholesky_decompose 12 8.2 17.744 18.486 17.744 18.486 cholesky_decomp 1 7.0 0.000 0.000 16.596 17.337 fft3d_s 5443 13.4 16.118 16.394 16.141 16.415 ao_to_mo_and_store_B_mult_1 272 7.0 10.772 15.314 10.772 15.314 calculate_wavefunction 272 8.0 5.428 5.557 12.338 13.055 rpa_num_int 1 6.0 0.004 0.063 10.713 10.770 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.695 10.698 calc_mat_Q 8 8.0 0.000 0.000 9.506 9.591 contract_S_to_Q 8 9.0 0.000 0.000 8.926 9.013 calc_potential_gpw 544 9.5 0.005 0.005 8.223 8.756 parallel_gemm_fm 14 9.1 0.000 0.000 8.514 8.602 parallel_gemm_fm_cosma 14 10.1 8.514 8.602 8.514 8.602 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.121 8.474 potential_pw2rs 545 10.0 0.107 0.108 7.528 8.223 create_integ_mat 1 6.0 0.022 0.026 7.702 7.702 collocate_single_gaussian 272 10.0 0.039 0.042 7.377 7.659 array2fm 1 7.0 0.000 0.000 6.716 7.044 pw_scatter_s 2720 13.7 4.315 4.398 4.315 4.398 pw_gather_s 2722 13.2 3.460 3.790 3.460 3.790 array2fm_buffer_send 1 8.0 2.958 3.125 2.958 3.125 pw_poisson_solve 545 10.5 1.122 1.207 2.230 2.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.868082, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 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 0.046 0.093 429.104 429.106 farming_run 1 2.0 427.176 427.222 429.006 429.019 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.248346E+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 25511415. 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 0.012 0.034 218.431 218.432 qs_energies 1 2.0 0.007 0.019 218.150 218.163 scf_env_do_scf 1 3.0 0.000 0.000 114.790 114.791 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.831 113.837 rebuild_ks_matrix 4 6.0 0.000 0.000 113.829 113.836 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 113.829 113.836 hfx_ks_matrix 4 8.0 0.001 0.001 113.470 113.475 integrate_four_center 4 9.0 0.152 0.470 113.469 113.474 integrate_four_center_main 4 10.0 0.122 0.648 102.017 104.795 integrate_four_center_bin 270 11.0 101.895 104.681 101.895 104.681 mp2_main 1 3.0 0.026 0.124 103.050 103.070 mp2_gpw_main 1 4.0 0.102 0.199 101.947 101.964 init_scf_loop 1 4.0 0.000 0.000 96.989 96.989 mp2_ri_gpw_compute_in 1 5.0 0.085 0.110 74.454 75.773 mp2_ri_gpw_compute_in_loop 1 6.0 0.006 0.053 54.059 55.352 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.164 41.604 46.526 integrate_v_rspace 95 8.0 0.397 0.561 28.084 32.853 pw_transfer 2240 10.6 0.146 0.177 29.489 29.891 mp2_ri_gpw_compute_en 1 5.0 0.079 0.260 27.249 29.610 ao_to_mo_and_store_B_mult_1 91 7.0 10.772 29.467 10.772 29.467 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.503 28.871 grid_integrate_task_list 95 9.0 23.494 28.464 23.494 28.464 fft_wrap_pw1pw2_100 730 12.4 1.269 1.488 26.262 26.613 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.888 24.908 24.921 get_2c_integrals 1 6.0 0.013 0.112 20.193 20.258 compute_2c_integrals 1 7.0 0.038 0.106 19.161 19.211 compute_2c_integrals_loop_lm 1 8.0 0.010 0.025 18.594 18.926 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.894 18.585 18.924 fft3d_s 1823 13.4 18.475 18.845 18.489 18.859 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.800 17.800 calculate_wavefunction 91 8.0 2.011 2.040 9.618 9.841 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.577 8.751 9.332 potential_pw2rs 186 10.0 0.033 0.034 8.427 8.931 local_gemm 172 8.0 8.194 8.778 8.194 8.778 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.123 8.483 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.520 7.924 8.283 calc_potential_gpw 182 9.5 0.002 0.002 7.878 8.090 collocate_single_gaussian 91 10.0 0.019 0.078 7.741 8.052 integrate_four_center_load 4 10.0 0.000 0.000 6.775 6.779 hfx_load_balance 1 11.0 0.000 0.000 6.775 6.779 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.416 6.346 6.416 mp_sendrecv_dm3 2068 8.0 5.958 6.310 5.958 6.310 mp_sync 37 10.5 3.490 5.755 3.490 5.755 pw_gather_s 912 13.2 4.504 4.901 4.504 4.901 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.942323, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.071424E+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 0.024 0.045 58.573 58.583 qs_mol_dyn_low 1 2.0 0.003 0.004 58.254 58.299 qs_forces 11 3.9 0.002 0.003 58.188 58.188 qs_energies 11 4.9 0.002 0.002 56.618 56.632 scf_env_do_scf 11 5.9 0.001 0.004 49.221 49.221 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 46.968 46.969 dbcsr_multiply_generic 2286 12.5 0.124 0.210 36.549 36.979 qs_scf_new_mos 108 7.5 0.000 0.001 36.242 36.545 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.241 36.544 ot_scf_mini 108 9.5 0.002 0.002 34.479 34.659 multiply_cannon 2286 13.5 0.189 0.197 28.231 29.732 multiply_cannon_loop 2286 14.5 1.825 1.945 27.456 28.954 velocity_verlet 10 3.0 0.001 0.002 27.638 27.638 ot_mini 108 10.5 0.001 0.001 20.970 21.213 qs_ot_get_derivative 108 11.5 0.001 0.001 17.930 18.110 mp_waitall_1 245248 16.5 9.688 15.990 9.688 15.990 multiply_cannon_metrocomm3 54864 15.5 0.074 0.081 6.549 13.543 multiply_cannon_multrec 54864 15.5 3.618 5.813 8.447 12.155 qs_ot_get_p 119 10.4 0.001 0.001 8.647 8.958 rebuild_ks_matrix 119 8.3 0.000 0.000 8.255 8.404 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.029 8.255 8.403 mp_sum_l 7287 12.8 5.903 7.648 5.903 7.648 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.260 7.401 dbcsr_mm_accdrv_process 76910 16.1 1.822 2.853 4.744 6.712 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.053 6.464 multiply_cannon_sync_h2d 54864 15.5 5.072 6.379 5.072 6.379 init_scf_run 11 5.9 0.000 0.001 6.082 6.083 scf_env_initial_rho_setup 11 6.9 0.007 0.008 6.082 6.082 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.862 5.970 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.665 5.724 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.845 4.846 sum_up_and_integrate 119 10.3 0.001 0.001 4.766 4.773 integrate_v_rspace 119 11.3 0.002 0.003 4.755 4.763 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.546 4.682 calculate_rho_elec 119 8.7 0.011 0.016 4.546 4.682 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.571 4.572 cp_fm_redistribute_end 50 14.0 2.325 4.529 2.334 4.533 cp_fm_diag_elpa_base 50 14.0 2.191 4.403 2.196 4.413 jit_kernel_multiply 13 15.8 2.857 3.971 2.857 3.971 calculate_dm_sparse 119 9.5 0.000 0.000 3.706 3.876 calculate_first_density_matrix 1 7.0 0.000 0.001 3.688 3.694 multiply_cannon_metrocomm1 54864 15.5 0.058 0.063 2.282 3.542 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.000 3.180 apply_single 119 13.6 0.000 0.000 3.000 3.180 acc_transpose_blocks 54864 15.5 0.239 0.266 2.347 2.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.862 2.869 ot_diis_step 108 11.5 0.006 0.006 2.757 2.758 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.653 2.744 density_rs2pw 119 9.7 0.004 0.004 2.508 2.605 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.380 2.442 wfi_extrapolate 11 7.9 0.001 0.001 2.311 2.311 init_scf_loop 11 6.9 0.000 0.001 2.221 2.222 make_m2s 4572 13.5 0.054 0.056 2.069 2.149 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 2.111 2.129 grid_integrate_task_list 119 12.3 2.028 2.119 2.028 2.119 potential_pw2rs 119 12.3 0.004 0.004 2.083 2.108 pw_transfer 1439 11.6 0.052 0.057 2.014 2.090 make_images 4572 14.5 0.134 0.141 1.986 2.066 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.939 2.017 mp_sum_d 4135 12.0 1.342 1.915 1.342 1.915 fft3d_ps 1201 14.6 0.374 0.479 1.655 1.727 mp_alltoall_d11v 2130 13.8 1.524 1.666 1.524 1.666 transfer_rs2pw 487 10.6 0.005 0.006 1.537 1.642 fft_wrap_pw1pw2_140 487 13.2 0.139 0.152 1.488 1.568 transfer_pw2rs 487 13.2 0.006 0.006 1.483 1.505 mp_waitany 12084 13.8 1.274 1.468 1.274 1.468 acc_transpose_blocks_sync 164592 16.5 1.205 1.437 1.205 1.437 grid_collocate_task_list 119 9.7 1.346 1.409 1.346 1.409 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.583000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.313408E+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 1601272. 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 0.029 46.059 46.060 qs_mol_dyn_low 1 2.0 0.003 0.003 45.807 45.816 qs_forces 11 3.9 0.002 0.003 45.645 45.647 qs_energies 11 4.9 0.001 0.002 43.772 43.777 scf_env_do_scf 11 5.9 0.000 0.001 37.992 37.992 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 34.617 34.617 dbcsr_multiply_generic 2286 12.5 0.103 0.105 23.883 24.339 qs_scf_new_mos 108 7.5 0.001 0.001 23.390 23.609 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.390 23.608 velocity_verlet 10 3.0 0.002 0.003 22.553 22.556 ot_scf_mini 108 9.5 0.002 0.003 22.391 22.551 multiply_cannon 2286 13.5 0.211 0.219 17.454 19.059 multiply_cannon_loop 2286 14.5 1.197 1.258 15.986 17.706 ot_mini 108 10.5 0.001 0.001 13.643 13.872 mp_waitall_1 200699 16.5 7.090 12.096 7.090 12.096 qs_ot_get_derivative 108 11.5 0.001 0.001 11.174 11.338 multiply_cannon_metrocomm3 27432 15.5 0.074 0.077 4.222 9.523 multiply_cannon_multrec 27432 15.5 1.817 4.118 6.199 9.115 rebuild_ks_matrix 119 8.3 0.000 0.000 8.269 8.440 qs_ks_build_kohn_sham_matrix 119 9.3 0.018 0.030 8.269 8.440 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.296 7.453 dbcsr_mm_accdrv_process 47894 16.0 3.494 6.038 4.301 6.444 qs_ot_get_p 119 10.4 0.001 0.001 5.587 5.796 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.193 5.227 calculate_rho_elec 119 8.7 0.021 0.024 5.192 5.227 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.240 5.098 sum_up_and_integrate 119 10.3 0.001 0.001 5.047 5.051 integrate_v_rspace 119 11.3 0.002 0.003 5.033 5.038 mp_sum_l 7287 12.8 2.376 4.455 2.376 4.455 init_scf_run 11 5.9 0.000 0.001 4.431 4.432 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.431 4.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.082 4.129 apply_single 119 13.6 0.000 0.000 3.082 4.129 make_m2s 4572 13.5 0.053 0.054 3.609 3.975 make_images 4572 14.5 0.207 0.244 3.519 3.888 density_rs2pw 119 9.7 0.004 0.004 3.447 3.525 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.459 3.484 init_scf_loop 11 6.9 0.001 0.006 3.330 3.330 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.996 2.997 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.770 2.853 potential_pw2rs 119 12.3 0.006 0.006 2.725 2.763 pw_transfer 1439 11.6 0.065 0.068 2.476 2.553 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.384 2.464 calculate_first_density_matrix 1 7.0 0.000 0.000 2.433 2.436 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.425 2.427 ot_diis_step 108 11.5 0.011 0.011 2.404 2.405 cp_fm_redistribute_end 50 14.0 1.225 2.370 1.231 2.374 transfer_rs2pw 487 10.6 0.005 0.005 2.277 2.353 cp_fm_diag_elpa_base 50 14.0 1.103 2.235 1.133 2.281 calculate_dm_sparse 119 9.5 0.000 0.001 2.196 2.277 prepare_preconditioner 11 7.9 0.000 0.000 2.237 2.269 make_preconditioner 11 8.9 0.000 0.000 2.237 2.269 make_images_data 4572 15.5 0.048 0.056 1.549 2.182 multiply_cannon_sync_h2d 27432 15.5 1.665 2.181 1.665 2.181 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.098 2.172 fft3d_ps 1201 14.6 0.523 0.581 2.022 2.094 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.033 2.036 transfer_pw2rs 487 13.2 0.004 0.005 1.941 1.980 acc_transpose_blocks 27432 15.5 0.119 0.123 1.629 1.975 mp_alltoall_d11v 2130 13.8 1.701 1.960 1.701 1.960 grid_integrate_task_list 119 12.3 1.831 1.957 1.831 1.957 hybrid_alltoall_any 4725 16.4 0.054 0.117 1.364 1.914 wfi_extrapolate 11 7.9 0.001 0.001 1.913 1.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.689 1.738 fft_wrap_pw1pw2_140 487 13.2 0.157 0.165 1.582 1.648 make_images_sizes 4572 15.5 0.005 0.005 1.189 1.592 mp_alltoall_i44 4572 16.5 1.184 1.587 1.184 1.587 jit_kernel_multiply 8 16.3 0.750 1.586 0.750 1.586 mp_sendrecv_dv 16779 12.7 1.475 1.520 1.475 1.520 mp_allgather_i34 2286 14.5 0.883 1.413 0.883 1.413 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.358 1.