=== 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: a01af14846922d2d741ffda4c698585069c13d48 ################# 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 # # 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: 04.10.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/01 job id: 49922843 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/02 job id: 49922845 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/03 job id: 49922847 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/04 job id: 49922849 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/05 job id: 49922851 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/06 job id: 49922853 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/07 job id: 49922855 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/08 job id: 49922858 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/09 job id: 49922860 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/10 job id: 49922862 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/11 job id: 49922863 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/12 job id: 49922864 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/13 job id: 49922865 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/14 job id: 49922866 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/15 job id: 49922870 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/16 job id: 49922878 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/17 job id: 49922887 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/18 job id: 49922890 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/19 job id: 49922897 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/20 job id: 49922904 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/21 job id: 49922912 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/22 job id: 49922917 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/23 job id: 49922923 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/24 job id: 49922925 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/25 job id: 49922928 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/26 job id: 49922931 --- 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/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/27 job id: 49922937 --- 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/a01af14846922d2d741ffda4c698585069c13d48_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 0.035 133.896 133.897 farming_run 1 2.0 133.305 133.329 133.867 133.870 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.504453E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 114.581 114.582 qs_energies 1 2.0 0.000 0.000 114.387 114.390 mp2_main 1 3.0 0.000 0.000 112.089 112.091 mp2_gpw_main 1 4.0 0.027 0.033 110.855 110.857 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.095 92.195 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.559 54.657 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 40.954 45.760 get_2c_integrals 1 6.0 0.008 0.009 36.791 37.362 integrate_v_rspace 273 8.0 0.441 0.457 24.567 29.040 pw_transfer 6555 10.6 0.384 0.401 26.714 26.921 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.316 25.501 grid_integrate_task_list 273 9.0 20.486 25.389 20.486 25.389 fft_wrap_pw1pw2_100 2178 12.4 1.093 1.129 22.872 23.053 compute_2c_integrals 1 7.0 0.002 0.002 18.923 18.925 rpa_ri_compute_en 1 5.0 0.023 0.025 18.650 18.751 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.554 18.655 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.414 18.553 18.654 cp_fm_cholesky_decompose 12 8.2 17.861 18.425 17.861 18.425 cholesky_decomp 1 7.0 0.000 0.000 16.704 17.274 fft3d_s 5443 13.4 16.077 16.288 16.100 16.311 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.319 10.771 15.319 calculate_wavefunction 272 8.0 5.400 5.494 12.290 12.996 rpa_num_int 1 6.0 0.000 0.000 10.734 10.743 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.684 10.699 calc_mat_Q 8 8.0 0.000 0.000 9.483 9.580 contract_S_to_Q 8 9.0 0.000 0.000 8.902 8.998 calc_potential_gpw 544 9.5 0.005 0.005 8.258 8.693 parallel_gemm_fm 14 9.1 0.000 0.000 8.506 8.595 parallel_gemm_fm_cosma 14 10.1 8.506 8.595 8.506 8.595 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.060 8.265 potential_pw2rs 545 10.0 0.107 0.109 7.435 8.065 create_integ_mat 1 6.0 0.015 0.027 7.718 7.728 collocate_single_gaussian 272 10.0 0.040 0.042 7.315 7.504 array2fm 1 7.0 0.000 0.000 6.724 7.181 pw_scatter_s 2720 13.7 4.346 4.477 4.346 4.477 pw_gather_s 2722 13.2 3.394 3.786 3.394 3.786 array2fm_buffer_send 1 8.0 2.949 3.141 2.949 3.141 pw_poisson_solve 545 10.5 1.120 1.197 2.220 2.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.857450, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2821.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 0.024 0.042 422.187 422.188 farming_run 1 2.0 421.269 421.273 422.147 422.149 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.242280E+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 2407093. 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 0.031 217.752 217.753 qs_energies 1 2.0 0.000 0.000 217.472 217.484 scf_env_do_scf 1 3.0 0.000 0.000 115.260 115.260 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.376 114.382 rebuild_ks_matrix 4 6.0 0.000 0.000 114.374 114.381 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.374 114.380 hfx_ks_matrix 4 8.0 0.001 0.001 114.031 114.036 integrate_four_center 4 9.0 0.154 0.505 114.030 114.035 integrate_four_center_main 4 10.0 0.094 0.529 101.857 105.948 integrate_four_center_bin 266 11.0 101.763 105.571 101.763 105.571 mp2_main 1 3.0 0.000 0.000 101.925 101.936 mp2_gpw_main 1 4.0 0.057 0.112 101.068 101.085 init_scf_loop 1 4.0 0.000 0.000 97.070 97.070 mp2_ri_gpw_compute_in 1 5.0 0.066 0.094 74.054 75.153 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.809 54.906 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.155 41.443 46.407 integrate_v_rspace 95 8.0 0.397 0.569 27.902 32.693 pw_transfer 2240 10.6 0.144 0.163 29.521 29.918 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.531 28.952 mp2_ri_gpw_compute_en 1 5.0 0.063 0.077 26.855 28.830 grid_integrate_task_list 95 9.0 23.315 28.302 23.315 28.302 fft_wrap_pw1pw2_100 730 12.4 1.267 1.453 26.294 26.650 ao_to_mo_and_store_B_mult_1 91 7.0 10.693 25.538 10.693 25.538 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.910 25.080 25.090 get_2c_integrals 1 6.0 0.000 0.000 20.129 20.161 compute_2c_integrals 1 7.0 0.002 0.003 19.118 19.131 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.610 19.004 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.832 18.609 19.004 fft3d_s 1823 13.4 18.516 18.866 18.531 18.880 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.189 18.189 calculate_wavefunction 91 8.0 2.014 2.049 9.627 9.864 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.571 8.859 9.302 potential_pw2rs 186 10.0 0.033 0.035 8.419 8.915 local_gemm 172 8.0 8.304 8.732 8.304 8.732 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.136 8.576 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.518 7.996 8.465 calc_potential_gpw 182 9.5 0.002 0.002 7.898 8.213 collocate_single_gaussian 91 10.0 0.017 0.021 7.758 8.011 mp_sync 37 10.5 4.154 7.074 4.154 7.074 integrate_four_center_load 4 10.0 0.000 0.000 6.748 6.751 hfx_load_balance 1 11.0 0.000 0.000 6.747 6.751 mp_sendrecv_dm3 2068 8.0 6.026 6.473 6.026 6.473 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.400 6.338 6.400 pw_gather_s 912 13.2 4.504 4.979 4.504 4.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.061609, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.722688E+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 0.007 0.030 54.943 54.944 qs_mol_dyn_low 1 2.0 0.005 0.006 54.670 54.678 qs_forces 11 3.9 0.002 0.003 54.607 54.608 qs_energies 11 4.9 0.002 0.002 53.098 53.124 scf_env_do_scf 11 5.9 0.000 0.001 46.517 46.517 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.474 44.474 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.500 34.968 qs_scf_new_mos 108 7.5 0.000 0.001 34.364 34.670 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.363 34.670 ot_scf_mini 108 9.5 0.002 0.002 32.678 32.868 multiply_cannon 2286 13.5 0.196 0.209 26.921 28.539 multiply_cannon_loop 2286 14.5 1.823 1.959 26.247 27.894 velocity_verlet 10 3.0 0.002 0.003 26.916 26.917 ot_mini 108 10.5 0.001 0.001 19.873 20.113 qs_ot_get_derivative 108 11.5 0.001 0.001 16.750 16.925 mp_waitall_1 245248 16.5 8.772 15.070 8.772 15.070 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.234 13.132 multiply_cannon_multrec 54864 15.5 3.657 5.705 7.986 11.447 qs_ot_get_p 119 10.4 0.001 0.001 8.113 8.408 rebuild_ks_matrix 119 8.3 0.000 0.000 7.863 8.019 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.863 8.019 mp_sum_l 7287 12.8 5.573 7.329 5.573 7.329 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.925 7.056 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.902 6.365 multiply_cannon_sync_h2d 54864 15.5 5.116 6.221 5.116 6.221 dbcsr_mm_accdrv_process 76910 16.1 1.817 2.915 4.241 6.064 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.384 5.477 init_scf_run 11 5.9 0.000 0.001 5.284 5.285 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.284 5.284 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.210 5.252 sum_up_and_integrate 119 10.3 0.001 0.001 4.466 4.472 integrate_v_rspace 119 11.3 0.002 0.002 4.456 4.463 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.414 4.414 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.213 4.383 calculate_rho_elec 119 8.7 0.011 0.016 4.212 4.383 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.212 4.212 cp_fm_redistribute_end 50 14.0 2.149 4.187 2.155 4.190 cp_fm_diag_elpa_base 50 14.0 2.028 4.078 2.032 4.086 calculate_dm_sparse 119 9.5 0.000 0.000 3.306 3.481 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.879 3.190 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.999 3.174 apply_single 119 13.6 0.000 0.000 2.999 3.174 jit_kernel_multiply 13 15.8 2.362 3.173 2.362 3.173 calculate_first_density_matrix 1 7.0 0.000 0.000 3.013 3.019 acc_transpose_blocks 54864 15.5 0.227 0.258 2.280 2.873 ot_diis_step 108 11.5 0.006 0.006 2.841 2.842 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.479 2.546 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.473 2.475 density_rs2pw 119 9.7 0.004 0.004 2.182 2.263 wfi_extrapolate 11 7.9 0.001 0.001 2.200 2.200 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.134 2.186 grid_integrate_task_list 119 12.3 2.020 2.126 2.020 2.126 mp_sum_d 4135 12.0 1.445 2.067 1.445 2.067 init_scf_loop 11 6.9 0.000 0.000 2.026 2.026 potential_pw2rs 119 12.3 0.004 0.004 1.831 1.844 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.793 1.826 pw_transfer 1439 11.6 0.051 0.058 1.689 1.761 make_m2s 4572 13.5 0.053 0.055 1.690 1.738 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.614 1.689 make_images 4572 14.5 0.134 0.139 1.609 1.656 mp_alltoall_d11v 2130 13.8 1.374 1.621 1.374 1.621 acc_transpose_blocks_sync 164592 16.5 1.196 1.465 1.196 1.465 transfer_rs2pw 487 10.6 0.005 0.006 1.382 1.451 mp_waitany 12084 13.8 1.262 1.426 1.262 1.426 grid_collocate_task_list 119 9.7 1.353 1.417 1.353 1.417 fft3d_ps 1201 14.6 0.370 0.478 1.330 1.400 transfer_pw2rs 487 13.2 0.006 0.006 1.333 1.342 fft_wrap_pw1pw2_140 487 13.2 0.140 0.153 1.240 1.319 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.819 1.220 acc_transpose_blocks_kernels 54864 16.5 0.245 0.383 0.810 1.101 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.944000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 488.595456E+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 910951. 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 0.018 0.049 38.816 38.817 qs_mol_dyn_low 1 2.0 0.003 0.003 38.487 38.495 qs_forces 11 3.9 0.002 0.003 38.425 38.426 qs_energies 11 4.9 0.002 0.002 36.753 36.758 scf_env_do_scf 11 5.9 0.000 0.001 31.682 31.682 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 29.222 29.222 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.027 22.383 qs_scf_new_mos 108 7.5 0.001 0.001 20.571 20.802 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.570 20.801 ot_scf_mini 108 9.5 0.002 0.003 19.650 19.816 velocity_verlet 10 3.0 0.001 0.001 18.465 18.466 multiply_cannon 2286 13.5 0.212 0.219 16.846 18.316 multiply_cannon_loop 2286 14.5 1.202 1.268 15.664 17.172 ot_mini 108 10.5 0.001 0.001 12.138 12.371 mp_waitall_1 200699 16.5 5.513 10.715 5.513 10.715 qs_ot_get_derivative 108 11.5 0.001 0.002 9.765 9.933 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 3.946 9.420 multiply_cannon_multrec 27432 15.5 1.841 4.055 6.158 9.018 rebuild_ks_matrix 119 8.3 0.000 0.000 6.789 6.930 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.789 6.930 dbcsr_mm_accdrv_process 47894 16.0 3.547 5.641 4.235 6.305 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.969 6.095 qs_ot_get_p 119 10.4 0.001 0.001 4.601 4.825 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.639 4.550 mp_sum_l 7287 12.8 2.082 4.087 2.082 4.087 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.078 4.068 apply_single 119 13.6 0.000 0.000 3.078 4.068 init_scf_run 11 5.9 0.000 0.001 3.871 3.871 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.871 3.871 sum_up_and_integrate 119 10.3 0.001 0.001 3.715 3.720 integrate_v_rspace 119 11.3 0.002 0.003 3.701 3.705 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.621 3.645 calculate_rho_elec 119 8.7 0.021 0.024 3.620 3.644 make_m2s 4572 13.5 0.051 0.053 2.657 3.005 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.973 2.990 make_images 4572 14.5 0.206 0.244 2.570 2.921 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.576 2.576 init_scf_loop 11 6.9 0.000 0.000 2.439 2.439 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.311 2.397 calculate_first_density_matrix 1 7.0 0.000 0.001 2.348 2.350 ot_diis_step 108 11.5 0.011 0.011 2.327 2.327 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.263 2.263 cp_fm_redistribute_end 50 14.0 1.147 2.239 1.150 2.241 multiply_cannon_sync_h2d 27432 15.5 1.696 2.211 1.696 2.211 cp_fm_diag_elpa_base 50 14.0 1.060 2.151 1.088 2.187 calculate_dm_sparse 119 9.5 0.000 0.001 2.085 2.163 density_rs2pw 119 9.7 0.004 0.004 1.920 2.005 acc_transpose_blocks 27432 15.5 0.114 0.120 1.621 1.962 grid_integrate_task_list 119 12.3 1.833 1.936 1.833 1.936 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.913 1.915 pw_transfer 1439 11.6 0.065 0.068 1.841 1.876 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.750 1.