=== 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: c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231206 # # 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: 13.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231206 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/01 job id: 50483105 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/02 job id: 50483106 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/03 job id: 50483107 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/04 job id: 50483108 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/05 job id: 50483109 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/06 job id: 50483110 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/07 job id: 50483111 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/08 job id: 50483112 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/09 job id: 50483113 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/10 job id: 50483114 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/11 job id: 50483115 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/12 job id: 50483116 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/13 job id: 50483117 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/14 job id: 50483118 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/15 job id: 50483119 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/16 job id: 50483120 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/17 job id: 50483121 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/18 job id: 50483122 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/19 job id: 50483123 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/20 job id: 50483124 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/21 job id: 50483125 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/22 job id: 50483126 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/23 job id: 50483127 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/24 job id: 50483128 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/25 job id: 50483129 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/26 job id: 50483130 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/27 job id: 50483132 --- 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/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 0.018 0.035 133.768 133.769 farming_run 1 2.0 133.127 133.128 133.741 133.745 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.482891E+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 0.005 0.025 115.007 115.007 qs_energies 1 2.0 0.000 0.000 114.818 114.821 mp2_main 1 3.0 0.000 0.000 112.763 112.765 mp2_gpw_main 1 4.0 0.027 0.033 111.778 111.780 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.945 93.109 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.560 54.723 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 40.969 45.785 get_2c_integrals 1 6.0 0.008 0.009 37.226 38.214 integrate_v_rspace 273 8.0 0.441 0.455 24.568 29.013 pw_transfer 6555 10.6 0.379 0.399 26.791 27.322 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.401 25.755 grid_integrate_task_list 273 9.0 20.475 25.346 20.475 25.346 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 22.946 23.298 compute_2c_integrals 1 7.0 0.002 0.002 19.393 19.394 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.593 19.132 mp2_eri_2c_integrate_gpw 1 9.0 2.374 2.420 18.591 19.131 rpa_ri_compute_en 1 5.0 0.020 0.040 18.730 18.792 cp_fm_cholesky_decompose 12 8.2 17.757 18.788 17.757 18.788 cholesky_decomp 1 7.0 0.000 0.000 16.675 17.698 fft3d_s 5443 13.4 16.083 16.379 16.103 16.400 ao_to_mo_and_store_B_mult_1 272 7.0 10.762 15.312 10.762 15.312 calculate_wavefunction 272 8.0 5.397 5.483 12.304 13.047 rpa_num_int 1 6.0 0.002 0.029 10.616 10.626 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.582 10.605 calc_mat_Q 8 8.0 0.000 0.000 9.455 9.554 contract_S_to_Q 8 9.0 0.000 0.000 8.882 8.978 calc_potential_gpw 544 9.5 0.004 0.005 8.259 8.731 parallel_gemm_fm 14 9.1 0.000 0.000 8.472 8.606 parallel_gemm_fm_cosma 14 10.1 8.472 8.606 8.472 8.606 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.082 8.429 potential_pw2rs 545 10.0 0.107 0.109 7.469 8.109 create_integ_mat 1 6.0 0.015 0.028 7.840 7.849 collocate_single_gaussian 272 10.0 0.038 0.041 7.332 7.661 array2fm 1 7.0 0.000 0.000 6.826 7.275 pw_scatter_s 2720 13.7 4.337 4.549 4.337 4.549 pw_gather_s 2722 13.2 3.455 3.861 3.455 3.861 array2fm_buffer_send 1 8.0 3.010 3.218 3.010 3.218 pw_poisson_solve 545 10.5 1.119 1.188 2.214 2.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.778147, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 0.025 0.036 421.141 421.142 farming_run 1 2.0 420.273 420.280 421.097 421.101 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.233818E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 217.434 217.435 qs_energies 1 2.0 0.005 0.042 217.214 217.228 scf_env_do_scf 1 3.0 0.000 0.000 115.199 115.199 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.292 114.299 rebuild_ks_matrix 4 6.0 0.000 0.000 114.290 114.297 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.290 114.297 hfx_ks_matrix 4 8.0 0.001 0.001 113.944 113.948 integrate_four_center 4 9.0 0.154 0.454 113.943 113.947 integrate_four_center_main 4 10.0 0.132 0.591 102.080 105.865 integrate_four_center_bin 264 11.0 101.948 105.863 101.948 105.863 mp2_main 1 3.0 0.006 0.061 101.719 101.731 mp2_gpw_main 1 4.0 0.054 0.106 100.816 100.829 init_scf_loop 1 4.0 0.000 0.000 96.971 96.972 mp2_ri_gpw_compute_in 1 5.0 0.070 0.110 73.812 74.891 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.018 53.786 54.865 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 41.560 46.471 integrate_v_rspace 95 8.0 0.399 0.568 28.025 32.771 pw_transfer 2240 10.6 0.146 0.165 29.492 29.868 ao_to_mo_and_store_B_mult_1 91 7.0 10.553 28.906 10.553 28.906 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.518 28.851 mp2_ri_gpw_compute_en 1 5.0 0.055 0.068 26.811 28.516 grid_integrate_task_list 95 9.0 23.440 28.416 23.440 28.416 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.273 26.625 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.884 25.084 25.095 get_2c_integrals 1 6.0 0.001 0.013 19.930 19.963 compute_2c_integrals 1 7.0 0.005 0.024 18.909 18.919 fft3d_s 1823 13.4 18.495 18.897 18.509 18.910 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.584 18.774 mp2_eri_2c_integrate_gpw 1 9.0 1.741 1.837 18.583 18.772 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.226 18.226 calculate_wavefunction 91 8.0 2.014 2.044 9.624 9.828 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.582 8.786 9.312 potential_pw2rs 186 10.0 0.033 0.034 8.418 8.887 local_gemm 172 8.0 8.227 8.731 8.227 8.731 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.518 8.085 8.589 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.128 8.485 calc_potential_gpw 182 9.5 0.002 0.002 7.884 8.043 collocate_single_gaussian 91 10.0 0.017 0.031 7.732 7.978 mp_sync 37 10.5 3.817 7.159 3.817 7.159 integrate_four_center_load 4 10.0 0.000 0.000 6.786 6.790 hfx_load_balance 1 11.0 0.000 0.000 6.786 6.790 mp_sendrecv_dm3 2068 8.0 6.115 6.640 6.115 6.640 mp2_ri_gpw_compute_en_ener 172 7.0 6.332 6.414 6.332 6.414 pw_gather_s 912 13.2 4.486 4.945 4.486 4.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.820684, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1493.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.144576E+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 592243. 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 0.012 0.032 53.340 53.342 qs_mol_dyn_low 1 2.0 0.003 0.004 52.987 52.996 qs_forces 11 3.9 0.002 0.002 52.922 52.923 qs_energies 11 4.9 0.001 0.001 51.394 51.410 scf_env_do_scf 11 5.9 0.000 0.001 45.480 45.480 scf_env_do_scf_inner_loop 108 6.5 0.008 0.019 43.383 43.383 qs_scf_new_mos 108 7.5 0.000 0.001 33.520 33.792 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.520 33.792 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.239 33.712 ot_scf_mini 108 9.5 0.002 0.002 31.927 32.104 multiply_cannon 2286 13.5 0.187 0.193 26.058 27.586 multiply_cannon_loop 2286 14.5 1.833 1.931 25.412 26.961 velocity_verlet 10 3.0 0.001 0.002 26.110 26.110 ot_mini 108 10.5 0.001 0.001 19.302 19.543 qs_ot_get_derivative 108 11.5 0.001 0.001 16.297 16.479 mp_waitall_1 245248 16.5 8.205 14.146 8.205 14.146 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 5.920 12.661 multiply_cannon_multrec 54864 15.5 3.668 5.762 7.669 11.045 qs_ot_get_p 119 10.4 0.001 0.001 8.058 8.312 rebuild_ks_matrix 119 8.3 0.000 0.000 7.796 7.920 qs_ks_build_kohn_sham_matrix 119 9.3 0.018 0.033 7.796 7.920 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.845 6.951 mp_sum_l 7287 12.8 5.270 6.777 5.270 6.777 multiply_cannon_sync_h2d 54864 15.5 5.203 6.644 5.203 6.644 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.675 6.159 dbcsr_mm_accdrv_process 76910 16.1 1.844 2.892 3.914 5.588 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.266 5.371 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.154 5.193 init_scf_run 11 5.9 0.000 0.001 4.586 4.586 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.585 4.586 sum_up_and_integrate 119 10.3 0.001 0.001 4.382 4.392 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.386 4.386 integrate_v_rspace 119 11.3 0.002 0.003 4.371 4.383 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.071 4.207 calculate_rho_elec 119 8.7 0.011 0.016 4.071 4.206 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.193 4.194 cp_fm_redistribute_end 50 14.0 2.140 4.168 2.145 4.171 cp_fm_diag_elpa_base 50 14.0 2.020 4.070 2.024 4.078 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.964 3.178 apply_single 119 13.6 0.000 0.000 2.964 3.178 calculate_dm_sparse 119 9.5 0.000 0.000 2.811 2.958 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.654 2.875 ot_diis_step 108 11.5 0.006 0.006 2.737 2.737 acc_transpose_blocks 54864 15.5 0.224 0.247 2.213 2.725 jit_kernel_multiply 13 15.8 2.006 2.618 2.006 2.618 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.400 2.490 calculate_first_density_matrix 1 7.0 0.000 0.000 2.353 2.361 density_rs2pw 119 9.7 0.004 0.004 2.102 2.201 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.191 2.193 wfi_extrapolate 11 7.9 0.001 0.001 2.159 2.159 grid_integrate_task_list 119 12.3 1.999 2.101 1.999 2.101 init_scf_loop 11 6.9 0.000 0.000 2.079 2.080 mp_sum_d 4137 12.0 1.363 1.962 1.363 1.962 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.885 1.952 potential_pw2rs 119 12.3 0.004 0.004 1.779 1.791 pw_transfer 1439 11.6 0.051 0.056 1.629 1.704 make_m2s 4572 13.5 0.053 0.055 1.602 1.647 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.554 1.630 make_images 4572 14.5 0.133 0.139 1.520 1.565 mp_waitany 12084 13.8 1.256 1.480 1.256 1.480 acc_transpose_blocks_sync 164592 16.5 1.213 1.443 1.213 1.443 transfer_rs2pw 487 10.6 0.005 0.006 1.332 1.443 mp_alltoall_d11v 2130 13.8 1.284 1.432 1.284 1.432 grid_collocate_task_list 119 9.7 1.346 1.426 1.346 1.426 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.378 1.407 transfer_pw2rs 487 13.2 0.006 0.006 1.323 1.340 fft3d_ps 1201 14.6 0.368 0.473 1.266 1.338 fft_wrap_pw1pw2_140 487 13.2 0.108 0.116 1.207 1.287 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.739 1.135 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.342000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.620032E+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 1809844. 