=== 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: 0165da4645489888fae55c9cbb9ccbdcebf313e1 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # 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. # # Author: Matthias Krack (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/01 job id: 45420141 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/02 job id: 45420143 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/03 job id: 45420145 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/04 job id: 45420146 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/05 job id: 45420148 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/06 job id: 45420150 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/07 job id: 45420152 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/08 job id: 45420153 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/09 job id: 45420154 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/10 job id: 45420155 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/11 job id: 45420156 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/12 job id: 45420159 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/13 job id: 45420161 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/14 job id: 45420162 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/15 job id: 45420164 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/16 job id: 45420167 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/17 job id: 45420169 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/18 job id: 45420171 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/19 job id: 45420175 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/20 job id: 45420178 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/21 job id: 45420181 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/22 job id: 45420183 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/23 job id: 45420185 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/24 job id: 45420187 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/25 job id: 45420189 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/26 job id: 45420190 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/27 job id: 45420191 --- 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/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 134.447 134.447 farming_run 1 2.0 133.945 133.949 134.419 134.422 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.464091E+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 1103589. MP_Allreduce 491 2254389. 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 0.007 0.025 115.869 115.869 qs_energies 1 2.0 0.000 0.000 115.673 115.676 mp2_main 1 3.0 0.000 0.000 113.602 113.605 mp2_gpw_main 1 4.0 0.021 0.028 112.567 112.570 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.460 93.955 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.552 56.046 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.868 47.347 get_2c_integrals 1 6.0 0.000 0.000 36.692 37.737 integrate_v_rspace 273 8.0 0.436 0.448 25.195 30.467 pw_transfer 6555 10.6 0.372 0.380 27.335 27.922 grid_integrate_task_list 273 9.0 21.006 26.780 21.006 26.780 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.007 26.560 fft_wrap_pw1pw2_100 2178 12.4 1.160 1.262 23.539 24.098 rpa_ri_compute_en 1 5.0 0.001 0.001 18.995 19.242 compute_2c_integrals 1 7.0 0.002 0.003 19.191 19.193 compute_2c_integrals_loop_lm 1 8.0 0.004 0.005 18.811 18.917 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.431 18.808 18.913 cp_fm_cholesky_decompose 12 8.2 17.454 18.490 17.454 18.490 cholesky_decomp 1 7.0 0.000 0.000 16.357 17.393 fft3d_s 5443 13.4 16.150 16.642 16.173 16.664 ao_to_mo_and_store_B_mult_1 272 7.0 10.865 15.602 10.865 15.602 calculate_wavefunction 272 8.0 5.481 5.647 12.590 13.217 rpa_num_int 1 6.0 0.000 0.000 10.523 10.524 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.507 10.513 calc_mat_Q 8 8.0 0.000 0.000 9.360 9.446 contract_S_to_Q 8 9.0 0.000 0.000 8.779 8.866 calc_potential_gpw 544 9.5 0.005 0.005 8.208 8.581 parallel_gemm_fm 14 9.1 0.000 0.000 8.351 8.443 parallel_gemm_fm_cosma 14 10.1 8.351 8.443 8.351 8.443 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.154 8.362 potential_pw2rs 545 10.0 0.107 0.110 7.679 8.312 create_integ_mat 1 6.0 0.021 0.028 7.881 7.882 collocate_single_gaussian 272 10.0 0.040 0.042 7.444 7.723 array2fm 1 7.0 0.000 0.000 6.852 7.324 pw_scatter_s 2720 13.7 4.426 4.584 4.426 4.584 pw_gather_s 2722 13.2 3.871 4.176 3.871 4.176 array2fm_buffer_send 1 8.0 2.979 3.164 2.979 3.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.568562, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.038 394.752 394.754 farming_run 1 2.0 394.019 394.025 394.712 394.715 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.230320E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 8736432. 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 0.010 0.039 209.377 209.378 qs_energies 1 2.0 0.000 0.000 209.159 209.165 scf_env_do_scf 1 3.0 0.000 0.000 106.340 106.341 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.461 105.468 rebuild_ks_matrix 4 6.0 0.000 0.000 105.438 105.446 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.438 105.446 hfx_ks_matrix 4 8.0 0.001 0.001 105.053 105.058 integrate_four_center 4 9.0 0.144 0.462 105.053 105.057 mp2_main 1 3.0 0.000 0.000 102.531 102.536 mp2_gpw_main 1 4.0 0.039 0.072 101.693 101.704 integrate_four_center_main 4 10.0 0.126 0.443 96.558 99.282 integrate_four_center_bin 266 11.0 96.433 99.219 96.433 99.219 init_scf_loop 1 4.0 0.000 0.000 91.979 91.980 mp2_ri_gpw_compute_in 1 5.0 0.066 0.072 74.830 75.880 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.421 55.468 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 41.970 47.150 integrate_v_rspace 95 8.0 0.398 0.566 28.342 33.388 pw_transfer 2240 10.6 0.144 0.162 29.932 30.271 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.939 29.318 grid_integrate_task_list 95 9.0 23.633 28.888 23.633 28.888 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 26.700 28.429 ao_to_mo_and_store_B_mult_1 91 7.0 10.772 27.303 10.772 27.303 fft_wrap_pw1pw2_100 730 12.4 1.285 1.462 26.660 27.042 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.847 1.904 25.010 25.020 get_2c_integrals 1 6.0 0.000 0.000 20.324 20.346 compute_2c_integrals 1 7.0 0.004 0.009 19.313 19.321 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.885 19.176 mp2_eri_2c_integrate_gpw 1 9.0 1.737 1.855 18.883 19.175 fft3d_s 1823 13.4 18.451 18.803 18.464 18.817 scf_env_do_scf_inner_loop 4 4.0 0.001 0.022 14.360 14.360 calculate_wavefunction 91 8.0 2.029 2.074 9.748 10.001 potential_pw2rs 186 10.0 0.034 0.035 8.672 9.175 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.583 8.813 9.120 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.258 8.658 local_gemm 172 8.0 8.259 8.577 8.259 8.577 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.517 7.961 8.221 collocate_single_gaussian 91 10.0 0.017 0.019 7.899 8.150 calc_potential_gpw 182 9.5 0.002 0.002 7.889 8.083 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.430 6.344 6.430 mp_sendrecv_dm3 2068 8.0 5.994 6.274 5.994 6.274 mp_sync 37 10.5 3.160 5.448 3.160 5.448 pw_gather_s 912 13.2 4.904 5.410 4.904 5.410 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.695300, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.547136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 588782. 