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.722 1.760 make_images_data 4572 15.5 0.047 0.054 1.266 1.630 prepare_preconditioner 11 7.9 0.000 0.000 1.554 1.580 make_preconditioner 11 8.9 0.000 0.000 1.554 1.580 jit_kernel_multiply 8 16.5 0.629 1.563 0.629 1.563 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.122 1.510 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.453 1.508 potential_pw2rs 119 12.3 0.006 0.006 1.462 1.467 wfi_extrapolate 11 7.9 0.001 0.001 1.465 1.465 fft_wrap_pw1pw2_140 487 13.2 0.160 0.169 1.386 1.423 fft3d_ps 1201 14.6 0.518 0.573 1.386 1.417 mp_alltoall_d11v 2130 13.8 1.207 1.378 1.207 1.378 grid_collocate_task_list 119 9.7 1.288 1.336 1.288 1.336 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.283 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.200 1.247 transfer_rs2pw 487 10.6 0.005 0.005 1.088 1.194 mp_allgather_i34 2286 14.5 0.600 1.021 0.600 1.021 mp_sum_d 4135 12.0 0.534 1.009 0.534 1.009 acc_transpose_blocks_sync 82296 16.5 0.826 0.955 0.826 0.955 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.932 0.933 transfer_pw2rs 487 13.2 0.004 0.005 0.926 0.931 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.878 0.892 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.654 0.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.817000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.363636, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 519.188480E+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 0.031 0.049 32.290 32.290 qs_mol_dyn_low 1 2.0 0.003 0.004 31.981 31.993 qs_forces 11 3.9 0.002 0.003 31.921 31.922 qs_energies 11 4.9 0.003 0.004 30.383 30.385 scf_env_do_scf 11 5.9 0.000 0.001 25.249 25.249 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.870 22.871 dbcsr_multiply_generic 2286 12.5 0.094 0.097 17.265 17.364 qs_scf_new_mos 108 7.5 0.001 0.001 15.293 15.306 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.293 15.305 velocity_verlet 10 3.0 0.002 0.002 15.012 15.013 multiply_cannon 2286 13.5 0.195 0.200 14.065 14.813 ot_scf_mini 108 9.5 0.002 0.002 14.537 14.546 multiply_cannon_loop 2286 14.5 0.859 0.896 13.289 14.069 ot_mini 108 10.5 0.001 0.001 9.014 9.028 qs_ot_get_derivative 108 11.5 0.001 0.001 7.528 7.538 multiply_cannon_multrec 18288 15.5 1.888 2.859 6.956 7.251 rebuild_ks_matrix 119 8.3 0.000 0.001 5.918 5.938 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.918 5.937 dbcsr_mm_accdrv_process 38222 16.0 4.916 5.776 4.974 5.841 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.214 5.232 mp_waitall_1 158411 16.6 2.862 4.107 2.862 4.107 init_scf_run 11 5.9 0.000 0.001 3.979 3.979 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.979 3.979 sum_up_and_integrate 119 10.3 0.001 0.001 3.507 3.512 integrate_v_rspace 119 11.3 0.003 0.003 3.494 3.501 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.830 3.421 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.205 3.329 calculate_rho_elec 119 8.7 0.031 0.031 3.205 3.328 qs_ot_get_p 119 10.4 0.001 0.001 3.282 3.302 calculate_first_density_matrix 1 7.0 0.000 0.000 2.768 2.774 multiply_cannon_metrocomm3 18288 15.5 0.047 0.050 1.676 2.655 init_scf_loop 11 6.9 0.000 0.000 2.363 2.364 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.993 2.334 apply_single 119 13.6 0.000 0.000 1.993 2.333 calculate_dm_sparse 119 9.5 0.000 0.001 2.282 2.295 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.188 2.195 acc_transpose_blocks 18288 15.5 0.080 0.082 1.763 2.060 make_m2s 4572 13.5 0.043 0.044 1.776 1.927 grid_integrate_task_list 119 12.3 1.804 1.920 1.804 1.920 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.894 1.895 density_rs2pw 119 9.7 0.004 0.004 1.770 1.871 make_images 4572 14.5 0.193 0.205 1.691 1.841 pw_transfer 1439 11.6 0.065 0.068 1.787 1.798 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.701 1.707 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.694 1.706 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.655 1.668 cp_fm_diag_elpa_base 50 14.0 1.632 1.647 1.653 1.667 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.658 1.663 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.622 1.631 prepare_preconditioner 11 7.9 0.000 0.000 1.612 1.614 make_preconditioner 11 8.9 0.000 0.000 1.612 1.614 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.473 1.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.511 1.516 ot_diis_step 108 11.5 0.011 0.012 1.470 1.470 mp_sum_l 7287 12.8 1.036 1.418 1.036 1.418 fft_wrap_pw1pw2_140 487 13.2 0.212 0.218 1.373 1.384 grid_collocate_task_list 119 9.7 1.232 1.318 1.232 1.318 potential_pw2rs 119 12.3 0.007 0.009 1.302 1.305 fft3d_ps 1201 14.6 0.526 0.542 1.263 1.273 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.922 1.236 multiply_cannon_sync_h2d 18288 15.5 1.046 1.177 1.046 1.177 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 transfer_rs2pw 487 10.6 0.005 0.005 0.941 1.051 jit_kernel_transpose 5 15.6 0.705 1.019 0.705 1.019 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.939 0.956 make_images_data 4572 15.5 0.048 0.051 0.770 0.940 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.935 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.672 0.877 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.466 0.874 acc_transpose_blocks_sync 54864 16.5 0.741 0.793 0.741 0.793 transfer_pw2rs 487 13.2 0.004 0.004 0.791 0.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.773 0.776 mp_alltoall_d11v 2130 13.8 0.659 0.757 0.659 0.757 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.674 0.737 cp_fm_cholesky_invert 11 10.9 0.681 0.685 0.681 0.685 mp_alltoall_z22v 1201 16.6 0.591 0.656 0.591 0.656 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.290000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.818182, yerr=0.385695 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 555.220992E+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 0.019 0.037 39.854 39.856 qs_mol_dyn_low 1 2.0 0.006 0.013 39.568 39.576 qs_forces 11 3.9 0.002 0.003 39.467 39.470 qs_energies 11 4.9 0.003 0.013 37.664 37.671 scf_env_do_scf 11 5.9 0.002 0.016 31.952 31.953 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.221 28.221 dbcsr_multiply_generic 2286 12.5 0.099 0.103 20.950 21.098 velocity_verlet 10 3.0 0.032 0.043 20.002 20.004 qs_scf_new_mos 108 7.5 0.001 0.001 19.453 19.518 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.453 19.517 ot_scf_mini 108 9.5 0.002 0.003 18.381 18.429 multiply_cannon 2286 13.5 0.219 0.227 16.425 17.014 multiply_cannon_loop 2286 14.5 1.526 1.606 15.324 15.745 ot_mini 108 10.5 0.001 0.001 10.989 11.050 multiply_cannon_multrec 27432 15.5 2.461 3.112 9.017 9.275 qs_ot_get_derivative 108 11.5 0.001 0.001 9.067 9.117 dbcsr_mm_accdrv_process 47916 15.9 5.960 7.280 6.454 7.847 rebuild_ks_matrix 119 8.3 0.000 0.000 6.747 6.803 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.746 6.802 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.973 6.023 qs_ot_get_p 119 10.4 0.001 0.001 4.431 4.502 init_scf_run 11 5.9 0.000 0.001 4.194 4.195 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.194 4.195 sum_up_and_integrate 119 10.3 0.001 0.001 3.840 3.847 integrate_v_rspace 119 11.3 0.003 0.003 3.828 3.836 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.755 3.790 calculate_rho_elec 119 8.7 0.040 0.046 3.754 3.790 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.242 3.719 init_scf_loop 11 6.9 0.001 0.002 3.684 3.685 make_m2s 4572 13.5 0.054 0.056 2.726 2.847 prepare_preconditioner 11 7.9 0.000 0.000 2.788 2.796 make_preconditioner 11 8.9 0.000 0.002 2.788 2.796 mp_waitall_1 137007 16.6 2.312 2.785 2.312 2.785 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.743 2.755 make_images 4572 14.5 0.275 0.333 2.618 2.737 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.392 2.717 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.212 2.698 apply_single 119 13.6 0.000 0.000 2.212 2.698 calculate_first_density_matrix 1 7.0 0.001 0.010 2.598 2.601 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.534 2.558 acc_transpose_blocks 27432 15.5 0.120 0.125 2.399 2.534 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.306 2.306 density_rs2pw 119 9.7 0.004 0.004 2.168 2.272 pw_transfer 1439 11.6 0.065 0.069 2.211 2.253 calculate_dm_sparse 119 9.5 0.000 0.000 2.197 2.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.193 2.194 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.119 2.164 mp_sum_l 7287 12.8 1.394 2.043 1.394 2.043 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.977 1.986 cp_fm_diag_elpa_base 50 14.0 1.936 1.955 1.974 1.983 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.938 1.963 grid_integrate_task_list 119 12.3 1.820 1.910 1.820 1.910 ot_diis_step 108 11.5 0.012 0.012 1.878 1.878 fft_wrap_pw1pw2_140 487 13.2 0.245 0.255 1.651 1.699 fft3d_ps 1201 14.6 0.557 0.606 1.639 1.668 potential_pw2rs 119 12.3 0.009 0.009 1.602 1.610 acc_transpose_blocks_sync 82296 16.5 1.418 1.544 1.418 1.544 wfi_extrapolate 11 7.9 0.001 0.001 1.538 1.538 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.361 1.375 grid_collocate_task_list 119 9.7 1.249 1.346 1.249 1.346 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.845 1.334 transfer_rs2pw 487 10.6 0.004 0.005 1.195 1.331 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.268 1.283 dbcsr_complete_redistribute 329 12.2 0.120 0.142 0.948 1.236 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.204 1.208 jit_kernel_multiply 6 16.1 0.429 1.178 0.429 1.178 cp_fm_upper_to_full 72 14.2 0.821 1.168 0.821 1.168 make_images_data 4572 15.5 0.048 0.052 1.003 1.167 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.857 1.073 mp_alltoall_d11v 2130 13.8 0.955 1.061 0.955 1.061 mp_alltoall_z22v 1201 16.6 0.964 1.001 0.964 1.001 transfer_pw2rs 487 13.2 0.004 0.004 0.951 0.958 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.652 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.871 0.876 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.875 cp_fm_cholesky_invert 11 10.9 0.868 0.872 0.868 0.872 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 0.832 0.841 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.278 0.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=39.856000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.181818, yerr=2.948595 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 628.875264E+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 0.009 0.028 28.757 28.758 qs_mol_dyn_low 1 2.0 0.004 0.013 28.571 28.580 qs_forces 11 3.9 0.003 0.008 28.511 28.512 qs_energies 11 4.9 0.012 0.026 26.792 26.796 scf_env_do_scf 11 5.9 0.000 0.001 22.001 22.001 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.538 19.539 velocity_verlet 10 3.0 0.002 0.002 14.816 14.819 dbcsr_multiply_generic 2286 12.5 0.091 0.096 13.041 13.109 qs_scf_new_mos 108 7.5 0.001 0.001 11.869 11.896 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.868 11.895 ot_scf_mini 108 9.5 0.002 0.002 11.155 11.180 multiply_cannon 2286 13.5 0.224 0.228 10.346 10.883 multiply_cannon_loop 2286 14.5 0.639 0.664 9.404 9.563 ot_mini 108 10.5 0.001 0.001 6.314 6.345 multiply_cannon_multrec 9144 15.5 1.655 1.837 5.909 6.095 rebuild_ks_matrix 119 8.3 0.000 0.000 5.803 5.830 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.803 5.830 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.156 5.179 qs_ot_get_derivative 108 11.5 0.001 0.001 5.022 5.047 dbcsr_mm_accdrv_process 12550 15.8 3.561 4.196 4.139 4.238 sum_up_and_integrate 119 10.3 0.001 0.001 3.472 3.477 integrate_v_rspace 119 11.3 0.003 0.003 3.462 3.466 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.356 3.360 calculate_rho_elec 119 8.7 0.060 0.061 3.356 3.360 init_scf_run 11 5.9 0.000 0.001 3.226 3.226 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.226 3.226 qs_ot_get_p 119 10.4 0.001 0.001 2.927 2.963 init_scf_loop 11 6.9 0.000 0.000 2.442 2.443 calculate_first_density_matrix 1 7.0 0.000 0.000 2.061 2.062 make_m2s 4572 13.5 0.033 0.035 1.873 2.030 pw_transfer 1439 11.6 0.066 0.068 2.010 2.019 grid_integrate_task_list 119 12.3 1.858 1.941 1.858 1.941 make_images 4572 14.5 0.269 0.301 1.785 1.939 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.917 1.928 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.910 1.913 mp_waitall_1 115863 16.7 1.388 1.873 1.388 1.873 density_rs2pw 119 9.7 0.003 0.004 1.750 1.804 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.749 1.772 calculate_dm_sparse 119 9.5 0.000 0.000 1.748 1.765 prepare_preconditioner 11 7.9 0.000 0.000 1.716 1.719 make_preconditioner 11 8.9 0.000 0.000 1.716 1.719 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.658 1.659 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.604 1.635 fft_wrap_pw1pw2_140 487 13.2 0.322 0.332 1.559 1.570 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.541 1.542 acc_transpose_blocks 9144 15.5 0.041 0.042 1.419 1.462 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.382 1.