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 0.021 0.103 39.013 39.013 qs_mol_dyn_low 1 2.0 0.003 0.003 38.664 38.672 qs_forces 11 3.9 0.002 0.003 38.605 38.605 qs_energies 11 4.9 0.002 0.002 36.941 36.945 scf_env_do_scf 11 5.9 0.000 0.001 31.711 31.711 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.284 29.284 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.267 22.616 qs_scf_new_mos 108 7.5 0.001 0.001 20.701 20.925 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.701 20.924 ot_scf_mini 108 9.5 0.002 0.003 19.774 19.936 multiply_cannon 2286 13.5 0.208 0.217 17.169 18.658 velocity_verlet 10 3.0 0.001 0.002 18.295 18.296 multiply_cannon_loop 2286 14.5 1.191 1.247 16.014 17.427 ot_mini 108 10.5 0.001 0.001 12.206 12.433 mp_waitall_1 200699 16.5 5.615 10.854 5.615 10.854 qs_ot_get_derivative 108 11.5 0.001 0.001 9.848 10.014 multiply_cannon_metrocomm3 27432 15.5 0.071 0.075 4.127 9.619 multiply_cannon_multrec 27432 15.5 1.820 4.226 6.312 9.078 rebuild_ks_matrix 119 8.3 0.000 0.000 6.764 6.903 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.763 6.902 dbcsr_mm_accdrv_process 47894 16.0 3.680 6.027 4.412 6.550 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.951 6.077 qs_ot_get_p 119 10.4 0.001 0.001 4.654 4.872 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.757 4.610 mp_sum_l 7287 12.8 2.098 4.180 2.098 4.180 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.002 4.056 apply_single 119 13.6 0.000 0.000 3.001 4.056 init_scf_run 11 5.9 0.000 0.001 4.050 4.050 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.049 4.050 sum_up_and_integrate 119 10.3 0.001 0.001 3.692 3.696 integrate_v_rspace 119 11.3 0.002 0.003 3.678 3.682 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.588 3.621 calculate_rho_elec 119 8.7 0.021 0.024 3.588 3.621 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.003 3.024 make_m2s 4572 13.5 0.051 0.053 2.608 2.939 make_images 4572 14.5 0.205 0.243 2.519 2.852 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.595 2.595 calculate_first_density_matrix 1 7.0 0.000 0.000 2.544 2.546 init_scf_loop 11 6.9 0.000 0.000 2.407 2.408 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.308 2.394 ot_diis_step 108 11.5 0.011 0.011 2.310 2.311 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.278 2.278 jit_kernel_multiply 11 16.2 0.674 2.257 0.674 2.257 cp_fm_redistribute_end 50 14.0 1.156 2.253 1.159 2.255 calculate_dm_sparse 119 9.5 0.000 0.001 2.158 2.234 multiply_cannon_sync_h2d 27432 15.5 1.689 2.211 1.689 2.211 cp_fm_diag_elpa_base 50 14.0 1.065 2.161 1.093 2.196 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.029 2.031 acc_transpose_blocks 27432 15.5 0.114 0.120 1.659 2.018 density_rs2pw 119 9.7 0.004 0.004 1.883 1.945 grid_integrate_task_list 119 12.3 1.841 1.942 1.841 1.942 pw_transfer 1439 11.6 0.065 0.068 1.861 1.896 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.834 1.874 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.769 1.807 make_images_data 4572 15.5 0.047 0.054 1.231 1.563 prepare_preconditioner 11 7.9 0.000 0.000 1.518 1.544 make_preconditioner 11 8.9 0.000 0.000 1.518 1.544 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.424 1.477 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.447 potential_pw2rs 119 12.3 0.006 0.006 1.439 1.443 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.097 1.442 fft3d_ps 1201 14.6 0.520 0.578 1.400 1.429 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.385 1.421 grid_collocate_task_list 119 9.7 1.291 1.346 1.291 1.346 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.338 1.346 mp_alltoall_d11v 2130 13.8 1.209 1.342 1.209 1.342 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.205 1.254 transfer_rs2pw 487 10.6 0.005 0.005 1.048 1.115 mp_sum_d 4137 12.0 0.534 1.018 0.534 1.018 mp_allgather_i34 2286 14.5 0.579 0.959 0.579 0.959 acc_transpose_blocks_sync 82296 16.5 0.822 0.948 0.822 0.948 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.925 0.925 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.697 0.925 transfer_pw2rs 487 13.2 0.004 0.005 0.884 0.888 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.869 0.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.013000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.212879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.137600E+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 0.006 0.023 32.285 32.286 qs_mol_dyn_low 1 2.0 0.007 0.010 32.070 32.079 qs_forces 11 3.9 0.003 0.004 32.006 32.010 qs_energies 11 4.9 0.002 0.005 30.457 30.463 scf_env_do_scf 11 5.9 0.000 0.001 25.523 25.523 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.067 23.067 dbcsr_multiply_generic 2286 12.5 0.094 0.096 17.056 17.145 qs_scf_new_mos 108 7.5 0.001 0.001 15.427 15.441 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.427 15.441 velocity_verlet 10 3.0 0.001 0.002 15.185 15.187 ot_scf_mini 108 9.5 0.002 0.002 14.666 14.678 multiply_cannon 2286 13.5 0.195 0.203 13.716 14.411 multiply_cannon_loop 2286 14.5 0.865 0.901 12.963 13.686 ot_mini 108 10.5 0.001 0.001 9.054 9.072 qs_ot_get_derivative 108 11.5 0.001 0.001 7.564 7.576 multiply_cannon_multrec 18288 15.5 1.872 2.876 6.993 7.338 rebuild_ks_matrix 119 8.3 0.000 0.000 5.949 5.961 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.949 5.961 dbcsr_mm_accdrv_process 38222 16.0 4.969 5.896 5.027 5.961 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.239 5.251 mp_waitall_1 158411 16.6 2.662 3.770 2.662 3.770 init_scf_run 11 5.9 0.000 0.001 3.687 3.687 scf_env_initial_rho_setup 11 6.9 0.024 0.040 3.687 3.687 sum_up_and_integrate 119 10.3 0.001 0.001 3.527 3.531 integrate_v_rspace 119 11.3 0.003 0.003 3.514 3.520 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.827 3.453 qs_ot_get_p 119 10.4 0.001 0.001 3.374 3.398 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.250 3.261 calculate_rho_elec 119 8.7 0.031 0.031 3.250 3.261 calculate_first_density_matrix 1 7.0 0.000 0.000 2.444 2.468 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.427 2.461 init_scf_loop 11 6.9 0.000 0.000 2.439 2.440 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.036 2.355 apply_single 119 13.6 0.000 0.000 2.036 2.355 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.237 2.243 make_m2s 4572 13.5 0.044 0.045 1.850 1.981 cp_dbcsr_syevd 50 12.0 0.009 0.029 1.940 1.941 grid_integrate_task_list 119 12.3 1.807 1.906 1.807 1.906 make_images 4572 14.5 0.192 0.205 1.765 1.895 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.886 1.887 pw_transfer 1439 11.6 0.065 0.069 1.870 1.880 density_rs2pw 119 9.7 0.004 0.004 1.824 1.880 calculate_dm_sparse 119 9.5 0.000 0.001 1.843 1.853 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.777 1.790 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.704 1.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.694 1.701 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.683 1.698 prepare_preconditioner 11 7.9 0.000 0.000 1.693 1.695 make_preconditioner 11 8.9 0.000 0.000 1.693 1.695 acc_transpose_blocks 18288 15.5 0.078 0.080 1.631 1.679 cp_fm_diag_elpa_base 50 14.0 1.639 1.650 1.661 1.672 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.543 1.632 ot_diis_step 108 11.5 0.011 0.012 1.473 1.474 mp_sum_l 7287 12.8 1.102 1.471 1.102 1.471 fft_wrap_pw1pw2_140 487 13.2 0.179 0.184 1.428 1.442 fft3d_ps 1201 14.6 0.546 0.564 1.331 1.343 potential_pw2rs 119 12.3 0.007 0.008 1.324 1.329 grid_collocate_task_list 119 9.7 1.234 1.306 1.234 1.306 multiply_cannon_sync_h2d 18288 15.5 1.043 1.184 1.043 1.184 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.168 1.169 transfer_rs2pw 487 10.6 0.005 0.005 0.966 1.068 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 1.002 1.007 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.953 0.971 make_images_data 4572 15.5 0.047 0.051 0.795 0.949 hybrid_alltoall_any 4725 16.4 0.059 0.117 0.689 0.860 transfer_pw2rs 487 13.2 0.004 0.004 0.794 0.798 mp_alltoall_d11v 2130 13.8 0.667 0.797 0.667 0.797 acc_transpose_blocks_sync 54864 16.5 0.746 0.797 0.746 0.797 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 0.787 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.734 cp_fm_cholesky_invert 11 10.9 0.715 0.719 0.715 0.719 multiply_cannon_metrocomm1 18288 15.5 0.029 0.031 0.485 0.712 mp_alltoall_z22v 1201 16.6 0.630 0.702 0.630 0.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.286000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.641406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 560.914432E+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 0.007 0.028 36.136 36.137 qs_mol_dyn_low 1 2.0 0.003 0.004 35.971 35.981 qs_forces 11 3.9 0.002 0.002 35.913 35.914 qs_energies 11 4.9 0.001 0.001 34.207 34.213 scf_env_do_scf 11 5.9 0.000 0.001 29.275 29.276 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.859 25.859 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.796 19.963 velocity_verlet 10 3.0 0.001 0.002 18.350 18.351 qs_scf_new_mos 108 7.5 0.001 0.001 17.841 17.894 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.840 17.894 multiply_cannon 2286 13.5 0.218 0.228 16.158 17.035 ot_scf_mini 108 9.5 0.002 0.003 16.815 16.873 multiply_cannon_loop 2286 14.5 1.499 1.600 15.210 15.928 ot_mini 108 10.5 0.001 0.001 10.379 10.445 multiply_cannon_multrec 27432 15.5 2.461 3.165 9.008 9.335 qs_ot_get_derivative 108 11.5 0.001 0.001 8.507 8.564 dbcsr_mm_accdrv_process 47916 15.9 6.172 7.707 6.446 7.791 rebuild_ks_matrix 119 8.3 0.000 0.000 6.297 6.354 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.297 6.353 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.577 5.627 qs_ot_get_p 119 10.4 0.001 0.001 3.605 3.677 init_scf_run 11 5.9 0.000 0.001 3.609 3.610 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.609 3.609 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.188 3.513 sum_up_and_integrate 119 10.3 0.001 0.001 3.457 3.465 integrate_v_rspace 119 11.3 0.003 0.003 3.445 3.454 init_scf_loop 11 6.9 0.000 0.000 3.396 3.397 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.327 3.360 calculate_rho_elec 119 8.7 0.040 0.046 3.326 3.360 prepare_preconditioner 11 7.9 0.000 0.000 2.560 2.568 make_preconditioner 11 8.9 0.000 0.000 2.560 2.568 acc_transpose_blocks 27432 15.5 0.119 0.121 2.367 2.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.157 2.492 make_m2s 4572 13.5 0.054 0.056 2.256 2.362 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.325 2.352 mp_waitall_1 137007 16.6 1.856 2.332 1.856 2.332 make_images 4572 14.5 0.273 0.335 2.148 2.252 calculate_first_density_matrix 1 7.0 0.000 0.000 2.235 2.236 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.035 2.223 apply_single 119 13.6 0.000 0.000 2.035 2.222 calculate_dm_sparse 119 9.5 0.000 0.001 2.144 2.203 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.187 2.201 pw_transfer 1439 11.6 0.064 0.069 1.976 2.010 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.884 1.921 density_rs2pw 119 9.7 0.004 0.004 1.796 1.905 grid_integrate_task_list 119 12.3 1.810 1.904 1.810 1.904 ot_diis_step 108 11.5 0.012 0.012 1.831 1.831 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.789 1.790 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.787 1.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.620 1.633 fft_wrap_pw1pw2_140 487 13.2 0.228 0.235 1.566 1.606 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.547 1.555 cp_fm_diag_elpa_base 50 14.0 1.513 1.529 1.545 1.553 acc_transpose_blocks_sync 82296 16.5 1.405 1.551 1.405 1.551 fft3d_ps 1201 14.6 0.592 0.650 1.372 1.400 mp_sum_l 7287 12.8 0.952 1.345 0.952 1.345 grid_collocate_task_list 119 9.7 1.249 1.343 1.249 1.343 wfi_extrapolate 11 7.9 0.001 0.001 1.321 1.322 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.280 1.299 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.905 1.289 potential_pw2rs 119 12.3 0.008 0.009 1.271 1.274 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.255 cp_fm_upper_to_full 72 14.2 0.828 1.181 0.828 1.181 jit_kernel_multiply 6 16.5 0.210 1.070 0.210 1.070 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.066 1.067 dbcsr_complete_redistribute 329 12.2 0.120 0.149 0.780 1.054 transfer_rs2pw 487 10.6 0.004 0.005 0.876 0.989 mp_alltoall_d11v 2130 13.8 0.726 0.957 0.726 0.957 make_images_data 4572 15.5 0.048 0.053 0.839 0.945 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.722 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.878 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.570 0.841 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 0.815 0.826 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.813 cp_fm_cholesky_invert 11 10.9 0.744 0.747 0.744 0.747 mp_alltoall_i22 627 13.8 0.443 0.734 0.443 0.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.137000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.000000, yerr=4.410731 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 622.510080E+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 0.014 0.113 28.641 28.722 qs_mol_dyn_low 1 2.0 0.003 0.003 28.345 28.476 qs_forces 11 3.9 0.002 0.002 28.276 28.277 qs_energies 11 4.9 0.002 0.002 26.569 26.571 scf_env_do_scf 11 5.9 0.000 0.001 21.891 21.891 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.383 19.383 velocity_verlet 10 3.0 0.002 0.002 14.751 14.753 dbcsr_multiply_generic 2286 12.5 0.094 0.099 12.891 12.961 qs_scf_new_mos 108 7.5 0.001 0.001 11.692 11.714 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.691 11.713 ot_scf_mini 108 9.5 0.002 0.002 10.985 11.018 multiply_cannon 2286 13.5 0.226 0.236 10.327 10.775 multiply_cannon_loop 2286 14.5 0.643 0.665 9.424 9.580 ot_mini 108 10.5 0.001 0.001 6.239 6.276 multiply_cannon_multrec 9144 15.5 1.661 1.852 5.916 6.153 rebuild_ks_matrix 119 8.3 0.000 0.000 5.797 5.814 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.796 5.814 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.157 5.172 qs_ot_get_derivative 108 11.5 0.001 0.001 4.953 4.985 dbcsr_mm_accdrv_process 12550 15.8 3.496 4.138 4.145 4.224 sum_up_and_integrate 119 10.3 0.001 0.001 3.447 3.453 integrate_v_rspace 119 11.3 0.003 0.003 3.437 3.443 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.381 3.386 calculate_rho_elec 119 8.7 0.060 0.061 3.380 3.385 init_scf_run 11 5.9 0.000 0.001 3.249 3.249 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.249 3.249 qs_ot_get_p 119 10.4 0.001 0.001 2.872 2.900 init_scf_loop 11 6.9 0.000 0.000 2.489 2.490 pw_transfer 1439 11.6 0.065 0.068 2.138 2.150 calculate_first_density_matrix 1 7.0 0.000 0.000 2.088 2.089 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.044 2.059 grid_integrate_task_list 119 12.3 1.868 1.943 1.868 1.943 make_m2s 4572 13.5 0.033 0.034 1.769 1.901 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.868 1.871 density_rs2pw 119 9.7 0.003 0.003 1.776 1.859 make_images 4572 14.5 0.269 0.299 1.680 1.811 mp_waitall_1 115863 16.7 1.298 1.796 1.298 1.796 calculate_dm_sparse 119 9.5 0.000 0.000 1.739 1.756 prepare_preconditioner 11 7.9 0.000 0.000 1.750 1.753 make_preconditioner 11 8.9 0.000 0.000 1.750 1.753 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.724 1.747 fft_wrap_pw1pw2_140 487 13.2 0.326 0.336 1.696 1.