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.346 1.359 grid_collocate_task_list 119 9.7 1.290 1.347 1.290 1.347 mp_alltoall_z22v 1201 16.6 1.094 1.213 1.094 1.213 mp_sum_d 4135 12.0 0.645 1.136 0.645 1.136 dbcsr_complete_redistribute 329 12.2 0.082 0.097 1.009 1.093 cp_fm_cholesky_invert 11 10.9 1.057 1.063 1.057 1.063 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.034 1.048 transfer_rs2pw_50 119 11.7 0.183 0.192 0.938 1.038 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.952 0.953 acc_transpose_blocks_sync 82296 16.5 0.820 0.948 0.820 0.948 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=46.060000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 520.376320E+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 0.258 0.693 45.660 45.713 qs_mol_dyn_low 1 2.0 0.003 0.010 43.793 43.801 qs_forces 11 3.9 0.003 0.004 43.293 43.294 qs_energies 11 4.9 0.100 0.273 41.526 41.528 scf_env_do_scf 11 5.9 0.002 0.012 30.712 30.729 scf_env_do_scf_inner_loop 108 6.5 0.020 0.049 25.497 25.531 dbcsr_multiply_generic 2286 12.5 0.175 0.225 18.876 19.097 qs_scf_new_mos 108 7.5 0.001 0.001 17.195 17.243 qs_scf_loop_do_ot 108 8.5 0.002 0.008 17.194 17.242 velocity_verlet 10 3.0 0.010 0.026 16.798 16.812 ot_scf_mini 108 9.5 0.004 0.011 16.390 16.437 multiply_cannon 2286 13.5 0.194 0.199 14.794 15.571 multiply_cannon_loop 2286 14.5 0.858 0.896 13.925 14.838 ot_mini 108 10.5 0.001 0.001 10.002 10.059 qs_ot_get_derivative 108 11.5 0.001 0.001 8.452 8.502 multiply_cannon_multrec 18288 15.5 1.976 3.191 7.739 8.162 rebuild_ks_matrix 119 8.3 0.000 0.000 6.626 6.676 qs_ks_build_kohn_sham_matrix 119 9.3 0.039 0.111 6.626 6.676 dbcsr_mm_accdrv_process 38222 16.0 5.597 6.531 5.672 6.596 init_scf_run 11 5.9 0.000 0.001 6.212 6.221 scf_env_initial_rho_setup 11 6.9 0.084 0.205 6.212 6.221 qs_ot_get_p 119 10.4 0.002 0.012 6.016 6.108 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.828 5.890 init_scf_loop 11 6.9 0.027 0.118 5.189 5.219 calculate_first_density_matrix 1 7.0 0.043 0.122 4.701 4.795 mp_waitall_1 158411 16.6 3.160 4.443 3.160 4.443 sum_up_and_integrate 119 10.3 0.001 0.001 3.992 4.026 integrate_v_rspace 119 11.3 0.003 0.003 3.979 4.013 transfer_rs2pw 487 10.6 0.005 0.005 2.885 3.832 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.171 3.776 qs_energies_init_hamiltonians 11 5.9 0.085 0.209 3.763 3.764 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.722 3.737 calculate_rho_elec 119 8.7 0.032 0.042 3.721 3.737 transfer_rs2pw_140 130 11.5 0.118 0.124 2.209 3.168 mp_waitany 9880 13.7 2.225 3.166 2.225 3.166 qs_env_update_s_mstruct 11 6.9 0.019 0.092 2.388 3.140 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.878 2.884 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.647 2.826 build_core_hamiltonian_matrix 11 6.9 0.023 0.064 1.060 2.805 qs_ot_p2m_diag 50 11.0 0.032 0.068 2.663 2.670 calculate_rho_core 11 7.9 0.099 0.278 1.857 2.648 arnoldi_extremal 119 11.4 0.007 0.043 2.578 2.628 arnoldi_normal_ev 119 12.4 0.030 0.221 2.571 2.614 density_rs2pw 119 9.7 0.004 0.004 2.234 2.390 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.997 2.385 apply_single 119 13.6 0.000 0.000 1.997 2.385 cp_dbcsr_syevd 50 12.0 0.008 0.016 2.324 2.346 make_m2s 4572 13.5 0.052 0.077 2.200 2.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.277 2.311 pw_transfer 1439 11.6 0.065 0.068 2.212 2.272 make_images 4572 14.5 0.194 0.206 2.107 2.233 ot_scf_init 11 7.9 0.003 0.008 2.179 2.232 dbcsr_matrix_vector_mult 1669 14.0 0.153 1.131 2.014 2.226 mp_sum_dv 6497 15.2 1.790 2.188 1.790 2.188 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.119 2.180 mp_sum_l 7287 12.8 1.581 2.127 1.581 2.127 calculate_dm_sparse 119 9.5 0.000 0.001 2.095 2.105 prepare_preconditioner 11 7.9 0.000 0.001 2.034 2.071 make_preconditioner 11 8.9 0.013 0.104 2.034 2.071 arnoldi_init 119 13.4 0.006 0.039 1.976 2.036 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.875 2.006 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.972 1.983 cp_fm_diag_elpa_base 50 14.0 1.947 1.960 1.970 1.981 grid_integrate_task_list 119 12.3 1.791 1.910 1.791 1.910 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.892 1.901 mp_sum_d 4135 12.0 1.281 1.841 1.281 1.841 potential_pw2rs 119 12.3 0.007 0.008 1.768 1.816 acc_transpose_blocks 18288 15.5 0.082 0.086 1.713 1.793 fft3d_ps 1201 14.6 0.529 0.549 1.690 1.750 dbcsr_col_vec_to_rep_row 1669 15.0 0.024 0.119 0.715 1.743 fft_wrap_pw1pw2_140 487 13.2 0.210 0.216 1.618 1.681 ot_diis_step 108 11.5 0.011 0.011 1.514 1.514 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.395 1.402 grid_collocate_task_list 119 9.7 1.246 1.378 1.246 1.378 wfi_extrapolate 11 7.9 0.001 0.001 1.267 1.267 dbcsr_dot_sd 1205 11.9 0.097 0.101 0.785 1.146 multiply_cannon_sync_h2d 18288 15.5 0.986 1.120 0.986 1.120 dbcsr_complete_redistribute 329 12.2 0.236 0.475 0.969 1.112 make_images_data 4572 15.5 0.049 0.054 0.887 1.083 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.450 1.055 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.033 1.048 transfer_pw2rs 487 13.2 0.004 0.004 1.029 1.045 mp_alltoall_z22v 1201 16.6 0.942 1.037 0.942 1.037 build_overlap_matrix_low 22 6.9 0.244 0.413 0.473 1.013 mp_alltoall_d11v 2130 13.8 0.853 0.979 0.853 0.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.953 0.957 hybrid_alltoall_any 4725 16.4 0.059 0.117 0.768 0.956 acc_transpose_blocks_kernels 18288 16.5 0.219 0.225 0.878 0.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=45.713000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.348400 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 559.353856E+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 0.014 0.027 42.797 42.797 qs_mol_dyn_low 1 2.0 0.003 0.003 42.036 42.054 qs_forces 11 3.9 0.003 0.011 41.958 41.958 qs_energies 11 4.9 0.001 0.001 40.148 40.154 scf_env_do_scf 11 5.9 0.000 0.001 34.165 34.166 scf_env_do_scf_inner_loop 108 6.5 0.036 0.096 29.693 29.694 dbcsr_multiply_generic 2286 12.5 0.118 0.121 21.716 21.854 velocity_verlet 10 3.0 0.018 0.028 20.994 20.996 qs_scf_new_mos 108 7.5 0.001 0.001 20.036 20.138 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.035 20.137 ot_scf_mini 108 9.5 0.002 0.002 18.947 19.040 multiply_cannon 2286 13.5 0.218 0.229 17.101 17.504 multiply_cannon_loop 2286 14.5 1.521 1.632 15.966 16.441 ot_mini 108 10.5 0.001 0.001 11.413 11.512 multiply_cannon_multrec 27432 15.5 2.575 3.268 9.550 9.830 qs_ot_get_derivative 108 11.5 0.001 0.001 9.495 9.589 dbcsr_mm_accdrv_process 47916 15.9 6.268 8.140 6.873 8.353 rebuild_ks_matrix 119 8.3 0.000 0.000 7.239 7.289 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.238 7.288 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.423 6.467 qs_ot_get_p 119 10.4 0.001 0.001 4.784 4.856 init_scf_run 11 5.9 0.000 0.001 4.567 4.567 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.566 4.567 init_scf_loop 11 6.9 0.006 0.010 4.444 4.445 sum_up_and_integrate 119 10.3 0.001 0.001 4.320 4.327 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.282 4.322 calculate_rho_elec 119 8.7 0.040 0.046 4.281 4.321 integrate_v_rspace 119 11.3 0.003 0.003 4.308 4.316 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.488 3.914 mp_waitall_1 137007 16.6 2.962 3.649 2.962 3.649 prepare_preconditioner 11 7.9 0.000 0.000 3.237 3.247 make_preconditioner 11 8.9 0.000 0.000 3.237 3.247 make_m2s 4572 13.5 0.055 0.057 3.015 3.166 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.817 3.151 make_images 4572 14.5 0.275 0.336 2.907 3.058 calculate_first_density_matrix 1 7.0 0.000 0.000 2.872 2.875 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.815 2.825 density_rs2pw 119 9.7 0.004 0.004 2.693 2.766 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.631 2.663 acc_transpose_blocks 27432 15.5 0.123 0.126 2.473 2.587 calculate_dm_sparse 119 9.5 0.000 0.000 2.505 2.557 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.371 2.372 pw_transfer 1439 11.6 0.065 0.069 2.289 2.330 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.139 2.284 apply_single 119 13.6 0.000 0.000 2.139 2.283 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.197 2.241 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.141 2.143 potential_pw2rs 119 12.3 0.009 0.009 2.078 2.089 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.020 2.027 cp_fm_diag_elpa_base 50 14.0 1.978 1.996 2.017 2.025 grid_integrate_task_list 119 12.3 1.822 1.922 1.822 1.922 ot_diis_step 108 11.5 0.012 0.012 1.875 1.875 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.860 1.874 fft3d_ps 1201 14.6 0.554 0.601 1.719 1.750 fft_wrap_pw1pw2_140 487 13.2 0.244 0.257 1.679 1.726 transfer_rs2pw 487 10.6 0.004 0.005 1.628 1.716 wfi_extrapolate 11 7.9 0.001 0.001 1.630 1.631 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.556 1.566 acc_transpose_blocks_sync 82296 16.5 1.423 1.537 1.423 1.537 mp_sum_l 7287 12.8 1.158 1.521 1.158 1.521 make_images_data 4572 15.5 0.049 0.053 1.208 1.457 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.834 1.448 transfer_pw2rs 487 13.2 0.004 0.004 1.392 1.400 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.359 1.375 grid_collocate_task_list 119 9.7 1.249 1.334 1.249 1.334 jit_kernel_multiply 6 16.0 0.540 1.298 0.540 1.298 hybrid_alltoall_any 4725 16.4 0.066 0.154 1.008 1.294 dbcsr_complete_redistribute 329 12.2 0.166 0.209 1.015 1.294 cp_fm_upper_to_full 72 14.2 0.852 1.215 0.852 1.215 mp_alltoall_d11v 2130 13.8 0.997 1.173 0.997 1.173 cp_fm_cholesky_invert 11 10.9 1.127 1.132 1.127 1.132 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.112 1.114 mp_alltoall_z22v 1201 16.6 1.047 1.077 1.047 1.077 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.722 0.997 mp_sendrecv_dv 8211 12.7 0.924 0.943 0.924 0.943 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 0.899 0.927 make_images_sizes 4572 15.5 0.005 0.005 0.668 0.924 mp_alltoall_i44 4572 16.5 0.663 0.919 0.663 0.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.911 multiply_cannon_metrocomm1 27432 15.5 0.036 0.037 0.315 0.893 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.874 arnoldi_extremal 119 11.4 0.002 0.002 0.844 0.873 arnoldi_normal_ev 119 12.4 0.149 0.238 0.842 0.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.797000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.727273, yerr=3.670296 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 627.056640E+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 0.090 0.146 32.370 32.394 qs_mol_dyn_low 1 2.0 0.003 0.004 31.924 31.955 qs_forces 11 3.9 0.003 0.004 31.735 31.736 qs_energies 11 4.9 0.009 0.010 30.007 30.010 scf_env_do_scf 11 5.9 0.000 0.001 23.558 23.558 scf_env_do_scf_inner_loop 108 6.5 0.009 0.013 20.746 20.747 velocity_verlet 10 3.0 0.080 0.081 15.607 15.613 dbcsr_multiply_generic 2286 12.5 0.104 0.108 14.528 14.610 qs_scf_new_mos 108 7.5 0.001 0.001 12.723 12.750 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.722 12.749 multiply_cannon 2286 13.5 0.224 0.233 11.581 12.171 ot_scf_mini 108 9.5 0.002 0.002 11.985 12.007 multiply_cannon_loop 2286 14.5 0.647 0.672 10.488 10.679 multiply_cannon_multrec 9144 15.5 1.820 2.013 6.782 6.991 ot_mini 108 10.5 0.001 0.001 6.822 6.848 rebuild_ks_matrix 119 8.3 0.000 0.000 6.056 6.084 qs_ks_build_kohn_sham_matrix 119 9.3 0.019 0.020 6.055 6.083 qs_ot_get_derivative 108 11.5 0.001 0.001 5.494 5.514 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.409 5.434 dbcsr_mm_accdrv_process 12550 15.8 3.875 4.898 4.852 4.978 init_scf_run 11 5.9 0.000 0.001 4.915 4.915 scf_env_initial_rho_setup 11 6.9 0.106 0.106 4.915 4.915 sum_up_and_integrate 119 10.3 0.001 0.001 3.619 3.626 integrate_v_rspace 119 11.3 0.003 0.003 3.609 3.615 calculate_first_density_matrix 1 7.0 0.000 0.000 3.605 3.605 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.544 3.553 calculate_rho_elec 119 8.7 0.060 0.061 3.543 3.553 qs_ot_get_p 119 10.4 0.001 0.001 3.241 3.279 init_scf_loop 11 6.9 0.007 0.008 2.786 2.787 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.445 2.446 make_m2s 4572 13.5 0.034 0.035 2.030 2.176 pw_transfer 1439 11.6 0.066 0.069 2.130 2.139 qs_ot_p2m_diag 50 11.0 0.023 0.027 2.106 2.109 mp_waitall_1 115863 16.7 1.642 2.096 1.642 2.096 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.077 2.087 make_images 4572 14.5 0.270 0.301 1.941 2.085 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.036 2.046 calculate_dm_sparse 119 9.5 0.000 0.000 1.988 2.006 density_rs2pw 119 9.7 0.004 0.004 1.908 1.980 grid_integrate_task_list 119 12.3 1.872 1.975 1.872 1.975 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.888 1.904 prepare_preconditioner 11 7.9 0.000 0.000 1.852 1.856 make_preconditioner 11 8.9 0.000 0.000 1.852 1.856 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.846 1.847 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.733 1.760 jit_kernel_multiply 6 15.3 0.938 1.716 0.938 1.716 fft_wrap_pw1pw2_140 487 13.2 0.322 0.334 1.623 1.635 acc_transpose_blocks 9144 15.5 0.042 0.044 1.518 1.560 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.532 1.541 cp_fm_diag_elpa_base 50 14.0 1.504 1.520 1.530 1.539 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.459 1.469 fft3d_ps 1201 14.6 0.561 0.573 1.451 1.460 potential_pw2rs 119 12.3 0.010 0.011 1.390 1.396 grid_collocate_task_list 119 9.7 1.303 1.384 1.303 1.384 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.355 1.359 ot_diis_step 108 11.5 0.013 0.013 1.310 1.310 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.236 1.276 apply_single 119 13.6 0.000 0.000 1.236 1.276 qs_energies_init_hamiltonians 11 5.9 0.009 0.011 1.249 1.251 wfi_extrapolate 11 7.9 0.001 0.001 1.150 1.150 hybrid_alltoall_any 4725 16.4 0.066 0.175 0.843 1.051 make_images_data 4572 15.5 0.043 0.047 0.880 1.047 mp_alltoall_d11v 2130 13.8 0.883 0.974 0.883 0.974 transfer_rs2pw 487 10.6 0.004 0.004 0.906 0.966 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.874 0.926 cp_fm_cholesky_invert 11 10.9 0.917 0.920 0.917 0.920 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.860 0.868 dbcsr_complete_redistribute 329 12.2 0.240 0.361 0.771 0.803 mp_alltoall_z22v 1201 16.6 0.763 0.799 0.763 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.772 0.774 acc_transpose_blocks_sync 27432 16.5 0.724 0.767 0.724 0.767 transfer_pw2rs 487 13.2 0.003 0.004 0.760 0.763 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.690 0.739 