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.372 1.381 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.356 1.364 cp_fm_diag_elpa_base 50 14.0 1.328 1.346 1.354 1.362 grid_collocate_task_list 119 9.7 1.299 1.344 1.299 1.344 fft3d_ps 1201 14.6 0.559 0.570 1.331 1.340 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.324 1.326 ot_diis_step 108 11.5 0.013 0.013 1.278 1.278 potential_pw2rs 119 12.3 0.010 0.011 1.265 1.267 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.256 apply_single 119 13.6 0.000 0.000 1.216 1.256 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.132 1.139 wfi_extrapolate 11 7.9 0.001 0.001 1.113 1.114 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.802 1.040 jit_kernel_multiply 6 15.5 0.539 1.022 0.539 1.022 make_images_data 4572 15.5 0.042 0.046 0.809 0.997 mp_alltoall_d11v 2130 13.8 0.819 0.990 0.819 0.990 transfer_rs2pw 487 10.6 0.004 0.004 0.865 0.936 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.930 cp_fm_cholesky_invert 11 10.9 0.845 0.848 0.845 0.848 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.836 0.843 qs_env_update_s_mstruct 11 6.9 0.014 0.016 0.761 0.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.770 0.773 acc_transpose_blocks_sync 27432 16.5 0.718 0.755 0.718 0.755 mp_allgather_i34 2286 14.5 0.250 0.694 0.250 0.694 transfer_pw2rs 487 13.2 0.003 0.004 0.681 0.682 mp_alltoall_z22v 1201 16.6 0.646 0.675 0.646 0.675 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.645 0.649 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.299 0.610 mp_waitany 5200 13.7 0.520 0.589 0.520 0.589 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.758000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=593.636364, yerr=7.795951 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 776.642560E+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 0.014 0.031 42.017 42.018 qs_mol_dyn_low 1 2.0 0.003 0.004 41.738 41.772 qs_forces 11 3.9 0.002 0.003 41.662 41.663 qs_energies 11 4.9 0.003 0.003 39.680 39.685 scf_env_do_scf 11 5.9 0.001 0.001 34.074 34.074 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.303 26.305 velocity_verlet 10 3.0 0.002 0.002 23.606 23.613 dbcsr_multiply_generic 2286 12.5 0.099 0.102 18.847 19.101 qs_scf_new_mos 108 7.5 0.001 0.001 17.076 17.172 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.076 17.172 ot_scf_mini 108 9.5 0.002 0.002 15.914 16.016 multiply_cannon 2286 13.5 0.303 0.308 15.002 15.995 multiply_cannon_loop 2286 14.5 0.868 0.892 13.761 14.784 ot_mini 108 10.5 0.001 0.001 9.738 9.853 multiply_cannon_multrec 9144 15.5 3.444 4.876 8.797 8.930 qs_ot_get_derivative 108 11.5 0.001 0.001 7.708 7.810 init_scf_loop 11 6.9 0.000 0.000 7.745 7.749 rebuild_ks_matrix 119 8.3 0.000 0.001 6.972 7.112 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.971 7.112 prepare_preconditioner 11 7.9 0.000 0.000 6.724 6.738 make_preconditioner 11 8.9 0.000 0.000 6.724 6.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.306 6.620 dbcsr_mm_accdrv_process 12550 15.8 4.577 6.551 5.213 6.583 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.285 6.412 cp_fm_upper_to_full 72 14.2 3.176 4.539 3.176 4.539 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.952 3.963 calculate_rho_elec 119 8.7 0.118 0.120 3.951 3.962 sum_up_and_integrate 119 10.3 0.001 0.001 3.694 3.699 integrate_v_rspace 119 11.3 0.003 0.003 3.683 3.689 init_scf_run 11 5.9 0.000 0.001 3.543 3.543 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.543 3.543 qs_ot_get_p 119 10.4 0.001 0.001 3.364 3.501 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.743 3.180 mp_waitall_1 94719 16.7 2.130 3.099 2.130 3.099 dbcsr_complete_redistribute 329 12.2 0.286 0.290 1.938 2.754 pw_transfer 1439 11.6 0.069 0.069 2.524 2.530 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.210 2.471 apply_single 119 13.6 0.000 0.000 2.210 2.471 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.642 2.441 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.425 2.431 make_m2s 4572 13.5 0.037 0.038 2.244 2.386 calculate_dm_sparse 119 9.5 0.000 0.000 2.286 2.301 make_images 4572 14.5 0.355 0.387 2.124 2.266 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.412 2.204 mp_alltoall_i22 627 13.8 1.394 2.194 1.394 2.194 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.282 2.185 calculate_first_density_matrix 1 7.0 0.000 0.000 2.116 2.120 grid_integrate_task_list 119 12.3 2.078 2.097 2.078 2.097 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.034 2.085 fft_wrap_pw1pw2_140 487 13.2 0.576 0.579 2.059 2.067 density_rs2pw 119 9.7 0.004 0.004 1.983 2.013 ot_diis_step 108 11.5 0.014 0.015 2.004 2.004 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.932 1.934 acc_transpose_blocks 9144 15.5 0.044 0.045 1.845 1.881 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.785 1.786 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.783 1.784 mp_sum_l 7287 12.8 1.055 1.707 1.055 1.707 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.577 1.578 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.539 1.575 grid_collocate_task_list 119 9.7 1.520 1.553 1.520 1.553 fft3d_ps 1201 14.6 0.595 0.605 1.520 1.524 cp_fm_cholesky_invert 11 10.9 1.391 1.394 1.391 1.394 wfi_extrapolate 11 7.9 0.001 0.001 1.363 1.363 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.300 1.322 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.301 1.301 cp_fm_diag_elpa_base 50 14.0 1.157 1.209 1.300 1.300 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.288 1.296 potential_pw2rs 119 12.3 0.014 0.014 1.215 1.215 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.013 1.211 acc_transpose_blocks_sync 27432 16.5 1.114 1.149 1.114 1.149 mp_alltoall_d11v 2130 13.8 1.122 1.144 1.122 1.144 qs_env_update_s_mstruct 11 6.9 0.002 0.002 1.107 1.120 make_images_data 4572 15.5 0.046 0.049 0.929 1.099 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.001 1.041 qs_create_task_list 11 7.9 0.023 0.025 0.948 0.962 generate_qs_task_list 11 8.9 0.367 0.387 0.925 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.896 0.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.018000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=729.545455, yerr=11.031885 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.923264E+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 2018798. 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 0.009 0.027 81.532 81.532 qs_mol_dyn_low 1 2.0 0.003 0.004 81.279 81.289 qs_forces 11 3.9 0.003 0.003 81.213 81.214 qs_energies 11 4.9 0.002 0.002 78.379 78.393 scf_env_do_scf 11 5.9 0.000 0.001 69.441 69.443 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.969 63.969 dbcsr_multiply_generic 2055 12.4 0.107 0.113 50.990 51.309 qs_scf_new_mos 99 7.5 0.000 0.001 47.064 47.186 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.063 47.186 ot_scf_mini 99 9.5 0.002 0.002 44.708 44.809 multiply_cannon 2055 13.4 0.198 0.210 42.609 43.361 velocity_verlet 10 3.0 0.003 0.006 43.305 43.306 multiply_cannon_loop 2055 14.4 1.801 1.844 41.607 42.463 ot_mini 99 10.5 0.001 0.001 26.578 26.682 qs_ot_get_derivative 99 11.5 0.001 0.001 19.763 19.858 multiply_cannon_multrec 49320 15.4 11.370 11.988 17.368 17.967 rebuild_ks_matrix 110 8.3 0.000 0.000 14.422 14.551 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.422 14.550 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.649 12.771 mp_waitall_1 220248 16.4 10.804 11.675 10.804 11.675 multiply_cannon_sync_h2d 49320 15.4 9.582 10.211 9.582 10.211 qs_ot_get_p 110 10.4 0.001 0.001 9.672 9.809 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.723 8.266 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.229 7.824 apply_single 110 13.6 0.000 0.001 7.229 7.823 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 6.543 7.671 sum_up_and_integrate 110 10.3 0.002 0.003 6.957 6.971 integrate_v_rspace 110 11.3 0.003 0.003 6.932 6.952 init_scf_run 11 5.9 0.000 0.001 6.885 6.885 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.885 6.885 ot_diis_step 99 11.5 0.006 0.006 6.601 6.601 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.399 6.420 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.249 6.387 calculate_rho_elec 110 8.6 0.020 0.024 6.249 6.387 dbcsr_mm_accdrv_process 87628 16.1 3.055 3.135 5.867 6.122 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.595 5.595 init_scf_loop 11 6.9 0.000 0.000 5.445 5.445 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.209 5.276 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.026 5.053 cp_fm_diag_elpa_base 48 14.0 5.013 5.040 5.024 5.051 mp_sum_l 6594 12.7 3.967 4.760 3.967 4.760 make_m2s 4110 13.4 0.060 0.064 3.952 4.088 wfi_extrapolate 11 7.9 0.001 0.001 4.063 4.063 make_images 4110 14.4 0.177 0.190 3.858 3.996 calculate_dm_sparse 110 9.5 0.001 0.001 3.731 3.847 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.426 3.692 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.674 3.682 density_rs2pw 110 9.6 0.004 0.005 3.299 3.444 grid_integrate_task_list 110 12.3 3.260 3.431 3.260 3.431 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.317 3.351 prepare_preconditioner 11 7.9 0.000 0.000 3.272 3.293 make_preconditioner 11 8.9 0.000 0.000 3.272 3.293 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.165 3.207 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.071 3.110 pw_transfer 1331 11.6 0.055 0.065 3.006 3.090 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.918 3.005 calculate_first_density_matrix 1 7.0 0.000 0.001 2.724 2.728 acc_transpose_blocks 49320 15.4 0.215 0.223 2.569 2.625 fft_wrap_pw1pw2_140 451 13.1 0.391 0.429 2.497 2.590 potential_pw2rs 110 12.3 0.005 0.006 2.560 2.579 jit_kernel_multiply 13 15.9 2.534 2.549 2.534 2.549 mp_alltoall_d11v 2046 13.8 2.020 2.420 2.020 2.420 grid_collocate_task_list 110 9.6 2.159 2.284 2.159 2.284 fft3d_ps 1111 14.6 0.793 0.886 2.213 2.268 mp_waitany 14300 13.8 1.820 2.148 1.820 2.148 transfer_rs2pw 451 10.6 0.005 0.006 1.897 2.059 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.921 mp_sum_d 3889 11.9 1.338 1.912 1.338 1.912 make_images_data 4110 15.4 0.043 0.047 1.761 1.897 hybrid_alltoall_any 4261 16.3 0.085 0.483 1.542 1.828 cp_fm_cholesky_invert 11 10.9 1.784 1.788 1.784 1.788 transfer_pw2rs 451 13.1 0.006 0.007 1.742 1.754 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.652 1.681 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.532000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=2.926087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 591.282176E+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 1870048. 