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.640 1.666 cp_dbcsr_syevd 50 12.0 0.005 0.005 1.623 1.623 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.532 1.533 acc_transpose_blocks 9144 15.5 0.042 0.044 1.431 1.458 jit_kernel_multiply 8 15.4 0.611 1.414 0.611 1.414 fft3d_ps 1201 14.6 0.652 0.666 1.395 1.405 grid_collocate_task_list 119 9.7 1.298 1.386 1.298 1.386 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.364 1.372 cp_fm_diag_elpa_base 50 14.0 1.335 1.354 1.362 1.370 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.362 1.369 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.349 1.367 ot_diis_step 108 11.5 0.013 0.013 1.273 1.273 potential_pw2rs 119 12.3 0.010 0.011 1.257 1.260 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.222 1.256 apply_single 119 13.6 0.000 0.000 1.222 1.256 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.220 1.221 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.127 1.132 wfi_extrapolate 11 7.9 0.001 0.001 1.109 1.109 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.926 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.731 0.914 make_images_data 4572 15.5 0.042 0.046 0.751 0.900 cp_fm_cholesky_invert 11 10.9 0.874 0.877 0.874 0.877 transfer_rs2pw 487 10.6 0.004 0.005 0.770 0.840 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.823 0.831 mp_alltoall_d11v 2130 13.8 0.747 0.823 0.747 0.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.760 0.763 acc_transpose_blocks_sync 27432 16.5 0.720 0.746 0.720 0.746 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.668 0.718 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.336 0.665 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.653 0.657 mp_alltoall_z22v 1201 16.6 0.615 0.653 0.615 0.653 transfer_pw2rs 487 13.2 0.003 0.004 0.648 0.651 mp_allgather_i34 2286 14.5 0.216 0.617 0.216 0.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.722000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.090909, yerr=3.848366 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 761.806848E+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 0.011 0.029 42.882 42.883 qs_mol_dyn_low 1 2.0 0.003 0.003 42.683 42.691 qs_forces 11 3.9 0.002 0.002 42.619 42.619 qs_energies 11 4.9 0.001 0.001 40.606 40.609 scf_env_do_scf 11 5.9 0.001 0.001 34.898 34.898 scf_env_do_scf_inner_loop 108 6.5 0.005 0.008 27.018 27.020 velocity_verlet 10 3.0 0.002 0.002 24.156 24.163 dbcsr_multiply_generic 2286 12.5 0.104 0.107 18.892 18.989 qs_scf_new_mos 108 7.5 0.001 0.001 17.357 17.459 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.357 17.458 ot_scf_mini 108 9.5 0.002 0.002 16.208 16.307 multiply_cannon 2286 13.5 0.296 0.300 15.035 15.863 multiply_cannon_loop 2286 14.5 0.863 0.878 13.790 14.650 ot_mini 108 10.5 0.001 0.001 9.964 10.078 multiply_cannon_multrec 9144 15.5 3.377 4.822 8.770 8.928 qs_ot_get_derivative 108 11.5 0.001 0.001 7.929 8.028 init_scf_loop 11 6.9 0.000 0.000 7.853 7.857 rebuild_ks_matrix 119 8.3 0.000 0.000 7.232 7.370 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.232 7.369 prepare_preconditioner 11 7.9 0.000 0.000 6.873 6.887 make_preconditioner 11 8.9 0.000 0.000 6.873 6.887 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.426 6.769 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.514 6.638 dbcsr_mm_accdrv_process 12550 15.8 4.476 5.782 5.254 6.505 cp_fm_upper_to_full 72 14.2 3.247 4.641 3.247 4.641 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.201 4.209 calculate_rho_elec 119 8.7 0.118 0.120 4.201 4.208 sum_up_and_integrate 119 10.3 0.001 0.001 3.843 3.851 integrate_v_rspace 119 11.3 0.003 0.004 3.833 3.841 init_scf_run 11 5.9 0.000 0.001 3.648 3.648 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.647 3.648 qs_ot_get_p 119 10.4 0.001 0.001 3.364 3.500 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.834 3.250 mp_waitall_1 94719 16.7 2.257 3.168 2.257 3.168 pw_transfer 1439 11.6 0.069 0.070 3.022 3.029 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.924 2.932 dbcsr_complete_redistribute 329 12.2 0.283 0.287 1.984 2.834 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.679 2.523 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.200 2.512 apply_single 119 13.6 0.000 0.000 2.200 2.512 fft_wrap_pw1pw2_140 487 13.2 0.651 0.653 2.493 2.502 make_m2s 4572 13.5 0.037 0.038 2.306 2.449 make_images 4572 14.5 0.356 0.389 2.185 2.328 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.442 2.279 mp_alltoall_i22 627 13.8 1.444 2.278 1.444 2.278 density_rs2pw 119 9.7 0.003 0.004 2.226 2.241 calculate_dm_sparse 119 9.5 0.000 0.000 2.216 2.234 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.379 2.209 calculate_first_density_matrix 1 7.0 0.000 0.000 2.179 2.180 grid_integrate_task_list 119 12.3 2.073 2.108 2.073 2.108 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.046 2.097 ot_diis_step 108 11.5 0.014 0.015 2.010 2.010 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.965 1.966 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.851 1.852 fft3d_ps 1201 14.6 0.869 0.887 1.843 1.850 acc_transpose_blocks 9144 15.5 0.044 0.045 1.808 1.818 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.779 1.779 mp_sum_l 7287 12.8 1.002 1.734 1.002 1.734 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.609 1.610 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.575 1.588 grid_collocate_task_list 119 9.7 1.516 1.543 1.516 1.543 cp_fm_cholesky_invert 11 10.9 1.411 1.414 1.411 1.414 wfi_extrapolate 11 7.9 0.001 0.001 1.405 1.405 potential_pw2rs 119 12.3 0.014 0.014 1.359 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.329 1.329 cp_fm_diag_elpa_base 50 14.0 1.182 1.236 1.327 1.327 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.320 hybrid_alltoall_any 4725 16.4 0.093 0.152 1.040 1.239 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.239 mp_alltoall_d11v 2130 13.8 1.179 1.207 1.179 1.207 make_images_data 4572 15.5 0.046 0.049 0.954 1.131 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.104 1.129 acc_transpose_blocks_sync 27432 16.5 1.103 1.119 1.103 1.119 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.996 1.056 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.927 0.941 qs_create_task_list 11 7.9 0.000 0.000 0.927 0.940 generate_qs_task_list 11 8.9 0.369 0.388 0.926 0.939 mp_alltoall_z22v 1201 16.6 0.837 0.859 0.837 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.883000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=715.909091, yerr=13.283308 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 504.422400E+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 687382. 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 0.008 0.027 80.933 80.934 qs_mol_dyn_low 1 2.0 0.003 0.003 80.683 80.694 qs_forces 11 3.9 0.003 0.003 80.615 80.615 qs_energies 11 4.9 0.002 0.002 77.781 77.794 scf_env_do_scf 11 5.9 0.000 0.001 68.956 68.958 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.556 63.556 dbcsr_multiply_generic 2055 12.4 0.114 0.117 50.773 50.943 qs_scf_new_mos 99 7.5 0.000 0.001 46.728 46.840 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.727 46.840 ot_scf_mini 99 9.5 0.002 0.002 44.356 44.454 multiply_cannon 2055 13.4 0.183 0.187 42.535 43.186 velocity_verlet 10 3.0 0.001 0.002 42.708 42.709 multiply_cannon_loop 2055 14.4 1.794 1.846 41.614 42.292 ot_mini 99 10.5 0.001 0.001 26.349 26.445 qs_ot_get_derivative 99 11.5 0.001 0.001 19.549 19.636 multiply_cannon_multrec 49320 15.4 11.402 12.111 17.507 18.335 rebuild_ks_matrix 110 8.3 0.000 0.000 14.368 14.487 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.368 14.486 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.596 12.705 mp_waitall_1 220248 16.4 10.555 11.410 10.555 11.410 multiply_cannon_sync_h2d 49320 15.4 9.661 10.178 9.661 10.178 qs_ot_get_p 110 10.4 0.001 0.001 9.565 9.689 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.416 7.980 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.228 7.713 apply_single 110 13.6 0.000 0.001 7.227 7.713 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.472 7.281 sum_up_and_integrate 110 10.3 0.002 0.003 6.906 6.919 integrate_v_rspace 110 11.3 0.003 0.003 6.881 6.901 init_scf_run 11 5.9 0.000 0.001 6.802 6.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.802 6.802 ot_diis_step 99 11.5 0.006 0.006 6.583 6.583 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.362 6.378 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.211 6.326 calculate_rho_elec 110 8.6 0.020 0.025 6.210 6.326 dbcsr_mm_accdrv_process 87628 16.1 3.057 3.149 5.974 6.258 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.571 5.571 init_scf_loop 11 6.9 0.000 0.000 5.373 5.374 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.311 5.371 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.020 5.043 cp_fm_diag_elpa_base 48 14.0 5.007 5.029 5.018 5.040 mp_sum_l 6594 12.7 3.895 4.543 3.895 4.543 wfi_extrapolate 11 7.9 0.001 0.001 3.990 3.990 make_m2s 4110 13.4 0.060 0.064 3.874 3.985 make_images 4110 14.4 0.177 0.191 3.780 3.894 calculate_dm_sparse 110 9.5 0.001 0.001 3.735 3.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.652 3.656 density_rs2pw 110 9.6 0.004 0.004 3.241 3.454 grid_integrate_task_list 110 12.3 3.265 3.423 3.265 3.423 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.321 3.378 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.302 3.295 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.186 3.268 prepare_preconditioner 11 7.9 0.000 0.000 3.203 3.222 make_preconditioner 11 8.9 0.000 0.000 3.203 3.222 pw_transfer 1331 11.6 0.055 0.065 2.991 3.070 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.962 3.006 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.903 2.981 calculate_first_density_matrix 1 7.0 0.000 0.000 2.709 2.717 jit_kernel_multiply 13 15.9 2.639 2.654 2.639 2.654 acc_transpose_blocks 49320 15.4 0.209 0.217 2.566 2.638 fft_wrap_pw1pw2_140 451 13.1 0.311 0.333 2.495 2.584 potential_pw2rs 110 12.3 0.005 0.006 2.505 2.519 mp_alltoall_d11v 2046 13.8 2.053 2.497 2.053 2.497 grid_collocate_task_list 110 9.6 2.160 2.264 2.160 2.264 fft3d_ps 1111 14.6 0.795 0.886 2.196 2.252 mp_waitany 14300 13.8 1.800 2.042 1.800 2.042 transfer_rs2pw 451 10.6 0.005 0.006 1.827 2.027 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.901 1.913 make_images_data 4110 15.4 0.043 0.047 1.758 1.892 mp_sum_d 3891 11.9 1.351 1.821 1.351 1.821 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.512 1.798 cp_fm_cholesky_invert 11 10.9 1.735 1.738 1.735 1.738 transfer_pw2rs 451 13.1 0.006 0.007 1.692 1.703 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.644 1.672 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.934000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=3.544289 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 591.802368E+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 1223189. 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 0.009 0.027 68.503 68.503 qs_mol_dyn_low 1 2.0 0.003 0.004 68.293 68.302 qs_forces 11 3.9 0.003 0.003 68.226 68.226 qs_energies 11 4.9 0.002 0.002 64.913 64.917 scf_env_do_scf 11 5.9 0.000 0.001 56.249 56.252 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.801 48.801 dbcsr_multiply_generic 2055 12.4 0.114 0.117 37.862 38.039 velocity_verlet 10 3.0 0.001 0.002 35.724 35.726 qs_scf_new_mos 99 7.5 0.001 0.001 33.066 33.201 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.065 33.200 multiply_cannon 2055 13.4 0.226 0.245 31.311 32.249 ot_scf_mini 99 9.5 0.003 0.003 31.396 31.516 multiply_cannon_loop 2055 14.4 1.162 1.185 30.180 31.345 ot_mini 99 10.5 0.001 0.001 18.354 18.483 multiply_cannon_multrec 24660 15.4 6.968 8.807 14.068 16.036 rebuild_ks_matrix 110 8.3 0.000 0.000 13.458 13.635 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.458 13.635 qs_ot_get_derivative 99 11.5 0.001 0.001 12.603 12.724 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.838 11.993 mp_waitall_1 176588 16.5 7.542 9.994 7.542 9.994 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.149 8.026 multiply_cannon_sync_h2d 24660 15.4 6.458 7.622 6.458 7.622 init_scf_loop 11 6.9 0.000 0.000 7.414 7.415 dbcsr_mm_accdrv_process 52282 16.1 5.488 6.322 6.930 7.218 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.448 7.125 apply_single 110 13.6 0.000 0.001 6.448 7.125 qs_ot_get_p 110 10.4 0.001 0.001 6.399 6.543 sum_up_and_integrate 110 10.3 0.001 0.003 6.311 6.323 init_scf_run 11 5.9 0.000 0.001 6.305 6.305 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.304 6.305 integrate_v_rspace 110 11.3 0.003 0.003 6.284 6.295 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.817 5.830 calculate_rho_elec 110 8.6 0.039 0.047 5.817 5.829 ot_diis_step 99 11.5 0.010 0.010 5.703 5.703 prepare_preconditioner 11 7.9 0.000 0.000 5.428 5.452 make_preconditioner 11 8.9 0.000 0.000 5.428 5.452 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.718 5.440 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.005 5.153 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.513 4.533 make_m2s 4110 13.4 0.057 0.060 4.195 4.515 make_images 4110 14.4 0.405 0.465 4.083 4.399 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.052 4.053 pw_transfer 1331 11.6 0.066 0.071 3.510 3.650 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.403 3.545 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.496 3.511 cp_fm_diag_elpa_base 48 14.0 3.450 3.466 3.493 3.508 wfi_extrapolate 11 7.9 0.001 0.001 3.474 3.474 grid_integrate_task_list 110 12.3 3.162 3.379 3.162 3.379 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.251 3.314 density_rs2pw 110 9.6 0.004 0.004 3.077 3.242 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.203 3.204 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 2.927 3.070 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.884 2.954 calculate_first_density_matrix 1 7.0 0.000 0.000 2.736 2.738 fft3d_ps 1111 14.6 1.115 1.349 2.495 2.653 make_images_data 4110 15.4 0.049 0.053 2.298 2.650 hybrid_alltoall_any 4261 16.3 0.105 0.450 1.983 2.614 cp_fm_cholesky_invert 11 10.9 2.518 2.525 2.518 2.525 mp_sum_l 6594 12.7 1.723 2.434 1.723 2.434 grid_collocate_task_list 110 9.6 2.169 2.360 2.169 2.360 potential_pw2rs 110 12.3 0.008 0.009 2.200 2.206 jit_kernel_multiply 11 16.3 1.079 2.104 1.079 2.104 mp_alltoall_d11v 2046 13.8 1.724 2.086 1.724 2.086 acc_transpose_blocks 24660 15.4 0.115 0.118 1.973 2.000 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.959 1.981 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.855 1.