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.736 0.738 mp_allgather_i34 2286 14.5 0.267 0.731 0.267 0.731 parallel_gemm_fm 81 9.0 0.000 0.000 0.724 0.725 parallel_gemm_fm_cosma 81 10.0 0.724 0.725 0.724 0.725 make_basis_sm 11 9.8 0.000 0.000 0.669 0.670 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.394000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.727273, yerr=7.484419 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 766.775296E+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 0.104 0.144 47.036 47.038 qs_mol_dyn_low 1 2.0 0.003 0.004 46.383 46.399 qs_forces 11 3.9 0.002 0.003 46.306 46.307 qs_energies 11 4.9 0.021 0.022 44.280 44.286 scf_env_do_scf 11 5.9 0.001 0.001 36.735 36.735 scf_env_do_scf_inner_loop 108 6.5 0.013 0.017 28.453 28.454 velocity_verlet 10 3.0 0.020 0.021 24.987 24.993 dbcsr_multiply_generic 2286 12.5 0.188 0.295 20.363 20.504 qs_scf_new_mos 108 7.5 0.001 0.001 18.746 18.847 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.745 18.846 ot_scf_mini 108 9.5 0.002 0.002 17.557 17.660 multiply_cannon 2286 13.5 0.299 0.313 15.813 16.683 multiply_cannon_loop 2286 14.5 0.868 0.889 14.515 15.360 ot_mini 108 10.5 0.001 0.001 10.666 10.789 multiply_cannon_multrec 9144 15.5 3.529 4.898 9.218 9.468 qs_ot_get_derivative 108 11.5 0.001 0.001 8.552 8.651 init_scf_loop 11 6.9 0.010 0.010 8.248 8.250 rebuild_ks_matrix 119 8.3 0.000 0.000 7.306 7.452 qs_ks_build_kohn_sham_matrix 119 9.3 0.026 0.027 7.306 7.451 prepare_preconditioner 11 7.9 0.000 0.000 7.234 7.247 make_preconditioner 11 8.9 0.000 0.000 7.234 7.247 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.771 7.123 dbcsr_mm_accdrv_process 12550 15.8 4.883 6.736 5.554 6.976 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.574 6.706 init_scf_run 11 5.9 0.000 0.001 5.216 5.216 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.216 5.216 cp_fm_upper_to_full 72 14.2 3.170 4.574 3.170 4.574 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.190 4.230 calculate_rho_elec 119 8.7 0.118 0.121 4.190 4.230 qs_ot_get_p 119 10.4 0.001 0.001 3.886 4.027 sum_up_and_integrate 119 10.3 0.001 0.001 3.940 3.948 integrate_v_rspace 119 11.3 0.004 0.004 3.930 3.938 mp_waitall_1 94719 16.7 2.744 3.745 2.744 3.745 calculate_first_density_matrix 1 7.0 0.000 0.000 3.665 3.673 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.060 3.516 dbcsr_complete_redistribute 329 12.2 0.298 0.305 2.096 2.922 make_m2s 4572 13.5 0.038 0.038 2.720 2.916 make_images 4572 14.5 0.353 0.388 2.598 2.795 pw_transfer 1439 11.6 0.069 0.069 2.737 2.743 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.294 2.658 apply_single 119 13.6 0.000 0.000 2.294 2.658 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.638 2.645 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.744 2.568 calculate_dm_sparse 119 9.5 0.000 0.000 2.513 2.530 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.549 2.449 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.380 2.381 mp_alltoall_i22 627 13.8 1.457 2.306 1.457 2.306 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.457 2.272 density_rs2pw 119 9.7 0.004 0.004 2.185 2.216 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.147 2.198 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.194 2.195 fft_wrap_pw1pw2_140 487 13.2 0.575 0.578 2.180 2.189 grid_integrate_task_list 119 12.3 2.106 2.157 2.106 2.157 ot_diis_step 108 11.5 0.014 0.014 2.078 2.078 mp_sum_l 7287 12.8 1.201 2.048 1.201 2.048 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.006 2.008 acc_transpose_blocks 9144 15.5 0.045 0.046 1.888 1.921 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.888 1.916 qs_energies_init_hamiltonians 11 5.9 0.021 0.021 1.825 1.826 cp_fm_cholesky_invert 11 10.9 1.748 1.752 1.748 1.752 fft3d_ps 1201 14.6 0.595 0.609 1.735 1.740 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.678 1.678 cp_fm_diag_elpa_base 50 14.0 1.529 1.581 1.676 1.676 grid_collocate_task_list 119 9.7 1.540 1.570 1.540 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.496 1.510 wfi_extrapolate 11 7.9 0.001 0.001 1.476 1.476 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.215 1.473 potential_pw2rs 119 12.3 0.014 0.014 1.398 1.401 make_images_data 4572 15.5 0.047 0.049 1.175 1.396 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.362 1.385 mp_alltoall_d11v 2130 13.8 1.264 1.314 1.264 1.314 make_basis_sm 11 9.8 0.000 0.000 1.293 1.294 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.135 1.151 parallel_gemm_fm 81 9.0 0.000 0.000 1.131 1.132 parallel_gemm_fm_cosma 81 10.0 1.131 1.132 1.131 1.132 acc_transpose_blocks_sync 27432 16.5 1.103 1.127 1.103 1.127 mp_alltoall_z22v 1201 16.6 1.004 1.026 1.004 1.026 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.994 1.021 jit_kernel_multiply 5 15.4 0.643 0.973 0.643 0.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.946 0.959 qs_create_task_list 11 7.9 0.006 0.008 0.945 0.955 generate_qs_task_list 11 8.9 0.368 0.387 0.938 0.949 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.038000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=721.090909, yerr=13.145719 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.497856E+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 1383689. 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 0.079 0.184 88.024 88.026 qs_mol_dyn_low 1 2.0 0.003 0.003 87.333 87.344 qs_forces 11 3.9 0.003 0.005 86.740 86.740 qs_energies 11 4.9 0.009 0.064 83.773 83.793 scf_env_do_scf 11 5.9 0.000 0.001 73.421 73.423 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.485 67.486 dbcsr_multiply_generic 2055 12.4 0.106 0.109 53.048 53.337 qs_scf_new_mos 99 7.5 0.000 0.001 49.581 49.716 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.580 49.716 ot_scf_mini 99 9.5 0.002 0.002 47.051 47.138 velocity_verlet 10 3.0 0.200 0.533 45.740 45.742 multiply_cannon 2055 13.4 0.185 0.190 44.067 44.642 multiply_cannon_loop 2055 14.4 1.796 1.837 42.960 43.597 ot_mini 99 10.5 0.001 0.001 27.831 27.925 qs_ot_get_derivative 99 11.5 0.001 0.001 20.977 21.063 multiply_cannon_multrec 49320 15.4 11.298 11.945 18.012 18.756 rebuild_ks_matrix 110 8.3 0.000 0.000 15.017 15.136 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.016 15.136 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.191 13.297 mp_waitall_1 220248 16.4 11.579 12.516 11.579 12.516 qs_ot_get_p 110 10.4 0.001 0.001 10.552 10.669 multiply_cannon_sync_h2d 49320 15.4 9.501 10.092 9.501 10.092 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.104 8.660 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.824 7.972 init_scf_run 11 5.9 0.000 0.001 7.928 7.928 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.928 7.928 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.231 7.801 apply_single 110 13.6 0.000 0.001 7.231 7.800 sum_up_and_integrate 110 10.3 0.001 0.003 7.382 7.396 integrate_v_rspace 110 11.3 0.003 0.003 7.357 7.377 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.162 7.187 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.823 6.958 calculate_rho_elec 110 8.6 0.021 0.024 6.822 6.958 dbcsr_mm_accdrv_process 87628 16.1 3.031 3.109 6.584 6.934 ot_diis_step 99 11.5 0.006 0.006 6.607 6.607 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.240 6.240 init_scf_loop 11 6.9 0.001 0.010 5.898 5.899 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.605 5.630 cp_fm_diag_elpa_base 48 14.0 5.589 5.611 5.603 5.628 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.391 5.434 mp_sum_l 6594 12.7 4.321 4.949 4.321 4.949 calculate_dm_sparse 110 9.5 0.001 0.001 4.423 4.523 make_m2s 4110 13.4 0.061 0.065 4.209 4.336 make_images 4110 14.4 0.178 0.191 4.113 4.241 wfi_extrapolate 11 7.9 0.001 0.001 4.216 4.216 density_rs2pw 110 9.6 0.004 0.005 3.774 3.918 multiply_cannon_metrocomm1 49320 15.4 0.066 0.068 2.616 3.728 prepare_preconditioner 11 7.9 0.000 0.001 3.645 3.666 make_preconditioner 11 8.9 0.000 0.000 3.645 3.666 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.660 3.664 pw_transfer 1331 11.6 0.056 0.066 3.520 3.619 calculate_first_density_matrix 1 7.0 0.000 0.002 3.606 3.616 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.430 3.530 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.370 3.444 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.392 3.441 grid_integrate_task_list 110 12.3 3.271 3.435 3.271 3.435 jit_kernel_multiply 13 15.9 3.273 3.358 3.273 3.358 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.185 3.233 acc_transpose_blocks 49320 15.4 0.222 0.230 2.890 2.963 potential_pw2rs 110 12.3 0.006 0.006 2.929 2.960 fft_wrap_pw1pw2_140 451 13.1 0.387 0.427 2.822 2.920 fft3d_ps 1111 14.6 0.796 0.884 2.727 2.798 mp_alltoall_d11v 2046 13.8 2.281 2.599 2.281 2.599 transfer_rs2pw 451 10.6 0.005 0.006 2.217 2.360 grid_collocate_task_list 110 9.6 2.157 2.327 2.157 2.327 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.183 2.224 mp_waitany 14300 13.8 1.947 2.217 1.947 2.217 mp_sum_d 3889 11.9 1.565 2.020 1.565 2.020 make_images_data 4110 15.4 0.044 0.048 1.879 2.018 cp_fm_cholesky_invert 11 10.9 1.997 2.002 1.997 2.002 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.959 1.975 transfer_pw2rs 451 13.1 0.006 0.007 1.946 1.969 hybrid_alltoall_any 4261 16.3 0.085 0.483 1.633 1.915 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=88.026000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.926087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 592.855040E+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 1436874. 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 0.079 0.268 73.930 73.975 qs_mol_dyn_low 1 2.0 0.003 0.004 72.967 73.401 qs_forces 11 3.9 0.003 0.003 72.888 72.889 qs_energies 11 4.9 0.019 0.139 69.311 69.314 scf_env_do_scf 11 5.9 0.004 0.029 60.125 60.136 scf_env_do_scf_inner_loop 99 6.5 0.082 0.635 51.901 51.912 dbcsr_multiply_generic 2055 12.4 0.115 0.118 39.066 39.264 velocity_verlet 10 3.0 0.001 0.002 36.957 36.958 qs_scf_new_mos 99 7.5 0.001 0.001 34.927 35.079 qs_scf_loop_do_ot 99 8.5 0.014 0.109 34.927 35.078 ot_scf_mini 99 9.5 0.008 0.047 33.228 33.377 multiply_cannon 2055 13.4 0.224 0.245 31.883 33.097 multiply_cannon_loop 2055 14.4 1.169 1.195 30.535 31.330 ot_mini 99 10.5 0.001 0.001 19.244 19.408 multiply_cannon_multrec 24660 15.4 6.997 8.925 14.108 15.856 rebuild_ks_matrix 110 8.3 0.000 0.000 13.930 14.036 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.930 14.035 qs_ot_get_derivative 99 11.5 0.001 0.001 13.449 13.608 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.177 12.280 mp_waitall_1 176588 16.5 7.993 10.401 7.993 10.401 init_scf_loop 11 6.9 0.004 0.037 8.181 8.194 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.355 8.078 qs_ot_get_p 110 10.4 0.010 0.073 7.506 7.672 dbcsr_mm_accdrv_process 52282 16.1 5.561 6.723 6.939 7.432 multiply_cannon_sync_h2d 24660 15.4 6.411 7.385 6.411 7.385 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.472 7.065 apply_single 110 13.6 0.000 0.001 6.472 7.065 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.751 6.836 calculate_rho_elec 110 8.6 0.039 0.047 6.751 6.836 sum_up_and_integrate 110 10.3 0.001 0.002 6.686 6.712 integrate_v_rspace 110 11.3 0.003 0.003 6.659 6.682 init_scf_run 11 5.9 0.000 0.001 6.555 6.556 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.555 6.556 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.301 6.028 prepare_preconditioner 11 7.9 0.000 0.001 5.798 5.813 make_preconditioner 11 8.9 0.006 0.045 5.798 5.813 ot_diis_step 99 11.5 0.010 0.010 5.746 5.746 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.381 5.552 qs_ot_p2m_diag 48 11.0 0.030 0.045 5.161 5.182 make_m2s 4110 13.4 0.057 0.060 4.528 4.935 make_images 4110 14.4 0.408 0.466 4.419 4.826 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.678 4.680 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.036 4.052 cp_fm_diag_elpa_base 48 14.0 3.984 3.999 4.033 4.048 pw_transfer 1331 11.6 0.066 0.072 3.874 4.021 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.767 3.915 density_rs2pw 110 9.6 0.004 0.005 3.418 3.681 wfi_extrapolate 11 7.9 0.001 0.001 3.578 3.578 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.443 3.517 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.497 3.501 grid_integrate_task_list 110 12.3 3.160 3.368 3.160 3.368 fft_wrap_pw1pw2_140 451 13.1 0.458 0.475 3.135 3.289 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.075 3.116 fft3d_ps 1111 14.6 1.115 1.347 2.864 3.036 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.027 mp_alltoall_d11v 2046 13.8 2.404 2.892 2.404 2.892 make_images_data 4110 15.4 0.050 0.054 2.434 2.880 calculate_first_density_matrix 1 7.0 0.003 0.021 2.850 2.854 hybrid_alltoall_any 4261 16.3 0.108 0.459 2.078 2.823 mp_sum_l 6594 12.7 2.161 2.753 2.161 2.753 cp_fm_cholesky_invert 11 10.9 2.693 2.700 2.693 2.700 potential_pw2rs 110 12.3 0.008 0.009 2.528 2.554 grid_collocate_task_list 110 9.6 2.174 2.324 2.174 2.324 acc_transpose_blocks 24660 15.4 0.118 0.122 2.106 2.151 jit_kernel_multiply 10 16.4 1.016 2.048 1.016 2.048 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.991 2.018 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.932 