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 0.029 70.102 70.103 qs_mol_dyn_low 1 2.0 0.003 0.004 69.486 69.829 qs_forces 11 3.9 0.003 0.003 69.413 69.413 qs_energies 11 4.9 0.002 0.003 66.092 66.095 scf_env_do_scf 11 5.9 0.000 0.001 57.517 57.520 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 49.929 49.929 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.189 38.393 velocity_verlet 10 3.0 0.001 0.001 36.179 36.180 qs_scf_new_mos 99 7.5 0.001 0.001 33.888 34.023 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.887 34.022 multiply_cannon 2055 13.4 0.225 0.248 31.380 32.469 ot_scf_mini 99 9.5 0.003 0.003 32.210 32.355 multiply_cannon_loop 2055 14.4 1.171 1.206 30.172 30.988 ot_mini 99 10.5 0.001 0.001 18.605 18.753 multiply_cannon_multrec 24660 15.4 7.003 8.949 14.046 15.926 rebuild_ks_matrix 110 8.3 0.000 0.000 13.618 13.733 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.618 13.732 qs_ot_get_derivative 99 11.5 0.001 0.001 12.793 12.941 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.991 12.092 mp_waitall_1 176588 16.5 7.653 9.905 7.653 9.905 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.186 8.038 multiply_cannon_sync_h2d 24660 15.4 6.445 7.922 6.445 7.922 init_scf_loop 11 6.9 0.000 0.000 7.552 7.552 dbcsr_mm_accdrv_process 52282 16.1 5.421 6.234 6.873 7.182 qs_ot_get_p 110 10.4 0.001 0.001 6.913 7.091 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.490 7.078 apply_single 110 13.6 0.000 0.001 6.490 7.077 sum_up_and_integrate 110 10.3 0.001 0.003 6.429 6.441 integrate_v_rspace 110 11.3 0.003 0.003 6.402 6.412 init_scf_run 11 5.9 0.000 0.001 6.198 6.198 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.197 6.198 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.997 6.004 calculate_rho_elec 110 8.6 0.039 0.047 5.996 6.004 ot_diis_step 99 11.5 0.010 0.010 5.761 5.761 prepare_preconditioner 11 7.9 0.000 0.000 5.540 5.562 make_preconditioner 11 8.9 0.000 0.000 5.540 5.562 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.755 5.475 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.120 5.275 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.935 4.957 make_m2s 4110 13.4 0.057 0.060 4.329 4.673 make_images 4110 14.4 0.407 0.466 4.220 4.559 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.454 4.455 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.825 3.836 cp_fm_diag_elpa_base 48 14.0 3.776 3.787 3.822 3.832 pw_transfer 1331 11.6 0.066 0.074 3.622 3.763 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.515 3.659 wfi_extrapolate 11 7.9 0.001 0.001 3.525 3.525 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.403 3.485 density_rs2pw 110 9.6 0.004 0.005 3.249 3.430 grid_integrate_task_list 110 12.3 3.164 3.313 3.164 3.313 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.244 3.245 fft_wrap_pw1pw2_140 451 13.1 0.460 0.473 2.964 3.109 calculate_dm_sparse 110 9.5 0.001 0.001 3.005 3.038 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.912 2.963 fft3d_ps 1111 14.6 1.111 1.334 2.610 2.765 make_images_data 4110 15.4 0.049 0.053 2.333 2.688 mp_sum_l 6594 12.7 1.852 2.616 1.852 2.616 cp_fm_cholesky_invert 11 10.9 2.596 2.603 2.596 2.603 hybrid_alltoall_any 4261 16.3 0.107 0.457 2.027 2.602 calculate_first_density_matrix 1 7.0 0.000 0.000 2.576 2.579 potential_pw2rs 110 12.3 0.008 0.009 2.303 2.317 jit_kernel_multiply 12 16.3 1.088 2.313 1.088 2.313 grid_collocate_task_list 110 9.6 2.167 2.305 2.167 2.305 mp_alltoall_d11v 2046 13.8 1.788 2.130 1.788 2.130 acc_transpose_blocks 24660 15.4 0.116 0.120 1.983 2.018 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.991 2.014 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.863 1.864 multiply_cannon_metrocomm4 22605 15.4 0.078 0.084 0.746 1.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.800 transfer_rs2pw 451 10.6 0.006 0.008 1.506 1.716 mp_irecv_dv 57340 16.2 0.618 1.602 0.618 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.565 1.576 mp_allgather_i34 2055 14.4 0.505 1.514 0.505 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.490 mp_waitany 10164 13.8 1.175 1.412 1.175 1.412 dbcsr_complete_redistribute 325 12.2 0.237 0.292 1.140 1.404 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.103000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.090909, yerr=5.743843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 668.176384E+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 0.008 0.030 61.223 61.225 qs_mol_dyn_low 1 2.0 0.003 0.004 60.938 60.948 qs_forces 11 3.9 0.003 0.003 60.871 60.872 qs_energies 11 4.9 0.002 0.002 57.656 57.659 scf_env_do_scf 11 5.9 0.000 0.001 49.624 49.625 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.722 40.723 velocity_verlet 10 3.0 0.001 0.002 33.315 33.317 dbcsr_multiply_generic 2055 12.4 0.108 0.114 29.156 29.361 qs_scf_new_mos 99 7.5 0.001 0.001 25.841 25.947 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.841 25.947 ot_scf_mini 99 9.5 0.002 0.003 24.558 24.675 multiply_cannon 2055 13.4 0.215 0.227 22.646 23.778 multiply_cannon_loop 2055 14.4 0.820 0.851 21.410 22.353 ot_mini 99 10.5 0.001 0.001 14.024 14.137 rebuild_ks_matrix 110 8.3 0.000 0.000 12.360 12.511 qs_ks_build_kohn_sham_matrix 110 9.3 0.040 0.093 12.360 12.511 multiply_cannon_multrec 16440 15.4 3.771 5.227 10.014 11.404 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.873 11.007 qs_ot_get_derivative 99 11.5 0.001 0.002 9.588 9.702 mp_waitall_1 139946 16.5 6.929 9.610 6.929 9.610 init_scf_loop 11 6.9 0.000 0.000 8.854 8.854 prepare_preconditioner 11 7.9 0.000 0.000 7.037 7.081 make_preconditioner 11 8.9 0.000 0.000 7.037 7.081 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.180 6.815 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.354 6.719 sum_up_and_integrate 110 10.3 0.001 0.002 6.410 6.427 integrate_v_rspace 110 11.3 0.003 0.003 6.384 6.401 dbcsr_mm_accdrv_process 34862 16.1 5.383 5.757 6.086 6.199 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.900 5.905 calculate_rho_elec 110 8.6 0.059 0.059 5.899 5.905 qs_ot_get_p 110 10.4 0.001 0.001 5.635 5.774 init_scf_run 11 5.9 0.000 0.001 5.613 5.614 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.613 5.613 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.872 5.382 apply_single 110 13.6 0.000 0.000 4.872 5.381 make_m2s 4110 13.4 0.049 0.051 4.402 4.805 make_images 4110 14.4 0.397 0.519 4.287 4.688 ot_diis_step 99 11.5 0.011 0.011 4.401 4.401 multiply_cannon_sync_h2d 16440 15.4 3.296 4.142 3.296 4.142 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.954 3.959 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.332 3.898 pw_transfer 1331 11.6 0.065 0.073 3.677 3.689 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.570 3.585 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.572 3.572 grid_integrate_task_list 110 12.3 3.209 3.449 3.209 3.449 density_rs2pw 110 9.6 0.004 0.005 3.099 3.254 wfi_extrapolate 11 7.9 0.001 0.001 3.102 3.102 fft_wrap_pw1pw2_140 451 13.1 0.579 0.588 3.063 3.080 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.980 2.990 cp_fm_diag_elpa_base 48 14.0 2.911 2.943 2.978 2.988 make_images_data 4110 15.4 0.046 0.050 2.431 2.960 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.134 2.904 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.820 2.821 cp_fm_cholesky_invert 11 10.9 2.806 2.813 2.806 2.813 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.563 2.631 calculate_dm_sparse 110 9.5 0.001 0.001 2.562 2.592 fft3d_ps 1111 14.6 1.091 1.100 2.513 2.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.468 2.508 calculate_first_density_matrix 1 7.0 0.000 0.001 2.409 2.411 grid_collocate_task_list 110 9.6 2.220 2.379 2.220 2.379 multiply_cannon_metrocomm4 14385 15.4 0.049 0.052 0.859 2.342 mp_alltoall_d11v 2046 13.8 1.910 2.295 1.910 2.295 potential_pw2rs 110 12.3 0.011 0.012 2.212 2.220 mp_irecv_dv 48980 15.7 0.785 2.205 0.785 2.205 mp_sum_l 6594 12.7 1.481 2.134 1.481 2.134 dbcsr_complete_redistribute 325 12.2 0.322 0.349 1.509 1.975 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.965 1.966 cp_fm_upper_to_full 70 14.2 1.417 1.802 1.417 1.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.683 1.700 cp_fm_cholesky_decompose 22 10.9 1.660 1.680 1.660 1.680 acc_transpose_blocks 16440 15.4 0.077 0.079 1.574 1.668 mp_allgather_i34 2055 14.4 0.545 1.569 0.545 1.569 transfer_rs2pw 451 10.6 0.005 0.006 1.350 1.514 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.467 1.480 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.018 1.473 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.361 1.473 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.316 1.330 mp_waitany 17072 13.8 1.165 1.328 1.165 1.328 rs_gather_matrices 110 12.3 0.138 0.151 0.900 1.304 mp_alltoall_z22v 1111 16.6 1.208 1.227 1.208 1.227 transfer_pw2rs 451 13.1 0.005 0.005 1.222 1.226 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.225000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.636364, yerr=9.048081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 734.294016E+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 0.009 0.028 65.884 65.886 qs_mol_dyn_low 1 2.0 0.003 0.004 65.656 65.673 qs_forces 11 3.9 0.003 0.003 65.529 65.531 qs_energies 11 4.9 0.001 0.002 62.141 62.145 scf_env_do_scf 11 5.9 0.000 0.001 53.942 53.945 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.392 42.392 velocity_verlet 10 3.0 0.001 0.002 37.266 37.271 dbcsr_multiply_generic 2055 12.4 0.114 0.120 30.931 31.193 qs_scf_new_mos 99 7.5 0.001 0.001 27.847 27.957 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.847 27.956 ot_scf_mini 99 9.5 0.002 0.003 26.151 26.255 multiply_cannon 2055 13.4 0.240 0.261 23.742 24.718 multiply_cannon_loop 2055 14.4 1.423 1.484 22.321 23.041 ot_mini 99 10.5 0.001 0.001 15.140 15.273 multiply_cannon_multrec 24660 15.4 4.095 6.712 12.981 14.130 rebuild_ks_matrix 110 8.3 0.000 0.000 11.930 12.067 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 11.930 12.066 init_scf_loop 11 6.9 0.000 0.000 11.511 11.512 qs_ot_get_derivative 99 11.5 0.001 0.001 10.940 11.047 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.538 10.661 prepare_preconditioner 11 7.9 0.000 0.000 9.786 9.804 make_preconditioner 11 8.9 0.000 0.000 9.786 9.804 dbcsr_mm_accdrv_process 52304 16.0 7.724 8.931 8.728 9.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.032 9.474 mp_waitall_1 121746 16.5 4.410 6.367 4.410 6.367 sum_up_and_integrate 110 10.3 0.001 0.002 6.212 6.225 integrate_v_rspace 110 11.3 0.003 0.003 6.186 6.198 qs_ot_get_p 110 10.4 0.001 0.001 5.875 6.023 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.794 5.805 calculate_rho_elec 110 8.6 0.078 0.081 5.794 5.805 make_m2s 4110 13.4 0.060 0.062 5.462 5.795 make_images 4110 14.4 0.577 0.704 5.321 5.650 init_scf_run 11 5.9 0.000 0.001 5.474 5.474 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.474 5.474 cp_fm_upper_to_full 70 14.2 3.321 4.810 3.321 4.810 ot_diis_step 99 11.5 0.011 0.011 4.158 4.158 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.064 4.130 apply_single 110 13.6 0.000 0.000 4.064 4.130 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.984 4.003 dbcsr_complete_redistribute 325 12.2 0.415 0.455 2.641 3.786 pw_transfer 1331 11.6 0.066 0.074 3.633 3.669 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.528 3.575 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.526 3.565 grid_integrate_task_list 110 12.3 3.281 3.480 3.281 3.480 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.401 3.457 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.416 3.416 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.559 3.290 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.096 3.227 fft_wrap_pw1pw2_140 451 13.1 0.611 0.633 3.038 3.076 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.353 3.068 density_rs2pw 110 9.6 0.004 0.004 2.917 3.061 make_images_data 4110 15.4 0.049 0.052 2.715 3.036 calculate_dm_sparse 110 9.5 0.001 0.001 2.994 3.025 wfi_extrapolate 11 7.9 0.001 0.001 3.000 3.001 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.871 2.880 cp_fm_diag_elpa_base 48 14.0 2.726 2.782 2.869 2.878 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.746 2.861 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.806 2.807 mp_alltoall_i22 605 13.7 1.645 2.803 1.645 2.803 multiply_cannon_sync_h2d 24660 15.4 2.379 2.712 2.379 2.712 cp_fm_cholesky_invert 11 10.9 2.676 2.685 2.676 2.685 acc_transpose_blocks 24660 15.4 0.114 0.117 2.419 2.559 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.445 2.493 grid_collocate_task_list 110 9.6 2.262 2.433 2.262 2.433 fft3d_ps 1111 14.6 1.084 1.114 2.406 2.420 calculate_first_density_matrix 1 7.0 0.000 0.000 2.380 2.383 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.231 2.233 mp_alltoall_d11v 2046 13.8 1.785 2.047 1.785 2.047 potential_pw2rs 110 12.3 0.012 0.013 2.026 2.030 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.812 1.843 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.709 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.651 cp_fm_cholesky_decompose 22 10.9 1.597 1.643 1.597 1.643 mp_sum_l 6594 12.7 1.038 1.622 1.038 1.622 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.561 acc_transpose_blocks_sync 73980 16.4 1.407 1.546 1.407 1.546 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.860 1.520 jit_kernel_multiply 8 15.9 0.661 1.444 0.661 1.444 mp_irecv_dv 62702 16.1 0.755 1.439 0.755 1.439 transfer_rs2pw 451 10.6 0.005 0.006 1.181 1.