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.775 1.794 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.761 1.755 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.567 1.577 transfer_rs2pw 451 10.6 0.006 0.007 1.392 1.559 mp_irecv_dv 57340 16.2 0.636 1.540 0.636 1.540 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.391 1.494 mp_allgather_i34 2055 14.4 0.432 1.459 0.432 1.459 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.503000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.545455, yerr=6.610785 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 664.199168E+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 0.026 0.047 59.521 59.537 qs_mol_dyn_low 1 2.0 0.004 0.004 59.201 59.213 qs_forces 11 3.9 0.003 0.003 59.132 59.133 qs_energies 11 4.9 0.001 0.002 55.931 55.935 scf_env_do_scf 11 5.9 0.000 0.001 48.082 48.082 scf_env_do_scf_inner_loop 99 6.5 0.004 0.010 39.616 39.617 velocity_verlet 10 3.0 0.001 0.002 32.328 32.329 dbcsr_multiply_generic 2055 12.4 0.108 0.110 28.565 28.812 qs_scf_new_mos 99 7.5 0.001 0.001 25.085 25.191 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.085 25.190 ot_scf_mini 99 9.5 0.002 0.003 23.838 23.952 multiply_cannon 2055 13.4 0.213 0.224 22.399 23.599 multiply_cannon_loop 2055 14.4 0.818 0.843 21.233 22.393 ot_mini 99 10.5 0.001 0.001 13.627 13.738 rebuild_ks_matrix 110 8.3 0.000 0.000 12.051 12.178 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.051 12.177 multiply_cannon_multrec 16440 15.4 3.691 5.199 9.885 11.302 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.599 10.711 mp_waitall_1 139946 16.5 6.670 9.931 6.670 9.931 qs_ot_get_derivative 99 11.5 0.001 0.001 9.207 9.321 init_scf_loop 11 6.9 0.000 0.000 8.431 8.431 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.193 7.225 prepare_preconditioner 11 7.9 0.000 0.000 6.679 6.697 make_preconditioner 11 8.9 0.000 0.000 6.679 6.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.036 6.406 sum_up_and_integrate 110 10.3 0.001 0.002 6.213 6.229 integrate_v_rspace 110 11.3 0.003 0.003 6.187 6.204 dbcsr_mm_accdrv_process 34862 16.1 5.269 5.633 6.036 6.194 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.701 5.709 calculate_rho_elec 110 8.6 0.059 0.059 5.700 5.708 qs_ot_get_p 110 10.4 0.001 0.001 5.409 5.551 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.944 5.401 apply_single 110 13.6 0.000 0.000 4.943 5.400 init_scf_run 11 5.9 0.000 0.001 5.375 5.375 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.375 5.375 make_m2s 4110 13.4 0.049 0.051 4.111 4.505 ot_diis_step 99 11.5 0.011 0.011 4.391 4.391 make_images 4110 14.4 0.398 0.518 3.995 4.390 multiply_cannon_sync_h2d 16440 15.4 3.285 4.082 3.285 4.082 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.136 3.778 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.764 3.767 pw_transfer 1331 11.6 0.065 0.072 3.606 3.618 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.499 3.514 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.398 3.398 grid_integrate_task_list 110 12.3 3.197 3.361 3.197 3.361 density_rs2pw 110 9.6 0.004 0.004 2.948 3.132 fft_wrap_pw1pw2_140 451 13.1 0.460 0.472 3.042 3.061 wfi_extrapolate 11 7.9 0.001 0.001 2.936 2.936 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.868 2.878 cp_fm_diag_elpa_base 48 14.0 2.801 2.833 2.866 2.875 make_images_data 4110 15.4 0.047 0.051 2.291 2.770 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.737 2.738 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.040 2.715 cp_fm_cholesky_invert 11 10.9 2.565 2.571 2.565 2.571 calculate_dm_sparse 110 9.5 0.001 0.001 2.517 2.553 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.419 2.486 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.429 2.475 fft3d_ps 1111 14.6 1.118 1.125 2.416 2.431 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.877 2.422 grid_collocate_task_list 110 9.6 2.213 2.407 2.213 2.407 calculate_first_density_matrix 1 7.0 0.000 0.000 2.350 2.351 mp_irecv_dv 48980 15.7 0.802 2.286 0.802 2.286 potential_pw2rs 110 12.3 0.011 0.011 2.069 2.076 qs_energies_init_hamiltonians 11 5.9 0.008 0.023 2.039 2.039 mp_sum_l 6594 12.7 1.431 2.022 1.431 2.022 mp_alltoall_d11v 2046 13.8 1.714 1.981 1.714 1.981 dbcsr_complete_redistribute 325 12.2 0.323 0.349 1.412 1.876 cp_fm_upper_to_full 70 14.2 1.452 1.828 1.452 1.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.643 1.658 acc_transpose_blocks 16440 15.4 0.079 0.081 1.584 1.637 cp_fm_cholesky_decompose 22 10.9 1.544 1.565 1.544 1.565 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.383 1.516 mp_allgather_i34 2055 14.4 0.479 1.495 0.479 1.495 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.456 1.470 transfer_rs2pw 451 10.6 0.005 0.006 1.253 1.454 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.928 1.391 mp_waitany 17072 13.8 1.115 1.349 1.115 1.349 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.263 1.272 qs_env_update_s_mstruct 11 6.9 0.012 0.016 1.114 1.210 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.537000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.454545, yerr=7.808662 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 743.555072E+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 0.010 0.031 65.496 65.497 qs_mol_dyn_low 1 2.0 0.003 0.003 65.272 65.283 qs_forces 11 3.9 0.003 0.003 65.201 65.203 qs_energies 11 4.9 0.001 0.001 61.803 61.808 scf_env_do_scf 11 5.9 0.000 0.001 53.616 53.619 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.985 41.985 velocity_verlet 10 3.0 0.001 0.002 37.208 37.210 dbcsr_multiply_generic 2055 12.4 0.116 0.122 30.457 30.612 qs_scf_new_mos 99 7.5 0.001 0.001 27.406 27.516 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.405 27.516 ot_scf_mini 99 9.5 0.002 0.003 25.727 25.830 multiply_cannon 2055 13.4 0.237 0.257 23.574 24.526 multiply_cannon_loop 2055 14.4 1.387 1.488 22.004 22.532 ot_mini 99 10.5 0.001 0.001 14.698 14.821 multiply_cannon_multrec 24660 15.4 4.145 6.824 13.064 14.264 rebuild_ks_matrix 110 8.3 0.000 0.000 11.936 12.030 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.935 12.030 init_scf_loop 11 6.9 0.000 0.000 11.590 11.591 qs_ot_get_derivative 99 11.5 0.001 0.001 10.558 10.664 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.541 10.623 prepare_preconditioner 11 7.9 0.000 0.000 9.873 9.889 make_preconditioner 11 8.9 0.000 0.000 9.873 9.889 dbcsr_mm_accdrv_process 52304 16.0 7.754 9.222 8.760 9.691 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.106 9.586 sum_up_and_integrate 110 10.3 0.001 0.002 6.203 6.215 integrate_v_rspace 110 11.3 0.003 0.003 6.177 6.189 qs_ot_get_p 110 10.4 0.001 0.001 5.923 6.085 mp_waitall_1 121746 16.5 3.955 5.893 3.955 5.893 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.834 5.845 calculate_rho_elec 110 8.6 0.078 0.081 5.834 5.845 make_m2s 4110 13.4 0.059 0.061 5.277 5.764 make_images 4110 14.4 0.574 0.697 5.135 5.620 init_scf_run 11 5.9 0.000 0.001 5.451 5.451 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.451 5.451 cp_fm_upper_to_full 70 14.2 3.395 4.924 3.395 4.924 ot_diis_step 99 11.5 0.011 0.011 4.104 4.104 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.030 4.100 apply_single 110 13.6 0.000 0.000 4.029 4.100 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.051 4.069 dbcsr_complete_redistribute 325 12.2 0.411 0.460 2.648 3.809 pw_transfer 1331 11.6 0.065 0.073 3.759 3.795 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.652 3.692 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.515 3.566 cp_dbcsr_syevd 48 12.0 0.007 0.008 3.485 3.485 grid_integrate_task_list 110 12.3 3.262 3.415 3.262 3.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.337 3.397 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.107 3.257 fft_wrap_pw1pw2_140 451 13.1 0.534 0.548 3.146 3.191 density_rs2pw 110 9.6 0.004 0.004 2.964 3.070 hybrid_alltoall_any 4261 16.3 0.124 0.464 2.214 3.034 make_images_data 4110 15.4 0.049 0.053 2.546 3.031 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.290 3.030 calculate_dm_sparse 110 9.5 0.001 0.001 2.983 3.018 wfi_extrapolate 11 7.9 0.001 0.001 2.987 2.987 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.946 2.957 cp_fm_diag_elpa_base 48 14.0 2.794 2.855 2.943 2.952 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.758 2.892 mp_alltoall_i22 605 13.7 1.667 2.845 1.667 2.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.779 2.781 cp_fm_cholesky_invert 11 10.9 2.670 2.678 2.670 2.678 acc_transpose_blocks 24660 15.4 0.111 0.113 2.391 2.498 fft3d_ps 1111 14.6 1.156 1.195 2.473 2.489 multiply_cannon_sync_h2d 24660 15.4 2.363 2.474 2.363 2.474 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.426 2.463 grid_collocate_task_list 110 9.6 2.264 2.398 2.264 2.398 calculate_first_density_matrix 1 7.0 0.000 0.000 2.369 2.372 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.246 2.248 potential_pw2rs 110 12.3 0.013 0.013 2.027 2.032 mp_alltoall_d11v 2046 13.8 1.731 1.844 1.731 1.844 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.784 1.832 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.616 1.715 cp_fm_cholesky_decompose 22 10.9 1.625 1.662 1.625 1.662 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.634 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.546 1.563 mp_allgather_i34 2055 14.4 0.644 1.539 0.644 1.539 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.833 1.508 acc_transpose_blocks_sync 73980 16.4 1.388 1.497 1.388 1.497 mp_sum_l 6594 12.7 0.916 1.477 0.916 1.477 jit_kernel_multiply 8 15.7 0.664 1.435 0.664 1.435 mp_irecv_dv 62702 16.1 0.730 1.427 0.730 1.427 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.497000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=702.909091, yerr=11.743118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 856.330240E+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 0.009 0.028 55.176 55.178 qs_mol_dyn_low 1 2.0 0.003 0.004 54.938 54.949 qs_forces 11 3.9 0.003 0.008 54.873 54.873 qs_energies 11 4.9 0.001 0.002 51.208 51.212 scf_env_do_scf 11 5.9 0.000 0.001 43.066 43.066 scf_env_do_scf_inner_loop 99 6.5 0.004 0.008 35.283 35.284 velocity_verlet 10 3.0 0.002 0.003 31.250 31.253 dbcsr_multiply_generic 2055 12.4 0.106 0.109 23.121 23.322 qs_scf_new_mos 99 7.5 0.001 0.001 20.616 20.689 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.615 20.689 ot_scf_mini 99 9.5 0.002 0.002 19.368 19.425 multiply_cannon 2055 13.4 0.239 0.251 17.647 18.810 multiply_cannon_loop 2055 14.4 0.604 0.628 16.421 16.791 rebuild_ks_matrix 110 8.3 0.000 0.000 11.674 11.748 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.674 11.748 ot_mini 99 10.5 0.001 0.001 10.722 10.782 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.384 10.453 multiply_cannon_multrec 8220 15.4 3.148 4.870 7.547 8.883 init_scf_loop 11 6.9 0.000 0.000 7.738 7.739 mp_waitall_1 103326 16.6 5.885 7.573 5.885 7.573 qs_ot_get_derivative 99 11.5 0.001 0.001 6.955 7.011 sum_up_and_integrate 110 10.3 0.001 0.002 6.253 6.266 integrate_v_rspace 110 11.3 0.003 0.003 6.226 6.239 prepare_preconditioner 11 7.9 0.000 0.000 6.093 6.100 make_preconditioner 11 8.9 0.000 0.000 6.092 6.099 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.068 6.087 calculate_rho_elec 110 8.6 0.113 0.113 6.067 6.086 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.669 5.747 dbcsr_mm_accdrv_process 