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 0.018 0.034 52.033 52.058 qs_mol_dyn_low 1 2.0 0.003 0.003 51.679 51.708 qs_forces 11 3.9 0.002 0.004 51.593 51.621 qs_energies 11 4.9 0.001 0.004 50.128 50.157 scf_env_do_scf 11 5.9 0.000 0.001 44.294 44.319 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.263 42.287 dbcsr_multiply_generic 2286 12.5 0.093 0.097 32.246 32.652 qs_scf_new_mos 108 7.5 0.000 0.001 32.335 32.644 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.335 32.644 ot_scf_mini 108 9.5 0.002 0.002 30.675 30.865 multiply_cannon 2286 13.5 0.194 0.205 25.558 27.054 multiply_cannon_loop 2286 14.5 1.462 1.562 24.890 26.398 velocity_verlet 10 3.0 0.001 0.001 26.329 26.345 ot_mini 108 10.5 0.001 0.001 18.293 18.538 qs_ot_get_derivative 108 11.5 0.001 0.001 15.407 15.585 mp_waitall_1 245248 16.5 8.100 14.005 8.100 14.005 multiply_cannon_metrocomm3 54864 15.5 0.070 0.076 5.837 12.356 multiply_cannon_multrec 54864 15.5 4.311 6.800 7.495 10.920 qs_ot_get_p 119 10.4 0.001 0.001 7.755 8.022 rebuild_ks_matrix 119 8.3 0.000 0.000 7.818 7.965 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.818 7.965 multiply_cannon_sync_h2d 54864 15.5 6.020 7.123 6.020 7.123 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.892 7.024 mp_sum_l 7207 12.9 4.876 6.450 4.876 6.450 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.837 6.279 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.055 5.081 init_scf_run 11 5.9 0.000 0.001 4.565 4.567 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.565 4.566 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.440 4.545 sum_up_and_integrate 119 10.3 0.012 0.014 4.464 4.471 integrate_v_rspace 119 11.3 0.002 0.002 4.452 4.459 dbcsr_mm_accdrv_process 76910 16.1 1.107 1.816 3.103 4.408 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.318 4.321 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.054 4.199 calculate_rho_elec 119 8.7 0.012 0.017 4.054 4.199 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.132 4.134 cp_fm_redistribute_end 50 14.0 2.109 4.108 2.115 4.111 cp_fm_diag_elpa_base 50 14.0 1.990 4.006 1.994 4.010 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.880 3.085 apply_single 119 13.6 0.000 0.000 2.880 3.085 calculate_dm_sparse 119 9.5 0.000 0.001 2.876 3.031 rs_pw_transfer 974 11.9 0.011 0.013 2.640 2.726 multiply_cannon_metrocomm1 54864 15.5 0.054 0.058 1.665 2.677 ot_diis_step 108 11.5 0.006 0.006 2.637 2.637 jit_kernel_multiply 13 15.8 1.939 2.515 1.939 2.515 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.458 2.509 calculate_first_density_matrix 1 7.0 0.000 0.003 2.462 2.470 density_rs2pw 119 9.7 0.004 0.004 2.100 2.228 acc_transpose_blocks 54864 15.5 0.223 0.253 1.690 2.199 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.192 2.194 grid_integrate_task_list 119 12.3 2.055 2.147 2.055 2.147 wfi_extrapolate 11 7.9 0.001 0.001 2.043 2.044 init_scf_loop 11 6.9 0.000 0.000 2.015 2.016 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.948 2.003 mp_sum_d 4135 12.0 1.241 1.939 1.241 1.939 potential_pw2rs 119 12.3 0.004 0.004 1.791 1.800 pw_transfer 1439 11.6 0.051 0.055 1.657 1.726 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.582 1.651 make_m2s 4572 13.5 0.053 0.056 1.510 1.554 make_images 4572 14.5 0.133 0.139 1.429 1.472 mp_alltoall_d11v 2130 13.8 1.206 1.442 1.206 1.442 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.372 1.407 mp_waitany 12084 13.8 1.241 1.407 1.241 1.407 grid_collocate_task_list 119 9.7 1.291 1.382 1.291 1.382 fft_wrap_pw1pw2_140 487 13.2 0.183 0.201 1.222 1.293 fft3d_ps 1201 14.6 0.373 0.480 1.226 1.291 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.678 1.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.058000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.485440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 1030466. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 38.162 38.164 qs_mol_dyn_low 1 2.0 0.003 0.003 37.963 37.970 qs_forces 11 3.9 0.003 0.005 37.893 37.894 qs_energies 11 4.9 0.001 0.001 36.216 36.219 scf_env_do_scf 11 5.9 0.000 0.001 31.186 31.187 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.678 28.678 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.029 21.378 qs_scf_new_mos 108 7.5 0.001 0.001 19.687 19.929 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.686 19.928 ot_scf_mini 108 9.5 0.002 0.003 18.796 18.968 velocity_verlet 10 3.0 0.001 0.001 17.945 17.946 multiply_cannon 2286 13.5 0.208 0.217 16.219 17.835 multiply_cannon_loop 2286 14.5 0.897 0.970 15.134 16.502 ot_mini 108 10.5 0.001 0.001 11.552 11.787 mp_waitall_1 200699 16.5 5.512 10.734 5.512 10.734 multiply_cannon_metrocomm3 27432 15.5 0.067 0.072 4.116 9.459 qs_ot_get_derivative 108 11.5 0.001 0.001 9.117 9.294 multiply_cannon_multrec 27432 15.5 1.974 4.504 5.851 8.591 rebuild_ks_matrix 119 8.3 0.000 0.000 7.154 7.291 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.153 7.290 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.324 6.448 dbcsr_mm_accdrv_process 47894 16.0 2.953 4.957 3.806 5.622 qs_ot_get_p 119 10.4 0.001 0.001 4.388 4.617 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.520 4.373 sum_up_and_integrate 119 10.3 0.024 0.027 4.130 4.158 integrate_v_rspace 119 11.3 0.002 0.003 4.106 4.135 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.033 4.107 apply_single 119 13.6 0.000 0.000 3.033 4.107 mp_sum_l 7207 12.9 1.994 3.936 1.994 3.936 init_scf_run 11 5.9 0.000 0.001 3.841 3.841 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.841 3.841 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.686 3.722 calculate_rho_elec 119 8.7 0.021 0.024 3.685 3.721 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.976 2.995 rs_pw_transfer 974 11.9 0.010 0.011 2.447 2.880 multiply_cannon_sync_h2d 27432 15.5 2.201 2.785 2.201 2.785 make_m2s 4572 13.5 0.052 0.054 2.368 2.591 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.547 2.547 make_images 4572 14.5 0.200 0.235 2.280 2.499 init_scf_loop 11 6.9 0.000 0.000 2.489 2.489 density_rs2pw 119 9.7 0.004 0.004 2.018 2.469 ot_diis_step 108 11.5 0.011 0.011 2.386 2.386 calculate_first_density_matrix 1 7.0 0.000 0.000 2.359 2.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.239 2.240 cp_fm_redistribute_end 50 14.0 1.135 2.215 1.138 2.217 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.083 2.174 cp_fm_diag_elpa_base 50 14.0 1.048 2.128 1.076 2.162 calculate_dm_sparse 119 9.5 0.000 0.001 2.025 2.098 pw_transfer 1439 11.6 0.065 0.069 1.908 1.960 grid_integrate_task_list 119 12.3 1.839 1.931 1.839 1.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.915 1.917 potential_pw2rs 119 12.3 0.006 0.006 1.860 1.890 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.815 1.870 jit_kernel_multiply 9 16.1 0.801 1.762 0.801 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.690 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.517 1.544 make_preconditioner 11 8.9 0.000 0.000 1.517 1.544 fft_wrap_pw1pw2_140 487 13.2 0.201 0.211 1.453 1.506 acc_transpose_blocks 27432 15.5 0.108 0.112 1.199 1.501 make_images_data 4572 15.5 0.044 0.051 1.112 1.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.420 1.476 wfi_extrapolate 11 7.9 0.001 0.001 1.432 1.432 fft3d_ps 1201 14.6 0.523 0.578 1.367 1.416 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.961 1.390 grid_collocate_task_list 119 9.7 1.228 1.367 1.228 1.367 mp_allgather_i34 2286 14.5 0.527 1.278 0.527 1.278 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.247 mp_alltoall_d11v 2130 13.8 1.126 1.228 1.126 1.228 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.166 1.216 mp_sum_d 4135 12.0 0.565 0.994 0.565 0.994 mp_waitany 5720 13.7 0.503 0.968 0.503 0.968 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.145 0.528 0.962 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.899 acc_transpose_blocks_kernels 27432 16.5 0.181 0.272 0.669 0.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.164000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.613312E+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 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1095582. 