1.945 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 1.905 1.907 transfer_rs2pw 451 10.6 0.006 0.007 1.623 1.838 mp_allgather_i34 2055 14.4 0.628 1.693 0.628 1.693 mp_sum_d 3889 11.9 1.145 1.581 1.145 1.581 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.563 mp_alltoall_z22v 1111 16.6 1.477 1.558 1.477 1.558 dbcsr_complete_redistribute 325 12.2 0.255 0.346 1.267 1.544 build_core_hamiltonian_matrix_ 11 4.9 0.011 0.072 1.406 1.537 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.777 1.494 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.975000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.545455, yerr=6.800583 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 667.394048E+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 0.102 0.187 67.012 67.028 qs_mol_dyn_low 1 2.0 0.003 0.004 66.327 66.354 qs_forces 11 3.9 0.003 0.003 66.227 66.228 qs_energies 11 4.9 0.001 0.003 62.971 62.975 scf_env_do_scf 11 5.9 0.000 0.001 53.273 53.273 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 43.635 43.635 velocity_verlet 10 3.0 0.003 0.004 35.151 35.154 dbcsr_multiply_generic 2055 12.4 0.119 0.125 30.864 31.110 qs_scf_new_mos 99 7.5 0.001 0.001 27.886 28.000 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.885 28.000 ot_scf_mini 99 9.5 0.002 0.003 26.609 26.741 multiply_cannon 2055 13.4 0.212 0.220 23.841 25.099 multiply_cannon_loop 2055 14.4 0.817 0.864 22.556 23.724 ot_mini 99 10.5 0.001 0.001 14.852 14.980 rebuild_ks_matrix 110 8.3 0.000 0.000 12.734 12.901 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.733 12.901 multiply_cannon_multrec 16440 15.4 3.717 4.886 10.757 11.906 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.239 11.394 qs_ot_get_derivative 99 11.5 0.001 0.001 10.415 10.542 mp_waitall_1 139946 16.5 7.522 10.282 7.522 10.282 init_scf_loop 11 6.9 0.000 0.001 9.595 9.596 prepare_preconditioner 11 7.9 0.000 0.000 7.781 7.801 make_preconditioner 11 8.9 0.000 0.000 7.781 7.801 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.073 7.435 init_scf_run 11 5.9 0.000 0.001 7.209 7.209 scf_env_initial_rho_setup 11 6.9 0.002 0.003 7.209 7.209 dbcsr_mm_accdrv_process 34862 16.1 5.918 6.591 6.884 7.030 qs_ot_get_p 110 10.4 0.001 0.001 6.808 6.934 multiply_cannon_metrocomm3 16440 15.4 0.047 0.049 4.229 6.910 sum_up_and_integrate 110 10.3 0.001 0.002 6.838 6.854 integrate_v_rspace 110 11.3 0.003 0.003 6.812 6.828 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.411 6.418 calculate_rho_elec 110 8.6 0.058 0.059 6.410 6.418 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.940 5.373 apply_single 110 13.6 0.000 0.000 4.940 5.373 make_m2s 4110 13.4 0.050 0.052 4.787 5.177 make_images 4110 14.4 0.397 0.520 4.672 5.063 qs_ot_p2m_diag 48 11.0 0.042 0.045 4.852 4.855 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.438 4.439 ot_diis_step 99 11.5 0.011 0.011 4.403 4.403 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.420 4.047 pw_transfer 1331 11.6 0.066 0.072 3.860 3.872 multiply_cannon_sync_h2d 16440 15.4 3.224 3.842 3.224 3.842 calculate_first_density_matrix 1 7.0 0.000 0.000 3.812 3.814 density_rs2pw 110 9.6 0.004 0.005 3.632 3.782 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.753 3.765 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.750 3.758 cp_fm_diag_elpa_base 48 14.0 3.674 3.712 3.748 3.756 grid_integrate_task_list 110 12.3 3.180 3.442 3.180 3.442 cp_fm_cholesky_invert 11 10.9 3.294 3.302 3.294 3.302 wfi_extrapolate 11 7.9 0.001 0.001 3.300 3.300 make_images_data 4110 15.4 0.046 0.051 2.611 3.195 calculate_dm_sparse 110 9.5 0.001 0.001 3.164 3.195 fft_wrap_pw1pw2_140 451 13.1 0.577 0.583 3.085 3.097 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.093 3.095 hybrid_alltoall_any 4261 16.3 0.109 0.378 2.291 3.020 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.701 2.778 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.710 2.761 fft3d_ps 1111 14.6 1.089 1.113 2.698 2.712 potential_pw2rs 110 12.3 0.011 0.011 2.647 2.672 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.879 2.463 grid_collocate_task_list 110 9.6 2.217 2.385 2.217 2.385 mp_irecv_dv 48980 15.7 0.804 2.328 0.804 2.328 mp_sum_l 6594 12.7 1.640 2.270 1.640 2.270 mp_alltoall_d11v 2046 13.8 1.992 2.255 1.992 2.255 dbcsr_complete_redistribute 325 12.2 0.322 0.363 1.626 2.101 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.086 2.098 acc_transpose_blocks 16440 15.4 0.079 0.081 1.977 2.037 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 1.983 1.984 cp_fm_cholesky_decompose 22 10.9 1.918 1.936 1.918 1.936 transfer_rs2pw 451 10.6 0.005 0.006 1.766 1.933 cp_fm_upper_to_full 70 14.2 1.413 1.814 1.413 1.814 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.689 1.704 mp_allgather_i34 2055 14.4 0.595 1.662 0.595 1.662 transfer_pw2rs 451 13.1 0.005 0.005 1.587 1.610 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.070 1.537 parallel_gemm_fm 81 9.0 0.000 0.000 1.526 1.529 parallel_gemm_fm_cosma 81 10.0 1.526 1.528 1.526 1.528 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.375 1.496 mp_alltoall_z22v 1111 16.6 1.395 1.416 1.395 1.416 make_basis_sm 11 9.8 0.000 0.000 1.407 1.411 mp_waitany 17072 13.8 1.221 1.407 1.221 1.407 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.351 1.358 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=67.028000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.818182, yerr=7.802309 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 726.323200E+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 0.077 0.431 75.699 75.707 qs_mol_dyn_low 1 2.0 0.003 0.004 73.183 73.194 qs_forces 11 3.9 0.006 0.007 73.107 73.107 qs_energies 11 4.9 0.083 0.435 69.583 69.589 scf_env_do_scf 11 5.9 0.001 0.001 56.761 56.765 scf_env_do_scf_inner_loop 99 6.5 0.023 0.076 44.512 44.512 velocity_verlet 10 3.0 0.015 0.017 38.197 38.199 dbcsr_multiply_generic 2055 12.4 0.177 0.362 32.663 32.921 qs_scf_new_mos 99 7.5 0.001 0.001 29.619 29.737 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.619 29.737 ot_scf_mini 99 9.5 0.003 0.004 27.917 28.032 multiply_cannon 2055 13.4 0.239 0.266 24.792 25.787 multiply_cannon_loop 2055 14.4 1.441 1.493 23.286 23.946 ot_mini 99 10.5 0.001 0.001 15.675 15.788 multiply_cannon_multrec 24660 15.4 4.128 6.929 13.823 14.916 rebuild_ks_matrix 110 8.3 0.000 0.000 12.107 12.214 qs_ks_build_kohn_sham_matrix 110 9.3 0.024 0.035 12.107 12.213 init_scf_loop 11 6.9 0.021 0.042 12.201 12.204 qs_ot_get_derivative 99 11.5 0.001 0.001 11.445 11.558 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.706 10.800 dbcsr_mm_accdrv_process 52304 16.0 8.092 9.804 9.536 10.565 prepare_preconditioner 11 7.9 0.000 0.002 10.353 10.393 make_preconditioner 11 8.9 0.005 0.040 10.352 10.393 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.625 10.056 init_scf_run 11 5.9 0.000 0.001 8.019 8.020 scf_env_initial_rho_setup 11 6.9 0.016 0.019 8.019 8.020 qs_ot_get_p 110 10.4 0.001 0.001 7.067 7.238 mp_waitall_1 121746 16.5 4.547 6.408 4.547 6.408 sum_up_and_integrate 110 10.3 0.001 0.002 6.381 6.395 integrate_v_rspace 110 11.3 0.003 0.003 6.355 6.369 make_m2s 4110 13.4 0.060 0.063 5.871 6.270 make_images 4110 14.4 0.579 0.705 5.732 6.128 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.089 6.106 calculate_rho_elec 110 8.6 0.078 0.081 6.088 6.106 qs_ot_p2m_diag 48 11.0 0.062 0.123 5.033 5.049 cp_fm_upper_to_full 70 14.2 3.328 4.791 3.328 4.791 calculate_first_density_matrix 1 7.0 0.001 0.005 4.761 4.778 cp_dbcsr_syevd 48 12.0 0.010 0.061 4.441 4.449 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.087 4.229 apply_single 110 13.6 0.000 0.000 4.087 4.229 ot_diis_step 99 11.5 0.011 0.011 4.187 4.187 dbcsr_complete_redistribute 325 12.2 0.449 0.606 2.803 4.117 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.891 3.893 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.773 3.823 pw_transfer 1331 11.6 0.066 0.074 3.768 3.805 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.695 3.705 cp_fm_diag_elpa_base 48 14.0 3.538 3.599 3.692 3.703 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.662 3.701 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.522 3.586 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.221 3.537 qs_energies_init_hamiltonians 11 5.9 0.026 0.164 3.458 3.470 grid_integrate_task_list 110 12.3 3.277 3.454 3.277 3.454 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.384 3.449 density_rs2pw 110 9.6 0.004 0.004 3.130 3.324 calculate_dm_sparse 110 9.5 0.001 0.001 3.291 3.323 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.566 3.183 fft_wrap_pw1pw2_140 451 13.1 0.610 0.629 3.088 3.132 make_images_data 4110 15.4 0.050 0.054 2.747 3.117 cp_fm_cholesky_invert 11 10.9 3.077 3.087 3.077 3.087 hybrid_alltoall_any 4261 16.3 0.126 0.464 2.383 3.082 wfi_extrapolate 11 7.9 0.001 0.001 3.079 3.080 mp_alltoall_i22 605 13.7 1.710 2.849 1.710 2.849 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.713 2.802 acc_transpose_blocks 24660 15.4 0.117 0.119 2.513 2.602 transfer_rs2pw 451 10.6 0.005 0.006 2.282 2.571 fft3d_ps 1111 14.6 1.086 1.116 2.544 2.560 multiply_cannon_sync_h2d 24660 15.4 2.383 2.536 2.383 2.536 grid_collocate_task_list 110 9.6 2.274 2.437 2.274 2.437 qs_env_update_s_mstruct 11 6.9 0.000 0.001 2.204 2.426 mp_waitany 13376 13.8 2.076 2.391 2.076 2.391 mp_alltoall_d11v 2046 13.8 1.954 2.309 1.954 2.309 potential_pw2rs 110 12.3 0.012 0.013 2.183 2.193 jit_kernel_multiply 8 15.5 1.098 2.092 1.098 2.092 transfer_rs2pw_140 121 11.5 0.167 0.172 1.795 2.088 mp_sum_l 6594 12.7 1.431 2.007 1.431 2.007 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.858 1.891 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.838 1.856 cp_fm_cholesky_decompose 22 10.9 1.772 1.831 1.772 1.831 build_core_hamiltonian_matrix_ 11 4.9 0.007 0.056 1.621 1.791 build_core_hamiltonian_matrix 11 6.9 0.003 0.024 0.948 1.757 qs_init_subsys 1 2.0 0.051 0.368 1.244 1.687 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.656 1.668 make_basis_sm 11 9.8 0.000 0.000 1.575 1.577 mp_allgather_i34 2055 14.4 0.578 1.546 0.578 1.546 multiply_cannon_metrocomm4 20550 15.4 0.063 0.065 0.846 1.527 mp_sum_d 3887 11.9 1.035 1.525 1.035 1.525 parallel_gemm_fm 81 9.0 0.000 0.000 1.513 1.522 parallel_gemm_fm_cosma 81 10.0 1.513 1.522 1.513 1.522 acc_transpose_blocks_sync 73980 16.4 1.447 1.521 1.447 1.521 mp_sum_iv 258 11.4 0.967 1.515 0.967 1.515 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=75.707000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.363636, yerr=5.756778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 855.044096E+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 0.043 0.097 59.026 59.052 qs_mol_dyn_low 1 2.0 0.003 0.005 58.174 58.199 qs_forces 11 3.9 0.003 0.004 57.687 57.688 qs_energies 11 4.9 0.001 0.002 54.010 54.015 scf_env_do_scf 11 5.9 0.000 0.001 44.959 44.959 scf_env_do_scf_inner_loop 99 6.5 0.051 0.076 36.991 36.992 velocity_verlet 10 3.0 0.371 0.425 32.341 32.344 dbcsr_multiply_generic 2055 12.4 0.106 0.110 24.031 24.260 qs_scf_new_mos 99 7.5 0.001 0.001 22.058 22.133 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.058 22.133 ot_scf_mini 99 9.5 0.002 0.002 20.794 20.854 multiply_cannon 2055 13.4 0.238 0.251 18.132 19.449 multiply_cannon_loop 2055 14.4 0.603 0.623 16.814 17.216 rebuild_ks_matrix 110 8.3 0.000 0.000 11.824 11.841 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.824 11.841 ot_mini 99 10.5 0.001 0.001 11.217 11.265 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.524 10.541 multiply_cannon_multrec 8220 15.4 3.233 4.377 7.899 8.731 init_scf_loop 11 6.9 0.010 0.011 7.918 7.920 mp_waitall_1 103326 16.6 6.051 7.568 6.051 7.568 qs_ot_get_derivative 99 11.5 0.001 0.001 7.508 7.566 sum_up_and_integrate 110 10.3 0.001 0.002 6.377 6.392 integrate_v_rspace 110 11.3 0.003 0.003 6.350 6.366 prepare_preconditioner 11 7.9 0.000 0.000 6.206 6.215 make_preconditioner 11 8.9 0.000 0.000 6.206 6.215 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.179 6.192 calculate_rho_elec 110 8.6 0.114 0.114 6.178 6.192 init_scf_run 11 5.9 0.000 0.001 5.884 5.884 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.884 5.884 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.786 5.860 qs_ot_get_p 110 10.4 0.001 0.001 5.559 5.579 dbcsr_mm_accdrv_process 17442 15.9 3.191 4.426 4.527 5.552 make_m2s 4110 13.4 0.038 0.039 4.455 4.746 make_images 4110 14.4 0.649 0.713 4.325 4.617 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.906 4.352 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.043 4.048 pw_transfer 1331 11.6 0.066 0.070 4.024 4.037 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.916 3.934 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.735 3.781 apply_single 110 13.6 0.000 0.000 3.735 3.780 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.688 3.689 ot_diis_step 99 11.5 0.012 0.012 3.685 3.686 grid_integrate_task_list 110 12.3 3.365 3.529 3.365 3.529 density_rs2pw 110 9.6 0.004 0.004 3.135 3.358 fft_wrap_pw1pw2_140 451 13.1 0.776 0.790 3.331 3.349 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.058 3.065 cp_fm_diag_elpa_base 48 14.0 2.999 3.019 3.056 3.063 cp_fm_cholesky_invert 11 10.9 3.010 3.014 3.010 3.014 calculate_first_density_matrix 1 7.0 0.000 0.000 2.956 2.960 hybrid_alltoall_any 4261 16.3 0.200 0.855 2.391 2.867 wfi_extrapolate 11 7.9 0.001 0.001 2.828 2.828 make_images_data 4110 15.4 0.041 0.047 2.397 2.