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.886000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.363636, yerr=9.441424 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 858.058752E+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 0.010 0.028 54.468 54.469 qs_mol_dyn_low 1 2.0 0.003 0.004 54.222 54.232 qs_forces 11 3.9 0.003 0.003 54.149 54.150 qs_energies 11 4.9 0.002 0.002 50.508 50.513 scf_env_do_scf 11 5.9 0.000 0.001 42.372 42.372 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 34.826 34.827 velocity_verlet 10 3.0 0.009 0.011 30.790 30.794 dbcsr_multiply_generic 2055 12.4 0.106 0.109 23.129 23.319 qs_scf_new_mos 99 7.5 0.001 0.001 20.487 20.548 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.486 20.547 ot_scf_mini 99 9.5 0.002 0.002 19.249 19.302 multiply_cannon 2055 13.4 0.238 0.247 17.658 18.762 multiply_cannon_loop 2055 14.4 0.604 0.628 16.411 16.739 rebuild_ks_matrix 110 8.3 0.000 0.000 11.477 11.535 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.477 11.535 ot_mini 99 10.5 0.001 0.001 10.723 10.773 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.209 10.261 multiply_cannon_multrec 8220 15.4 3.158 4.498 7.536 8.696 mp_waitall_1 103326 16.6 5.882 7.592 5.882 7.592 init_scf_loop 11 6.9 0.000 0.000 7.500 7.501 qs_ot_get_derivative 99 11.5 0.001 0.001 7.000 7.052 sum_up_and_integrate 110 10.3 0.001 0.002 6.113 6.127 integrate_v_rspace 110 11.3 0.003 0.003 6.087 6.100 prepare_preconditioner 11 7.9 0.000 0.000 5.874 5.880 make_preconditioner 11 8.9 0.000 0.000 5.873 5.880 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.869 5.879 calculate_rho_elec 110 8.6 0.113 0.114 5.869 5.879 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.461 5.530 dbcsr_mm_accdrv_process 17442 15.9 3.128 4.642 4.239 5.159 init_scf_run 11 5.9 0.000 0.001 5.016 5.016 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.015 5.016 qs_ot_get_p 110 10.4 0.001 0.001 4.572 4.636 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.147 4.549 make_m2s 4110 13.4 0.038 0.039 4.148 4.401 make_images 4110 14.4 0.648 0.708 4.019 4.270 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.769 3.832 apply_single 110 13.6 0.000 0.000 3.768 3.832 pw_transfer 1331 11.6 0.066 0.071 3.770 3.784 ot_diis_step 99 11.5 0.012 0.012 3.702 3.702 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.662 3.680 grid_integrate_task_list 110 12.3 3.398 3.507 3.398 3.507 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.241 3.244 fft_wrap_pw1pw2_140 451 13.1 0.778 0.788 3.186 3.211 density_rs2pw 110 9.6 0.004 0.004 2.851 2.961 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.906 2.907 cp_fm_cholesky_invert 11 10.9 2.831 2.834 2.831 2.834 wfi_extrapolate 11 7.9 0.001 0.001 2.667 2.667 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.665 2.665 hybrid_alltoall_any 4261 16.3 0.200 0.854 2.234 2.651 make_images_data 4110 15.4 0.041 0.046 2.212 2.544 multiply_cannon_sync_h2d 8220 15.4 2.350 2.525 2.350 2.525 grid_collocate_task_list 110 9.6 2.367 2.522 2.367 2.522 calculate_dm_sparse 110 9.5 0.001 0.001 2.459 2.501 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.479 2.481 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.345 2.352 cp_fm_diag_elpa_base 48 14.0 2.290 2.311 2.343 2.351 fft3d_ps 1111 14.6 1.133 1.155 2.327 2.343 calculate_first_density_matrix 1 7.0 0.000 0.000 2.254 2.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.122 2.141 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.992 2.029 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 2.014 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.889 1.920 potential_pw2rs 110 12.3 0.015 0.015 1.917 1.920 mp_alltoall_d11v 2046 13.8 1.685 1.822 1.685 1.822 cp_fm_cholesky_decompose 22 10.9 1.611 1.624 1.611 1.624 qs_env_update_s_mstruct 11 6.9 0.002 0.006 1.488 1.607 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.596 1.603 dbcsr_complete_redistribute 325 12.2 0.566 0.629 1.410 1.491 mp_allgather_i34 2055 14.4 0.413 1.468 0.413 1.468 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.425 1.437 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.479 1.410 acc_transpose_blocks 8220 15.4 0.039 0.040 1.367 1.401 mp_irecv_dv 24056 15.7 0.452 1.365 0.452 1.365 qs_create_task_list 11 7.9 0.001 0.001 1.203 1.301 generate_qs_task_list 11 8.9 0.372 0.440 1.202 1.300 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.187 1.208 transfer_rs2pw 451 10.6 0.005 0.005 1.090 1.206 mp_waitany 9240 13.8 1.044 1.176 1.044 1.176 jit_kernel_multiply 7 15.6 0.799 1.173 0.799 1.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.469000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.454545, yerr=11.586541 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 1.363964E+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 0.059 0.083 87.412 87.418 qs_mol_dyn_low 1 2.0 0.003 0.004 87.065 87.074 qs_forces 11 3.9 0.003 0.003 86.977 86.979 qs_energies 11 4.9 0.002 0.002 82.872 82.873 scf_env_do_scf 11 5.9 0.001 0.001 72.926 72.927 velocity_verlet 10 3.0 0.012 0.025 55.703 55.710 scf_env_do_scf_inner_loop 99 6.5 0.004 0.008 44.619 44.620 dbcsr_multiply_generic 2055 12.4 0.120 0.126 29.496 29.584 init_scf_loop 11 6.9 0.000 0.000 28.234 28.234 qs_scf_new_mos 99 7.5 0.001 0.001 27.232 27.267 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.231 27.266 prepare_preconditioner 11 7.9 0.000 0.000 26.250 26.257 make_preconditioner 11 8.9 0.000 0.000 26.250 26.257 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.519 25.731 ot_scf_mini 99 9.5 0.002 0.002 25.425 25.448 multiply_cannon 2055 13.4 0.333 0.356 22.347 23.134 multiply_cannon_loop 2055 14.4 0.821 0.834 20.557 20.848 cp_fm_upper_to_full 70 14.2 12.649 18.113 12.649 18.113 ot_mini 99 10.5 0.001 0.001 14.175 14.200 rebuild_ks_matrix 110 8.3 0.000 0.001 13.637 13.669 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.636 13.668 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.373 12.402 dbcsr_complete_redistribute 325 12.2 1.016 1.039 7.314 10.461 multiply_cannon_multrec 8220 15.4 4.072 4.204 9.735 9.880 qs_ot_get_derivative 99 11.5 0.001 0.001 9.600 9.625 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.286 9.434 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.716 8.842 mp_alltoall_i22 605 13.7 5.343 8.505 5.343 8.505 mp_waitall_1 84994 16.7 7.474 8.304 7.474 8.304 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.146 7.180 calculate_rho_elec 110 8.6 0.223 0.223 7.146 7.179 sum_up_and_integrate 110 10.3 0.002 0.002 6.812 6.828 integrate_v_rspace 110 11.3 0.004 0.004 6.784 6.800 qs_ot_get_p 110 10.4 0.001 0.001 5.953 5.982 make_m2s 4110 13.4 0.043 0.044 5.426 5.913 make_images 4110 14.4 0.884 0.937 5.238 5.723 dbcsr_mm_accdrv_process 11614 15.7 3.836 4.107 5.515 5.719 init_scf_run 11 5.9 0.000 0.001 5.670 5.670 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.669 5.670 cp_fm_cholesky_invert 11 10.9 5.305 5.310 5.305 5.310 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.711 5.178 apply_single 110 13.6 0.000 0.000 4.711 5.178 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.794 5.130 pw_transfer 1331 11.6 0.074 0.075 4.968 4.973 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.851 4.857 ot_diis_step 99 11.5 0.015 0.016 4.561 4.561 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.340 4.347 fft_wrap_pw1pw2_140 451 13.1 1.280 1.283 4.214 4.219 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.814 3.815 grid_integrate_task_list 110 12.3 3.712 3.805 3.712 3.805 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.659 3.661 density_rs2pw 110 9.6 0.004 0.004 3.576 3.618 hybrid_alltoall_any 4261 16.3 0.262 0.562 2.830 3.557 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.030 3.480 make_images_data 4110 15.4 0.045 0.049 2.824 3.464 wfi_extrapolate 11 7.9 0.001 0.001 3.321 3.322 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.200 3.200 cp_fm_diag_elpa_base 48 14.0 2.660 2.863 3.198 3.198 calculate_dm_sparse 110 9.5 0.001 0.001 3.168 3.191 multiply_cannon_sync_h2d 8220 15.4 3.129 3.159 3.129 3.159 fft3d_ps 1111 14.6 1.294 1.302 2.905 2.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.867 2.871 grid_collocate_task_list 110 9.6 2.697 2.729 2.697 2.729 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.678 2.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.368 2.381 potential_pw2rs 110 12.3 0.021 0.021 2.279 2.282 calculate_first_density_matrix 1 7.0 0.000 0.000 2.236 2.236 qs_env_update_s_mstruct 11 6.9 0.002 0.002 2.168 2.226 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.106 2.173 mp_alltoall_d11v 2046 13.8 2.082 2.173 2.082 2.173 cp_fm_cholesky_decompose 22 10.9 2.034 2.052 2.034 2.052 qs_create_task_list 11 7.9 0.001 0.001 1.881 1.929 generate_qs_task_list 11 8.9 0.730 0.783 1.880 1.928 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.862 1.866 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.771 1.805 acc_transpose_blocks 8220 15.4 0.041 0.042 1.718 1.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.418000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1233.363636, yerr=48.694358 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 631.533568E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1954140. 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 0.009 0.027 207.482 207.485 qs_mol_dyn_low 1 2.0 0.004 0.004 206.997 207.011 qs_forces 11 3.9 0.005 0.005 206.462 206.464 qs_energies 11 4.9 0.002 0.003 200.880 200.896 scf_env_do_scf 11 5.9 0.001 0.001 184.550 184.555 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.050 164.054 dbcsr_multiply_generic 2507 12.6 0.180 0.183 124.658 125.169 qs_scf_new_mos 117 7.6 0.001 0.001 124.433 124.622 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.432 124.621 velocity_verlet 10 3.0 0.001 0.002 124.018 124.024 ot_scf_mini 117 9.6 0.003 0.003 117.735 117.890 multiply_cannon 2507 13.6 0.240 0.250 101.249 102.713 multiply_cannon_loop 2507 14.6 2.405 2.453 99.056 100.410 ot_mini 117 10.6 0.001 0.001 65.271 65.413 multiply_cannon_multrec 60168 15.6 31.740 33.261 41.496 43.016 qs_ot_get_derivative 117 11.6 0.001 0.002 40.623 40.786 rebuild_ks_matrix 128 8.3 0.001 0.001 33.619 33.951 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.619 33.951 mp_waitall_1 267128 16.5 28.443 30.593 28.443 30.593 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.189 30.492 qs_ot_get_p 128 10.4 0.001 0.001 30.263 30.466 multiply_cannon_sync_h2d 60168 15.6 26.463 28.120 26.463 28.120 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.290 25.186 apply_single 128 13.6 0.001 0.001 24.289 25.185 ot_diis_step 117 11.6 0.008 0.008 24.405 24.407 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.694 23.766 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.914 20.914 init_scf_loop 11 6.9 0.000 0.001 20.425 20.426 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.185 19.346 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.887 17.919 cp_fm_diag_elpa_base 83 14.4 17.815 17.863 17.883 17.916 multiply_cannon_metrocomm3 60168 15.6 0.119 0.123 15.814 17.287 prepare_preconditioner 11 7.9 0.000 0.000 15.824 15.884 make_preconditioner 11 8.9 0.000 0.000 15.824 15.884 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.071 15.228 make_m2s 5014 13.6 0.104 0.114 13.750 14.146 sum_up_and_integrate 128 10.3 0.002 0.004 14.083 14.102 integrate_v_rspace 128 11.3 0.004 0.004 14.025 14.045 make_images 5014 14.6 0.397 0.417 13.567 13.975 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.300 13.490 calculate_rho_elec 128 8.7 0.045 0.064 13.299 13.489 init_scf_run 11 5.9 0.000 0.001 12.207 12.207 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.207 12.207 dbcsr_mm_accdrv_process 124484 16.2 4.719 4.881 9.323 9.864 mp_sum_l 7950 12.9 8.396 9.369 8.396 9.369 cp_fm_cholesky_invert 11 10.9 9.095 9.103 9.095 9.103 wfi_extrapolate 11 7.9 0.001 0.001 9.007 9.008 calculate_dm_sparse 128 9.5 0.001 0.001 8.498 8.621 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.151 8.291 multiply_cannon_metrocomm1 60168 15.6 0.095 0.100 6.147 8.261 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.143 8.215 pw_transfer 1547 11.6 0.074 0.089 7.637 7.819 make_images_data 5014 15.6 0.067 0.073 6.693 7.664 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.436 7.620 grid_integrate_task_list 128 12.3 7.101 7.462 7.101 7.462 density_rs2pw 128 9.7 0.006 0.007 6.906 7.457 hybrid_alltoall_any 5200 16.5 0.297 2.284 5.846 7.210 fft_wrap_pw1pw2_140 523 13.2 1.124 1.174 6.496 6.680 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.612 6.623 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.767 5.858 fft3d_ps 1291 14.7 2.193 2.844 5.444 5.793 mp_alltoall_d11v 2415 14.1 4.326 5.328 4.326 5.328 grid_collocate_task_list 128 9.7 4.839 5.184 4.839 5.184 potential_pw2rs 128 12.3 0.009 0.010 4.566 4.593 cp_fm_cholesky_decompose 22 10.9 4.570 4.586 4.570 4.586 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.485000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.909091, yerr=6.430814 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 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 11613089636352 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 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 844.918784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57818. MP_Allreduce 11153 957. MP_Sync 87 MP_Alltoall 1983 5778090. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.067 193.504 193.536 qs_mol_dyn_low 1 2.0 0.003 0.008 193.118 193.140 qs_forces 11 3.9 0.004 0.005 192.944 192.946 qs_energies 11 4.9 0.002 0.003 186.151 186.162 scf_env_do_scf 11 5.9 0.001 0.001 169.633 169.644 scf_env_do_scf_inner_loop 118 6.6 0.036 0.063 135.794 135.798 velocity_verlet 10 3.0 0.004 0.006 122.613 122.617 dbcsr_multiply_generic 2527 12.6 0.193 0.199 98.019 99.271 qs_scf_new_mos 118 7.6 0.001 0.001 96.603 97.162 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.602 97.161 ot_scf_mini 118 9.6 0.004 0.004 91.733 92.362 multiply_cannon 2527 13.6 0.508 0.573 77.384 81.775 multiply_cannon_loop 2527 14.6 1.593 1.689 74.358 77.391 ot_mini 118 10.6 0.001 0.001 50.839 51.383 mp_waitall_1 216438 16.6 24.479 38.412 24.479 38.412 multiply_cannon_multrec 30324 15.6 21.076 26.087 31.702 37.088 init_scf_loop 11 6.9 0.000 0.000 33.736 33.737 rebuild_ks_matrix 129 8.3 0.001 0.001 32.612 33.324 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 32.612 33.323 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.341 29.988 prepare_preconditioner 11 7.9 0.000 0.000 29.379 29.464 make_preconditioner 11 8.9 0.000 0.000 29.379 29.464 qs_ot_get_derivative 118 11.6 0.001 0.002 28.637 29.256 multiply_cannon_metrocomm3 30324 15.6 0.098 0.105 15.883 28.778 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.001 28.556 qs_ot_get_p 129 10.4 0.001 0.001 22.830 23.541 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.270 23.357 apply_single 129 13.6 0.001 0.001 22.270 23.356 ot_diis_step 118 11.6 0.015 0.015 22.014 22.016 multiply_cannon_sync_h2d 30324 15.6 18.034 20.590 18.034 20.590 qs_ot_p2m_diag 83 11.4 0.189 0.217 17.725 17.766 cp_fm_cholesky_invert 11 10.9 17.154 17.167 17.154 17.167 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.498 16.498 make_m2s 5054 13.6 0.091 0.095 14.420 15.649 make_images 5054 14.6 1.177 1.388 14.205 15.434 sum_up_and_integrate 129 10.3 0.002 0.004 14.279 14.308 integrate_v_rspace 129 11.3 0.003 0.004 14.218 14.252 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.735 13.769 calculate_rho_elec 129 8.7 0.088 0.104 13.734 13.768 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.237 13.272 cp_fm_diag_elpa_base 83 14.4 12.977 13.072 13.230 13.261 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.310 11.820 init_scf_run 11 5.9 0.000 0.001 11.696 11.698 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.696 11.698 multiply_cannon_metrocomm4 27797 15.6 0.106 0.119 3.903 11.163 mp_irecv_dv 70031 16.3 3.696 10.753 3.696 10.753 dbcsr_mm_accdrv_process 62734 16.2 5.469 6.039 10.077 10.640 make_images_data 5054 15.6 0.068 0.077 8.493 10.017 hybrid_alltoall_any 5240 16.5 0.354 1.530 7.224 9.451 pw_transfer 1559 11.6 0.085 0.095 8.835 8.906 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.610 8.680 wfi_extrapolate 11 7.9 0.001 0.001 8.508 8.508 density_rs2pw 129 9.7 0.006 0.007 7.357 7.725 grid_integrate_task_list 129 12.3 7.236 7.677 7.236 7.677 fft_wrap_pw1pw2_140 527 13.2 1.221 1.240 7.562 7.647 cp_fm_cholesky_decompose 22 10.9 7.210 7.303 7.210 7.303 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.453 7.230 calculate_dm_sparse 129 9.5 0.001 0.001 6.531 6.666 mp_sum_l 8010 12.9 4.402 6.636 4.402 6.636 fft3d_ps 1301 14.7 2.863 3.041 6.152 6.207 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.107 6.114 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.472 5.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.165 5.382 grid_collocate_task_list 129 9.7 5.046 5.375 5.046 5.375 mp_alltoall_d11v 2423 14.1 4.381 4.988 4.381 4.988 potential_pw2rs 129 12.3 0.016 0.019 4.675 4.691 mp_allgather_i34 2527 14.6 1.534 4.664 1.534 4.664 dbcsr_complete_redistribute 395 12.7 0.773 0.871 3.356 4.202 mp_sum_d 4502 12.1 2.695 4.113 2.695 4.113 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.536000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.545455, yerr=3.893202 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 949.436416E+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 0.020 0.041 173.377 173.378 qs_mol_dyn_low 1 2.0 0.003 0.004 172.969 172.983 qs_forces 11 3.9 0.004 0.004 172.781 172.784 qs_energies 11 4.9 0.002 0.003 166.267 166.276 scf_env_do_scf 11 5.9 0.001 0.001 150.834 150.835 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.187 116.187 velocity_verlet 10 3.0 0.002 0.002 111.479 111.484 dbcsr_multiply_generic 2507 12.6 0.180 0.186 81.610 82.716 qs_scf_new_mos 117 7.6 0.001 0.001 79.418 79.723 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.417 79.722 ot_scf_mini 117 9.6 0.003 0.004 75.160 75.511 multiply_cannon 2507 13.6 0.512 0.527 61.989 67.041 multiply_cannon_loop 2507 14.6 1.134 1.192 59.065 61.390 ot_mini 117 10.6 0.001 0.001 42.413 42.778 init_scf_loop 11 6.9 0.000 0.000 34.551 34.552 mp_waitall_1 170520 16.6 24.341 33.862 24.341 33.862 prepare_preconditioner 11 7.9 0.000 0.000 30.605 30.655 make_preconditioner 11 8.9 0.000 0.000 30.605 30.655 rebuild_ks_matrix 128 8.3 0.001 0.001 29.998 30.467 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.998 30.466 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.169 29.647 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.016 27.452 multiply_cannon_multrec 20056 15.6 13.219 16.225 22.526 25.586 multiply_cannon_metrocomm3 20056 15.6 0.063 0.066 15.077 24.969 qs_ot_get_derivative 117 11.6 0.002 0.002 22.916 23.277 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.539 20.602 apply_single 128 13.6 0.001 0.001 19.538 20.602 ot_diis_step 117 11.6 0.018 0.018 19.386 19.386 qs_ot_get_p 128 10.4 0.001 0.001 18.222 18.687 make_m2s 5014 13.6 0.079 0.084 14.553 15.754 make_images 5014 14.6 1.144 1.226 14.319 15.517 multiply_cannon_sync_h2d 20056 15.6 13.605 15.488 13.605 15.488 cp_fm_cholesky_invert 11 10.9 14.058 14.067 14.058 14.067 qs_ot_p2m_diag 83 11.4 0.266 0.272 13.927 13.936 sum_up_and_integrate 128 10.3 0.002 0.003 13.823 13.847 integrate_v_rspace 128 11.3 0.003 0.004 13.764 13.787 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.425 13.456 calculate_rho_elec 128 8.7 0.130 0.145 13.424 13.455 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.877 12.878 init_scf_run 11 5.9 0.000 0.001 10.433 10.434 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.433 10.433 make_images_data 5014 15.6 0.063 0.072 8.704 10.262 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.753 9.773 cp_fm_diag_elpa_base 83 14.4 9.346 9.495 9.750 9.770 hybrid_alltoall_any 5200 16.5 0.449 2.031 7.544 9.448 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.054 9.312 multiply_cannon_metrocomm4 17549 15.6 0.067 0.075 3.421 9.098 dbcsr_mm_accdrv_process 41502 16.2 5.632 5.938 8.761 8.876 mp_irecv_dv 50230 16.2 3.293 8.836 3.293 8.836 pw_transfer 1547 11.6 0.085 0.103 8.603 8.710 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.379 8.494 fft_wrap_pw1pw2_140 523 13.2 1.305 1.352 7.477 7.598 grid_integrate_task_list 128 12.3 7.275 7.588 7.275 7.588 cp_fm_upper_to_full 105 14.8 5.833 7.440 5.833 7.440 wfi_extrapolate 11 7.9 0.001 0.001 7.381 7.381 cp_fm_cholesky_decompose 22 10.9 7.207 7.235 7.207 7.235 density_rs2pw 128 9.7 0.006 0.006 6.767 7.212 dbcsr_complete_redistribute 395 12.7 1.170 1.199 4.618 6.396 calculate_dm_sparse 128 9.5 0.001 0.001 5.848 5.957 fft3d_ps 1291 14.7 2.760 2.987 5.752 5.843 grid_collocate_task_list 128 9.7 5.160 5.605 5.160 5.605 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.479 5.488 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.619 5.272 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.482 5.258 mp_alltoall_d11v 2415 14.1 4.319 4.782 4.319 4.782 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.585 4.716 mp_sum_l 7950 12.9 3.230 4.603 3.230 4.603 mp_allgather_i34 2507 14.6 1.420 4.549 1.420 4.549 potential_pw2rs 128 12.3 0.020 0.022 4.246 4.255 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.418 4.160 mp_alltoall_i22 716 14.1 2.029 3.943 2.029 3.943 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.880 3.912 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.723 3.724 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.480 3.519 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.378000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.727273, yerr=2.987578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.142063E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57598. MP_Allreduce 11189 1079. MP_Sync 88 MP_Alltoall 1724 12509605. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.047 188.036 188.037 qs_mol_dyn_low 1 2.0 0.003 0.004 187.582 187.595 qs_forces 11 3.9 0.004 0.004 187.473 187.480 qs_energies 11 4.9 0.003 0.010 180.366 180.379 scf_env_do_scf 11 5.9 0.001 0.002 163.747 163.759 velocity_verlet 10 3.0 0.002 0.002 125.154 125.157 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.442 117.444 dbcsr_multiply_generic 2529 12.6 0.199 0.215 81.154 81.951 qs_scf_new_mos 118 7.6 0.001 0.001 81.056 81.360 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.055 81.359 ot_scf_mini 118 9.6 0.003 0.004 76.424 76.747 multiply_cannon 2529 13.6 0.557 0.589 56.133 58.900 multiply_cannon_loop 2529 14.6 1.837 1.937 52.344 54.341 init_scf_loop 11 6.9 0.001 0.004 46.181 46.182 ot_mini 118 10.6 0.001 0.001 43.211 43.514 prepare_preconditioner 11 7.9 0.000 0.000 42.076 42.097 make_preconditioner 11 8.9 0.000 0.003 42.076 42.097 make_full_inverse_cholesky 11 9.9 0.011 0.022 35.731 40.676 multiply_cannon_multrec 30348 15.6 13.740 19.063 26.693 31.706 rebuild_ks_matrix 129 8.3 0.001 0.001 29.457 29.741 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 29.456 29.741 mp_waitall_1 149172 16.7 17.288 27.258 17.288 27.258 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.602 26.856 qs_ot_get_derivative 118 11.6 0.002 0.002 23.366 23.692 make_m2s 5058 13.6 0.097 0.102 20.585 21.759 make_images 5058 14.6 1.952 2.248 20.273 21.446 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.228 19.756 apply_single 129 13.6 0.001 0.001 19.227 19.756 ot_diis_step 118 11.6 0.018 0.018 19.715 19.716 qs_ot_get_p 129 10.4 0.001 0.001 19.210 19.542 cp_fm_upper_to_full 106 14.8 11.447 16.790 11.447 16.790 cp_fm_cholesky_invert 11 10.9 16.079 16.087 16.079 16.087 qs_ot_p2m_diag 84 11.4 0.347 0.395 15.033 15.090 multiply_cannon_metrocomm3 30348 15.6 0.050 0.054 6.291 14.967 sum_up_and_integrate 129 10.3 0.002 0.003 14.025 14.052 integrate_v_rspace 129 11.3 0.003 0.005 13.965 13.992 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.697 13.744 calculate_rho_elec 129 8.7 0.174 0.190 13.696 13.743 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.617 13.619 make_images_data 5058 15.6 0.067 0.074 11.141 13.054 dbcsr_mm_accdrv_process 62780 16.2 8.527 9.525 12.522 12.993 dbcsr_complete_redistribute 397 12.7 1.495 1.618 8.936 12.677 hybrid_alltoall_any 5245 16.5 0.534 2.241 9.865 12.227 multiply_cannon_sync_h2d 30348 15.6 10.598 11.363 10.598 11.363 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.582 11.317 init_scf_run 11 5.9 0.000 0.001 10.726 10.727 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.726 10.727 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.495 10.506 cp_fm_diag_elpa_base 84 14.4 9.526 9.843 10.488 10.499 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.859 10.106 transfer_fm_to_dbcsr 11 9.9 0.002 0.015 6.322 9.987 mp_alltoall_i22 720 14.1 5.556 9.397 5.556 9.397 pw_transfer 1559 11.6 0.086 0.100 8.961 9.047 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.734 8.825 grid_integrate_task_list 129 12.3 7.617 7.909 7.617 7.909 fft_wrap_pw1pw2_140 527 13.2 1.455 1.495 7.799 7.906 wfi_extrapolate 11 7.9 0.001 0.001 7.652 7.652 cp_fm_cholesky_decompose 22 10.9 7.488 7.577 7.488 7.577 density_rs2pw 129 9.7 0.006 0.007 6.744 7.201 multiply_cannon_metrocomm4 25290 15.6 0.085 0.095 2.842 7.085 mp_irecv_dv 76751 16.2 2.686 6.811 2.686 6.811 calculate_dm_sparse 129 9.5 0.001 0.001 6.321 6.406 fft3d_ps 1301 14.7 2.875 2.951 5.890 5.958 grid_collocate_task_list 129 9.7 5.330 5.672 5.330 5.672 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.501 5.546 mp_alltoall_d11v 2429 14.1 4.715 5.295 4.715 5.295 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.459 4.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.418 4.536 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.385 4.386 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.275 4.346 potential_pw2rs 129 12.3 0.023 0.024 4.253 4.262 dbcsr_special_finalize 12645 15.6 0.067 0.070 3.545 3.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.037000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.363636, yerr=18.006427 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.559577E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3989 58388. MP_Allreduce 11050 1168. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.055 168.609 168.614 qs_mol_dyn_low 1 2.0 0.003 0.004 168.196 168.210 qs_forces 11 3.9 0.004 0.005 168.095 168.099 qs_energies 11 4.9 0.002 0.002 160.713 160.721 scf_env_do_scf 11 5.9 0.001 0.001 143.601 143.615 velocity_verlet 10 3.0 0.002 0.002 110.430 110.434 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 108.133 108.134 dbcsr_multiply_generic 2485 12.5 0.189 0.199 71.658 72.118 qs_scf_new_mos 116 7.6 0.001 0.001 71.947 72.050 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.947 72.049 ot_scf_mini 116 9.6 0.003 0.004 67.573 67.675 multiply_cannon 2485 13.5 0.557 0.582 53.238 56.567 multiply_cannon_loop 2485 14.5 0.800 0.841 50.129 51.233 ot_mini 116 10.6 0.001 0.001 37.535 37.638 init_scf_loop 11 6.9 0.000 0.000 35.318 35.319 prepare_preconditioner 11 7.9 0.000 0.000 31.503 31.534 make_preconditioner 11 8.9 0.000 0.000 31.503 31.534 mp_waitall_1 124680 16.7 24.655 31.124 24.655 31.124 make_full_inverse_cholesky 11 9.9 0.014 0.024 29.464 29.714 rebuild_ks_matrix 127 8.3 0.001 0.001 28.657 28.794 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.657 28.793 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.094 26.218 multiply_cannon_multrec 9940 15.5 10.249 16.331 17.776 22.389 qs_ot_get_derivative 116 11.6 0.001 0.002 20.387 20.484 multiply_cannon_metrocomm3 9940 15.5 0.026 0.028 12.951 20.331 cp_fm_cholesky_invert 11 10.9 18.073 18.078 18.073 18.078 qs_ot_get_p 127 10.4 0.001 0.001 17.409 17.571 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.221 17.381 apply_single 127 13.6 0.001 0.001 17.221 17.381 ot_diis_step 116 11.6 0.020 0.021 17.078 17.078 make_m2s 4970 13.5 0.065 0.068 14.577 15.594 make_images 4970 14.5 2.147 2.566 14.274 15.285 sum_up_and_integrate 127 10.3 0.002 0.003 13.932 13.976 integrate_v_rspace 127 11.3 0.004 0.004 13.872 13.918 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.828 13.853 calculate_rho_elec 127 8.7 0.253 0.264 13.828 13.853 qs_ot_p2m_diag 82 11.4 0.489 0.496 13.670 13.687 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.522 