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 0.069 32.065 32.067 qs_mol_dyn_low 1 2.0 0.003 0.003 31.757 31.766 qs_forces 11 3.9 0.002 0.003 31.420 31.421 qs_energies 11 4.9 0.002 0.009 29.820 29.823 scf_env_do_scf 11 5.9 0.001 0.001 25.231 25.231 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 22.753 22.753 dbcsr_multiply_generic 2286 12.5 0.092 0.094 16.126 16.206 velocity_verlet 10 3.0 0.001 0.001 15.339 15.349 qs_scf_new_mos 108 7.5 0.001 0.001 14.656 14.683 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.656 14.683 ot_scf_mini 108 9.5 0.002 0.003 13.937 13.955 multiply_cannon 2286 13.5 0.195 0.202 12.957 13.750 multiply_cannon_loop 2286 14.5 0.631 0.662 12.211 13.015 ot_mini 108 10.5 0.001 0.001 8.587 8.603 qs_ot_get_derivative 108 11.5 0.001 0.001 7.097 7.117 multiply_cannon_multrec 18288 15.5 1.943 2.947 6.759 7.049 rebuild_ks_matrix 119 8.3 0.000 0.000 6.319 6.333 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.319 6.332 dbcsr_mm_accdrv_process 38222 16.0 3.966 5.338 4.730 5.636 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.573 5.585 sum_up_and_integrate 119 10.3 0.031 0.032 3.916 3.920 integrate_v_rspace 119 11.3 0.002 0.003 3.885 3.893 mp_waitall_1 158411 16.6 2.526 3.728 2.526 3.728 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.417 3.426 calculate_rho_elec 119 8.7 0.031 0.032 3.417 3.425 init_scf_run 11 5.9 0.000 0.001 3.411 3.412 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.411 3.411 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.659 3.284 qs_ot_get_p 119 10.4 0.001 0.001 3.205 3.232 rs_pw_transfer 974 11.9 0.009 0.010 2.231 2.521 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.439 2.477 init_scf_loop 11 6.9 0.001 0.004 2.461 2.462 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.035 2.349 apply_single 119 13.6 0.000 0.000 2.034 2.349 density_rs2pw 119 9.7 0.004 0.004 1.937 2.234 calculate_first_density_matrix 1 7.0 0.000 0.001 2.205 2.207 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.170 2.175 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.909 1.910 grid_integrate_task_list 119 12.3 1.797 1.897 1.797 1.897 pw_transfer 1439 11.6 0.065 0.069 1.878 1.886 jit_kernel_multiply 10 16.2 0.713 1.860 0.713 1.860 make_m2s 4572 13.5 0.044 0.045 1.711 1.837 calculate_dm_sparse 119 9.5 0.000 0.001 1.801 1.811 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.785 1.795 make_images 4572 14.5 0.189 0.201 1.626 1.752 potential_pw2rs 119 12.3 0.007 0.008 1.682 1.689 prepare_preconditioner 11 7.9 0.000 0.000 1.684 1.686 make_preconditioner 11 8.9 0.000 0.003 1.684 1.686 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.669 1.677 cp_fm_diag_elpa_base 50 14.0 1.646 1.656 1.667 1.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.656 1.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.541 1.623 multiply_cannon_sync_h2d 18288 15.5 1.388 1.595 1.388 1.595 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.521 1.533 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.525 1.532 ot_diis_step 108 11.5 0.011 0.011 1.469 1.469 fft_wrap_pw1pw2_140 487 13.2 0.253 0.259 1.442 1.452 mp_sum_l 7207 12.9 1.109 1.418 1.109 1.418 grid_collocate_task_list 119 9.7 1.213 1.368 1.213 1.368 fft3d_ps 1201 14.6 0.530 0.548 1.272 1.284 acc_transpose_blocks 18288 15.5 0.074 0.076 1.246 1.274 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.173 1.178 wfi_extrapolate 11 7.9 0.001 0.001 1.145 1.145 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.956 0.958 make_images_data 4572 15.5 0.044 0.048 0.774 0.922 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.869 0.888 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.665 0.841 mp_waitany 9880 13.7 0.533 0.835 0.533 0.835 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.809 0.828 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.121 0.531 0.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.818 0.820 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.739 mp_alltoall_d11v 2130 13.8 0.643 0.733 0.643 0.733 cp_fm_cholesky_invert 11 10.9 0.693 0.696 0.693 0.696 mp_alltoall_z22v 1201 16.6 0.601 0.666 0.601 0.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.067000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.727273, yerr=1.762793 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 550.498304E+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 63496. MP_Allreduce 10074 349. 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 0.062 35.384 35.397 qs_mol_dyn_low 1 2.0 0.005 0.018 35.130 35.137 qs_forces 11 3.9 0.005 0.027 35.030 35.049 qs_energies 11 4.9 0.002 0.009 33.309 33.318 scf_env_do_scf 11 5.9 0.001 0.004 27.681 27.683 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.300 24.300 dbcsr_multiply_generic 2286 12.5 0.096 0.101 18.434 18.540 velocity_verlet 10 3.0 0.001 0.002 17.565 17.567 qs_scf_new_mos 108 7.5 0.001 0.001 15.945 15.993 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.944 15.992 multiply_cannon 2286 13.5 0.224 0.257 14.890 15.340 ot_scf_mini 108 9.5 0.002 0.004 14.982 15.032 multiply_cannon_loop 2286 14.5 0.935 0.968 13.983 14.415 ot_mini 108 10.5 0.001 0.002 9.024 9.091 multiply_cannon_multrec 27432 15.5 2.337 3.025 8.589 8.952 dbcsr_mm_accdrv_process 47916 15.9 5.274 7.156 6.157 7.395 qs_ot_get_derivative 108 11.5 0.002 0.012 7.220 7.275 rebuild_ks_matrix 119 8.3 0.000 0.000 6.530 6.592 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.020 6.529 6.591 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.802 5.851 init_scf_run 11 5.9 0.000 0.001 4.251 4.251 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.250 4.251 sum_up_and_integrate 119 10.3 0.036 0.041 3.778 3.788 integrate_v_rspace 119 11.3 0.003 0.003 3.742 3.750 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.481 3.507 calculate_rho_elec 119 8.7 0.040 0.046 3.481 3.506 qs_ot_get_p 119 10.4 0.001 0.001 3.354 3.434 init_scf_loop 11 6.9 0.001 0.010 3.361 3.361 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.635 3.097 calculate_first_density_matrix 1 7.0 0.000 0.003 2.894 2.895 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.100 2.554 apply_single 119 13.6 0.000 0.000 2.100 2.553 prepare_preconditioner 11 7.9 0.000 0.001 2.510 2.519 make_preconditioner 11 8.9 0.000 0.001 2.510 2.519 mp_waitall_1 137007 16.6 1.920 2.503 1.920 2.503 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.120 2.445 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.419 2.420 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.254 2.267 make_m2s 4572 13.5 0.054 0.056 2.091 2.192 rs_pw_transfer 974 11.9 0.009 0.010 1.959 2.155 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.118 2.127 calculate_dm_sparse 119 9.5 0.000 0.001 2.047 2.103 acc_transpose_blocks 27432 15.5 0.112 0.115 1.786 2.100 density_rs2pw 119 9.7 0.004 0.005 1.905 2.090 make_images 4572 14.5 0.270 0.332 1.984 2.084 pw_transfer 1439 11.6 0.065 0.071 1.998 2.037 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.906 1.950 grid_integrate_task_list 119 12.3 1.823 1.908 1.823 1.908 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.871 1.898 cp_dbcsr_syevd 50 12.0 0.003 0.006 1.806 1.806 ot_diis_step 108 11.5 0.012 0.015 1.764 1.764 fft_wrap_pw1pw2_140 487 13.2 0.289 0.307 1.573 1.619 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.562 1.570 cp_fm_diag_elpa_base 50 14.0 1.526 1.542 1.558 1.566 mp_sum_l 7207 12.9 1.041 1.554 1.041 1.554 potential_pw2rs 119 12.3 0.008 0.009 1.530 1.534 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.835 1.531 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 1.161 1.468 fft3d_ps 1201 14.6 0.558 0.607 1.343 1.370 grid_collocate_task_list 119 9.7 1.220 1.351 1.220 1.351 jit_kernel_multiply 7 15.9 0.823 1.308 0.823 1.308 wfi_extrapolate 11 7.9 0.001 0.005 1.302 1.302 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.201 1.213 jit_kernel_transpose 5 15.6 0.894 1.194 0.894 1.194 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.103 1.152 cp_fm_upper_to_full 72 14.2 0.807 1.148 0.807 1.148 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.116 1.116 multiply_cannon_sync_h2d 27432 15.5 1.001 1.066 1.001 1.066 dbcsr_complete_redistribute 329 12.2 0.117 0.143 0.752 1.023 make_images_data 4572 15.5 0.045 0.049 0.813 0.915 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.873 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.709 0.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.822 0.827 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.556 0.819 multiply_cannon_metrocomm1 27432 15.5 0.033 0.034 0.318 0.819 mp_alltoall_d11v 2130 13.8 0.700 0.790 0.700 0.790 cp_fm_cholesky_invert 11 10.9 0.731 0.734 0.731 0.734 mp_alltoall_i22 627 13.8 0.431 0.721 0.431 0.721 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.397000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=523.818182, yerr=2.208174 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 607.830016E+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 63495. MP_Allreduce 10074 348. 