17442 15.9 3.189 4.134 4.259 5.171 init_scf_run 11 5.9 0.000 0.001 5.019 5.019 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.019 5.019 qs_ot_get_p 110 10.4 0.001 0.001 4.703 4.781 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.127 4.703 make_m2s 4110 13.4 0.038 0.039 4.120 4.305 make_images 4110 14.4 0.645 0.711 3.990 4.171 pw_transfer 1331 11.6 0.066 0.070 4.107 4.118 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.999 4.013 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.802 3.861 apply_single 110 13.6 0.000 0.000 3.801 3.861 ot_diis_step 99 11.5 0.012 0.012 3.745 3.745 grid_integrate_task_list 110 12.3 3.383 3.504 3.383 3.504 fft_wrap_pw1pw2_140 451 13.1 0.721 0.735 3.469 3.490 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.331 3.335 density_rs2pw 110 9.6 0.004 0.004 3.042 3.123 cp_fm_cholesky_invert 11 10.9 2.990 2.993 2.990 2.993 cp_dbcsr_syevd 48 12.0 0.004 0.004 2.986 2.986 wfi_extrapolate 11 7.9 0.001 0.001 2.690 2.690 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.664 2.665 hybrid_alltoall_any 4261 16.3 0.201 0.855 2.202 2.575 fft3d_ps 1111 14.6 1.294 1.303 2.540 2.551 calculate_dm_sparse 110 9.5 0.001 0.001 2.476 2.516 multiply_cannon_sync_h2d 8220 15.4 2.337 2.483 2.337 2.483 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.476 2.477 grid_collocate_task_list 110 9.6 2.371 2.449 2.371 2.449 make_images_data 4110 15.4 0.040 0.047 2.181 2.442 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.416 2.424 cp_fm_diag_elpa_base 48 14.0 2.361 2.383 2.415 2.422 calculate_first_density_matrix 1 7.0 0.000 0.000 2.234 2.235 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.117 2.144 potential_pw2rs 110 12.3 0.015 0.016 2.055 2.057 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.998 2.041 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 2.005 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.899 1.931 mp_alltoall_d11v 2046 13.8 1.718 1.851 1.718 1.851 cp_fm_cholesky_decompose 22 10.9 1.669 1.691 1.669 1.691 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.619 1.624 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.492 1.602 dbcsr_complete_redistribute 325 12.2 0.554 0.578 1.415 1.505 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.487 1.490 mp_irecv_dv 24056 15.7 0.461 1.447 0.461 1.447 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.436 1.446 mp_allgather_i34 2055 14.4 0.393 1.412 0.393 1.412 acc_transpose_blocks 8220 15.4 0.041 0.042 1.361 1.398 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.743 1.340 qs_create_task_list 11 7.9 0.001 0.001 1.202 1.300 generate_qs_task_list 11 8.9 0.373 0.440 1.201 1.299 transfer_rs2pw 451 10.6 0.005 0.005 1.115 1.216 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.180 1.212 mp_waitany 9240 13.8 1.067 1.173 1.067 1.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.178000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=808.727273, yerr=10.127289 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 1.381605E+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 0.016 0.056 87.824 87.847 qs_mol_dyn_low 1 2.0 0.003 0.004 87.522 87.536 qs_forces 11 3.9 0.003 0.003 87.438 87.438 qs_energies 11 4.9 0.001 0.002 83.241 83.243 scf_env_do_scf 11 5.9 0.000 0.001 73.222 73.222 velocity_verlet 10 3.0 0.002 0.002 56.156 56.165 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.430 44.432 dbcsr_multiply_generic 2055 12.4 0.121 0.126 29.243 29.332 init_scf_loop 11 6.9 0.000 0.000 28.722 28.726 qs_scf_new_mos 99 7.5 0.001 0.001 26.568 26.602 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.567 26.601 prepare_preconditioner 11 7.9 0.000 0.000 26.566 26.574 make_preconditioner 11 8.9 0.000 0.000 26.566 26.574 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.531 26.041 ot_scf_mini 99 9.5 0.002 0.002 24.788 24.815 multiply_cannon 2055 13.4 0.334 0.363 22.288 23.030 multiply_cannon_loop 2055 14.4 0.824 0.840 20.529 20.825 cp_fm_upper_to_full 70 14.2 12.909 18.673 12.909 18.673 rebuild_ks_matrix 110 8.3 0.001 0.001 14.066 14.100 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.065 14.100 ot_mini 99 10.5 0.001 0.001 13.977 14.005 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.773 12.804 dbcsr_complete_redistribute 325 12.2 1.013 1.035 7.576 10.769 multiply_cannon_multrec 8220 15.4 4.073 4.223 9.738 9.856 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.574 9.766 qs_ot_get_derivative 99 11.5 0.001 0.001 9.422 9.451 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.019 9.185 mp_alltoall_i22 605 13.7 5.635 8.849 5.635 8.849 mp_waitall_1 84994 16.7 7.309 8.122 7.309 8.122 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.443 7.480 calculate_rho_elec 110 8.6 0.223 0.223 7.442 7.480 sum_up_and_integrate 110 10.3 0.002 0.002 6.903 6.917 integrate_v_rspace 110 11.3 0.004 0.004 6.874 6.888 pw_transfer 1331 11.6 0.075 0.075 5.802 5.809 make_m2s 4110 13.4 0.043 0.043 5.268 5.743 dbcsr_mm_accdrv_process 11614 15.7 3.757 4.107 5.517 5.715 init_scf_run 11 5.9 0.000 0.001 5.713 5.713 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.713 5.713 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.684 5.692 qs_ot_get_p 110 10.4 0.001 0.001 5.544 5.578 make_images 4110 14.4 0.886 0.945 5.079 5.551 cp_fm_cholesky_invert 11 10.9 5.176 5.181 5.176 5.181 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.702 5.167 apply_single 110 13.6 0.000 0.000 4.702 5.167 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.748 5.063 fft_wrap_pw1pw2_140 451 13.1 1.368 1.373 5.002 5.012 ot_diis_step 99 11.5 0.015 0.015 4.542 4.542 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.996 4.002 density_rs2pw 110 9.6 0.004 0.004 3.914 3.943 grid_integrate_task_list 110 12.3 3.679 3.752 3.679 3.752 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.683 3.683 hybrid_alltoall_any 4261 16.3 0.271 0.574 2.789 3.516 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.479 3.481 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.978 3.432 fft3d_ps 1111 14.6 1.895 1.913 3.392 3.402 make_images_data 4110 15.4 0.045 0.048 2.751 3.361 wfi_extrapolate 11 7.9 0.001 0.001 3.323 3.323 calculate_dm_sparse 110 9.5 0.001 0.001 3.164 3.191 multiply_cannon_sync_h2d 8220 15.4 3.127 3.157 3.127 3.157 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.896 2.896 cp_fm_diag_elpa_base 48 14.0 2.348 2.550 2.894 2.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.852 2.854 grid_collocate_task_list 110 9.6 2.668 2.684 2.668 2.684 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.630 2.647 potential_pw2rs 110 12.3 0.021 0.022 2.440 2.444 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.355 2.367 calculate_first_density_matrix 1 7.0 0.000 0.000 2.273 2.274 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.193 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.119 2.223 mp_alltoall_d11v 2046 13.8 2.007 2.064 2.007 2.064 cp_fm_cholesky_decompose 22 10.9 1.941 1.962 1.941 1.962 qs_create_task_list 11 7.9 0.001 0.001 1.878 1.921 generate_qs_task_list 11 8.9 0.731 0.786 1.877 1.920 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.891 1.895 acc_transpose_blocks 8220 15.4 0.041 0.042 1.735 1.779 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.734 1.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.847000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1245.181818, yerr=56.875971 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.693312E+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 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2529087. 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 0.063 0.087 205.678 205.679 qs_mol_dyn_low 1 2.0 0.004 0.004 205.183 205.197 qs_forces 11 3.9 0.004 0.005 205.094 205.095 qs_energies 11 4.9 0.001 0.002 199.443 199.455 scf_env_do_scf 11 5.9 0.001 0.001 182.978 182.981 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.814 162.816 dbcsr_multiply_generic 2507 12.6 0.178 0.182 125.855 126.550 qs_scf_new_mos 117 7.6 0.001 0.001 123.669 123.827 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.668 123.827 velocity_verlet 10 3.0 0.001 0.002 122.812 122.814 ot_scf_mini 117 9.6 0.003 0.003 116.943 117.157 multiply_cannon 2507 13.6 0.237 0.245 101.906 103.182 multiply_cannon_loop 2507 14.6 2.400 2.457 99.774 101.048 ot_mini 117 10.6 0.001 0.001 66.223 66.455 multiply_cannon_multrec 60168 15.6 32.006 33.908 41.775 43.158 qs_ot_get_derivative 117 11.6 0.001 0.001 41.221 41.468 rebuild_ks_matrix 128 8.3 0.001 0.001 33.442 33.761 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.442 33.760 mp_waitall_1 267128 16.5 28.846 31.624 28.846 31.624 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.027 30.345 qs_ot_get_p 128 10.4 0.001 0.001 28.365 28.720 multiply_cannon_sync_h2d 60168 15.6 26.428 28.112 26.428 28.112 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.517 25.710 apply_single 128 13.6 0.001 0.001 24.517 25.709 ot_diis_step 117 11.6 0.008 0.008 24.754 24.755 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.730 21.786 init_scf_loop 11 6.9 0.000 0.001 20.091 20.093 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.626 19.823 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.988 18.989 multiply_cannon_metrocomm3 60168 15.6 0.115 0.119 16.201 18.459 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.096 16.138 cp_fm_diag_elpa_base 83 14.4 16.030 16.070 16.092 16.134 prepare_preconditioner 11 7.9 0.000 0.000 15.504 15.555 make_preconditioner 11 8.9 0.000 0.000 15.504 15.555 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.732 14.938 make_m2s 5014 13.6 0.104 0.112 13.746 14.126 make_images 5014 14.6 0.397 0.417 13.561 13.947 sum_up_and_integrate 128 10.3 0.002 0.004 13.841 13.856 integrate_v_rspace 128 11.3 0.004 0.010 13.783 13.800 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.945 13.078 calculate_rho_elec 128 8.7 0.045 0.064 12.944 13.078 init_scf_run 11 5.9 0.000 0.001 12.269 12.270 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.269 12.270 mp_sum_l 7950 12.9 8.938 10.450 8.938 10.450 dbcsr_mm_accdrv_process 124484 16.2 4.725 4.866 9.332 9.854 wfi_extrapolate 11 7.9 0.001 0.001 9.004 9.004 cp_fm_cholesky_invert 11 10.9 8.747 8.754 8.747 8.754 calculate_dm_sparse 128 9.5 0.001 0.001 8.543 8.655 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.159 8.295 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.132 8.261 multiply_cannon_metrocomm1 60168 15.6 0.091 0.095 6.292 8.151 pw_transfer 1547 11.6 0.073 0.089 7.485 7.653 make_images_data 5014 15.6 0.066 0.073 6.725 7.651 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.284 7.452 grid_integrate_task_list 128 12.3 7.025 7.411 7.025 7.411 hybrid_alltoall_any 5200 16.5 0.293 2.256 5.894 7.203 density_rs2pw 128 9.7 0.006 0.006 6.589 7.146 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.699 6.711 fft_wrap_pw1pw2_140 523 13.2 0.858 0.898 6.434 6.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.807 5.876 fft3d_ps 1291 14.7 2.193 2.839 5.291 5.606 mp_alltoall_d11v 2415 14.1 4.333 5.471 4.333 5.471 grid_collocate_task_list 128 9.7 4.836 5.276 4.836 5.276 cp_fm_cholesky_decompose 22 10.9 4.571 4.586 4.571 4.586 potential_pw2rs 128 12.3 0.009 0.010 4.336 4.352 mp_sum_d 4481 12.1 3.531 4.273 3.531 4.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.679000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.090909, yerr=7.038266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 838.520832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5381322. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.060 188.754 188.755 qs_mol_dyn_low 1 2.0 0.004 0.004 188.255 188.268 qs_forces 11 3.9 0.004 0.005 188.146 188.148 qs_energies 11 4.9 0.005 0.033 181.320 181.341 scf_env_do_scf 11 5.9 0.001 0.002 164.988 164.998 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.487 132.490 velocity_verlet 10 3.0 0.020 0.031 118.552 118.554 dbcsr_multiply_generic 2507 12.6 0.187 0.192 97.295 98.502 qs_scf_new_mos 117 7.6 0.001 0.001 94.007 94.649 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.006 94.649 ot_scf_mini 117 9.6 0.004 0.004 89.182 89.899 multiply_cannon 2507 13.6 0.522 0.575 77.224 81.597 multiply_cannon_loop 2507 14.6 1.567 1.637 73.887 76.735 ot_mini 117 10.6 0.001 0.001 49.923 50.566 mp_waitall_1 214728 16.6 23.813 37.360 23.813 37.360 multiply_cannon_multrec 30084 15.6 21.083 25.142 31.664 36.047 rebuild_ks_matrix 128 8.3 0.001 0.001 32.277 33.000 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.043 32.277 32.999 init_scf_loop 11 6.9 0.001 0.004 32.413 32.414 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.016 29.667 qs_ot_get_derivative 117 11.6 0.001 0.002 28.039 28.741 multiply_cannon_metrocomm3 30084 15.6 0.098 0.105 15.427 28.555 prepare_preconditioner 11 7.9 0.000 0.000 28.102 28.168 make_preconditioner 11 8.9 0.000 0.001 28.102 28.168 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.834 27.374 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.018 23.076 apply_single 128 13.6 0.001 0.001 22.018 23.076 qs_ot_get_p 128 10.4 0.002 0.012 21.187 21.966 ot_diis_step 117 11.6 0.014 0.015 21.709 21.711 multiply_cannon_sync_h2d 30084 15.6 18.188 20.371 18.188 20.371 cp_fm_cholesky_invert 11 10.9 16.462 16.475 16.462 16.475 qs_ot_p2m_diag 83 11.4 0.193 0.252 16.304 16.358 make_m2s 5014 13.6 0.086 0.091 14.126 15.768 make_images 5014 14.6 1.159 1.366 13.911 15.554 cp_dbcsr_syevd 83 12.4 0.007 0.008 15.102 15.104 sum_up_and_integrate 128 10.3 0.002 0.004 13.910 13.945 integrate_v_rspace 128 11.3 0.003 0.004 13.850 13.889 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.310 13.345 calculate_rho_elec 128 8.7 0.086 0.103 13.309 13.344 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.947 11.979 cp_fm_diag_elpa_base 83 14.4 11.695 11.786 11.941 11.968 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.134 11.648 init_scf_run 11 5.9 0.000 0.001 11.481 11.482 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.481 11.482 multiply_cannon_metrocomm4 27577 15.6 0.099 0.112 3.814 10.870 dbcsr_mm_accdrv_process 62242 16.2 5.478 6.181 10.030 10.585 mp_irecv_dv 69486 16.3 3.613 10.481 3.613 10.481 make_images_data 5014 15.6 0.067 0.077 8.389 10.424 hybrid_alltoall_any 5200 16.5 0.350 1.558 7.102 9.925 pw_transfer 1547 11.6 0.085 0.096 8.505 8.559 wfi_extrapolate 11 7.9 0.001 0.001 8.346 8.346 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.281 8.339 grid_integrate_task_list 128 12.3 7.195 7.506 7.195 7.506 density_rs2pw 128 9.7 0.006 0.006 7.008 7.499 fft_wrap_pw1pw2_140 523 13.2 0.934 0.971 7.334 7.411 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.208 6.967 cp_fm_cholesky_decompose 22 10.9 6.834 6.913 6.834 6.913 calculate_dm_sparse 128 9.5 0.001 0.001 6.481 6.615 mp_sum_l 7950 12.9 4.177 6.299 4.177 6.299 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.114 6.124 fft3d_ps 1291 14.7 2.829 2.999 5.840 5.896 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.390 5.469 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.214 5.412 grid_collocate_task_list 128 9.7 5.006 5.367 5.006 5.367 mp_allgather_i34 2507 14.6 1.839 5.089 1.839 5.089 mp_alltoall_d11v 2415 14.1 4.044 4.471 4.044 4.471 potential_pw2rs 128 12.3 0.015 0.018 4.428 4.446 mp_sum_d 4472 12.1 2.640 3.951 2.640 3.951 dbcsr_complete_redistribute 395 12.7 0.772 0.858 3.031 3.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.755000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.363636, yerr=1.720081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 971.988992E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. 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 0.051 0.137 175.075 175.080 qs_mol_dyn_low 1 2.0 0.004 0.006 174.441 174.508 qs_forces 11 3.9 0.004 0.004 174.335 174.343 qs_energies 11 4.9 0.004 0.023 167.770 167.788 scf_env_do_scf 11 5.9 0.003 0.018 152.128 152.131 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.098 117.099 velocity_verlet 10 3.0 0.002 0.003 112.107 112.109 dbcsr_multiply_generic 2507 12.6 0.184 0.189 81.905 83.292 qs_scf_new_mos 117 7.6 0.001 0.001 80.195 80.532 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.194 80.531 ot_scf_mini 117 9.6 0.003 0.004 75.938 76.329 multiply_cannon 2507 13.6 0.507 0.522 62.214 66.296 multiply_cannon_loop 2507 14.6 1.133 1.192 59.169 62.355 ot_mini 117 10.6 0.001 0.001 42.522 42.901 init_scf_loop 11 6.9 0.001 0.003 34.921 34.925 mp_waitall_1 170520 16.6 24.384 33.636 24.384 33.636 prepare_preconditioner 11 7.9 0.000 0.000 30.911 30.965 make_preconditioner 11 8.9 0.000 0.002 30.911 30.965 rebuild_ks_matrix 128 8.3 0.001 0.001 30.092 30.644 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.028 30.091 30.644 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.494 29.946 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.117 27.619 multiply_cannon_multrec 20056 15.6 13.127 16.045 22.495 25.450 multiply_cannon_metrocomm3 20056 15.6 0.061 0.064 15.135 24.897 qs_ot_get_derivative 117 11.6 0.001 0.002 22.967 23.349 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.631 20.524 apply_single 128 13.6 0.001 0.001 19.630 20.524 ot_diis_step 117 11.6 0.018 0.018 19.444 19.444 qs_ot_get_p 128 10.4 0.002 0.010 18.780 19.289 make_m2s 5014 13.6 0.078 0.082 14.602 15.894 make_images 5014 14.6 1.146 1.231 14.365 15.654 multiply_cannon_sync_h2d 20056 15.6 13.711 15.535 13.711 15.535 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.449 14.457 cp_fm_cholesky_invert 11 10.9 14.220 14.228 14.220 14.228 sum_up_and_integrate 128 10.3 0.002 0.004 13.804 13.827 integrate_v_rspace 128 11.3 0.003 0.004 13.744 13.770 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.519 13.560 calculate_rho_elec 128 8.7 0.128 0.143 13.519 13.559 cp_dbcsr_syevd 83 12.4 0.006 0.008 13.401 13.402 init_scf_run 11 5.9 0.000 0.001 10.486 10.487 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.486 10.486 make_images_data 5014 15.6 0.064 0.073 8.725 10.485 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.251 10.277 cp_fm_diag_elpa_base 83 14.4 9.830 9.992 10.247 10.272 hybrid_alltoall_any 5200 16.5 0.449 2.039 7.602 9.871 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.051 9.337 multiply_cannon_metrocomm4 17549 15.6 0.065 0.074 3.451 9.113 dbcsr_mm_accdrv_process 41502 16.2 5.697 6.008 8.824 8.950 mp_irecv_dv 50230 16.2 3.322 8.858 3.322 8.858 pw_transfer 1547 11.6 0.084 0.102 8.715 8.813 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.492 8.596 grid_integrate_task_list 128 12.3 7.320 7.789 7.320 7.789 fft_wrap_pw1pw2_140 523 13.2 1.063 1.102 7.559 7.675 cp_fm_upper_to_full 105 14.8 5.913 7.478 5.913 7.478 wfi_extrapolate 11 7.9 0.003 0.018 7.446 7.446 cp_fm_cholesky_decompose 22 10.9 7.300 7.364 7.300 7.364 density_rs2pw 128 9.7 0.006 0.006 6.883 7.193 dbcsr_complete_redistribute 395 12.7 1.160 1.188 4.623 6.456 calculate_dm_sparse 128 9.5 0.001 0.001 5.854 5.941 fft3d_ps 1291 14.7 2.822 3.053 5.806 5.899 grid_collocate_task_list 128 9.7 5.151 5.697 5.151 5.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.460 5.468 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.479 5.311 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.554 5.213 mp_sum_l 7950 12.9 3.276 5.050 3.276 5.050 mp_alltoall_d11v 2415 14.1 4.263 4.795 4.263 4.795 mp_allgather_i34 2507 14.6 1.539 4.723 1.539 4.723 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.563 4.714 potential_pw2rs 128 12.3 0.020 0.022 4.262 4.270 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.398 4.214 mp_alltoall_i22 716 14.1 2.024 4.043 2.024 4.043 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.924 3.960 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.809 3.817 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.471 3.522 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.080000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=923.000000, yerr=10.846030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/18/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 4.320339E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.167483E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 186.428 186.430 qs_mol_dyn_low 1 2.0 0.003 0.003 186.067 186.081 qs_forces 11 3.9 0.004 0.004 185.956 185.962 qs_energies 11 4.9 0.002 0.002 178.825 178.831 scf_env_do_scf 11 5.9 0.001 0.001 162.125 162.134 velocity_verlet 10 3.0 0.002 0.002 123.226 123.229 scf_env_do_scf_inner_loop 116 6.6 0.006 0.012 115.732 115.733 dbcsr_multiply_generic 2485 12.5 0.190 0.195 79.598 80.292 qs_scf_new_mos 116 7.6 0.001 0.001 79.838 80.185 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.837 80.185 ot_scf_mini 116 9.6 0.003 0.003 75.316 75.662 multiply_cannon 2485 13.5 0.548 0.581 55.293 57.503 multiply_cannon_loop 2485 14.5 1.778 1.918 51.581 53.302 init_scf_loop 11 6.9 0.000 0.000 46.273 46.274 ot_mini 116 10.6 0.001 0.001 42.499 42.836 prepare_preconditioner 11 7.9 0.000 0.000 42.185 42.209 make_preconditioner 11 8.9 0.000 0.000 42.185 42.208 make_full_inverse_cholesky 11 9.9 0.012 0.024 35.692 40.725 multiply_cannon_multrec 29820 15.5 13.511 18.921 26.315 31.430 rebuild_ks_matrix 127 8.3 0.001 0.001 29.048 29.313 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.048 29.313 mp_waitall_1 146592 16.7 16.784 26.885 16.784 26.885 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.143 26.384 qs_ot_get_derivative 116 11.6 0.001 0.002 22.980 23.324 make_m2s 4970 13.5 0.091 0.094 19.992 20.959 make_images 4970 14.5 1.927 2.244 19.691 20.661 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.978 19.458 apply_single 127 13.6 0.001 0.001 18.977 19.458 qs_ot_get_p 127 10.4 0.001 0.001 19.094 19.441 ot_diis_step 116 11.6 0.017 0.018 19.383 19.384 cp_fm_upper_to_full 104 14.8 11.538 16.957 11.538 16.957 cp_fm_cholesky_invert 11 10.9 16.001 16.010 16.001 16.010 multiply_cannon_metrocomm3 29820 15.5 0.048 0.051 6.245 15.143 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.008 15.061 sum_up_and_integrate 127 10.3 0.002 0.003 13.908 13.932 integrate_v_rspace 127 11.3 0.003 0.004 13.848 13.874 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.650 13.687 calculate_rho_elec 127 8.7 0.169 0.184 13.650 13.686 cp_dbcsr_syevd 82 12.4 0.011 0.012 13.628 13.630 dbcsr_complete_redistribute 393 12.7 1.505 1.616 9.214 13.004 dbcsr_mm_accdrv_process 61748 16.2 8.415 9.382 12.376 12.815 make_images_data 4970 15.5 0.065 0.071 10.690 12.388 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.863 11.669 multiply_cannon_sync_h2d 29820 15.5 10.425 11.441 10.425 11.441 hybrid_alltoall_any 5155 16.4 0.526 2.192 9.669 11.038 init_scf_run 11 5.9 0.000 0.001 10.814 10.816 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.814 10.815 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.474 10.489 cp_fm_diag_elpa_base 82 14.4 9.513 9.827 10.467 10.485 transfer_fm_to_dbcsr 11 9.9 0.001 0.009 6.470 10.219 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.570 9.830 mp_alltoall_i22 712 14.1 5.614 9.391 5.614 9.391 pw_transfer 1535 11.6 0.085 0.101 9.165 9.251 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.942 9.030 fft_wrap_pw1pw2_140 519 13.2 1.243 1.280 7.960 8.058 multiply_cannon_metrocomm4 24850 15.5 0.079 0.089 2.822 7.793 grid_integrate_task_list 127 12.3 7.435 7.718 7.435 7.718 wfi_extrapolate 11 7.9 0.001 0.001 7.658 7.658 cp_fm_cholesky_decompose 22 10.9 7.449 7.574 7.449 7.574 mp_irecv_dv 75445 16.2 2.670 7.509 2.670 7.509 density_rs2pw 127 9.7 0.005 0.006 6.790 7.122 calculate_dm_sparse 127 9.5 0.001 0.001 6.201 6.278 fft3d_ps 1281 14.7 2.976 3.057 6.009 6.076 grid_collocate_task_list 127 9.7 5.244 5.642 5.244 5.642 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.444 5.492 mp_alltoall_d11v 2401 14.1 4.944 5.487 4.944 5.487 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.436 4.519 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.358 4.438 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.401 4.402 potential_pw2rs 127 12.3 0.023 0.024 4.277 4.285 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.223 4.279 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.430000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1092.454545, yerr=26.158133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.533157E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591807936 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57807. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.049 170.368 170.384 qs_mol_dyn_low 1 2.0 0.003 0.004 169.957 169.996 qs_forces 11 3.9 0.004 0.005 169.863 169.865 qs_energies 11 4.9 0.002 0.002 162.431 162.436 scf_env_do_scf 11 5.9 0.001 0.001 145.323 145.336 velocity_verlet 10 3.0 0.002 0.002 111.752 111.756 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 110.041 110.042 dbcsr_multiply_generic 2527 12.6 0.184 0.191 72.475 72.900 qs_scf_new_mos 118 7.6 0.001 0.001 72.610 72.717 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.610 72.716 ot_scf_mini 118 9.6 0.003 0.004 68.160 68.259 multiply_cannon 2527 13.6 0.565 0.576 53.938 57.232 multiply_cannon_loop 2527 14.6 0.817 0.849 50.809 51.772 ot_mini 118 10.6 0.001 0.001 37.997 38.093 init_scf_loop 11 6.9 0.000 0.000 35.131 35.133 mp_waitall_1 126780 16.7 24.933 31.328 24.933 31.328 prepare_preconditioner 11 7.9 0.000 0.000 31.290 31.316 make_preconditioner 11 8.9 0.000 0.000 31.290 31.316 rebuild_ks_matrix 129 8.3 0.001 0.001 29.418 29.528 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.417 29.527 make_full_inverse_cholesky 11 9.9 0.015 0.028 29.252 29.518 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.851 26.953 multiply_cannon_multrec 10108 15.6 10.427 15.825 18.006 22.286 qs_ot_get_derivative 118 11.6 0.002 0.002 20.589 20.682 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 13.074 20.273 cp_fm_cholesky_invert 11 10.9 17.887 17.893 17.887 17.893 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.447 17.632 apply_single 129 13.6 0.001 0.001 17.447 17.631 qs_ot_get_p 129 10.4 0.001 0.001 17.345 17.477 ot_diis_step 118 11.6 0.020 0.021 17.333 17.333 make_m2s 5054 13.6 0.064 0.067 14.681 15.573 make_images 5054 14.6 2.176 2.592 14.371 15.257 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.348 14.387 calculate_rho_elec 129 8.7 0.255 0.266 14.347 14.386 sum_up_and_integrate 129 10.3 0.002 0.002 14.311 14.356 integrate_v_rspace 129 11.3 0.004 0.004 14.249 14.297 