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.752 2.753 qs_energies_init_hamiltonians 11 5.9 0.006 0.015 2.692 2.693 calculate_dm_sparse 110 9.5 0.001 0.001 2.613 2.657 fft3d_ps 1111 14.6 1.135 1.160 2.583 2.592 grid_collocate_task_list 110 9.6 2.362 2.520 2.362 2.520 multiply_cannon_sync_h2d 8220 15.4 2.376 2.435 2.376 2.435 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.313 2.330 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.256 2.304 mp_alltoall_d11v 2046 13.8 1.904 2.157 1.904 2.157 potential_pw2rs 110 12.3 0.015 0.015 2.106 2.111 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.057 2.073 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 1.996 cp_fm_cholesky_decompose 22 10.9 1.745 1.770 1.745 1.770 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.633 1.636 dbcsr_complete_redistribute 325 12.2 0.606 0.631 1.538 1.636 jit_kernel_multiply 8 15.7 1.018 1.634 1.018 1.634 mp_allgather_i34 2055 14.4 0.481 1.625 0.481 1.625 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.503 1.614 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.556 1.570 acc_transpose_blocks 8220 15.4 0.040 0.042 1.498 1.550 transfer_rs2pw 451 10.6 0.005 0.005 1.238 1.445 qs_create_task_list 11 7.9 0.001 0.001 1.215 1.314 generate_qs_task_list 11 8.9 0.373 0.440 1.214 1.314 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.850 1.313 mp_waitany 9240 13.8 1.093 1.294 1.093 1.294 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.266 1.288 mp_alltoall_z22v 1111 16.6 1.213 1.231 1.213 1.231 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.052000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=806.636364, yerr=11.617172 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 1.397142E+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 0.068 0.115 95.221 95.242 qs_mol_dyn_low 1 2.0 0.003 0.003 94.484 94.509 qs_forces 11 3.9 0.003 0.003 94.214 94.215 qs_energies 11 4.9 0.001 0.002 90.018 90.020 scf_env_do_scf 11 5.9 0.001 0.001 78.311 78.312 velocity_verlet 10 3.0 0.002 0.002 59.080 59.111 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.439 48.441 dbcsr_multiply_generic 2055 12.4 0.120 0.124 31.805 31.958 init_scf_loop 11 6.9 0.001 0.001 29.790 29.792 qs_scf_new_mos 99 7.5 0.001 0.001 29.750 29.765 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.749 29.764 ot_scf_mini 99 9.5 0.002 0.002 27.916 27.929 prepare_preconditioner 11 7.9 0.000 0.000 27.697 27.703 make_preconditioner 11 8.9 0.000 0.000 27.697 27.703 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.830 27.144 multiply_cannon 2055 13.4 0.336 0.372 23.771 24.754 multiply_cannon_loop 2055 14.4 0.815 0.832 21.816 22.147 cp_fm_upper_to_full 70 14.2 12.598 18.183 12.598 18.183 ot_mini 99 10.5 0.001 0.001 15.438 15.452 rebuild_ks_matrix 110 8.3 0.000 0.000 14.494 14.518 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.493 14.517 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.152 13.174 dbcsr_complete_redistribute 325 12.2 1.014 1.038 7.600 10.858 multiply_cannon_multrec 8220 15.4 4.244 4.348 10.412 10.744 qs_ot_get_derivative 99 11.5 0.001 0.001 10.713 10.730 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.495 9.763 mp_waitall_1 84994 16.7 8.679 9.670 8.679 9.670 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.851 9.091 mp_alltoall_i22 605 13.7 5.458 8.720 5.458 8.720 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.829 7.880 calculate_rho_elec 110 8.6 0.225 0.225 7.828 7.880 sum_up_and_integrate 110 10.3 0.002 0.002 7.433 7.448 integrate_v_rspace 110 11.3 0.003 0.003 7.405 7.420 init_scf_run 11 5.9 0.000 0.001 7.330 7.330 scf_env_initial_rho_setup 11 6.9 0.002 0.003 7.330 7.330 qs_ot_get_p 110 10.4 0.001 0.001 6.988 7.004 make_m2s 4110 13.4 0.043 0.044 6.228 6.769 make_images 4110 14.4 0.884 0.937 6.038 6.579 dbcsr_mm_accdrv_process 11614 15.7 3.836 4.102 6.020 6.294 cp_fm_cholesky_invert 11 10.9 6.218 6.223 6.218 6.223 pw_transfer 1331 11.6 0.075 0.075 5.640 5.648 multiply_cannon_metrocomm3 8220 15.4 0.020 0.020 5.082 5.587 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.523 5.531 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.868 5.307 apply_single 110 13.6 0.000 0.000 4.868 5.306 qs_ot_p2m_diag 48 11.0 0.151 0.155 5.209 5.218 fft_wrap_pw1pw2_140 451 13.1 1.274 1.278 4.711 4.717 ot_diis_step 99 11.5 0.015 0.015 4.697 4.697 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.664 4.665 hybrid_alltoall_any 4261 16.3 0.263 0.565 3.302 4.289 make_images_data 4110 15.4 0.046 0.049 3.392 4.281 density_rs2pw 110 9.6 0.004 0.004 4.165 4.192 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.942 3.943 cp_fm_diag_elpa_base 48 14.0 3.377 3.596 3.940 3.940 calculate_dm_sparse 110 9.5 0.001 0.001 3.824 3.860 grid_integrate_task_list 110 12.3 3.700 3.793 3.700 3.793 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.332 3.784 qs_energies_init_hamiltonians 11 5.9 0.004 0.009 3.709 3.710 calculate_first_density_matrix 1 7.0 0.000 0.000 3.613 3.623 fft3d_ps 1111 14.6 1.301 1.306 3.581 3.588 wfi_extrapolate 11 7.9 0.001 0.001 3.578 3.579 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.177 3.180 multiply_cannon_sync_h2d 8220 15.4 3.142 3.169 3.142 3.169 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.902 2.911 potential_pw2rs 110 12.3 0.021 0.021 2.811 2.816 grid_collocate_task_list 110 9.6 2.686 2.706 2.686 2.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.611 2.619 mp_alltoall_d11v 2046 13.8 2.494 2.568 2.494 2.568 cp_fm_cholesky_decompose 22 10.9 2.486 2.499 2.486 2.499 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.307 2.333 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.219 2.276 jit_kernel_multiply 10 15.2 1.985 2.243 1.985 2.243 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.171 acc_transpose_blocks 8220 15.4 0.041 0.042 1.975 2.043 mp_alltoall_z22v 1111 16.6 2.028 2.036 2.028 2.036 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.944 1.946 qs_create_task_list 11 7.9 0.000 0.001 1.891 1.937 generate_qs_task_list 11 8.9 0.731 0.785 1.891 1.936 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.886 1.915 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=95.242000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1231.363636, yerr=64.330500 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 630.820864E+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 3265480. 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 0.052 0.107 212.068 212.081 qs_mol_dyn_low 1 2.0 0.003 0.004 211.563 211.590 qs_forces 11 3.9 0.005 0.005 211.348 211.349 qs_energies 11 4.9 0.002 0.002 205.608 205.629 scf_env_do_scf 11 5.9 0.001 0.002 188.622 188.626 scf_env_do_scf_inner_loop 117 6.6 0.016 0.030 167.434 167.436 qs_scf_new_mos 117 7.6 0.001 0.001 127.424 127.676 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.424 127.676 dbcsr_multiply_generic 2507 12.6 0.185 0.195 126.751 127.230 velocity_verlet 10 3.0 0.073 0.137 126.412 126.423 ot_scf_mini 117 9.6 0.003 0.003 120.664 120.847 multiply_cannon 2507 13.6 0.240 0.250 102.308 103.692 multiply_cannon_loop 2507 14.6 2.413 2.467 100.011 101.484 ot_mini 117 10.6 0.001 0.001 67.104 67.334 multiply_cannon_multrec 60168 15.6 31.792 33.663 41.672 43.575 qs_ot_get_derivative 117 11.6 0.001 0.001 42.213 42.376 rebuild_ks_matrix 128 8.3 0.001 0.001 33.952 34.191 qs_ks_build_kohn_sham_matrix 128 9.3 0.022 0.075 33.951 34.190 mp_waitall_1 267128 16.5 29.329 32.636 29.329 32.636 qs_ot_get_p 128 10.4 0.001 0.001 31.206 31.469 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.424 30.668 multiply_cannon_sync_h2d 60168 15.6 26.214 28.108 26.214 28.108 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.398 24.895 apply_single 128 13.6 0.001 0.001 24.398 24.895 ot_diis_step 117 11.6 0.008 0.008 24.605 24.607 qs_ot_p2m_diag 83 11.4 0.079 0.092 24.533 24.592 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.764 21.765 init_scf_loop 11 6.9 0.000 0.001 21.104 21.106 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.077 20.231 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.658 18.696 cp_fm_diag_elpa_base 83 14.4 18.574 18.625 18.653 18.691 multiply_cannon_metrocomm3 60168 15.6 0.118 0.122 16.186 18.005 prepare_preconditioner 11 7.9 0.000 0.000 16.465 16.517 make_preconditioner 11 8.9 0.000 0.000 16.465 16.517 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.699 15.890 make_m2s 5014 13.6 0.105 0.113 14.142 14.472 sum_up_and_integrate 128 10.3 0.002 0.004 14.287 14.344 make_images 5014 14.6 0.400 0.419 13.959 14.299 integrate_v_rspace 128 11.3 0.004 0.005 14.229 14.287 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.526 13.683 calculate_rho_elec 128 8.7 0.045 0.064 13.526 13.683 init_scf_run 11 5.9 0.000 0.001 12.815 12.816 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.815 12.816 mp_sum_l 7950 12.9 9.059 10.315 9.059 10.315 dbcsr_mm_accdrv_process 124484 16.2 4.688 4.845 9.452 9.968 cp_fm_cholesky_invert 11 10.9 9.461 9.469 9.461 9.469 wfi_extrapolate 11 7.9 0.001 0.001 9.195 9.195 calculate_dm_sparse 128 9.5 0.001 0.001 8.586 8.689 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.380 8.522 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.177 8.329 pw_transfer 1547 11.6 0.075 0.098 7.929 8.127 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.725 7.922 multiply_cannon_metrocomm1 60168 15.6 0.095 0.098 6.357 7.825 make_images_data 5014 15.6 0.068 0.074 6.949 7.823 density_rs2pw 128 9.7 0.006 0.007 7.125 7.766 grid_integrate_task_list 128 12.3 7.043 7.431 7.043 7.431 hybrid_alltoall_any 5200 16.5 0.295 2.274 6.056 7.134 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.955 6.967 fft_wrap_pw1pw2_140 523 13.2 1.122 1.175 6.669 6.872 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.055 6.146 fft3d_ps 1291 14.7 2.193 2.827 5.735 6.079 mp_alltoall_d11v 2415 14.1 4.532 5.853 4.532 5.853 grid_collocate_task_list 128 9.7 4.841 5.213 4.841 5.213 cp_fm_cholesky_decompose 22 10.9 4.805 4.819 4.805 4.819 potential_pw2rs 128 12.3 0.009 0.010 4.725 4.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=212.081000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.000000, yerr=6.768913 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 841.097216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58227. MP_Allreduce 11081 960. MP_Sync 86 MP_Alltoall 1955 5813279. 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 0.075 0.387 195.477 195.489 qs_mol_dyn_low 1 2.0 0.005 0.019 194.035 194.051 qs_forces 11 3.9 0.004 0.005 193.912 193.914 qs_energies 11 4.9 0.004 0.013 187.104 187.116 scf_env_do_scf 11 5.9 0.001 0.001 170.138 170.148 scf_env_do_scf_inner_loop 116 6.6 0.016 0.037 136.283 136.285 velocity_verlet 10 3.0 0.010 0.019 120.789 120.791 dbcsr_multiply_generic 2485 12.5 0.205 0.210 97.307 98.627 qs_scf_new_mos 116 7.6 0.001 0.001 97.536 98.157 qs_scf_loop_do_ot 116 8.6 0.001 0.001 97.536 98.156 ot_scf_mini 116 9.6 0.004 0.004 92.746 93.440 multiply_cannon 2485 13.5 0.499 0.562 76.850 81.446 multiply_cannon_loop 2485 14.5 1.565 1.632 73.778 76.679 ot_mini 116 10.6 0.001 0.001 50.888 51.520 mp_waitall_1 212858 16.6 24.687 37.970 24.687 37.970 multiply_cannon_multrec 29820 15.5 20.701 24.407 31.218 35.355 init_scf_loop 11 6.9 0.000 0.001 33.760 33.763 rebuild_ks_matrix 127 8.3 0.001 0.001 32.200 32.938 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 32.199 32.937 qs_ot_get_derivative 116 11.6 0.001 0.002 29.025 29.698 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.936 29.610 prepare_preconditioner 11 7.9 0.000 0.000 29.427 29.500 make_preconditioner 11 8.9 0.000 0.000 29.427 29.500 multiply_cannon_metrocomm3 29820 15.5 0.097 0.104 16.112 29.008 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.070 28.626 qs_ot_get_p 127 10.4 0.001 0.001 24.103 24.873 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.944 23.106 apply_single 127 13.6 0.001 0.001 21.944 23.105 ot_diis_step 116 11.6 0.014 0.015 21.683 21.685 multiply_cannon_sync_h2d 29820 15.5 17.727 19.811 17.727 19.811 qs_ot_p2m_diag 82 11.4 0.187 0.217 19.072 19.111 cp_dbcsr_syevd 82 12.4 0.006 0.007 17.827 17.829 cp_fm_cholesky_invert 11 10.9 17.252 17.264 17.252 17.264 make_m2s 4970 13.5 0.089 0.094 14.325 15.579 make_images 4970 14.5 1.168 1.366 14.115 15.369 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.457 14.491 cp_fm_diag_elpa_base 82 14.4 14.195 14.294 14.451 14.480 sum_up_and_integrate 127 10.3 0.002 0.004 14.169 14.197 integrate_v_rspace 127 11.3 0.003 0.004 14.109 14.140 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.699 13.727 calculate_rho_elec 127 8.7 0.087 0.103 13.699 13.726 init_scf_run 11 5.9 0.000 0.001 12.085 12.086 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.085 12.086 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.323 11.808 multiply_cannon_metrocomm4 27335 15.5 0.105 0.119 3.863 10.878 dbcsr_mm_accdrv_process 61726 16.2 5.487 6.138 9.974 10.555 mp_irecv_dv 68888 16.3 3.658 10.472 3.658 10.472 make_images_data 4970 15.5 0.067 0.076 8.388 9.846 hybrid_alltoall_any 5155 16.4 0.352 1.544 7.186 9.177 pw_transfer 1535 11.6 0.084 0.099 8.754 8.821 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.532 8.605 wfi_extrapolate 11 7.9 0.001 0.001 8.475 8.475 density_rs2pw 127 9.7 0.006 0.007 7.322 7.881 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.853 7.599 grid_integrate_task_list 127 12.3 7.111 7.574 7.111 7.574 fft_wrap_pw1pw2_140 519 13.2 1.204 1.238 7.435 7.526 cp_fm_cholesky_decompose 22 10.9 7.194 7.274 7.194 7.274 mp_sum_l 7884 12.9 4.625 6.630 4.625 6.630 calculate_dm_sparse 127 9.5 0.001 0.001 6.450 6.596 fft3d_ps 1281 14.7 2.820 2.995 6.112 6.167 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.134 6.143 mp_alltoall_d11v 2401 14.1 4.567 5.664 4.567 5.664 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.387 5.465 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.207 5.385 grid_collocate_task_list 127 9.7 4.962 5.350 4.962 5.350 potential_pw2rs 127 12.3 0.016 0.019 4.719 4.744 mp_allgather_i34 2485 14.5 1.595 4.688 1.595 4.688 dbcsr_complete_redistribute 393 12.7 0.767 0.852 3.309 4.155 mp_sum_d 4451 12.1 2.685 3.970 2.685 3.