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 0.037 0.057 36.111 36.113 qs_mol_dyn_low 1 2.0 0.003 0.003 35.844 35.851 qs_forces 11 3.9 0.002 0.002 35.780 35.781 qs_energies 11 4.9 0.001 0.001 33.870 33.875 scf_env_do_scf 11 5.9 0.000 0.001 28.290 28.290 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.527 24.528 velocity_verlet 10 3.0 0.001 0.001 18.910 18.913 dbcsr_multiply_generic 2286 12.5 0.091 0.095 14.341 14.728 qs_scf_new_mos 108 7.5 0.001 0.001 14.111 14.143 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.110 14.142 ot_scf_mini 108 9.5 0.002 0.002 13.284 13.317 multiply_cannon 2286 13.5 0.234 0.240 10.589 12.035 multiply_cannon_loop 2286 14.5 0.329 0.341 9.520 10.505 rebuild_ks_matrix 119 8.3 0.000 0.000 7.513 7.614 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.513 7.614 ot_mini 108 10.5 0.001 0.001 7.164 7.211 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.697 6.783 multiply_cannon_multrec 9144 15.5 1.676 1.912 5.855 6.093 qs_ot_get_derivative 108 11.5 0.001 0.001 5.625 5.663 sum_up_and_integrate 119 10.3 0.038 0.041 4.951 4.969 integrate_v_rspace 119 11.3 0.003 0.003 4.913 4.932 mp_waitall_1 115863 16.7 3.233 4.923 3.233 4.923 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.854 4.861 calculate_rho_elec 119 8.7 0.060 0.061 4.853 4.861 dbcsr_mm_accdrv_process 12550 15.8 3.133 4.094 4.077 4.151 qs_ot_get_p 119 10.4 0.001 0.001 4.034 4.090 init_scf_run 11 5.9 0.000 0.002 4.063 4.063 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.062 4.062 init_scf_loop 11 6.9 0.000 0.000 3.735 3.737 rs_pw_transfer 974 11.9 0.008 0.008 3.521 3.694 density_rs2pw 119 9.7 0.004 0.004 3.036 3.189 prepare_preconditioner 11 7.9 0.000 0.000 2.792 2.797 make_preconditioner 11 8.9 0.000 0.000 2.792 2.797 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.597 2.660 pw_transfer 1439 11.6 0.066 0.069 2.609 2.651 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.644 2.647 make_m2s 4572 13.5 0.034 0.036 2.355 2.569 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.515 2.559 potential_pw2rs 119 12.3 0.010 0.011 2.499 2.519 make_images 4572 14.5 0.269 0.304 2.264 2.479 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.698 2.411 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.398 2.399 calculate_first_density_matrix 1 7.0 0.000 0.000 2.282 2.283 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.011 2.020 cp_fm_diag_elpa_base 50 14.0 1.981 1.999 2.009 2.018 calculate_dm_sparse 119 9.5 0.000 0.000 1.957 2.010 grid_integrate_task_list 119 12.3 1.903 2.010 1.903 2.010 fft_wrap_pw1pw2_140 487 13.2 0.365 0.372 1.931 1.977 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.951 1.965 fft3d_ps 1201 14.6 0.566 0.579 1.849 1.895 jit_kernel_multiply 10 15.9 0.905 1.790 0.905 1.790 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.719 1.722 wfi_extrapolate 11 7.9 0.001 0.001 1.681 1.681 cp_fm_cholesky_invert 11 10.9 1.634 1.643 1.634 1.643 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.536 1.557 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.751 1.550 make_images_data 4572 15.5 0.039 0.042 0.958 1.529 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.915 1.485 ot_diis_step 108 11.5 0.013 0.013 1.464 1.464 grid_collocate_task_list 119 9.7 1.314 1.452 1.314 1.452 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.406 1.450 mp_sum_l 7207 12.9 0.975 1.442 0.975 1.442 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.268 1.378 apply_single 119 13.6 0.000 0.000 1.268 1.378 mp_alltoall_d11v 2130 13.8 1.214 1.367 1.214 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.235 1.250 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.242 1.244 mp_alltoall_z22v 1201 16.6 1.156 1.216 1.156 1.216 mp_sendrecv_dv 5355 12.7 1.032 1.063 1.032 1.063 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.938 0.952 make_images_sizes 4572 15.5 0.005 0.005 0.732 0.950 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.912 0.950 mp_alltoall_i44 4572 16.5 0.726 0.946 0.726 0.946 acc_transpose_blocks 9144 15.5 0.037 0.038 0.895 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.916 dbcsr_complete_redistribute 329 12.2 0.159 0.171 0.815 0.854 mp_allgather_i34 2286 14.5 0.397 0.850 0.397 0.850 rs_pw_transfer_RS2PW_50 119 11.7 0.201 0.208 0.731 0.841 rs_pw_transfer_PW2RS_50 119 14.3 0.217 0.225 0.694 0.807 multiply_cannon_sync_h2d 9144 15.5 0.712 0.793 0.712 0.793 mp_waitany 5200 13.7 0.615 0.766 0.615 0.766 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.696 0.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=36.113000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=573.090909, yerr=6.775015 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 757.989376E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_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 0.016 0.038 50.352 50.353 qs_mol_dyn_low 1 2.0 0.003 0.003 50.077 50.084 qs_forces 11 3.9 0.002 0.002 49.951 49.953 qs_energies 11 4.9 0.001 0.001 47.748 47.758 scf_env_do_scf 11 5.9 0.001 0.001 41.100 41.100 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 31.504 31.506 velocity_verlet 10 3.0 0.001 0.001 28.518 28.532 dbcsr_multiply_generic 2286 12.5 0.100 0.102 20.084 20.661 qs_scf_new_mos 108 7.5 0.001 0.001 19.441 19.549 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.440 19.548 ot_scf_mini 108 9.5 0.002 0.002 18.183 18.299 multiply_cannon 2286 13.5 0.312 0.322 15.024 16.947 multiply_cannon_loop 2286 14.5 0.346 0.348 13.621 15.449 ot_mini 108 10.5 0.001 0.001 10.619 10.804 init_scf_loop 11 6.9 0.000 0.000 9.559 9.562 rebuild_ks_matrix 119 8.3 0.000 0.000 8.759 9.013 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.759 9.012 multiply_cannon_multrec 9144 15.5 3.494 5.098 8.737 8.872 prepare_preconditioner 11 7.9 0.000 0.000 8.406 8.425 make_preconditioner 11 8.9 0.000 0.000 8.406 8.425 qs_ot_get_derivative 108 11.5 0.001 0.001 8.249 8.354 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.968 8.284 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.901 8.125 dbcsr_mm_accdrv_process 12550 15.8 3.974 5.763 5.116 6.587 mp_waitall_1 94719 16.7 4.241 5.815 4.241 5.815 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.454 5.500 calculate_rho_elec 119 8.7 0.118 0.121 5.453 5.499 sum_up_and_integrate 119 10.3 0.064 0.066 5.177 5.184 integrate_v_rspace 119 11.3 0.003 0.003 5.112 5.121 qs_ot_get_p 119 10.4 0.001 0.001 4.504 4.711 cp_fm_upper_to_full 72 14.2 3.169 4.537 3.169 4.537 init_scf_run 11 5.9 0.000 0.001 4.473 4.473 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.473 4.473 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 2.359 4.114 pw_transfer 1439 11.6 0.069 0.069 3.492 3.513 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.994 3.448 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.393 3.415 mp_sum_l 7207 12.9 1.950 3.318 1.950 3.318 density_rs2pw 119 9.7 0.004 0.004 3.174 3.208 dbcsr_complete_redistribute 329 12.2 0.330 0.336 2.187 3.009 rs_pw_transfer 974 11.9 0.009 0.009 2.818 2.855 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.777 2.786 fft_wrap_pw1pw2_140 487 13.2 0.618 0.621 2.759 2.780 make_m2s 4572 13.5 0.038 0.039 2.566 2.716 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.784 2.611 cp_fm_cholesky_invert 11 10.9 2.602 2.610 2.602 2.610 make_images 4572 14.5 0.351 0.382 2.444 2.593 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.384 2.562 apply_single 119 13.6 0.000 0.000 2.384 2.562 calculate_first_density_matrix 1 7.0 0.000 0.000 2.510 2.559 calculate_dm_sparse 119 9.5 0.000 0.000 2.493 2.532 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.461 2.462 mp_alltoall_i22 627 13.8 1.581 2.448 1.581 2.448 fft3d_ps 1201 14.6 0.601 0.617 2.409 2.435 potential_pw2rs 119 12.3 0.014 0.014 2.341 2.357 ot_diis_step 108 11.5 0.014 0.014 2.253 2.254 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.433 2.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.209 2.212 grid_integrate_task_list 119 12.3 2.115 2.168 2.115 2.168 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.101 2.152 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.067 2.068 cp_fm_diag_elpa_base 50 14.0 1.918 1.960 2.065 2.065 jit_kernel_multiply 10 15.5 1.114 1.942 1.114 1.942 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 1.839 1.842 wfi_extrapolate 11 7.9 0.001 0.001 1.806 1.806 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.745 1.752 mp_alltoall_z22v 1201 16.6 1.669 1.714 1.669 1.714 mp_alltoall_d11v 2130 13.8 1.592 1.639 1.592 1.639 grid_collocate_task_list 119 9.7 1.530 1.558 1.530 1.558 make_images_data 4572 15.5 0.043 0.046 1.287 1.494 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.284 1.448 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.402 1.442 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.357 1.425 qs_env_update_s_mstruct 11 6.9 0.002 0.003 1.173 1.187 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.071 1.099 multiply_cannon_sync_h2d 9144 15.5 1.048 1.056 1.048 1.056 cp_fm_cholesky_decompose 22 10.9 1.029 1.035 1.029 1.035 yz_to_x 606 15.1 0.057 0.059 0.999 1.034 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.987 1.029 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=50.353000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.090909, yerr=14.304828 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 502.652928E+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 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1244853. 