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.581 13.598 cp_dbcsr_syevd 83 12.4 0.099 0.114 12.418 12.419 multiply_cannon_sync_h2d 10108 15.6 10.819 11.276 10.819 11.276 init_scf_run 11 5.9 0.000 0.001 10.324 10.324 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.324 10.324 pw_transfer 1559 11.6 0.086 0.093 10.010 10.042 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.784 9.820 make_images_data 5054 15.6 0.056 0.064 8.395 9.748 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.444 9.545 cp_fm_diag_elpa_base 83 14.4 9.192 9.263 9.424 9.435 hybrid_alltoall_any 5240 16.5 0.850 3.817 8.250 9.424 fft_wrap_pw1pw2_140 527 13.2 1.696 1.742 8.667 8.709 grid_integrate_task_list 129 12.3 7.824 8.207 7.824 8.207 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.014 8.070 cp_fm_cholesky_decompose 22 10.9 7.815 7.947 7.815 7.947 dbcsr_mm_accdrv_process 20926 16.1 3.432 4.433 7.225 7.945 density_rs2pw 129 9.7 0.005 0.005 7.022 7.407 wfi_extrapolate 11 7.9 0.001 0.001 7.280 7.280 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.237 7.019 fft3d_ps 1301 14.7 3.174 3.283 6.192 6.218 calculate_dm_sparse 129 9.5 0.001 0.001 6.100 6.181 grid_collocate_task_list 129 9.7 5.604 5.899 5.604 5.899 dbcsr_complete_redistribute 395 12.7 2.095 2.154 5.154 5.550 multiply_cannon_metrocomm4 7581 15.6 0.026 0.029 1.879 5.430 mp_irecv_dv 29086 15.9 1.838 5.340 1.838 5.340 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.210 5.211 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.070 5.078 mp_alltoall_d11v 2423 14.1 4.556 4.955 4.556 4.955 mp_allgather_i34 2527 14.6 1.303 4.611 1.303 4.611 potential_pw2rs 129 12.3 0.027 0.028 4.411 4.417 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.999 4.035 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.901 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.685 3.780 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.679 3.727 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.312 3.627 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.440 3.525 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.479 3.486 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.384000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1441.181818, yerr=17.298545 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.087757E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59814. MP_Allreduce 10965 1506. MP_Sync 86 MP_Alltoall 1700 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.051 289.490 289.503 qs_mol_dyn_low 1 2.0 0.003 0.003 288.948 288.983 qs_forces 11 3.9 0.005 0.005 288.847 288.851 qs_energies 11 4.9 0.076 0.079 279.947 279.954 scf_env_do_scf 11 5.9 0.001 0.001 257.222 257.228 velocity_verlet 10 3.0 0.003 0.006 209.313 209.321 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 132.531 132.532 init_scf_loop 11 6.9 0.000 0.000 124.419 124.420 prepare_preconditioner 11 7.9 0.000 0.000 119.581 119.600 make_preconditioner 11 8.9 0.000 0.000 119.581 119.600 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.828 116.742 qs_scf_new_mos 116 7.6 0.001 0.001 88.168 88.221 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.167 88.220 ot_scf_mini 116 9.6 0.004 0.004 83.379 83.385 dbcsr_multiply_generic 2485 12.5 0.214 0.224 80.907 81.286 cp_fm_upper_to_full 104 14.8 53.789 76.508 53.789 76.508 multiply_cannon 2485 13.5 0.682 0.740 58.241 59.851 multiply_cannon_loop 2485 14.5 1.032 1.051 54.245 55.401 ot_mini 116 10.6 0.001 0.001 43.521 43.524 dbcsr_complete_redistribute 393 12.7 3.982 4.016 29.633 42.282 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.172 38.816 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.711 36.221 rebuild_ks_matrix 127 8.3 0.001 0.001 34.421 34.426 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.420 34.425 mp_alltoall_i22 712 14.1 21.466 34.267 21.466 34.267 cp_fm_cholesky_invert 11 10.9 33.071 33.078 33.071 33.078 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.891 31.905 mp_waitall_1 102768 16.8 27.455 31.893 27.455 31.893 qs_ot_get_p 127 10.4 0.001 0.001 24.881 24.918 qs_ot_get_derivative 116 11.6 0.002 0.002 24.382 24.388 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.883 20.912 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.550 19.686 make_m2s 4970 13.5 0.074 0.077 18.203 19.684 make_images 4970 14.5 3.032 3.225 17.723 19.205 cp_dbcsr_syevd 82 12.4 0.007 0.007 19.120 19.122 ot_diis_step 116 11.6 0.022 0.023 19.113 19.113 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.460 18.545 apply_single 127 13.6 0.001 0.001 18.460 18.545 multiply_cannon_multrec 9940 15.5 10.201 12.023 17.939 18.046 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.148 17.155 calculate_rho_elec 127 8.7 0.474 0.475 17.148 17.154 sum_up_and_integrate 127 10.3 0.002 0.002 16.303 16.396 integrate_v_rspace 127 11.3 0.005 0.005 16.240 16.333 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.025 16.025 cp_fm_diag_elpa_base 82 14.4 11.641 13.249 16.021 16.022 multiply_cannon_sync_h2d 9940 15.5 14.203 14.234 14.203 14.234 pw_transfer 1535 11.6 0.095 0.096 13.067 13.086 hybrid_alltoall_any 5155 16.4 1.302 3.020 10.519 12.847 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.827 12.846 make_images_data 4970 15.5 0.063 0.069 10.221 12.399 init_scf_run 11 5.9 0.000 0.001 12.172 12.174 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.172 12.174 fft_wrap_pw1pw2_140 519 13.2 3.208 3.268 11.432 11.456 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.548 9.557 dbcsr_mm_accdrv_process 20590 16.0 4.220 5.982 7.490 9.380 wfi_extrapolate 11 7.9 0.001 0.001 8.961 8.961 cp_fm_cholesky_decompose 22 10.9 8.800 8.821 8.800 8.821 grid_integrate_task_list 127 12.3 8.497 8.693 8.497 8.693 density_rs2pw 127 9.7 0.005 0.005 8.256 8.317 qs_energies_init_hamiltonians 11 5.9 0.012 0.014 8.294 8.295 fft3d_ps 1281 14.7 3.947 3.979 7.298 7.376 mp_alltoall_d11v 2401 14.1 6.780 6.914 6.780 6.914 calculate_dm_sparse 127 9.5 0.001 0.001 6.638 6.704 grid_collocate_task_list 127 9.7 6.366 6.445 6.366 6.445 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.312 6.344 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.087 6.126 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.503000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2841.363636, yerr=135.992829 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 1.260909E+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 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 0.012 0.044 84.996 84.996 qs_energies 1 2.0 0.000 0.000 84.515 84.548 ls_scf 1 3.0 0.000 0.000 83.492 83.525 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.514 72.645 multiply_cannon 111 7.7 0.018 0.021 55.844 56.869 multiply_cannon_loop 111 8.7 0.228 0.240 52.417 53.750 ls_scf_main 1 4.0 0.000 0.000 52.280 52.280 density_matrix_trs4 2 5.0 0.002 0.003 46.765 46.850 ls_scf_init_scf 1 4.0 0.000 0.000 28.162 28.163 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.033 27.091 mp_waitall_1 11031 10.9 22.478 25.053 22.478 25.053 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.932 24.953 multiply_cannon_multrec 2664 9.7 8.137 8.851 15.519 17.193 multiply_cannon_sync_h2d 2664 9.7 13.443 15.437 13.443 15.437 make_m2s 222 7.7 0.009 0.011 12.964 13.486 make_images 222 8.7 0.099 0.109 12.942 13.466 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.751 12.309 make_images_data 222 9.7 0.004 0.005 7.567 8.180 dbcsr_mm_accdrv_process 4760 10.4 0.590 0.699 6.999 7.920 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.490 7.787 hybrid_alltoall_any 227 10.6 0.216 1.845 6.522 7.529 dbcsr_mm_accdrv_process_sort 4760 11.4 6.210 7.069 6.210 7.069 calculate_norms 4752 9.8 5.494 6.127 5.494 6.127 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.038 5.182 mp_sum_l 887 5.1 3.166 4.522 3.166 4.522 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.418 3.486 make_images_sizes 222 9.7 0.000 0.000 0.660 3.428 mp_alltoall_i44 222 10.7 0.659 3.428 0.659 3.428 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.067 3.353 mp_irecv_dv 6231 10.9 2.049 3.322 2.049 3.322 arnoldi_extremal 4 6.8 0.000 0.000 3.244 3.275 arnoldi_normal_ev 4 7.8 0.001 0.003 3.244 3.275 build_subspace 16 8.4 0.009 0.012 3.152 3.153 ls_scf_post 1 4.0 0.000 0.000 3.051 3.083 ls_scf_store_result 1 5.0 0.000 0.000 2.844 2.896 dbcsr_special_finalize 555 9.7 0.005 0.006 2.381 2.753 dbcsr_merge_single_wm 555 10.7 0.464 0.599 2.373 2.745 make_images_pack 222 9.7 2.208 2.622 2.210 2.624 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.339 2.578 dbcsr_sort_data 658 11.4 2.167 2.498 2.167 2.498 dbcsr_matrix_vector_mult_local 304 10.0 2.072 2.473 2.074 2.475 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.274 2.346 buffer_matrices_ensure_size 222 8.7 1.751 2.045 1.751 2.045 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.761 1.762 rebuild_ks_matrix 3 7.3 0.000 0.000 1.752 1.753 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.752 1.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.996000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 2.173686E+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 0.029 0.049 91.369 91.371 qs_energies 1 2.0 0.000 0.000 89.274 89.278 ls_scf 1 3.0 0.000 0.000 87.885 87.888 dbcsr_multiply_generic 111 6.7 0.015 0.017 73.984 74.296 multiply_cannon 111 7.7 0.027 0.035 52.519 56.805 ls_scf_main 1 4.0 0.000 0.000 54.289 54.295 multiply_cannon_loop 111 8.7 0.136 0.148 49.787 53.080 density_matrix_trs4 2 5.0 0.002 0.003 48.582 48.779 mp_waitall_1 9105 10.9 20.763 30.292 20.763 30.292 ls_scf_init_scf 1 4.0 0.000 0.000 29.841 29.843 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.642 28.755 multiply_cannon_multrec 1332 9.7 13.251 17.094 22.543 27.520 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.289 26.301 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.455 20.774 make_m2s 222 7.7 0.006 0.008 15.143 15.696 make_images 222 8.7 1.369 1.692 15.113 15.666 dbcsr_mm_accdrv_process 4041 10.4 0.334 0.504 8.889 10.397 dbcsr_mm_accdrv_process_sort 4041 11.4 8.407 9.883 8.407 9.883 make_images_data 222 9.7 0.004 0.005 8.822 9.736 hybrid_alltoall_any 227 10.6 0.543 2.568 8.190 9.503 mp_sum_l 887 5.1 4.854 7.839 4.854 7.839 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.219 7.709 mp_irecv_dv 3311 11.0 3.199 7.660 3.199 7.660 calculate_norms 2376 9.8 6.055 6.817 6.055 6.817 multiply_cannon_sync_h2d 1332 9.7 4.890 6.258 4.890 6.258 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.612 6.231 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.954 5.208 arnoldi_extremal 4 6.8 0.000 0.000 4.704 4.720 arnoldi_normal_ev 4 7.8 0.001 0.004 4.704 4.720 build_subspace 16 8.4 0.014 0.021 4.455 4.457 ls_scf_post 1 4.0 0.000 0.000 3.754 3.759 ls_scf_store_result 1 5.0 0.000 0.000 3.442 3.579 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.166 3.409 dbcsr_matrix_vector_mult_local 304 10.0 2.764 3.256 2.766 3.258 mp_allgather_i34 111 8.7 0.724 2.937 0.724 2.937 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.606 2.708 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.134 2.508 dbcsr_data_new 4174 10.1 2.114 2.402 2.114 2.402 make_images_pack 222 9.7 1.821 2.132 1.824 2.134 dbcsr_sort_data 436 11.2 1.834 2.067 1.834 2.067 qs_init_subsys 1 2.0 0.088 0.470 1.851 1.875 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.844 1.846 rebuild_ks_matrix 3 7.3 0.000 0.000 1.832 1.834 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.831 1.834 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.371000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1780.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 2.892890E+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 0.062 0.237 93.828 93.830 qs_energies 1 2.0 0.001 0.011 92.951 92.958 ls_scf 1 3.0 0.001 0.009 91.303 91.364 dbcsr_multiply_generic 111 6.7 0.017 0.024 74.659 74.960 ls_scf_main 1 4.0 0.000 0.005 55.900 55.907 multiply_cannon 111 7.7 0.041 0.119 51.833 55.899 multiply_cannon_loop 111 8.7 0.117 0.130 49.103 52.956 density_matrix_trs4 2 5.0 0.002 0.013 49.949 50.190 mp_waitall_1 7281 11.0 23.293 33.094 23.293 33.094 ls_scf_init_scf 1 4.0 0.000 0.001 31.573 31.590 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.241 30.313 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 27.813 27.827 multiply_cannon_multrec 888 9.7 12.744 15.976 21.302 25.170 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.683 22.662 make_m2s 222 7.7 0.006 0.007 16.153 16.866 make_images 222 8.7 1.584 1.843 16.115 16.826 make_images_data 222 9.7 0.004 0.004 9.464 10.547 hybrid_alltoall_any 227 10.6 0.643 2.955 9.056 10.224 dbcsr_mm_accdrv_process 3754 10.4 0.303 0.489 8.095 9.325 dbcsr_mm_accdrv_process_sort 3754 11.4 7.661 8.836 7.661 8.836 mp_sum_l 887 5.1 5.032 8.322 5.032 8.322 multiply_cannon_sync_h2d 888 9.7 6.051 7.283 6.051 7.283 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.478 7.209 mp_irecv_dv 2335 11.1 2.462 7.150 2.462 7.150 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.791 7.016 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.746 6.455 arnoldi_extremal 4 6.8 0.000 0.000 5.089 5.108 arnoldi_normal_ev 4 7.8 0.001 0.005 5.089 5.108 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.845 5.086 build_subspace 16 8.4 0.014 0.020 4.779 4.787 calculate_norms 1584 9.8 4.338 4.700 4.338 4.700 ls_scf_post 1 4.0 0.008 0.076 3.829 3.883 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.444 3.770 mp_allgather_i34 111 8.7 0.837 3.679 0.837 3.679 dbcsr_matrix_vector_mult_local 304 10.0 3.031 3.607 3.033 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.438 3.525 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.859 2.944 dbcsr_data_new 4116 9.9 2.098 2.426 2.098 2.426 dbcsr_sort_data 325 11.1 1.853 2.062 1.853 2.062 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.968 1.986 rebuild_ks_matrix 3 7.3 0.000 0.000 1.951 1.968 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.022 1.951 1.968 make_images_sizes 222 9.7 0.000 0.000 0.870 1.934 mp_alltoall_i44 222 10.7 0.870 1.934 0.870 1.934 compute_matrix_preconditioner 1 6.0 0.000 0.001 1.911 1.918 make_images_pack 222 9.7 1.630 1.877 1.633 1.