12.523 multiply_cannon_sync_h2d 9940 15.5 10.629 11.119 10.629 11.119 init_scf_run 11 5.9 0.000 0.001 10.326 10.326 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.326 10.326 make_images_data 4970 15.5 0.055 0.064 8.362 9.826 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.656 9.667 cp_fm_diag_elpa_base 82 14.4 9.423 9.505 9.653 9.663 hybrid_alltoall_any 5155 16.4 0.837 3.750 8.197 9.414 pw_transfer 1535 11.6 0.084 0.092 9.138 9.165 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 8.918 8.951 grid_integrate_task_list 127 12.3 7.705 8.071 7.705 8.071 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.940 8.016 cp_fm_cholesky_decompose 22 10.9 7.813 7.955 7.813 7.955 fft_wrap_pw1pw2_140 519 13.2 1.788 1.831 7.910 7.944 dbcsr_mm_accdrv_process 20590 16.1 3.479 4.516 7.182 7.853 wfi_extrapolate 11 7.9 0.001 0.001 7.264 7.264 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.101 7.184 density_rs2pw 127 9.7 0.005 0.006 6.595 6.773 calculate_dm_sparse 127 9.5 0.001 0.001 6.035 6.112 multiply_cannon_metrocomm4 7455 15.5 0.026 0.029 1.858 6.078 mp_irecv_dv 28618 15.9 1.818 5.989 1.818 5.989 fft3d_ps 1281 14.7 2.742 2.806 5.675 5.705 grid_collocate_task_list 127 9.7 5.539 5.701 5.539 5.701 dbcsr_complete_redistribute 393 12.7 2.095 2.169 5.174 5.527 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.230 5.231 mp_alltoall_d11v 2401 14.1 4.556 5.193 4.556 5.193 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.146 5.154 mp_allgather_i34 2485 14.5 1.312 4.612 1.312 4.612 potential_pw2rs 127 12.3 0.026 0.026 4.166 4.171 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.065 4.095 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.577 3.876 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.610 3.699 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.622 3.684 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.306 3.612 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.447 3.554 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.454 3.467 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.614000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1480.636364, yerr=20.759544 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 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 3.208835E+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 11.342275E+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 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 59207. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954383. 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 0.034 0.057 291.068 291.073 qs_mol_dyn_low 1 2.0 0.003 0.004 290.423 290.450 qs_forces 11 3.9 0.004 0.004 290.316 290.319 qs_energies 11 4.9 0.002 0.002 281.493 281.499 scf_env_do_scf 11 5.9 0.001 0.001 258.331 258.338 velocity_verlet 10 3.0 0.002 0.003 209.966 209.974 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 133.732 133.733 init_scf_loop 11 6.9 0.000 0.000 124.321 124.323 prepare_preconditioner 11 7.9 0.000 0.000 119.528 119.548 make_preconditioner 11 8.9 0.000 0.000 119.528 119.548 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.878 116.639 qs_scf_new_mos 116 7.6 0.001 0.001 90.726 90.805 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.725 90.804 ot_scf_mini 116 9.6 0.004 0.004 85.875 85.922 dbcsr_multiply_generic 2485 12.5 0.212 0.224 82.117 82.676 cp_fm_upper_to_full 104 14.8 52.680 75.216 52.680 75.216 multiply_cannon 2485 13.5 0.685 0.739 58.794 60.056 multiply_cannon_loop 2485 14.5 1.045 1.068 54.729 56.265 ot_mini 116 10.6 0.001 0.001 44.806 44.875 dbcsr_complete_redistribute 393 12.7 3.986 4.022 29.655 42.276 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.080 38.732 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.608 36.139 cp_fm_cholesky_invert 11 10.9 33.889 33.895 33.889 33.895 mp_alltoall_i22 712 14.1 21.348 33.886 21.348 33.886 rebuild_ks_matrix 127 8.3 0.001 0.001 33.644 33.706 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.644 33.705 mp_waitall_1 102768 16.8 28.199 32.312 28.199 32.312 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.186 31.255 qs_ot_get_p 127 10.4 0.001 0.001 26.259 26.367 qs_ot_get_derivative 116 11.6 0.002 0.002 24.889 24.933 qs_ot_p2m_diag 82 11.4 0.868 0.873 22.103 22.134 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 18.837 20.377 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.314 20.316 make_m2s 4970 13.5 0.075 0.078 18.672 20.186 ot_diis_step 116 11.6 0.022 0.022 19.882 19.882 make_images 4970 14.5 3.040 3.217 18.196 19.714 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.235 19.406 apply_single 127 13.6 0.001 0.001 19.234 19.405 multiply_cannon_multrec 9940 15.5 10.242 12.053 18.082 18.136 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.103 17.104 cp_fm_diag_elpa_base 82 14.4 12.802 14.344 17.099 17.099 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.400 16.414 calculate_rho_elec 127 8.7 0.476 0.477 16.399 16.414 sum_up_and_integrate 127 10.3 0.002 0.002 15.881 15.968 integrate_v_rspace 127 11.3 0.004 0.004 15.819 15.906 multiply_cannon_sync_h2d 9940 15.5 14.182 14.214 14.182 14.214 hybrid_alltoall_any 5155 16.4 1.301 3.030 10.777 12.932 make_images_data 4970 15.5 0.064 0.069 10.614 12.824 init_scf_run 11 5.9 0.000 0.001 12.777 12.777 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.776 12.777 pw_transfer 1535 11.6 0.095 0.095 11.302 11.308 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.065 11.073 fft_wrap_pw1pw2_140 519 13.2 3.014 3.068 9.850 9.862 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.702 9.734 dbcsr_mm_accdrv_process 20590 16.0 4.260 6.063 7.591 9.488 cp_fm_cholesky_decompose 22 10.9 9.027 9.051 9.027 9.051 wfi_extrapolate 11 7.9 0.001 0.001 9.008 9.008 grid_integrate_task_list 127 12.3 8.493 8.690 8.493 8.690 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.212 8.212 density_rs2pw 127 9.7 0.005 0.005 7.481 7.535 mp_alltoall_d11v 2401 14.1 6.893 7.063 6.893 7.063 calculate_dm_sparse 127 9.5 0.001 0.001 6.763 6.819 grid_collocate_task_list 127 9.7 6.366 6.446 6.366 6.446 fft3d_ps 1281 14.7 2.826 2.832 6.359 6.410 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.358 6.410 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.245 6.300 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.073000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2917.000000, yerr=159.866990 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 1.259852E+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 44744890. 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 0.074 0.286 85.597 85.598 qs_energies 1 2.0 0.000 0.000 84.405 84.415 ls_scf 1 3.0 0.000 0.000 83.504 83.514 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.531 72.680 multiply_cannon 111 7.7 0.018 0.020 55.756 56.998 multiply_cannon_loop 111 8.7 0.227 0.241 52.380 53.748 ls_scf_main 1 4.0 0.000 0.000 52.291 52.292 density_matrix_trs4 2 5.0 0.002 0.003 46.767 46.851 ls_scf_init_scf 1 4.0 0.000 0.000 28.168 28.170 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.040 27.095 mp_waitall_1 11031 10.9 22.488 25.064 22.488 25.064 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.952 24.972 multiply_cannon_multrec 2664 9.7 8.124 8.867 15.491 17.196 multiply_cannon_sync_h2d 2664 9.7 13.473 15.042 13.473 15.042 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.775 13.641 make_m2s 222 7.7 0.008 0.011 13.120 13.613 make_images 222 8.7 0.098 0.108 13.098 13.592 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.420 8.551 make_images_data 222 9.7 0.004 0.005 7.627 8.245 hybrid_alltoall_any 227 10.6 0.216 1.834 6.530 8.005 dbcsr_mm_accdrv_process 4760 10.4 0.587 0.699 6.984 7.979 dbcsr_mm_accdrv_process_sort 4760 11.4 6.198 7.114 6.198 7.114 calculate_norms 4752 9.8 5.507 6.149 5.507 6.149 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.004 5.130 mp_sum_l 887 5.1 3.319 4.520 3.319 4.520 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.049 3.712 mp_irecv_dv 6231 10.9 2.033 3.685 2.033 3.685 make_images_sizes 222 9.7 0.000 0.000 0.789 3.678 mp_alltoall_i44 222 10.7 0.788 3.678 0.788 3.678 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.354 3.479 arnoldi_extremal 4 6.8 0.000 0.000 3.230 3.251 arnoldi_normal_ev 4 7.8 0.001 0.002 3.230 3.250 build_subspace 16 8.4 0.009 0.012 3.139 3.140 ls_scf_post 1 4.0 0.000 0.000 3.045 3.054 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.352 2.865 dbcsr_merge_single_wm 555 10.7 0.457 0.589 2.343 2.856 make_images_pack 222 9.7 2.207 2.632 2.209 2.633 dbcsr_sort_data 658 11.4 2.143 2.596 2.143 2.596 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.305 2.549 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.454 2.063 2.456 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.313 2.377 buffer_matrices_ensure_size 222 8.7 1.744 2.067 1.744 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.785 1.786 rebuild_ks_matrix 3 7.3 0.000 0.000 1.775 1.777 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.775 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.598000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 2.172936E+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 0.025 0.037 90.137 90.139 qs_energies 1 2.0 0.000 0.000 89.666 89.670 ls_scf 1 3.0 0.000 0.000 88.296 88.300 dbcsr_multiply_generic 111 6.7 0.018 0.038 74.418 74.744 multiply_cannon 111 7.7 0.033 0.043 52.746 57.008 ls_scf_main 1 4.0 0.000 0.000 54.413 54.419 multiply_cannon_loop 111 8.7 0.136 0.146 49.918 52.887 density_matrix_trs4 2 5.0 0.002 0.003 48.739 48.961 ls_scf_init_scf 1 4.0 0.000 0.000 30.136 30.137 mp_waitall_1 9105 10.9 20.754 29.631 20.754 29.631 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.950 29.038 multiply_cannon_multrec 1332 9.7 13.260 17.160 22.507 27.512 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.427 26.436 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.550 20.210 make_m2s 222 7.7 0.006 0.008 15.174 15.767 make_images 222 8.7 1.368 1.692 15.143 15.738 dbcsr_mm_accdrv_process 4041 10.4 0.344 0.535 8.842 10.416 dbcsr_mm_accdrv_process_sort 4041 11.4 8.361 9.880 8.361 9.880 make_images_data 222 9.7 0.004 0.005 8.754 9.643 hybrid_alltoall_any 227 10.6 0.541 2.550 8.139 9.448 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.214 8.019 mp_irecv_dv 3311 11.0 3.194 7.966 3.194 7.966 mp_sum_l 887 5.1 5.026 7.560 5.026 7.560 calculate_norms 2376 9.8 6.065 6.858 6.065 6.858 multiply_cannon_sync_h2d 1332 9.7 4.925 6.139 4.925 6.139 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.778 5.995 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.053 5.263 arnoldi_extremal 4 6.8 0.000 0.000 4.593 4.606 arnoldi_normal_ev 4 7.8 0.001 0.004 4.593 4.606 build_subspace 16 8.4 0.014 0.021 4.343 4.345 ls_scf_post 1 4.0 0.000 0.000 3.747 3.751 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.568 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.120 3.346 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.214 2.743 3.216 mp_allgather_i34 111 8.7 0.817 3.170 0.817 3.170 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.098 2.756 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.620 2.706 dbcsr_data_new 4174 10.1 2.113 2.401 2.113 2.401 make_images_pack 222 9.7 1.877 2.198 1.879 2.201 dbcsr_sort_data 436 11.2 1.829 2.102 1.829 2.102 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.852 1.854 rebuild_ks_matrix 3 7.3 0.000 0.000 1.840 1.842 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.840 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.139000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1789.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 2.860233E+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 0.040 0.058 92.717 92.719 qs_energies 1 2.0 0.001 0.006 92.111 92.125 ls_scf 1 3.0 0.000 0.001 90.479 90.490 dbcsr_multiply_generic 111 6.7 0.016 0.020 74.911 75.226 multiply_cannon 111 7.7 0.046 0.115 51.846 56.258 ls_scf_main 1 4.0 0.000 0.001 56.247 56.250 multiply_cannon_loop 111 8.7 0.117 0.129 49.077 52.820 density_matrix_trs4 2 5.0 0.003 0.022 50.393 50.584 mp_waitall_1 7281 11.0 23.337 32.120 23.337 32.120 ls_scf_init_scf 1 4.0 0.000 0.001 30.583 30.587 ls_scf_init_matrix_S 1 5.0 0.008 0.066 29.371 29.463 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.808 26.831 multiply_cannon_multrec 888 9.7 12.648 15.418 21.209 24.247 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.049 21.522 make_m2s 222 7.7 0.006 0.007 16.482 17.235 make_images 222 8.7 1.581 1.848 16.444 17.193 make_images_data 222 9.7 0.004 0.004 9.544 10.718 hybrid_alltoall_any 227 10.6 0.640 2.961 9.166 10.418 dbcsr_mm_accdrv_process 3754 10.4 0.316 0.493 8.070 9.322 dbcsr_mm_accdrv_process_sort 3754 11.4 7.636 8.829 7.636 8.829 mp_sum_l 887 5.1 4.869 8.545 4.869 8.545 multiply_cannon_sync_h2d 888 9.7 6.113 7.538 6.113 7.538 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.498 7.021 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.387 7.007 mp_irecv_dv 2335 11.1 2.482 6.987 2.482 6.987 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.745 6.619 arnoldi_extremal 4 6.8 0.000 0.000 5.150 5.169 arnoldi_normal_ev 4 7.8 0.001 0.005 5.150 5.