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 0.013 0.028 94.632 94.633 qs_mol_dyn_low 1 2.0 0.003 0.003 94.356 94.365 qs_forces 11 3.9 0.004 0.004 94.281 94.282 qs_energies 11 4.9 0.001 0.001 90.996 91.008 scf_env_do_scf 11 5.9 0.000 0.001 81.355 81.357 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 73.391 73.393 dbcsr_multiply_generic 2055 12.4 0.107 0.110 53.513 53.707 qs_scf_new_mos 99 7.5 0.000 0.001 52.485 52.601 qs_scf_loop_do_ot 99 8.5 0.001 0.001 52.484 52.600 ot_scf_mini 99 9.5 0.002 0.002 49.993 50.118 velocity_verlet 10 3.0 0.001 0.001 48.688 48.688 multiply_cannon 2055 13.4 0.187 0.192 42.941 43.743 multiply_cannon_loop 2055 14.4 1.504 1.537 41.417 42.245 ot_mini 99 10.5 0.001 0.001 28.203 28.341 qs_ot_get_derivative 99 11.5 0.001 0.001 21.370 21.502 multiply_cannon_multrec 49320 15.4 12.249 13.096 17.112 17.899 rebuild_ks_matrix 110 8.3 0.000 0.000 16.855 16.994 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 16.854 16.993 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.675 14.803 mp_waitall_1 220248 16.4 12.779 14.121 12.779 14.121 qs_ot_get_p 110 10.4 0.001 0.001 12.924 13.061 multiply_cannon_sync_h2d 49320 15.4 10.031 10.719 10.031 10.719 qs_ot_p2m_diag 48 11.0 0.012 0.019 9.074 9.129 sum_up_and_integrate 110 10.3 0.036 0.043 8.873 8.890 integrate_v_rspace 110 11.3 0.003 0.004 8.837 8.861 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.610 8.725 calculate_rho_elec 110 8.6 0.021 0.026 8.610 8.724 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.622 8.153 init_scf_loop 11 6.9 0.000 0.000 7.928 7.930 multiply_cannon_metrocomm3 49320 15.4 0.079 0.082 6.546 7.820 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.805 7.806 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.268 7.797 apply_single 110 13.6 0.000 0.001 7.268 7.797 init_scf_run 11 5.9 0.000 0.002 7.446 7.446 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.446 7.446 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.892 6.918 cp_fm_diag_elpa_base 48 14.0 6.867 6.896 6.889 6.915 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.650 6.705 rs_pw_transfer 902 11.9 0.011 0.013 5.919 6.632 ot_diis_step 99 11.5 0.006 0.006 6.631 6.631 density_rs2pw 110 9.6 0.004 0.005 5.349 6.038 mp_sum_l 6514 12.8 4.857 5.932 4.857 5.932 make_m2s 4110 13.4 0.061 0.064 5.253 5.507 prepare_preconditioner 11 7.9 0.000 0.000 5.444 5.472 make_preconditioner 11 8.9 0.000 0.000 5.444 5.472 make_images 4110 14.4 0.178 0.191 5.156 5.411 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.186 5.261 pw_transfer 1331 11.6 0.056 0.068 4.917 5.101 dbcsr_mm_accdrv_process 87628 16.1 1.926 2.032 4.740 5.058 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.828 5.015 wfi_extrapolate 11 7.9 0.001 0.001 4.673 4.673 potential_pw2rs 110 12.3 0.006 0.007 4.260 4.367 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 3.015 4.202 fft3d_ps 1111 14.6 0.796 0.889 4.027 4.188 fft_wrap_pw1pw2_140 451 13.1 0.447 0.497 3.859 4.049 calculate_dm_sparse 110 9.5 0.001 0.001 3.899 3.992 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.685 3.691 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.538 3.593 grid_integrate_task_list 110 12.3 3.228 3.440 3.228 3.440 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.266 3.306 mp_alltoall_d11v 2046 13.8 2.573 3.075 2.573 3.075 cp_fm_cholesky_invert 11 10.9 3.050 3.055 3.050 3.055 mp_alltoall_z22v 1111 16.6 2.758 3.026 2.758 3.026 mp_waitany 14300 13.8 2.068 2.821 2.068 2.821 make_images_data 4110 15.4 0.042 0.045 2.157 2.741 calculate_first_density_matrix 1 7.0 0.000 0.000 2.671 2.674 hybrid_alltoall_any 4261 16.3 0.081 0.472 1.785 2.555 jit_kernel_multiply 13 15.9 2.540 2.554 2.540 2.554 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.310 2.323 grid_collocate_task_list 110 9.6 2.081 2.315 2.081 2.315 acc_transpose_blocks 49320 15.4 0.211 0.219 2.109 2.170 make_images_sizes 4110 15.4 0.004 0.005 1.551 2.004 mp_alltoall_i44 4110 16.4 1.547 1.999 1.547 1.999 mp_sum_d 3889 11.9 1.412 1.911 1.412 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=94.633000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=2.219374 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.948608E+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 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 1420965. 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 0.017 0.045 84.216 84.230 qs_mol_dyn_low 1 2.0 0.003 0.005 83.944 83.953 qs_forces 11 3.9 0.003 0.003 83.870 83.872 qs_energies 11 4.9 0.005 0.036 80.221 80.251 scf_env_do_scf 11 5.9 0.000 0.001 70.818 70.822 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 59.777 59.778 velocity_verlet 10 3.0 0.001 0.001 44.292 44.294 dbcsr_multiply_generic 2055 12.4 0.117 0.122 42.187 43.364 qs_scf_new_mos 99 7.5 0.001 0.001 40.356 40.516 qs_scf_loop_do_ot 99 8.5 0.001 0.001 40.355 40.516 ot_scf_mini 99 9.5 0.003 0.003 38.436 38.590 multiply_cannon 2055 13.4 0.222 0.244 32.722 36.637 multiply_cannon_loop 2055 14.4 0.923 0.947 30.954 34.551 ot_mini 99 10.5 0.001 0.001 22.043 22.210 mp_waitall_1 176588 16.5 10.389 17.238 10.389 17.238 qs_ot_get_derivative 99 11.5 0.001 0.001 15.792 15.953 multiply_cannon_multrec 24660 15.4 7.434 9.545 13.711 15.929 rebuild_ks_matrix 110 8.3 0.000 0.000 15.648 15.790 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 15.647 15.789 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.760 13.873 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 6.395 13.366 init_scf_loop 11 6.9 0.001 0.005 10.996 10.999 qs_ot_get_p 110 10.4 0.001 0.001 9.207 9.381 prepare_preconditioner 11 7.9 0.000 0.000 8.629 8.653 make_preconditioner 11 8.9 0.000 0.000 8.629 8.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.122 8.343 sum_up_and_integrate 110 10.3 0.052 0.059 8.286 8.300 integrate_v_rspace 110 11.3 0.003 0.003 8.234 8.248 multiply_cannon_sync_h2d 24660 15.4 6.929 8.079 6.929 8.079 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.941 7.960 calculate_rho_elec 110 8.6 0.040 0.048 7.940 7.960 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.670 7.306 apply_single 110 13.6 0.000 0.001 6.670 7.306 init_scf_run 11 5.9 0.000 0.002 6.843 6.843 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.842 6.843 qs_ot_p2m_diag 48 11.0 0.028 0.044 6.629 6.675 dbcsr_mm_accdrv_process 52282 16.1 4.765 5.703 6.123 6.426 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.437 6.211 rs_pw_transfer 902 11.9 0.012 0.013 5.353 6.142 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.101 6.102 ot_diis_step 99 11.5 0.010 0.010 6.006 6.006 make_m2s 4110 13.4 0.056 0.059 5.241 5.740 make_images 4110 14.4 0.397 0.442 5.131 5.625 density_rs2pw 110 9.6 0.004 0.005 4.994 5.607 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.278 5.290 cp_fm_diag_elpa_base 48 14.0 5.213 5.239 5.275 5.287 pw_transfer 1331 11.6 0.067 0.079 4.629 4.877 mp_sum_l 6514 12.8 3.597 4.774 3.597 4.774 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.521 4.772 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.557 4.640 cp_fm_cholesky_invert 11 10.9 4.623 4.639 4.623 4.639 wfi_extrapolate 11 7.9 0.001 0.001 4.187 4.187 potential_pw2rs 110 12.3 0.008 0.008 3.972 4.055 fft3d_ps 1111 14.6 1.111 1.337 3.504 3.775 fft_wrap_pw1pw2_140 451 13.1 0.512 0.531 3.368 3.619 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.461 3.467 grid_integrate_task_list 110 12.3 3.128 3.353 3.128 3.353 calculate_dm_sparse 110 9.5 0.001 0.001 3.239 3.277 make_images_data 4110 15.4 0.046 0.050 2.688 3.244 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.990 3.041 hybrid_alltoall_any 4261 16.3 0.103 0.447 2.300 3.023 mp_alltoall_d11v 2046 13.8 2.226 2.691 2.226 2.691 calculate_first_density_matrix 1 7.0 0.003 0.025 2.546 2.549 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.427 2.543 grid_collocate_task_list 110 9.6 2.054 2.491 2.054 2.491 mp_alltoall_z22v 1111 16.6 2.061 2.250 2.061 2.250 cp_fm_cholesky_decompose 22 10.9 2.226 2.234 2.226 2.234 mp_allgather_i34 2055 14.4 1.096 2.229 1.096 2.229 mp_waitany 10164 13.8 1.604 2.157 1.604 2.157 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.048 2.072 multiply_cannon_metrocomm4 22605 15.4 0.075 0.079 0.795 2.056 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.947 1.949 mp_sum_d 3889 11.9 1.450 1.905 1.450 1.905 mp_irecv_dv 57340 16.2 0.671 1.837 0.671 1.837 dbcsr_complete_redistribute 325 12.2 0.234 0.293 1.471 1.754 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.222 1.066 1.692 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=84.230000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.818182, yerr=5.859230 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 660.279296E+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 66428. MP_Allreduce 9694 567. 