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.489000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.909091, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 971.501568E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58094. MP_Allreduce 11102 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.158 0.648 182.630 182.631 qs_mol_dyn_low 1 2.0 0.003 0.003 180.905 180.918 qs_forces 11 3.9 0.004 0.005 180.806 180.809 qs_energies 11 4.9 0.027 0.212 174.017 174.116 scf_env_do_scf 11 5.9 0.004 0.025 156.848 156.859 scf_env_do_scf_inner_loop 117 6.6 0.067 0.457 121.068 121.071 velocity_verlet 10 3.0 0.002 0.006 114.180 114.182 dbcsr_multiply_generic 2507 12.6 0.181 0.187 83.058 84.050 qs_scf_new_mos 117 7.6 0.001 0.001 83.069 83.403 qs_scf_loop_do_ot 117 8.6 0.005 0.033 83.068 83.402 ot_scf_mini 117 9.6 0.004 0.006 78.805 79.194 multiply_cannon 2507 13.6 0.508 0.527 63.137 67.221 multiply_cannon_loop 2507 14.6 1.131 1.204 59.881 62.633 ot_mini 117 10.6 0.001 0.001 43.143 43.550 init_scf_loop 11 6.9 0.008 0.063 35.668 35.679 mp_waitall_1 170520 16.6 24.912 33.770 24.912 33.770 prepare_preconditioner 11 7.9 0.000 0.001 31.245 31.297 make_preconditioner 11 8.9 0.002 0.014 31.245 31.297 rebuild_ks_matrix 128 8.3 0.001 0.001 30.540 31.027 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.050 30.539 31.027 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.908 30.256 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.718 28.185 multiply_cannon_multrec 20056 15.6 13.091 16.126 22.710 25.728 multiply_cannon_metrocomm3 20056 15.6 0.063 0.066 15.432 24.549 qs_ot_get_derivative 117 11.6 0.002 0.002 23.625 24.021 qs_ot_get_p 128 10.4 0.001 0.001 20.979 21.432 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.554 20.578 apply_single 128 13.6 0.001 0.001 19.554 20.578 ot_diis_step 117 11.6 0.018 0.018 19.407 19.408 qs_ot_p2m_diag 83 11.4 0.271 0.320 16.558 16.565 make_m2s 5014 13.6 0.080 0.085 14.750 16.238 make_images 5014 14.6 1.157 1.254 14.517 15.999 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.448 15.456 multiply_cannon_sync_h2d 20056 15.6 13.681 15.393 13.681 15.393 cp_fm_cholesky_invert 11 10.9 14.723 14.735 14.723 14.735 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.261 14.340 calculate_rho_elec 128 8.7 0.130 0.145 14.260 14.340 sum_up_and_integrate 128 10.3 0.002 0.002 14.139 14.161 integrate_v_rspace 128 11.3 0.003 0.004 14.079 14.103 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.047 12.066 cp_fm_diag_elpa_base 83 14.4 11.627 11.772 12.043 12.063 init_scf_run 11 5.9 0.000 0.001 11.461 11.461 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.461 11.461 make_images_data 5014 15.6 0.063 0.073 8.709 10.632 hybrid_alltoall_any 5200 16.5 0.450 2.030 7.699 9.995 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.188 9.461 dbcsr_mm_accdrv_process 41502 16.2 5.912 6.211 9.080 9.214 multiply_cannon_metrocomm4 17549 15.6 0.067 0.076 3.468 9.173 pw_transfer 1547 11.6 0.085 0.104 8.946 9.063 mp_irecv_dv 50230 16.2 3.339 8.923 3.339 8.923 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.723 8.845 grid_integrate_task_list 128 12.3 7.292 7.831 7.292 7.831 fft_wrap_pw1pw2_140 523 13.2 1.299 1.332 7.669 7.805 wfi_extrapolate 11 7.9 0.001 0.001 7.587 7.587 cp_fm_cholesky_decompose 22 10.9 7.453 7.543 7.453 7.543 density_rs2pw 128 9.7 0.006 0.006 7.201 7.524 cp_fm_upper_to_full 105 14.8 5.631 7.115 5.631 7.115 dbcsr_complete_redistribute 395 12.7 1.177 1.227 4.782 6.491 calculate_dm_sparse 128 9.5 0.001 0.001 6.153 6.264 fft3d_ps 1291 14.7 2.760 2.999 6.101 6.186 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.146 5.775 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.704 5.710 mp_alltoall_d11v 2415 14.1 4.876 5.560 4.876 5.560 grid_collocate_task_list 128 9.7 5.168 5.548 5.168 5.548 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.520 5.226 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.709 4.864 mp_sum_l 7950 12.9 3.348 4.740 3.348 4.740 mp_allgather_i34 2507 14.6 1.747 4.699 1.747 4.699 potential_pw2rs 128 12.3 0.020 0.022 4.553 4.565 qs_energies_init_hamiltonians 11 5.9 0.020 0.147 4.340 4.341 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.316 4.006 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.945 3.976 mp_alltoall_i22 716 14.1 2.015 3.892 2.015 3.892 mp_sum_d 4472 12.1 2.785 3.659 2.785 3.659 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=182.631000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=913.818182, yerr=15.764276 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.174864E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669921808880 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509633. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.141 196.367 196.376 qs_mol_dyn_low 1 2.0 0.003 0.003 195.253 195.279 qs_forces 11 3.9 0.004 0.004 194.545 194.553 qs_energies 11 4.9 0.010 0.065 187.431 187.442 scf_env_do_scf 11 5.9 0.001 0.001 168.714 168.727 velocity_verlet 10 3.0 0.021 0.025 128.390 128.418 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 121.667 121.669 qs_scf_new_mos 118 7.6 0.001 0.001 84.649 84.941 qs_scf_loop_do_ot 118 8.6 0.001 0.001 84.648 84.940 dbcsr_multiply_generic 2529 12.6 0.195 0.203 82.431 83.073 ot_scf_mini 118 9.6 0.003 0.004 80.004 80.312 multiply_cannon 2529 13.6 0.553 0.582 56.931 59.328 multiply_cannon_loop 2529 14.6 1.880 1.946 53.189 54.796 init_scf_loop 11 6.9 0.002 0.016 46.917 46.918 ot_mini 118 10.6 0.001 0.001 43.890 44.161 prepare_preconditioner 11 7.9 0.000 0.002 42.801 42.833 make_preconditioner 11 8.9 0.000 0.000 42.801 42.833 make_full_inverse_cholesky 11 9.9 0.010 0.023 36.441 41.467 multiply_cannon_multrec 30348 15.6 13.921 19.163 27.440 32.312 rebuild_ks_matrix 129 8.3 0.001 0.001 29.780 30.111 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.779 30.110 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.926 27.220 mp_waitall_1 149172 16.7 17.547 27.172 17.547 27.172 qs_ot_get_derivative 118 11.6 0.002 0.002 24.188 24.490 qs_ot_get_p 129 10.4 0.001 0.001 22.054 22.384 make_m2s 5058 13.6 0.096 0.098 20.947 21.661 make_images 5058 14.6 1.954 2.267 20.647 21.359 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.051 19.587 apply_single 129 13.6 0.001 0.001 19.051 19.587 ot_diis_step 118 11.6 0.018 0.021 19.567 19.569 qs_ot_p2m_diag 84 11.4 0.346 0.394 17.661 17.713 cp_fm_upper_to_full 106 14.8 11.465 16.907 11.465 16.907 cp_fm_cholesky_invert 11 10.9 16.558 16.567 16.558 16.567 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.206 16.208 multiply_cannon_metrocomm3 30348 15.6 0.051 0.054 6.098 14.401 sum_up_and_integrate 129 10.3 0.002 0.003 14.286 14.310 integrate_v_rspace 129 11.3 0.003 0.004 14.225 14.251 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.025 14.089 calculate_rho_elec 129 8.7 0.174 0.190 14.024 14.088 dbcsr_mm_accdrv_process 62780 16.2 8.809 9.730 13.082 13.549 dbcsr_complete_redistribute 397 12.7 1.552 1.786 9.513 13.300 cp_fm_diag_elpa 84 13.4 0.000 0.000 12.796 12.812 cp_fm_diag_elpa_base 84 14.4 11.795 12.145 12.789 12.803 make_images_data 5058 15.6 0.068 0.076 11.348 12.663 init_scf_run 11 5.9 0.000 0.001 12.060 12.062 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.060 12.061 copy_fm_to_dbcsr 210 11.7 0.001 0.002 8.061 11.870 multiply_cannon_sync_h2d 30348 15.6 10.483 11.631 10.483 11.631 hybrid_alltoall_any 5245 16.5 0.536 2.217 9.939 11.498 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.201 10.418 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.337 10.022 mp_alltoall_i22 720 14.1 5.603 9.324 5.603 9.324 pw_transfer 1559 11.6 0.086 0.102 9.193 9.273 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.965 9.051 fft_wrap_pw1pw2_140 527 13.2 1.456 1.491 7.922 8.026 grid_integrate_task_list 129 12.3 7.565 7.954 7.565 7.954 cp_fm_cholesky_decompose 22 10.9 7.776 7.857 7.776 7.857 wfi_extrapolate 11 7.9 0.001 0.001 7.770 7.770 density_rs2pw 129 9.7 0.006 0.006 7.000 7.455 multiply_cannon_metrocomm4 25290 15.6 0.086 0.097 2.761 7.166 mp_irecv_dv 76751 16.2 2.604 6.884 2.604 6.884 calculate_dm_sparse 129 9.5 0.001 0.001 6.618 6.731 fft3d_ps 1301 14.7 2.870 2.938 6.123 6.173 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.056 6.135 mp_alltoall_d11v 2429 14.1 5.096 6.050 5.096 6.050 grid_collocate_task_list 129 9.7 5.331 5.719 5.331 5.719 qs_energies_init_hamiltonians 11 5.9 0.004 0.014 5.044 5.046 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.637 4.729 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.541 4.635 potential_pw2rs 129 12.3 0.023 0.023 4.470 4.481 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.316 4.385 calculate_first_density_matrix 1 7.0 0.000 0.001 4.090 4.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=196.376000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1101.818182, yerr=27.643241 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 5.820059E+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 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 1.530143E+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 716108613552 ------------------------------------------------------------------------------- - - - 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 3983 58474. MP_Allreduce 11035 1168. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.119 183.175 183.178 qs_mol_dyn_low 1 2.0 0.003 0.003 182.559 182.578 qs_forces 11 3.9 0.004 0.004 182.078 182.080 qs_energies 11 4.9 0.038 0.053 174.603 174.609 scf_env_do_scf 11 5.9 0.001 0.001 154.873 154.881 velocity_verlet 10 3.0 0.386 0.395 117.248 117.252 scf_env_do_scf_inner_loop 116 6.6 0.080 0.098 116.026 116.026 qs_scf_new_mos 116 7.6 0.001 0.001 78.194 78.262 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.193 78.262 dbcsr_multiply_generic 2485 12.5 0.181 0.192 74.826 75.072 ot_scf_mini 116 9.6 0.003 0.004 73.763 73.794 multiply_cannon 2485 13.5 0.556 0.594 55.287 59.054 multiply_cannon_loop 2485 14.5 0.802 0.835 51.788 52.354 ot_mini 116 10.6 0.001 0.001 39.172 39.205 init_scf_loop 11 6.9 0.013 0.014 38.687 38.689 prepare_preconditioner 11 7.9 0.000 0.000 34.217 34.243 make_preconditioner 11 8.9 0.000 0.000 34.217 34.243 make_full_inverse_cholesky 11 9.9 0.015 0.028 32.071 32.387 mp_waitall_1 124680 16.7 25.876 32.068 25.876 32.068 rebuild_ks_matrix 127 8.3 0.001 0.001 29.698 29.727 qs_ks_build_kohn_sham_matrix 127 9.3 0.078 0.082 29.697 29.727 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.098 27.120 qs_ot_get_derivative 116 11.6 0.001 0.002 22.186 22.215 qs_ot_get_p 127 10.4 0.001 0.001 21.916 21.991 multiply_cannon_multrec 9940 15.5 10.370 14.705 18.552 21.743 cp_fm_cholesky_invert 11 10.9 19.561 19.567 19.561 19.567 multiply_cannon_metrocomm3 9940 15.5 0.026 0.027 12.437 19.427 qs_ot_p2m_diag 82 11.4 0.490 0.495 17.825 17.842 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.078 17.233 apply_single 127 13.6 0.001 0.001 17.077 17.233 ot_diis_step 116 11.6 0.020 0.020 16.919 16.919 make_m2s 4970 13.5 0.065 0.070 15.552 16.586 cp_dbcsr_syevd 82 12.4 0.005 0.006 16.561 16.562 make_images 4970 14.5 2.148 2.555 15.248 16.284 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.815 14.835 calculate_rho_elec 127 8.7 0.254 0.265 14.814 14.835 sum_up_and_integrate 127 10.3 0.002 0.002 14.708 14.759 integrate_v_rspace 127 11.3 0.004 0.004 14.648 14.701 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.326 13.337 cp_fm_diag_elpa_base 82 14.4 13.049 13.161 13.322 13.333 init_scf_run 11 5.9 0.000 0.001 12.675 12.675 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.675 12.675 multiply_cannon_sync_h2d 9940 15.5 10.774 11.083 10.774 11.083 make_images_data 4970 15.5 0.056 0.065 8.794 10.513 hybrid_alltoall_any 5155 16.4 0.835 3.750 8.535 9.927 pw_transfer 1535 11.6 0.085 0.094 9.646 9.679 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.426 9.465 cp_fm_cholesky_decompose 22 10.9 8.887 8.989 8.887 8.989 dbcsr_mm_accdrv_process 20590 16.1 3.364 4.843 7.836 8.800 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.563 8.582 fft_wrap_pw1pw2_140 519 13.2 1.786 1.827 8.183 8.227 grid_integrate_task_list 127 12.3 7.700 8.081 7.700 8.081 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 5.031 7.927 wfi_extrapolate 11 7.9 0.001 0.001 7.719 7.719 density_rs2pw 127 9.7 0.006 0.006 7.349 7.583 calculate_dm_sparse 127 9.5 0.001 0.001 6.676 6.743 fft3d_ps 1281 14.7 2.750 2.820 6.186 6.226 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.104 6.114 dbcsr_complete_redistribute 393 12.7 2.146 2.254 5.633 6.055 grid_collocate_task_list 127 9.7 5.529 5.866 5.529 5.866 mp_alltoall_d11v 2401 14.1 5.176 5.721 5.176 5.721 qs_energies_init_hamiltonians 11 5.9 0.008 0.011 5.353 5.353 mp_allgather_i34 2485 14.5 1.672 5.312 1.672 5.312 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.804 4.846 potential_pw2rs 127 12.3 0.026 0.026 4.780 4.798 calculate_first_density_matrix 1 7.0 0.000 0.000 4.686 4.689 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.382 4.464 multiply_cannon_metrocomm4 7455 15.5 0.027 0.029 1.759 4.079 mp_irecv_dv 28618 15.9 1.719 4.003 1.719 4.003 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.631 3.992 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.587 3.898 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.655 3.748 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.709 3.