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.830000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2261.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 3.380224E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.044 96.889 96.890 qs_energies 1 2.0 0.000 0.000 95.866 95.871 ls_scf 1 3.0 0.000 0.000 94.195 94.200 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.958 78.250 ls_scf_main 1 4.0 0.000 0.000 58.499 58.500 multiply_cannon 111 7.7 0.042 0.079 51.449 55.822 density_matrix_trs4 2 5.0 0.002 0.003 52.341 52.472 multiply_cannon_loop 111 8.7 0.152 0.166 46.414 48.707 ls_scf_init_scf 1 4.0 0.000 0.000 32.452 32.454 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.240 31.313 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.773 28.787 mp_waitall_1 6369 11.0 22.293 28.528 22.293 28.528 multiply_cannon_multrec 1332 9.7 14.233 17.449 22.156 24.666 make_m2s 222 7.7 0.006 0.007 20.992 22.444 make_images 222 8.7 3.138 3.596 20.942 22.396 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.936 16.861 make_images_data 222 9.7 0.004 0.004 11.694 13.280 hybrid_alltoall_any 227 10.6 0.798 3.795 11.035 12.972 dbcsr_mm_accdrv_process 3641 10.4 0.287 0.482 7.559 9.080 dbcsr_mm_accdrv_process_sort 3641 11.4 7.118 8.586 7.118 8.586 mp_sum_l 887 5.1 3.929 7.531 3.929 7.531 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.085 6.171 mp_irecv_dv 3229 10.9 2.060 6.078 2.060 6.078 multiply_cannon_sync_h2d 1332 9.7 5.465 6.034 5.465 6.034 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.022 6.000 arnoldi_extremal 4 6.8 0.000 0.000 5.248 5.261 arnoldi_normal_ev 4 7.8 0.001 0.005 5.247 5.261 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.494 5.053 build_subspace 16 8.4 0.014 0.021 4.913 4.920 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.573 4.758 calculate_norms 2376 9.8 4.178 4.516 4.178 4.516 mp_allgather_i34 111 8.7 2.146 4.505 2.146 4.505 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.603 3.928 dbcsr_matrix_vector_mult_local 304 10.0 3.198 3.723 3.201 3.725 dbcsr_sort_data 658 11.4 3.131 3.418 3.131 3.418 ls_scf_post 1 4.0 0.000 0.000 3.244 3.249 dbcsr_special_finalize 555 9.7 0.006 0.007 2.885 3.229 dbcsr_merge_single_wm 555 10.7 0.545 0.671 2.876 3.221 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.036 3.067 ls_scf_store_result 1 5.0 0.000 0.000 2.999 3.047 dbcsr_data_release 10477 10.7 1.591 2.385 1.591 2.385 dbcsr_finalize 304 7.8 0.049 0.061 1.804 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.890000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 4.742513E+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 0.035 0.050 92.700 92.701 qs_energies 1 2.0 0.000 0.000 90.821 90.826 ls_scf 1 3.0 0.000 0.000 88.852 88.858 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.464 70.623 ls_scf_main 1 4.0 0.000 0.000 56.206 56.206 multiply_cannon 111 7.7 0.080 0.153 52.509 56.034 multiply_cannon_loop 111 8.7 0.089 0.105 49.936 51.147 density_matrix_trs4 2 5.0 0.002 0.003 49.258 49.328 ls_scf_init_scf 1 4.0 0.000 0.000 29.246 29.246 mp_waitall_1 5436 11.0 24.257 28.815 24.257 28.815 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.991 28.020 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.896 25.914 multiply_cannon_multrec 444 9.7 13.653 16.443 20.731 22.103 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.912 15.634 make_m2s 222 7.7 0.005 0.005 13.309 14.232 make_images 222 8.7 2.033 2.469 13.242 14.165 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.064 13.769 make_images_data 222 9.7 0.003 0.004 8.096 9.586 hybrid_alltoall_any 227 10.6 0.802 3.813 8.036 9.325 multiply_cannon_sync_h2d 444 9.7 6.749 8.298 6.749 8.298 dbcsr_mm_accdrv_process 3003 10.4 0.352 0.398 6.775 7.875 dbcsr_mm_accdrv_process_sort 3003 11.4 6.411 7.477 6.411 7.477 arnoldi_extremal 4 6.8 0.000 0.000 5.821 5.838 arnoldi_normal_ev 4 7.8 0.002 0.005 5.821 5.838 build_subspace 16 8.4 0.015 0.020 5.428 5.443 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.413 4.634 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.146 4.348 mp_sum_l 887 5.1 2.643 4.348 2.643 4.348 dbcsr_matrix_vector_mult_local 304 10.0 3.692 4.159 3.694 4.161 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.532 3.854 mp_irecv_dv 1241 11.2 1.519 3.826 1.519 3.826 calculate_norms 792 9.8 3.610 3.700 3.610 3.700 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.891 3.577 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.516 3.575 mp_allgather_i34 111 8.7 1.147 3.532 1.147 3.532 ls_scf_post 1 4.0 0.000 0.000 3.401 3.407 ls_scf_store_result 1 5.0 0.000 0.000 3.191 3.228 make_images_sizes 222 9.7 0.000 0.000 0.838 3.199 mp_alltoall_i44 222 10.7 0.838 3.198 0.838 3.198 dbcsr_data_new 4608 9.7 1.785 2.292 1.785 2.292 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.266 dbcsr_merge_all 275 8.9 0.479 0.524 2.054 2.118 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.042 2.043 rebuild_ks_matrix 3 7.3 0.000 0.000 2.010 2.010 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.010 2.010 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.952 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.701000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 8.829702E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.090 105.543 105.543 qs_energies 1 2.0 0.000 0.000 104.182 104.187 ls_scf 1 3.0 0.000 0.000 101.194 101.198 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.817 74.952 ls_scf_main 1 4.0 0.000 0.000 63.868 63.869 density_matrix_trs4 2 5.0 0.002 0.003 54.777 54.835 multiply_cannon 111 7.7 0.159 0.241 48.560 50.922 multiply_cannon_loop 111 8.7 0.098 0.100 45.467 46.102 ls_scf_init_scf 1 4.0 0.000 0.000 33.544 33.545 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.992 32.014 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.288 29.297 mp_waitall_1 4527 11.1 21.944 25.752 21.944 25.752 make_m2s 222 7.7 0.005 0.005 22.629 23.807 make_images 222 8.7 3.562 3.855 22.522 23.698 multiply_cannon_multrec 444 9.7 17.791 18.417 22.432 23.041 hybrid_alltoall_any 227 10.6 1.653 3.612 12.798 15.703 make_images_data 222 9.7 0.003 0.004 13.038 15.267 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.324 10.525 multiply_cannon_sync_h2d 444 9.7 8.792 8.851 8.792 8.851 arnoldi_extremal 4 6.8 0.000 0.000 7.319 7.331 arnoldi_normal_ev 4 7.8 0.003 0.008 7.319 7.331 build_subspace 16 8.4 0.026 0.035 6.765 6.776 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.425 5.564 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.233 5.330 dbcsr_matrix_vector_mult_local 304 10.0 5.034 5.308 5.037 5.310 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.856 5.112 dbcsr_mm_accdrv_process 1814 10.4 0.243 0.349 4.458 4.583 dbcsr_mm_accdrv_process_sort 1814 11.4 4.123 4.255 4.123 4.255 mp_allgather_i34 111 8.7 1.217 3.912 1.217 3.912 ls_scf_post 1 4.0 0.000 0.000 3.782 3.787 make_images_sizes 222 9.7 0.000 0.000 1.477 3.650 mp_alltoall_i44 222 10.7 1.477 3.649 1.477 3.649 ls_scf_store_result 1 5.0 0.000 0.000 3.502 3.540 calculate_norms 792 9.8 3.232 3.271 3.232 3.271 dbcsr_finalize 304 7.8 0.082 0.089 3.067 3.123 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.958 2.958 dbcsr_merge_all 275 8.9 0.886 0.918 2.851 2.903 dbcsr_complete_redistribute 5 7.6 1.451 1.492 2.752 2.901 matrix_ls_to_qs 2 6.0 0.000 0.000 2.424 2.575 dbcsr_sort_data 325 11.1 2.436 2.501 2.436 2.501 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.457 2.459 rebuild_ks_matrix 3 7.3 0.000 0.000 2.389 2.392 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.389 2.392 dbcsr_data_new 6591 9.6 1.835 2.383 1.835 2.383 dbcsr_new_transposed 4 7.5 0.242 0.252 2.290 2.298 dbcsr_frobenius_norm 74 6.6 2.055 2.137 2.211 2.246 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.203 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.202 dbcsr_data_release 12724 10.6 1.972 2.158 1.972 2.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.543000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7030.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3_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 593.993728E+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 4806724. 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 0.022 0.106 230.134 230.135 qs_mol_dyn_low 1 2.0 0.005 0.060 229.067 229.092 qs_forces 5 3.8 0.004 0.011 228.878 228.881 qs_energies 5 4.8 0.002 0.015 225.819 225.856 scf_env_do_scf 5 5.8 0.000 0.003 211.787 211.790 scf_env_do_scf_inner_loop 105 6.6 0.003 0.018 184.761 184.770 qs_scf_new_mos 105 7.6 0.000 0.001 143.352 143.533 qs_scf_loop_do_ot 105 8.6 0.001 0.001 143.351 143.533 ot_scf_mini 105 9.6 0.003 0.005 133.448 133.576 dbcsr_multiply_generic 1445 12.2 0.129 0.137 132.959 133.477 multiply_cannon 1445 13.2 0.277 0.297 113.783 116.129 multiply_cannon_loop 1445 14.2 2.842 2.992 112.120 113.641 velocity_verlet 4 3.0 0.001 0.021 108.565 108.566 ot_mini 105 10.6 0.001 0.010 60.575 60.685 multiply_cannon_multrec 69360 15.2 29.692 34.375 39.532 44.654 mp_waitall_1 488190 16.1 35.498 43.414 35.498 43.414 qs_ot_get_p 112 10.4 0.001 0.002 42.945 43.221 qs_ot_get_derivative 55 11.6 0.001 0.002 38.778 38.887 multiply_cannon_metrocomm3 69360 15.2 0.199 0.210 25.870 33.625 multiply_cannon_sync_h2d 69360 15.2 29.105 33.096 29.105 33.096 qs_ot_p2m_diag 40 11.0 0.020 0.031 31.601 31.675 rebuild_ks_matrix 110 8.4 0.000 0.000 29.437 29.597 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.028 29.437 29.597 cp_dbcsr_syevd 40 12.0 0.002 0.004 28.414 28.415 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.987 27.130 init_scf_loop 7 6.6 0.000 0.005 26.994 26.996 cp_fm_syevd 40 13.0 0.000 0.002 23.251 23.394 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.964 23.222 apply_single 62 13.6 0.000 0.000 22.964 23.221 prepare_preconditioner 7 7.6 0.000 0.001 22.038 22.072 make_preconditioner 7 8.6 0.000 0.003 22.038 22.072 ot_new_cg_direction 55 11.6 0.001 0.003 21.012 21.012 cp_fm_redistribute_end 40 14.0 9.259 18.467 9.263 18.468 cp_fm_syevd_base 40 14.0 9.199 18.410 9.199 18.410 qs_rho_update_rho_low 110 7.6 0.001 0.005 17.927 18.303 calculate_rho_elec 110 8.6 0.030 0.033 17.926 18.303 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.782 14.933 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.714 14.776 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.981 14.112 mp_sum_l 4764 12.2 12.333 13.214 12.333 13.214 density_rs2pw 110 9.6 0.005 0.006 12.271 12.864 pw_transfer 1645 12.4 0.080 0.106 11.721 11.938 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.581 11.800 calculate_dm_sparse 110 9.5 0.000 0.001 11.471 11.650 init_scf_run 5 5.8 0.000 0.001 11.300 11.301 scf_env_initial_rho_setup 5 6.8 0.002 0.004 11.300 11.301 qs_vxc_create 110 10.4 0.002 0.005 10.456 10.496 dbcsr_mm_accdrv_process 154766 15.8 6.231 6.410 9.707 10.487 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.387 10.450 fft_wrap_pw1pw2_240 915 15.0 0.847 0.930 10.194 10.401 cp_fm_cholesky_invert 7 10.6 10.303 10.312 10.303 10.312 check_diag 80 13.5 8.615 8.888 9.553 9.695 transfer_rs2pw 445 10.6 0.007 0.008 8.941 9.500 sum_up_and_integrate 60 10.3 0.001 0.003 8.985 9.002 integrate_v_rspace 60 11.3 0.002 0.004 8.968 8.986 fft3d_pb 915 16.0 2.402 2.688 8.350 8.622 xc_rho_set_and_dset_create 110 12.4 0.076 0.096 7.254 7.502 acc_transpose_blocks 69360 15.2 0.359 0.374 6.982 7.412 multiply_cannon_metrocomm1 69360 15.2 0.094 0.101 4.694 7.184 make_full_single_inverse 7 9.6 0.001 0.004 7.049 7.083 xc_vxc_pw_create 60 11.3 0.039 0.051 7.016 7.055 make_m2s 2890 13.2 0.078 0.088 6.387 6.953 make_images 2890 14.2 0.240 0.260 6.280 6.846 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.770 6.786 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.351 6.401 calculate_first_density_matrix 1 7.0 0.000 0.005 6.368 6.381 xc_pw_derive 510 13.4 0.006 0.007 6.208 6.280 mp_alltoall_z22v 2340 17.7 5.330 5.656 5.330 5.656 potential_pw2rs 60 12.3 0.002 0.003 5.187 5.248 mp_waitany 7680 13.5 4.380 5.090 4.380 5.090 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.135000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c75b5083bd09b6d5662fcbd9e72d3d3fb7c342b3 Summary: empty Status: OK