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 0.037 0.107 59.891 59.892 qs_mol_dyn_low 1 2.0 0.003 0.004 59.459 59.468 qs_forces 11 3.9 0.005 0.007 59.157 59.158 qs_energies 11 4.9 0.004 0.024 55.905 55.910 scf_env_do_scf 11 5.9 0.001 0.003 47.932 47.932 scf_env_do_scf_inner_loop 99 6.5 0.004 0.006 39.403 39.405 velocity_verlet 10 3.0 0.001 0.004 32.520 32.528 dbcsr_multiply_generic 2055 12.4 0.104 0.108 28.101 28.388 qs_scf_new_mos 99 7.5 0.001 0.001 24.410 24.513 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.409 24.513 ot_scf_mini 99 9.5 0.002 0.003 23.198 23.314 multiply_cannon 2055 13.4 0.212 0.223 21.922 23.153 multiply_cannon_loop 2055 14.4 0.614 0.627 20.775 21.880 ot_mini 99 10.5 0.001 0.001 13.269 13.386 rebuild_ks_matrix 110 8.3 0.000 0.000 12.399 12.556 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.398 12.556 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.886 11.029 multiply_cannon_multrec 16440 15.4 3.918 4.966 9.650 10.583 mp_waitall_1 139946 16.5 6.837 9.946 6.837 9.946 qs_ot_get_derivative 99 11.5 0.001 0.001 8.875 8.989 init_scf_loop 11 6.9 0.002 0.015 8.496 8.506 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.220 7.202 prepare_preconditioner 11 7.9 0.000 0.000 6.644 6.667 make_preconditioner 11 8.9 0.000 0.001 6.644 6.667 sum_up_and_integrate 110 10.3 0.060 0.061 6.550 6.573 integrate_v_rspace 110 11.3 0.003 0.003 6.489 6.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.955 6.316 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.929 5.942 calculate_rho_elec 110 8.6 0.059 0.059 5.929 5.941 dbcsr_mm_accdrv_process 34862 16.1 4.482 5.175 5.585 5.775 init_scf_run 11 5.9 0.000 0.002 5.492 5.492 scf_env_initial_rho_setup 11 6.9 0.002 0.004 5.491 5.492 qs_ot_get_p 110 10.4 0.001 0.001 5.217 5.351 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.928 5.343 apply_single 110 13.6 0.000 0.000 4.928 5.343 make_m2s 4110 13.4 0.050 0.052 4.135 4.503 make_images 4110 14.4 0.390 0.510 4.020 4.386 ot_diis_step 99 11.5 0.011 0.011 4.365 4.365 multiply_cannon_sync_h2d 16440 15.4 3.690 4.363 3.690 4.363 density_rs2pw 110 9.6 0.004 0.005 3.040 4.251 rs_pw_transfer 902 11.9 0.010 0.011 2.674 3.885 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.045 3.686 pw_transfer 1331 11.6 0.065 0.073 3.646 3.659 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.597 3.600 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.539 3.554 grid_integrate_task_list 110 12.3 3.162 3.404 3.162 3.404 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.266 3.266 fft_wrap_pw1pw2_140 451 13.1 0.636 0.642 3.075 3.096 wfi_extrapolate 11 7.9 0.001 0.001 2.940 2.940 make_images_data 4110 15.4 0.043 0.048 2.426 2.878 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.183 2.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.824 2.826 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.743 2.755 cp_fm_diag_elpa_base 48 14.0 2.676 2.713 2.742 2.753 cp_fm_cholesky_invert 11 10.9 2.570 2.575 2.570 2.575 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.483 2.544 calculate_dm_sparse 110 9.5 0.001 0.001 2.481 2.515 mp_waitany 17072 13.8 1.235 2.508 1.235 2.508 calculate_first_density_matrix 1 7.0 0.001 0.008 2.466 2.468 grid_collocate_task_list 110 9.6 2.083 2.466 2.083 2.466 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.889 2.396 fft3d_ps 1111 14.6 1.092 1.101 2.364 2.382 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.268 2.330 potential_pw2rs 110 12.3 0.011 0.011 2.287 2.302 mp_irecv_dv 48980 15.7 0.818 2.273 0.818 2.273 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.182 0.943 2.145 mp_alltoall_d11v 2046 13.8 1.753 2.140 1.753 2.140 mp_sum_l 6514 12.8 1.494 2.137 1.494 2.137 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.012 2.014 dbcsr_complete_redistribute 325 12.2 0.324 0.353 1.424 1.880 cp_fm_upper_to_full 70 14.2 1.387 1.779 1.387 1.779 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.707 1.723 cp_fm_cholesky_decompose 22 10.9 1.546 1.564 1.546 1.564 mp_allgather_i34 2055 14.4 0.471 1.554 0.471 1.554 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.360 1.497 jit_kernel_multiply 8 16.3 0.715 1.495 0.715 1.495 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.449 1.465 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.001 1.439 rs_gather_matrices 110 12.3 0.234 0.262 0.962 1.264 acc_transpose_blocks 16440 15.4 0.072 0.073 1.225 1.246 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.224 1.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.892000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.272727, yerr=8.832697 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 737.529856E+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 66426. MP_Allreduce 9694 608. 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 0.018 0.039 64.649 64.653 qs_mol_dyn_low 1 2.0 0.003 0.004 64.324 64.331 qs_forces 11 3.9 0.003 0.005 64.254 64.255 qs_energies 11 4.9 0.002 0.007 60.845 60.848 scf_env_do_scf 11 5.9 0.000 0.001 52.585 52.588 scf_env_do_scf_inner_loop 99 6.5 0.008 0.019 40.960 40.961 velocity_verlet 10 3.0 0.001 0.002 36.671 36.673 dbcsr_multiply_generic 2055 12.4 0.111 0.117 29.076 29.301 qs_scf_new_mos 99 7.5 0.001 0.001 25.985 26.115 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.985 26.114 ot_scf_mini 99 9.5 0.003 0.003 24.367 24.500 multiply_cannon 2055 13.4 0.243 0.269 22.291 23.260 multiply_cannon_loop 2055 14.4 0.884 0.906 20.920 21.482 ot_mini 99 10.5 0.001 0.001 13.842 14.000 multiply_cannon_multrec 24660 15.4 4.207 6.797 12.627 13.766 rebuild_ks_matrix 110 8.3 0.000 0.000 12.167 12.271 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.166 12.271 init_scf_loop 11 6.9 0.002 0.013 11.585 11.586 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.753 10.847 qs_ot_get_derivative 99 11.5 0.001 0.001 9.703 9.842 prepare_preconditioner 11 7.9 0.000 0.001 9.812 9.829 make_preconditioner 11 8.9 0.000 0.001 9.812 9.829 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.004 9.480 dbcsr_mm_accdrv_process 52304 16.0 6.921 7.954 8.272 9.166 sum_up_and_integrate 110 10.3 0.068 0.071 6.477 6.491 integrate_v_rspace 110 11.3 0.003 0.003 6.409 6.422 mp_waitall_1 121746 16.5 4.112 6.096 4.112 6.096 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.053 6.063 calculate_rho_elec 110 8.6 0.078 0.081 6.053 6.063 qs_ot_get_p 110 10.4 0.001 0.001 5.581 5.732 make_m2s 4110 13.4 0.060 0.063 5.218 5.559 init_scf_run 11 5.9 0.000 0.001 5.457 5.457 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.457 5.457 make_images 4110 14.4 0.574 0.690 5.077 5.415 cp_fm_upper_to_full 70 14.2 3.352 4.876 3.352 4.876 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.009 4.110 apply_single 110 13.6 0.000 0.000 4.009 4.110 ot_diis_step 99 11.5 0.011 0.011 4.104 4.104 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.838 3.852 dbcsr_complete_redistribute 325 12.2 0.418 0.456 2.653 3.799 pw_transfer 1331 11.6 0.066 0.075 3.724 3.754 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.617 3.652 density_rs2pw 110 9.6 0.004 0.006 2.995 3.510 grid_integrate_task_list 110 12.3 3.270 3.469 3.270 3.469 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.368 3.368 multiply_cannon_sync_h2d 24660 15.4 3.181 3.361 3.181 3.361 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.434 3.329 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.244 3.323 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.135 3.268 fft_wrap_pw1pw2_140 451 13.1 0.666 0.691 3.115 3.152 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.957 3.014 hybrid_alltoall_any 4261 16.3 0.120 0.462 2.288 2.957 make_images_data 4110 15.4 0.046 0.051 2.618 2.945 wfi_extrapolate 11 7.9 0.001 0.001 2.937 2.937 rs_pw_transfer 902 11.9 0.010 0.011 2.406 2.937 calculate_dm_sparse 110 9.5 0.001 0.001 2.900 2.934 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.800 2.919 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.835 2.844 cp_fm_diag_elpa_base 48 14.0 2.689 2.750 2.833 2.842 mp_alltoall_i22 605 13.7 1.661 2.840 1.661 2.840 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.749 2.750 cp_fm_cholesky_invert 11 10.9 2.620 2.628 2.620 2.