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=183.178000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1431.636364, yerr=34.150130 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_performance_tests/20/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 11.696234E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.196383E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350232272 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 58881. MP_Allreduce 11127 1502. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.063 293.628 293.635 qs_mol_dyn_low 1 2.0 0.003 0.004 292.779 292.796 qs_forces 11 3.9 0.018 0.019 292.676 292.679 qs_energies 11 4.9 0.002 0.002 283.866 283.874 scf_env_do_scf 11 5.9 0.001 0.001 259.875 259.878 velocity_verlet 10 3.0 0.003 0.010 210.660 210.667 scf_env_do_scf_inner_loop 117 6.6 0.017 0.021 136.084 136.086 init_scf_loop 11 6.9 0.001 0.001 123.515 123.518 prepare_preconditioner 11 7.9 0.000 0.000 118.735 118.752 make_preconditioner 11 8.9 0.000 0.000 118.735 118.752 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.087 115.859 qs_scf_new_mos 117 7.6 0.001 0.001 92.223 92.287 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.222 92.286 ot_scf_mini 117 9.6 0.004 0.004 87.283 87.312 dbcsr_multiply_generic 2507 12.6 0.214 0.226 84.187 84.693 cp_fm_upper_to_full 105 14.8 51.884 74.483 51.884 74.483 multiply_cannon 2507 13.6 0.689 0.732 60.993 62.039 multiply_cannon_loop 2507 14.6 1.054 1.064 56.921 58.352 ot_mini 117 10.6 0.001 0.001 45.357 45.391 dbcsr_complete_redistribute 395 12.7 3.999 4.034 29.557 41.969 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.057 38.457 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.605 35.850 rebuild_ks_matrix 128 8.3 0.001 0.001 33.904 33.938 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.019 33.904 33.938 cp_fm_cholesky_invert 11 10.9 33.900 33.907 33.900 33.907 mp_alltoall_i22 716 14.1 21.330 33.872 21.330 33.872 mp_waitall_1 103674 16.8 28.413 32.991 28.413 32.991 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.487 31.528 qs_ot_get_p 128 10.4 0.001 0.001 26.711 26.768 qs_ot_get_derivative 117 11.6 0.002 0.002 25.601 25.631 qs_ot_p2m_diag 83 11.4 0.879 0.885 22.580 22.614 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 19.422 20.769 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.741 20.743 make_m2s 5014 13.6 0.077 0.082 18.404 19.904 ot_diis_step 117 11.6 0.022 0.022 19.721 19.722 multiply_cannon_multrec 10028 15.6 10.553 12.618 19.489 19.603 make_images 5014 14.6 3.063 3.249 17.924 19.426 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.196 19.361 apply_single 128 13.6 0.001 0.001 19.196 19.360 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.487 17.487 cp_fm_diag_elpa_base 83 14.4 13.082 14.689 17.482 17.483 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.552 16.566 calculate_rho_elec 128 8.7 0.480 0.481 16.551 16.565 sum_up_and_integrate 128 10.3 0.002 0.002 16.069 16.157 integrate_v_rspace 128 11.3 0.004 0.005 16.006 16.095 multiply_cannon_sync_h2d 10028 15.6 14.297 14.310 14.297 14.310 init_scf_run 11 5.9 0.000 0.001 13.654 13.655 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.654 13.654 hybrid_alltoall_any 5200 16.5 1.315 3.052 10.375 12.576 make_images_data 5014 15.6 0.066 0.072 10.247 12.474 pw_transfer 1547 11.6 0.095 0.096 11.413 11.426 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 11.174 11.187 dbcsr_mm_accdrv_process 20762 16.1 4.407 6.582 8.686 10.782 fft_wrap_pw1pw2_140 523 13.2 3.020 3.059 9.936 9.947 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.921 9.945 cp_fm_cholesky_decompose 22 10.9 9.166 9.186 9.166 9.186 wfi_extrapolate 11 7.9 0.001 0.001 9.001 9.001 grid_integrate_task_list 128 12.3 8.555 8.742 8.555 8.742 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.246 8.246 density_rs2pw 128 9.7 0.005 0.006 7.541 7.596 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.172 7.187 mp_alltoall_d11v 2415 14.1 6.921 7.056 6.921 7.056 calculate_dm_sparse 128 9.5 0.001 0.001 6.939 7.013 grid_collocate_task_list 128 9.7 6.410 6.492 6.410 6.492 fft3d_ps 1291 14.7 2.851 2.859 6.447 6.482 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.143 6.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.635000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2881.545455, yerr=170.636969 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 1.259639E+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 16074971. 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 0.035 0.073 87.534 87.556 qs_energies 1 2.0 0.000 0.001 86.854 86.866 ls_scf 1 3.0 0.000 0.001 85.879 85.890 dbcsr_multiply_generic 111 6.7 0.015 0.022 73.105 73.298 multiply_cannon 111 7.7 0.018 0.023 56.174 57.243 multiply_cannon_loop 111 8.7 0.227 0.244 52.750 54.060 ls_scf_main 1 4.0 0.007 0.110 52.619 52.633 density_matrix_trs4 2 5.0 0.024 0.230 47.018 47.133 ls_scf_init_scf 1 4.0 0.000 0.001 29.532 29.547 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.224 28.284 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.947 25.968 mp_waitall_1 11031 10.9 22.331 25.329 22.331 25.329 multiply_cannon_multrec 2664 9.7 8.171 8.837 15.636 17.492 multiply_cannon_sync_h2d 2664 9.7 13.704 15.222 13.704 15.222 make_m2s 222 7.7 0.009 0.013 13.113 13.574 make_images 222 8.7 0.099 0.109 13.090 13.555 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.586 12.276 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.458 8.711 dbcsr_mm_accdrv_process 4760 10.4 0.581 0.682 7.085 8.237 make_images_data 222 9.7 0.004 0.005 7.613 8.183 hybrid_alltoall_any 227 10.6 0.216 1.849 6.517 7.992 dbcsr_mm_accdrv_process_sort 4760 11.4 6.286 7.198 6.286 7.198 calculate_norms 4752 9.8 5.523 6.141 5.523 6.141 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.021 5.218 mp_sum_l 887 5.1 3.291 4.968 3.291 4.968 ls_scf_post 1 4.0 0.003 0.021 3.728 3.740 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.371 3.727 make_images_sizes 222 9.7 0.000 0.000 0.803 3.662 mp_alltoall_i44 222 10.7 0.803 3.661 0.803 3.661 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.051 3.523 mp_irecv_dv 6231 10.9 2.034 3.513 2.034 3.513 arnoldi_extremal 4 6.8 0.000 0.003 3.431 3.457 arnoldi_normal_ev 4 7.8 0.002 0.010 3.431 3.457 build_subspace 16 8.4 0.009 0.015 3.302 3.312 dbcsr_special_finalize 555 9.7 0.005 0.006 2.344 2.903 dbcsr_merge_single_wm 555 10.7 0.455 0.587 2.335 2.894 ls_scf_store_result 1 5.0 0.000 0.000 2.845 2.888 dbcsr_sort_data 658 11.4 2.138 2.634 2.138 2.634 make_images_pack 222 9.7 2.205 2.619 2.207 2.620 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.362 2.593 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.465 2.072 2.467 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.341 2.441 buffer_matrices_ensure_size 222 8.7 1.750 2.049 1.750 2.049 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.863 1.878 rebuild_ks_matrix 3 7.3 0.000 0.000 1.853 1.868 qs_ks_build_kohn_sham_matrix 3 8.3 0.020 0.046 1.853 1.868 compute_matrix_preconditioner 1 6.0 0.000 0.002 1.824 1.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.556000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 2.148131E+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 0.074 0.162 92.339 92.340 qs_energies 1 2.0 0.000 0.001 91.353 91.514 ls_scf 1 3.0 0.008 0.068 89.979 90.134 dbcsr_multiply_generic 111 6.7 0.015 0.017 75.207 75.543 multiply_cannon 111 7.7 0.028 0.040 53.088 57.881 ls_scf_main 1 4.0 0.005 0.075 55.239 55.247 multiply_cannon_loop 111 8.7 0.135 0.145 50.225 53.723 density_matrix_trs4 2 5.0 0.003 0.017 49.513 49.773 mp_waitall_1 9105 10.9 21.061 31.212 21.061 31.212 ls_scf_init_scf 1 4.0 0.000 0.001 30.980 30.986 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.548 29.653 multiply_cannon_multrec 1332 9.7 13.323 17.343 22.723 27.973 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.189 27.201 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.593 21.358 make_m2s 222 7.7 0.006 0.007 15.265 15.876 make_images 222 8.7 1.368 1.692 15.235 15.847 dbcsr_mm_accdrv_process 4041 10.4 0.342 0.673 8.998 10.513 dbcsr_mm_accdrv_process_sort 4041 11.4 8.478 9.979 8.478 9.979 make_images_data 222 9.7 0.004 0.005 8.880 9.798 hybrid_alltoall_any 227 10.6 0.542 2.534 8.221 9.625 mp_sum_l 887 5.1 5.456 8.767 5.456 8.767 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.162 7.691 mp_irecv_dv 3311 11.0 3.142 7.638 3.142 7.638 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.088 6.992 calculate_norms 2376 9.8 6.050 6.587 6.050 6.587 multiply_cannon_sync_h2d 1332 9.7 4.906 6.237 4.906 6.237 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.971 5.229 arnoldi_extremal 4 6.8 0.000 0.000 4.750 4.769 arnoldi_normal_ev 4 7.8 0.001 0.004 4.750 4.769 build_subspace 16 8.4 0.014 0.021 4.495 4.499 ls_scf_post 1 4.0 0.018 0.143 3.751 3.850 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.559 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.193 3.419 dbcsr_matrix_vector_mult_local 304 10.0 2.783 3.265 2.786 3.267 mp_allgather_i34 111 8.7 0.850 3.102 0.850 3.102 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.617 2.713 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.230 2.662 dbcsr_data_new 4174 10.1 2.109 2.408 2.109 2.408 make_images_pack 222 9.7 1.817 2.117 1.820 2.119 dbcsr_sort_data 436 11.2 1.847 2.089 1.847 2.089 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.869 1.876 rebuild_ks_matrix 3 7.3 0.000 0.000 1.855 1.862 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.008 1.855 1.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.340000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 2.917216E+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 0.050 0.141 93.554 93.573 qs_energies 1 2.0 0.000 0.000 92.598 92.606 ls_scf 1 3.0 0.000 0.001 91.149 91.154 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.386 75.733 ls_scf_main 1 4.0 0.000 0.003 56.596 56.599 multiply_cannon 111 7.7 0.032 0.076 52.036 56.521 multiply_cannon_loop 111 8.7 0.117 0.129 49.223 53.573 density_matrix_trs4 2 5.0 0.002 0.004 50.643 50.831 mp_waitall_1 7281 11.0 23.611 33.048 23.611 33.048 ls_scf_init_scf 1 4.0 0.000 0.001 30.866 30.869 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.654 29.758 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.251 27.263 multiply_cannon_multrec 888 9.7 12.734 15.146 21.387 24.449 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.798 22.469 make_m2s 222 7.7 0.006 0.007 16.609 17.419 make_images 222 8.7 1.582 1.849 16.571 17.378 make_images_data 222 9.7 0.004 0.005 9.711 10.844 hybrid_alltoall_any 227 10.6 0.639 2.935 9.149 10.476 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.525 8.182 9.438 dbcsr_mm_accdrv_process_sort 3754 11.4 7.740 8.913 7.740 8.913 mp_sum_l 887 5.1 5.048 8.782 5.048 8.782 multiply_cannon_sync_h2d 888 9.7 6.036 7.589 6.036 7.589 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.745 7.088 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.467 6.889 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.877 6.844 mp_irecv_dv 2335 11.1 2.451 6.825 2.451 6.825 arnoldi_extremal 4 6.8 0.000 0.000 5.182 5.201 arnoldi_normal_ev 4 7.8 0.001 0.005 5.182 5.200 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.827 5.084 build_subspace 16 8.4 0.014 0.020 4.882 4.889 calculate_norms 1584 9.8 4.323 4.748 4.323 4.748 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.504 3.842 mp_allgather_i34 111 8.7 0.927 3.829 0.927 3.829 ls_scf_post 1 4.0 0.000 0.001 3.687 3.693 dbcsr_matrix_vector_mult_local 304 10.0 3.062 3.646 3.064 3.648 ls_scf_store_result 1 5.0 0.000 0.000 3.422 3.514 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.856 2.958 dbcsr_data_new 4116 9.9 2.109 2.452 2.109 2.452 make_images_sizes 222 9.7 0.000 0.000 1.052 2.280 mp_alltoall_i44 222 10.7 1.051 2.280 1.051 2.280 dbcsr_sort_data 325 11.1 1.887 2.187 1.887 2.187 dbcsr_finalize 304 7.8 0.026 0.032 1.613 1.887 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.883 1.884 make_images_pack 222 9.7 1.628 1.875 1.631 1.