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.701 4.926 build_subspace 16 8.4 0.014 0.020 4.776 4.782 calculate_norms 1584 9.8 4.333 4.739 4.333 4.739 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.438 3.776 mp_allgather_i34 111 8.7 0.874 3.738 0.874 3.738 ls_scf_post 1 4.0 0.001 0.011 3.649 3.661 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.599 3.021 3.601 ls_scf_store_result 1 5.0 0.000 0.000 3.387 3.448 ls_scf_dm_to_ks 2 5.0 0.000 0.008 2.843 2.934 dbcsr_data_new 4116 9.9 2.101 2.455 2.101 2.455 make_images_sizes 222 9.7 0.000 0.000 1.114 2.239 mp_alltoall_i44 222 10.7 1.114 2.238 1.114 2.238 dbcsr_sort_data 325 11.1 1.880 2.145 1.880 2.145 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.891 make_images_pack 222 9.7 1.622 1.876 1.625 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.871 1.872 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.871 1.872 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.719000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2266.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 3.377119E+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 0.028 0.045 97.657 97.658 qs_energies 1 2.0 0.000 0.000 97.033 97.039 ls_scf 1 3.0 0.000 0.000 95.198 95.203 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.985 79.244 ls_scf_main 1 4.0 0.000 0.000 59.358 59.360 multiply_cannon 111 7.7 0.040 0.071 51.868 56.643 density_matrix_trs4 2 5.0 0.002 0.003 53.218 53.363 multiply_cannon_loop 111 8.7 0.153 0.168 46.856 50.257 ls_scf_init_scf 1 4.0 0.000 0.000 32.569 32.571 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.366 31.432 mp_waitall_1 6369 11.0 23.149 29.077 23.149 29.077 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.911 28.922 multiply_cannon_multrec 1332 9.7 14.077 17.359 21.956 25.269 make_m2s 222 7.7 0.006 0.008 21.265 22.684 make_images 222 8.7 3.136 3.593 21.215 22.636 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.461 17.738 make_images_data 222 9.7 0.004 0.004 11.869 13.382 hybrid_alltoall_any 227 10.6 0.797 3.730 11.261 13.005 dbcsr_mm_accdrv_process 3641 10.4 0.296 0.491 7.520 9.032 dbcsr_mm_accdrv_process_sort 3641 11.4 7.079 8.535 7.079 8.535 mp_sum_l 887 5.1 4.268 8.053 4.268 8.053 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.334 6.366 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.098 6.228 mp_irecv_dv 3229 10.9 2.074 6.146 2.074 6.146 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.650 6.020 multiply_cannon_sync_h2d 1332 9.7 5.412 5.996 5.412 5.996 arnoldi_extremal 4 6.8 0.000 0.000 5.259 5.274 arnoldi_normal_ev 4 7.8 0.001 0.005 5.259 5.274 build_subspace 16 8.4 0.014 0.021 4.923 4.933 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.562 4.740 calculate_norms 2376 9.8 4.179 4.497 4.179 4.497 mp_allgather_i34 111 8.7 2.136 4.415 2.136 4.415 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.596 3.901 dbcsr_matrix_vector_mult_local 304 10.0 3.195 3.696 3.197 3.698 dbcsr_sort_data 658 11.4 3.108 3.462 3.108 3.462 ls_scf_post 1 4.0 0.000 0.000 3.271 3.278 dbcsr_special_finalize 555 9.7 0.006 0.007 2.863 3.262 dbcsr_merge_single_wm 555 10.7 0.543 0.669 2.855 3.254 ls_scf_store_result 1 5.0 0.000 0.000 2.995 3.076 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.989 3.027 dbcsr_data_release 10477 10.7 1.581 2.408 1.581 2.408 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.991 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.658000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2770.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 4.742504E+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 0.037 0.054 92.747 92.748 qs_energies 1 2.0 0.000 0.000 91.845 91.850 ls_scf 1 3.0 0.000 0.000 89.782 89.786 dbcsr_multiply_generic 111 6.7 0.018 0.020 71.355 71.566 ls_scf_main 1 4.0 0.000 0.000 56.969 56.969 multiply_cannon 111 7.7 0.090 0.169 52.811 55.639 multiply_cannon_loop 111 8.7 0.088 0.093 50.199 51.929 density_matrix_trs4 2 5.0 0.002 0.003 50.066 50.141 ls_scf_init_scf 1 4.0 0.000 0.000 29.365 29.366 mp_waitall_1 5436 11.0 24.907 29.136 24.907 29.136 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.124 28.154 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.018 26.029 multiply_cannon_multrec 444 9.7 13.654 16.226 20.706 21.868 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.925 16.091 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.307 14.940 make_m2s 222 7.7 0.005 0.005 13.733 14.737 make_images 222 8.7 2.037 2.480 13.665 14.668 hybrid_alltoall_any 227 10.6 0.801 3.821 8.267 10.019 make_images_data 222 9.7 0.003 0.004 8.490 9.867 multiply_cannon_sync_h2d 444 9.7 6.745 8.656 6.745 8.656 dbcsr_mm_accdrv_process 3003 10.4 0.352 0.406 6.748 7.855 dbcsr_mm_accdrv_process_sort 3003 11.4 6.385 7.454 6.385 7.454 arnoldi_extremal 4 6.8 0.000 0.000 5.879 5.895 arnoldi_normal_ev 4 7.8 0.001 0.005 5.879 5.895 build_subspace 16 8.4 0.015 0.019 5.475 5.483 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.460 4.655 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.198 4.413 dbcsr_matrix_vector_mult_local 304 10.0 3.707 4.191 3.709 4.194 mp_sum_l 887 5.1 2.812 4.066 2.812 4.066 mp_allgather_i34 111 8.7 1.198 3.811 1.198 3.811 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.563 3.755 calculate_norms 792 9.8 3.614 3.725 3.614 3.725 mp_irecv_dv 1241 11.2 1.551 3.725 1.551 3.725 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.472 3.517 ls_scf_post 1 4.0 0.000 0.000 3.448 3.454 make_images_sizes 222 9.7 0.000 0.000 0.869 3.386 mp_alltoall_i44 222 10.7 0.869 3.386 0.869 3.386 ls_scf_store_result 1 5.0 0.000 0.000 3.228 3.268 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.994 3.247 dbcsr_finalize 304 7.8 0.062 0.077 2.201 2.307 dbcsr_data_new 4608 9.7 1.779 2.252 1.779 2.252 dbcsr_merge_all 275 8.9 0.480 0.529 2.059 2.143 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.046 2.046 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.011 2.012 rebuild_ks_matrix 3 7.3 0.000 0.000 1.979 1.980 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.979 1.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.748000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3778.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 8.991142E+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 0.072 0.093 105.590 105.592 qs_energies 1 2.0 0.000 0.000 104.166 104.172 ls_scf 1 3.0 0.000 0.000 101.196 101.201 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.771 74.935 ls_scf_main 1 4.0 0.000 0.000 63.782 63.782 density_matrix_trs4 2 5.0 0.002 0.003 54.744 54.808 multiply_cannon 111 7.7 0.148 0.244 48.653 50.575 multiply_cannon_loop 111 8.7 0.098 0.099 45.644 46.311 ls_scf_init_scf 1 4.0 0.000 0.000 33.620 33.620 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.122 32.146 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.424 29.432 mp_waitall_1 4527 11.1 21.866 25.764 21.866 25.764 make_m2s 222 7.7 0.005 0.005 22.399 23.492 make_images 222 8.7 3.571 3.859 22.292 23.383 multiply_cannon_multrec 444 9.7 17.793 18.420 22.426 22.994 hybrid_alltoall_any 227 10.6 1.652 3.614 12.558 15.337 make_images_data 222 9.7 0.003 0.004 12.777 14.959 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.491 10.820 multiply_cannon_sync_h2d 444 9.7 8.791 8.839 8.791 8.839 arnoldi_extremal 4 6.8 0.000 0.000 7.515 7.524 arnoldi_normal_ev 4 7.8 0.002 0.009 7.515 7.524 build_subspace 16 8.4 0.026 0.036 6.963 6.973 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.530 5.674 dbcsr_matrix_vector_mult_local 304 10.0 5.063 5.380 5.065 5.383 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.063 5.155 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.866 5.119 dbcsr_mm_accdrv_process 1814 10.4 0.220 0.348 4.443 4.584 dbcsr_mm_accdrv_process_sort 1814 11.4 4.104 4.244 4.104 4.244 ls_scf_post 1 4.0 0.000 0.000 3.794 3.801 make_images_sizes 222 9.7 0.000 0.000 1.496 3.648 mp_alltoall_i44 222 10.7 1.496 3.648 1.496 3.648 ls_scf_store_result 1 5.0 0.000 0.000 3.516 3.549 mp_allgather_i34 111 8.7 1.127 3.523 1.127 3.523 calculate_norms 792 9.8 3.229 3.272 3.229 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.124 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 2.938 2.938 dbcsr_merge_all 275 8.9 0.885 0.912 2.854 2.893 dbcsr_complete_redistribute 5 7.6 1.429 1.463 2.732 2.836 matrix_ls_to_qs 2 6.0 0.000 0.000 2.383 2.487 dbcsr_sort_data 325 11.1 2.436 2.479 2.436 2.479 dbcsr_data_new 6591 9.6 1.882 2.470 1.882 2.470 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.381 2.383 rebuild_ks_matrix 3 7.3 0.000 0.000 2.315 2.317 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.315 2.317 dbcsr_new_transposed 4 7.5 0.241 0.251 2.262 2.276 dbcsr_add_d 103 6.2 0.000 0.000 2.149 2.220 dbcsr_add_anytype 103 7.2 0.858 0.890 2.148 2.220 dbcsr_data_release 12724 10.6 1.988 2.211 1.988 2.211 dbcsr_frobenius_norm 74 6.6 2.056 2.130 2.183 2.211 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.592000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7083.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a01af14846922d2d741ffda4c698585069c13d48_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 588.689408E+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 4334532. 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 0.046 0.113 239.291 239.299 qs_mol_dyn_low 1 2.0 0.004 0.013 238.110 238.154 qs_forces 5 3.8 0.005 0.053 237.899 237.901 qs_energies 5 4.8 0.004 0.033 234.635 234.675 scf_env_do_scf 5 5.8 0.000 0.002 219.502 219.505 scf_env_do_scf_inner_loop 105 6.6 0.002 0.017 191.598 191.612 qs_scf_new_mos 105 7.6 0.000 0.001 149.584 149.706 qs_scf_loop_do_ot 105 8.6 0.001 0.001 149.583 149.705 ot_scf_mini 105 9.6 0.003 0.004 139.744 139.863 dbcsr_multiply_generic 1445 12.2 0.127 0.135 134.617 134.974 multiply_cannon 1445 13.2 0.276 0.287 115.151 117.301 multiply_cannon_loop 1445 14.2 2.862 3.011 113.305 114.394 velocity_verlet 4 3.0 0.003 0.013 108.362 108.363 ot_mini 105 10.6 0.001 0.001 61.015 61.119 qs_ot_get_p 112 10.4 0.001 0.001 48.715 49.043 multiply_cannon_multrec 69360 15.2 29.655 34.361 39.502 44.118 mp_waitall_1 488190 16.1 35.834 42.484 35.834 42.484 qs_ot_get_derivative 55 11.6 0.001 0.001 39.090 39.212 qs_ot_p2m_diag 40 11.0 0.020 0.031 37.413 37.492 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.077 34.078 multiply_cannon_sync_h2d 69360 15.2 29.039 33.133 29.039 33.133 multiply_cannon_metrocomm3 69360 15.2 0.202 0.217 26.030 32.976 rebuild_ks_matrix 110 8.4 0.000 0.000 30.143 30.300 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.015 30.143 30.300 cp_fm_syevd 40 13.0 0.000 0.000 28.766 28.901 init_scf_loop 7 6.6 0.000 0.005 27.863 27.864 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.582 27.722 cp_fm_redistribute_end 40 14.0 11.959 23.869 11.965 23.871 cp_fm_syevd_base 40 14.0 11.897 23.807 11.897 23.807 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.146 23.435 apply_single 62 13.6 0.000 0.000 23.146 23.434 prepare_preconditioner 7 7.6 0.000 0.000 22.803 22.829 make_preconditioner 7 8.6 0.000 0.004 22.803 22.829 ot_new_cg_direction 55 11.6 0.001 0.001 21.169 21.169 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.908 18.225 calculate_rho_elec 110 8.6 0.030 0.032 17.908 18.224 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.094 15.163 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.764 14.950 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.776 13.918 mp_sum_l 4764 12.2 12.258 13.058 12.258 13.058 pw_transfer 1645 12.4 0.079 0.100 12.533 12.783 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.395 12.649 density_rs2pw 110 9.6 0.005 0.006 12.139 12.486 init_scf_run 5 5.8 0.000 0.001 12.395 12.396 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.395 12.396 calculate_dm_sparse 110 9.5 0.000 0.001 11.396 11.516 fft_wrap_pw1pw2_240 915 15.0 1.128 1.232 10.906 11.138 qs_vxc_create 110 10.4 0.002 0.004 10.924 10.971 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.802 10.864 dbcsr_mm_accdrv_process 154766 15.8 6.223 6.397 9.714 10.551 cp_fm_cholesky_invert 7 10.6 10.512 10.522 10.512 10.522 check_diag 80 13.5 8.684 8.937 9.741 9.876 fft3d_pb 915 16.0 2.380 2.666 9.143 9.421 sum_up_and_integrate 60 10.3 0.001 0.002 9.123 9.136 integrate_v_rspace 60 11.3 0.001 0.002 9.106 9.119 transfer_rs2pw 445 10.6 0.007 0.008 8.623 8.997 acc_transpose_blocks 69360 15.2 0.356 0.369 8.081 8.582 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.940 7.958 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.616 7.878 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.502 7.560 calculate_first_density_matrix 1 7.0 0.000 0.010 7.495 7.507 make_m2s 2890 13.2 0.079 0.088 6.808 7.470 make_full_single_inverse 7 9.6 0.001 0.003 7.408 7.434 multiply_cannon_metrocomm1 69360 15.2 0.096 0.103 4.681 7.408 xc_vxc_pw_create 60 11.3 0.039 0.050 7.332 7.379 make_images 2890 14.2 0.241 0.260 6.700 7.362 xc_pw_derive 510 13.4 0.005 0.007 6.653 6.736 mp_alltoall_z22v 2340 17.7 6.227 6.528 6.227 6.528 acc_transpose_blocks_kernels 69360 16.2 0.847 0.887 5.300 5.695 potential_pw2rs 60 12.3 0.002 0.003 5.266 5.299 mp_waitany 7680 13.5 4.571 5.042 4.571 5.042 multiply_cannon_metrocomm4 67915 15.2 0.184 0.200 2.023 5.022 jit_kernel_transpose 5 15.0 4.453 4.817 4.453 4.817 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=239.299000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a01af14846922d2d741ffda4c698585069c13d48 Summary: empty Status: OK