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.407 2.441 calculate_first_density_matrix 1 7.0 0.000 0.002 2.425 2.427 grid_collocate_task_list 110 9.6 2.181 2.427 2.181 2.427 fft3d_ps 1111 14.6 1.084 1.116 2.382 2.399 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.316 2.317 potential_pw2rs 110 12.3 0.012 0.013 2.144 2.152 mp_alltoall_d11v 2046 13.8 1.702 2.025 1.702 2.025 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.682 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.602 1.702 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.663 1.675 cp_fm_cholesky_decompose 22 10.9 1.617 1.673 1.617 1.673 jit_kernel_multiply 9 15.8 1.022 1.648 1.022 1.648 mp_waitany 13376 13.8 1.100 1.602 1.100 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.516 1.532 acc_transpose_blocks 24660 15.4 0.106 0.109 1.506 1.521 mp_sum_l 6514 12.8 0.888 1.512 0.888 1.512 mp_allgather_i34 2055 14.4 0.473 1.459 0.473 1.459 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.853 1.445 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.183 0.865 1.399 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.259 1.367 mp_irecv_dv 62702 16.1 0.749 1.365 0.749 1.365 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.653000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.363636, yerr=11.841934 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 845.336576E+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 66424. MP_Allreduce 9694 649. 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 0.016 0.059 55.241 55.257 qs_mol_dyn_low 1 2.0 0.003 0.003 54.877 54.884 qs_forces 11 3.9 0.003 0.003 54.801 54.802 qs_energies 11 4.9 0.001 0.001 51.104 51.109 scf_env_do_scf 11 5.9 0.000 0.001 42.710 42.710 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.073 35.074 velocity_verlet 10 3.0 0.001 0.001 30.927 30.930 dbcsr_multiply_generic 2055 12.4 0.103 0.108 22.956 23.115 qs_scf_new_mos 99 7.5 0.001 0.001 20.104 20.169 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.104 20.169 multiply_cannon 2055 13.4 0.247 0.261 17.648 18.921 ot_scf_mini 99 9.5 0.002 0.002 18.877 18.921 multiply_cannon_loop 2055 14.4 0.319 0.331 16.310 16.663 rebuild_ks_matrix 110 8.3 0.000 0.000 11.894 11.937 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.894 11.937 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.549 10.587 ot_mini 99 10.5 0.001 0.001 10.418 10.460 multiply_cannon_multrec 8220 15.4 3.253 4.899 7.661 8.864 mp_waitall_1 103326 16.6 5.971 7.689 5.971 7.689 init_scf_loop 11 6.9 0.000 0.000 7.591 7.592 qs_ot_get_derivative 99 11.5 0.001 0.001 6.569 6.614 sum_up_and_integrate 110 10.3 0.079 0.080 6.495 6.508 integrate_v_rspace 110 11.3 0.003 0.003 6.416 6.430 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.236 6.250 calculate_rho_elec 110 8.6 0.114 0.115 6.235 6.249 prepare_preconditioner 11 7.9 0.000 0.000 5.933 5.938 make_preconditioner 11 8.9 0.000 0.000 5.933 5.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.526 5.598 init_scf_run 11 5.9 0.000 0.001 5.279 5.279 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.278 5.279 dbcsr_mm_accdrv_process 17442 15.9 2.862 3.897 4.278 5.246 qs_ot_get_p 110 10.4 0.001 0.001 4.578 4.635 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.123 4.624 make_m2s 4110 13.4 0.039 0.040 4.073 4.326 make_images 4110 14.4 0.635 0.690 3.944 4.194 pw_transfer 1331 11.6 0.066 0.073 3.920 3.939 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.811 3.834 ot_diis_step 99 11.5 0.012 0.012 3.829 3.829 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.732 3.794 apply_single 110 13.6 0.000 0.000 3.731 3.794 grid_integrate_task_list 110 12.3 3.356 3.543 3.356 3.543 density_rs2pw 110 9.6 0.004 0.005 2.993 3.360 fft_wrap_pw1pw2_140 451 13.1 0.831 0.842 3.298 3.327 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.261 3.265 multiply_cannon_sync_h2d 8220 15.4 2.909 3.032 2.909 3.032 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.965 2.966 cp_fm_cholesky_invert 11 10.9 2.862 2.866 2.862 2.866 make_images_data 4110 15.4 0.038 0.043 2.317 2.712 wfi_extrapolate 11 7.9 0.001 0.001 2.697 2.697 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.670 2.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.643 2.645 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.253 2.603 calculate_dm_sparse 110 9.5 0.001 0.001 2.536 2.574 rs_pw_transfer 902 11.9 0.010 0.010 2.143 2.544 grid_collocate_task_list 110 9.6 2.278 2.541 2.278 2.541 calculate_first_density_matrix 1 7.0 0.000 0.000 2.488 2.489 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.458 2.466 cp_fm_diag_elpa_base 48 14.0 2.404 2.425 2.457 2.464 fft3d_ps 1111 14.6 1.138 1.190 2.361 2.380 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.301 2.316 potential_pw2rs 110 12.3 0.015 0.015 2.048 2.052 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.778 1.990 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.938 1.961 mp_alltoall_d11v 2046 13.8 1.586 1.834 1.586 1.834 jit_kernel_multiply 9 15.8 1.111 1.770 1.111 1.770 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.742 1.764 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.678 cp_fm_cholesky_decompose 22 10.9 1.654 1.671 1.654 1.671 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.505 1.630 mp_allgather_i34 2055 14.4 0.525 1.591 0.525 1.591 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.508 1.518 dbcsr_complete_redistribute 325 12.2 0.557 0.583 1.395 1.486 mp_waitany 9240 13.8 1.052 1.471 1.052 1.471 qs_create_task_list 11 7.9 0.000 0.001 1.219 1.322 generate_qs_task_list 11 8.9 0.374 0.443 1.218 1.322 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.485 1.297 mp_irecv_dv 24056 15.7 0.461 1.256 0.461 1.256 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.794 1.194 rs_gather_matrices 110 12.3 0.327 0.371 0.923 1.142 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.109 1.140 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.257000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=799.545455, yerr=10.611860 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 1.337848E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_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 0.020 0.068 87.707 87.727 qs_mol_dyn_low 1 2.0 0.003 0.003 87.378 87.388 qs_forces 11 3.9 0.003 0.004 87.275 87.276 qs_energies 11 4.9 0.001 0.001 83.126 83.129 scf_env_do_scf 11 5.9 0.001 0.001 72.175 72.175 velocity_verlet 10 3.0 0.001 0.001 55.591 55.602 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 43.844 43.845 dbcsr_multiply_generic 2055 12.4 0.118 0.124 28.824 28.911 init_scf_loop 11 6.9 0.000 0.000 28.259 28.262 prepare_preconditioner 11 7.9 0.000 0.000 26.241 26.248 make_preconditioner 11 8.9 0.000 0.000 26.241 26.248 qs_scf_new_mos 99 7.5 0.001 0.001 25.764 25.796 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.763 25.795 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.403 25.718 ot_scf_mini 99 9.5 0.002 0.002 24.013 24.040 multiply_cannon 2055 13.4 0.344 0.379 21.824 22.517 multiply_cannon_loop 2055 14.4 0.339 0.342 20.050 20.492 cp_fm_upper_to_full 70 14.2 12.691 18.253 12.691 18.253 rebuild_ks_matrix 110 8.3 0.001 0.001 14.017 14.059 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.017 14.058 ot_mini 99 10.5 0.001 0.001 13.366 13.387 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.693 12.730 dbcsr_complete_redistribute 