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.573000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2229.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 3.354309E+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 0.102 0.172 99.986 99.996 qs_energies 1 2.0 0.000 0.000 98.977 98.998 ls_scf 1 3.0 0.001 0.005 97.230 97.253 dbcsr_multiply_generic 111 6.7 0.018 0.022 79.434 79.748 ls_scf_main 1 4.0 0.003 0.032 59.636 59.638 multiply_cannon 111 7.7 0.043 0.097 52.038 56.431 density_matrix_trs4 2 5.0 0.014 0.059 53.484 53.600 multiply_cannon_loop 111 8.7 0.153 0.168 47.004 50.661 ls_scf_init_scf 1 4.0 0.000 0.001 34.158 34.161 ls_scf_init_matrix_S 1 5.0 0.006 0.046 32.720 32.811 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.976 30.216 mp_waitall_1 6369 11.0 23.177 29.019 23.177 29.019 multiply_cannon_multrec 1332 9.7 14.127 17.218 22.110 24.682 make_m2s 222 7.7 0.007 0.009 21.365 22.779 make_images 222 8.7 3.137 3.594 21.314 22.729 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.316 17.474 make_images_data 222 9.7 0.004 0.005 11.969 13.656 hybrid_alltoall_any 227 10.6 0.797 3.720 11.332 12.930 dbcsr_mm_accdrv_process 3641 10.4 0.293 0.479 7.612 9.139 dbcsr_mm_accdrv_process_sort 3641 11.4 7.174 8.647 7.174 8.647 mp_sum_l 887 5.1 4.462 8.554 4.462 8.554 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.447 6.794 arnoldi_extremal 4 6.8 0.000 0.000 6.041 6.314 arnoldi_normal_ev 4 7.8 0.258 0.497 6.041 6.314 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.095 6.128 multiply_cannon_sync_h2d 1332 9.7 5.478 6.072 5.478 6.072 mp_irecv_dv 3229 10.9 2.071 6.043 2.071 6.043 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.722 5.241 build_subspace 16 8.4 0.014 0.021 5.034 5.040 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.572 4.751 calculate_norms 2376 9.8 4.177 4.510 4.177 4.510 mp_allgather_i34 111 8.7 2.134 4.491 2.134 4.491 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.661 3.965 dbcsr_matrix_vector_mult_local 304 10.0 3.237 3.739 3.239 3.741 ls_scf_post 1 4.0 0.005 0.022 3.435 3.456 dbcsr_sort_data 658 11.4 3.019 3.352 3.019 3.352 dbcsr_special_finalize 555 9.7 0.006 0.007 2.764 3.157 dbcsr_merge_single_wm 555 10.7 0.534 0.656 2.756 3.149 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.034 3.110 ls_scf_store_result 1 5.0 0.000 0.000 3.011 3.076 dbcsr_finalize 304 7.8 0.049 0.061 2.027 2.879 dbcsr_data_release 10477 10.7 1.598 2.473 1.598 2.473 dbcsr_merge_all 275 8.9 0.323 0.382 1.776 2.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.996000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 4.729954E+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 0.063 0.114 93.417 93.442 qs_energies 1 2.0 0.000 0.000 91.541 91.545 ls_scf 1 3.0 0.000 0.000 89.585 89.590 dbcsr_multiply_generic 111 6.7 0.018 0.019 69.686 69.882 ls_scf_main 1 4.0 0.000 0.000 55.890 55.890 multiply_cannon 111 7.7 0.080 0.140 51.917 55.164 multiply_cannon_loop 111 8.7 0.088 0.093 49.280 50.893 density_matrix_trs4 2 5.0 0.002 0.003 48.930 49.007 ls_scf_init_scf 1 4.0 0.000 0.000 30.108 30.110 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.306 28.340 mp_waitall_1 5436 11.0 23.196 26.831 23.196 26.831 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.200 26.213 multiply_cannon_multrec 444 9.7 13.705 16.226 20.862 21.797 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.105 14.521 make_m2s 222 7.7 0.005 0.005 13.134 14.111 make_images 222 8.7 2.038 2.475 13.066 14.042 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.027 13.929 make_images_data 222 9.7 0.003 0.004 7.881 9.419 hybrid_alltoall_any 227 10.6 0.806 3.844 7.758 9.395 dbcsr_mm_accdrv_process 3003 10.4 0.346 0.439 6.848 7.990 multiply_cannon_sync_h2d 444 9.7 6.762 7.854 6.762 7.854 dbcsr_mm_accdrv_process_sort 3003 11.4 6.474 7.588 6.474 7.588 arnoldi_extremal 4 6.8 0.000 0.000 6.337 6.402 arnoldi_normal_ev 4 7.8 0.007 0.030 6.337 6.402 build_subspace 16 8.4 0.014 0.019 5.876 5.887 mp_sum_l 887 5.1 2.653 4.719 2.653 4.719 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.438 4.623 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.389 4.617 dbcsr_matrix_vector_mult_local 304 10.0 3.757 4.251 3.759 4.254 calculate_norms 792 9.8 3.620 3.779 3.620 3.779 ls_scf_post 1 4.0 0.000 0.000 3.588 3.592 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.513 3.562 mp_allgather_i34 111 8.7 1.185 3.522 1.185 3.522 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.801 3.482 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.535 3.397 mp_irecv_dv 1241 11.2 1.523 3.358 1.523 3.358 ls_scf_store_result 1 5.0 0.000 0.000 3.202 3.235 make_images_sizes 222 9.7 0.000 0.000 0.872 3.023 mp_alltoall_i44 222 10.7 0.871 3.023 0.871 3.023 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.534 2.535 rebuild_ks_matrix 3 7.3 0.000 0.000 2.501 2.502 qs_ks_build_kohn_sham_matrix 3 8.3 0.014 0.015 2.501 2.502 dbcsr_finalize 304 7.8 0.062 0.077 2.198 2.312 dbcsr_data_new 4608 9.7 1.793 2.291 1.793 2.291 dbcsr_merge_all 275 8.9 0.480 0.534 2.056 2.150 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.940 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.442000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3772.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 8.862601E+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 0.088 0.119 106.428 106.437 qs_energies 1 2.0 0.000 0.000 104.470 104.473 ls_scf 1 3.0 0.000 0.000 101.552 101.555 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.436 74.609 ls_scf_main 1 4.0 0.000 0.000 63.647 63.647 density_matrix_trs4 2 5.0 0.002 0.003 54.670 54.731 multiply_cannon 111 7.7 0.160 0.365 47.944 50.496 multiply_cannon_loop 111 8.7 0.098 0.100 44.849 45.175 ls_scf_init_scf 1 4.0 0.000 0.000 34.122 34.122 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.332 32.353 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.606 29.618 mp_waitall_1 4527 11.1 21.418 25.012 21.418 25.012 make_m2s 222 7.7 0.005 0.005 22.882 24.071 make_images 222 8.7 3.568 3.855 22.774 23.962 multiply_cannon_multrec 444 9.7 17.812 18.444 22.489 22.972 hybrid_alltoall_any 227 10.6 1.653 3.609 12.867 16.183 make_images_data 222 9.7 0.003 0.004 13.197 15.517 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.633 10.067 multiply_cannon_sync_h2d 444 9.7 8.795 8.848 8.795 8.848 arnoldi_extremal 4 6.8 0.000 0.000 7.654 7.733 arnoldi_normal_ev 4 7.8 0.031 0.090 7.653 7.733 build_subspace 16 8.4 0.026 0.036 7.045 7.053 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.596 5.770 dbcsr_matrix_vector_mult_local 304 10.0 5.118 5.445 5.120 5.448 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.104 5.196 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.857 5.102 dbcsr_mm_accdrv_process 1814 10.4 0.274 0.374 4.502 4.620 dbcsr_mm_accdrv_process_sort 1814 11.4 4.159 4.297 4.159 4.297 make_images_sizes 222 9.7 0.000 0.000 1.551 3.977 mp_alltoall_i44 222 10.7 1.550 3.976 1.550 3.976 ls_scf_post 1 4.0 0.000 0.000 3.784 3.786 mp_allgather_i34 111 8.7 1.196 3.754 1.196 3.754 ls_scf_store_result 1 5.0 0.000 0.000 3.492 3.532 calculate_norms 792 9.8 3.226 3.270 3.226 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.160 dbcsr_merge_all 275 8.9 0.883 0.918 2.860 2.939 dbcsr_complete_redistribute 5 7.6 1.429 1.470 2.779 2.891 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.888 2.888 matrix_ls_to_qs 2 6.0 0.000 0.000 2.402 2.525 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.512 2.514 dbcsr_sort_data 325 11.1 2.438 2.497 2.438 2.497 rebuild_ks_matrix 3 7.3 0.000 0.000 2.446 2.448 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.011 2.446 2.448 dbcsr_new_transposed 4 7.5 0.241 0.251 2.360 2.369 dbcsr_data_new 6591 9.6 1.858 2.351 1.858 2.351 dbcsr_frobenius_norm 74 6.6 2.057 2.139 2.200 2.235 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.203 dbcsr_add_anytype 103 7.2 0.859 0.891 2.125 2.203 dbcsr_data_release 12724 10.6 1.975 2.202 1.975 2.202 mp_alltoall_d11v 48 9.2 2.140 2.176 2.140 2.176 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.437000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7083.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3bcd6e6d97c8df411d49f3e0e700dc328631558d_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 592.994304E+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 4809200. 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 0.058 0.200 277.416 277.418 qs_mol_dyn_low 1 2.0 0.005 0.132 275.087 275.109 qs_forces 5 3.8 0.004 0.006 275.012 275.016 qs_energies 5 4.8 0.018 0.277 271.356 271.546 scf_env_do_scf 5 5.8 0.001 0.023 254.814 254.819 scf_env_do_scf_inner_loop 105 6.6 0.006 0.043 220.203 220.224 qs_scf_new_mos 105 7.6 0.000 0.001 169.909 170.141 qs_scf_loop_do_ot 105 8.6 0.001 0.008 169.909 170.141 ot_scf_mini 105 9.6 0.004 0.020 159.851 160.047 dbcsr_multiply_generic 1445 12.2 0.148 0.158 137.712 138.350 velocity_verlet 4 3.0 0.006 0.017 118.457 118.457 multiply_cannon 1445 13.2 0.278 0.291 116.022 117.917 multiply_cannon_loop 1445 14.2 2.850 3.020 113.628 115.542 qs_ot_get_p 112 10.4 0.001 0.001 65.800 66.118 ot_mini 105 10.6 0.001 0.001 63.214 63.366 qs_ot_p2m_diag 40 11.0 0.020 0.030 53.415 53.544 cp_dbcsr_syevd 40 12.0 0.003 0.029 49.614 49.617 mp_waitall_1 488190 16.1 39.713 47.562 39.713 47.562 multiply_cannon_multrec 69360 15.2 29.549 34.572 39.668 44.708 cp_fm_syevd 40 13.0 0.007 0.451 43.019 43.175 qs_ot_get_derivative 55 11.6 0.001 0.001 41.316 41.476 cp_fm_redistribute_end 40 14.0 18.392 36.657 18.402 36.664 cp_fm_syevd_base 40 14.0 18.244 36.509 18.244 36.509 multiply_cannon_metrocomm3 69360 15.2 0.204 0.214 26.702 36.449 init_scf_loop 7 6.6 0.007 0.076 34.547 34.552 rebuild_ks_matrix 110 8.4 0.000 0.000 33.948 34.107 qs_ks_build_kohn_sham_matrix 110 9.4 0.019 0.366 33.948 34.106 multiply_cannon_sync_h2d 69360 15.2 28.952 33.274 28.952 33.274 qs_ks_update_qs_env 112 7.6 0.001 0.001 31.594 31.754 prepare_preconditioner 7 7.6 0.000 0.011 27.771 27.809 make_preconditioner 7 8.6 0.002 0.051 27.771 27.809 qs_rho_update_rho_low 110 7.6 0.001 0.001 24.199 24.554 calculate_rho_elec 110 8.6 0.030 0.042 24.199 24.553 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.130 23.364 apply_single 62 13.6 0.000 0.000 23.130 23.364 ot_new_cg_direction 55 11.6 0.001 0.001 21.117 21.117 make_full_inverse_cholesky 7 9.6 0.001 0.029 19.093 19.165 density_rs2pw 110 9.6 0.005 0.007 17.926 18.395 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.496 15.775 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.792 14.900 transfer_rs2pw 445 10.6 0.007 0.008 14.048 14.552 pw_transfer 1645 12.4 0.080 0.109 13.210 13.742 mp_sum_l 4764 12.2 12.989 13.721 12.989 13.721 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 13.071 13.603 init_scf_run 5 5.8 0.000 0.001 13.453 13.454 scf_env_initial_rho_setup 5 6.8 0.001 0.003 13.453 13.454 cp_fm_cholesky_invert 7 10.6 12.408 12.424 12.408 12.424 sum_up_and_integrate 60 10.3 0.001 0.002 12.150 12.161 integrate_v_rspace 60 11.3 0.002 0.002 12.133 12.143 check_diag 80 13.5 8.689 8.987 11.971 12.117 calculate_dm_sparse 110 9.5 0.000 0.001 11.733 11.955 fft_wrap_pw1pw2_240 915 15.0 1.140 1.230 11.176 11.708 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.354 11.433 qs_vxc_create 110 10.4 0.022 0.222 11.213 11.271 dbcsr_mm_accdrv_process 154766 15.8 6.143 6.340 9.987 10.996 fft3d_pb 915 16.0 2.385 2.673 9.382 10.050 multiply_cannon_metrocomm1 69360 15.2 0.097 0.105 5.146 9.001 make_m2s 2890 13.2 0.079 0.087 8.288 8.923 make_images 2890 14.2 0.243 0.267 8.180 8.816 transfer_rs2pw_30 110 11.6 1.265 1.335 7.294 8.547 make_full_single_inverse 7 9.6 0.001 0.001 8.259 8.297 mp_sendrecv_dv 168740 12.6 7.949 8.217 7.949 8.217 potential_pw2rs 60 12.3 0.003 0.004 8.030 8.136 xc_rho_set_and_dset_create 110 12.4 0.077 0.111 7.815 8.098 acc_transpose_blocks 69360 15.2 0.365 0.381 7.186 7.852 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.741 7.758 calculate_first_density_matrix 1 7.0 0.000 0.004 7.731 7.749 xc_vxc_pw_create 60 11.3 0.038 0.050 7.623 7.699 mp_alltoall_z22v 2340 17.7 6.483 7.273 6.483 7.273 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.086 7.129 xc_pw_derive 510 13.4 0.005 0.007 6.855 6.961 transfer_pw2rs 245 13.2 0.003 0.004 6.809 6.915 cp_fm_cholesky_decompose 14 10.2 6.068 6.076 6.068 6.076 mp_alltoall_d11v 1300 13.8 4.744 5.696 4.744 5.696 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=277.418000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.800000, yerr=3.310589 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 3bcd6e6d97c8df411d49f3e0e700dc328631558d Summary: empty Status: OK