325 12.2 1.019 1.044 7.389 10.684 multiply_cannon_multrec 8220 15.4 4.360 4.549 9.754 10.110 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.372 9.663 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.822 9.089 qs_ot_get_derivative 99 11.5 0.001 0.001 8.804 8.831 mp_alltoall_i22 605 13.7 5.478 8.720 5.478 8.720 mp_waitall_1 84994 16.7 7.752 8.489 7.752 8.489 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.634 7.673 calculate_rho_elec 110 8.6 0.225 0.226 7.633 7.672 sum_up_and_integrate 110 10.3 0.150 0.151 7.255 7.269 integrate_v_rspace 110 11.3 0.004 0.004 7.105 7.119 init_scf_run 11 5.9 0.000 0.001 6.708 6.708 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.707 6.708 make_m2s 4110 13.4 0.043 0.043 5.303 5.805 dbcsr_mm_accdrv_process 11614 15.7 3.249 3.593 5.251 5.790 make_images 4110 14.4 0.874 0.919 5.116 5.617 qs_ot_get_p 110 10.4 0.001 0.001 5.467 5.493 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.845 5.338 cp_fm_cholesky_invert 11 10.9 5.213 5.217 5.213 5.217 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.672 5.127 apply_single 110 13.6 0.000 0.000 4.672 5.127 pw_transfer 1331 11.6 0.075 0.075 4.989 4.996 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.872 4.879 ot_diis_step 99 11.5 0.015 0.015 4.539 4.539 fft_wrap_pw1pw2_140 451 13.1 1.335 1.340 4.252 4.261 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.958 3.965 multiply_cannon_sync_h2d 8220 15.4 3.952 3.960 3.952 3.960 hybrid_alltoall_any 4261 16.3 0.257 0.556 2.981 3.753 grid_integrate_task_list 110 12.3 3.661 3.722 3.661 3.722 density_rs2pw 110 9.6 0.004 0.004 3.682 3.713 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.652 3.653 make_images_data 4110 15.4 0.042 0.045 2.977 3.600 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.523 3.524 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.422 3.426 calculate_first_density_matrix 1 7.0 0.000 0.000 3.304 3.306 wfi_extrapolate 11 7.9 0.001 0.001 3.283 3.283 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.803 3.252 calculate_dm_sparse 110 9.5 0.001 0.001 3.110 3.129 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.911 2.958 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.943 2.943 cp_fm_diag_elpa_base 48 14.0 2.405 2.600 2.941 2.941 fft3d_ps 1111 14.6 1.300 1.308 2.805 2.811 grid_collocate_task_list 110 9.6 2.633 2.656 2.633 2.656 potential_pw2rs 110 12.3 0.021 0.021 2.372 2.377 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.348 2.363 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.178 2.242 rs_pw_transfer 902 11.9 0.010 0.011 2.229 2.241 jit_kernel_multiply 10 15.3 1.803 2.212 1.803 2.212 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.176 mp_alltoall_d11v 2046 13.8 1.912 2.004 1.912 2.004 cp_fm_cholesky_decompose 22 10.9 1.958 1.986 1.958 1.986 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.781 1.956 qs_create_task_list 11 7.9 0.001 0.001 1.889 1.936 generate_qs_task_list 11 8.9 0.731 0.784 1.888 1.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.924 1.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.727000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1207.727273, yerr=55.409371 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/15/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 1.094965E+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 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 633.626624E+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 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 2399688. 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 0.019 0.030 203.874 203.876 qs_mol_dyn_low 1 2.0 0.003 0.003 203.512 203.526 qs_forces 11 3.9 0.005 0.005 203.379 203.381 qs_energies 11 4.9 0.001 0.001 197.823 197.837 scf_env_do_scf 11 5.9 0.001 0.001 181.344 181.347 scf_env_do_scf_inner_loop 117 6.6 0.020 0.113 160.531 160.533 dbcsr_multiply_generic 2507 12.6 0.179 0.182 122.461 123.105 velocity_verlet 10 3.0 0.001 0.001 122.217 122.219 qs_scf_new_mos 117 7.6 0.001 0.001 120.745 120.983 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.744 120.983 ot_scf_mini 117 9.6 0.003 0.003 114.183 114.405 multiply_cannon 2507 13.6 0.242 0.251 100.391 102.200 multiply_cannon_loop 2507 14.6 2.004 2.086 98.245 100.190 ot_mini 117 10.6 0.001 0.001 64.208 64.450 multiply_cannon_multrec 60168 15.6 33.755 35.508 41.823 43.655 qs_ot_get_derivative 117 11.6 0.001 0.001 39.466 39.700 rebuild_ks_matrix 128 8.3 0.001 0.001 33.611 33.923 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 33.611 33.923 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.187 30.458 mp_waitall_1 267128 16.5 26.777 30.210 26.777 30.210 multiply_cannon_sync_h2d 60168 15.6 28.321 30.115 28.321 30.115 qs_ot_get_p 128 10.4 0.001 0.001 28.468 28.764 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.473 25.039 apply_single 128 13.6 0.001 0.001 24.473 25.039 ot_diis_step 117 11.6 0.008 0.008 24.510 24.511 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.924 22.024 init_scf_loop 11 6.9 0.000 0.000 20.740 20.741 cp_dbcsr_syevd 83 12.4 0.004 0.005 19.591 19.591 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.124 18.253 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.661 16.694 cp_fm_diag_elpa_base 83 14.4 16.595 16.626 16.657 16.692 multiply_cannon_metrocomm3 60168 15.6 0.116 0.121 14.849 16.688 prepare_preconditioner 11 7.9 0.000 0.000 16.048 16.085 make_preconditioner 11 8.9 0.000 0.000 16.048 16.085 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.274 15.452 sum_up_and_integrate 128 10.3 0.089 0.108 14.147 14.160 integrate_v_rspace 128 11.3 0.004 0.004 14.057 14.075 make_m2s 5014 13.6 0.104 0.111 13.572 13.959 make_images 5014 14.6 0.399 0.417 13.391 13.787 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.519 13.599 calculate_rho_elec 128 8.7 0.046 0.065 13.519 13.598 init_scf_run 11 5.9 0.000 0.001 12.325 12.325 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.325 12.325 density_rs2pw 128 9.7 0.006 0.007 6.953 10.147 cp_fm_cholesky_invert 11 10.9 9.269 9.277 9.269 9.277 mp_sum_l 7870 13.0 7.211 9.161 7.211 9.161 wfi_extrapolate 11 7.9 0.001 0.001 9.049 9.049 rs_pw_transfer 1046 11.9 0.016 0.018 5.629 8.816 calculate_dm_sparse 128 9.5 0.001 0.001 8.446 8.545 dbcsr_mm_accdrv_process 124484 16.2 2.936 3.207 7.600 8.347 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.006 8.226 pw_transfer 1547 11.6 0.074 0.089 7.690 7.861 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.596 7.706 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.487 7.652 make_images_data 5014 15.6 0.064 0.071 6.678 7.635 grid_integrate_task_list 128 12.3 7.077 7.534 7.077 7.534 multiply_cannon_metrocomm1 60168 15.6 0.090 0.094 5.556 7.532 hybrid_alltoall_any 5200 16.5 0.292 2.262 5.854 7.209 fft_wrap_pw1pw2_140 523 13.2 1.273 1.320 6.596 6.765 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.647 6.656 mp_waitany 16020 13.9 2.687 5.964 2.687 5.964 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.836 5.905 grid_collocate_task_list 128 9.7 4.560 5.709 4.560 5.709 fft3d_ps 1291 14.7 2.141 2.788 5.255 5.591 mp_alltoall_d11v 2415 14.1 4.113 5.461 4.113 5.461 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.300 2.107 5.303 cp_fm_cholesky_decompose 22 10.9 4.586 4.601 4.586 4.601 potential_pw2rs 128 12.3 0.009 0.010 4.529 4.547 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.876000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=600.545455, yerr=7.808662 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 827.285504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 5608649. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.040 218.505 218.508 qs_mol_dyn_low 1 2.0 0.004 0.004 218.062 218.075 qs_forces 11 3.9 0.005 0.005 217.961 217.963 qs_energies 11 4.9 0.008 0.018 210.907 210.938 scf_env_do_scf 11 5.9 0.003 0.016 193.562 193.572 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 152.167 152.169 velocity_verlet 10 3.0 0.005 0.019 137.150 137.152 qs_scf_new_mos 117 7.6 0.001 0.001 107.863 108.530 qs_scf_loop_do_ot 117 8.6 0.001 0.001 107.863 108.529 dbcsr_multiply_generic 2507 12.6 0.190 0.197 103.474 106.444 ot_scf_mini 117 9.6 0.004 0.005 102.861 103.621 multiply_cannon 2507 13.6 0.478 0.530 79.480 89.467 multiply_cannon_loop 2507 14.6 1.249 1.293 75.695 82.358 ot_mini 117 10.6 0.001 0.001 55.155 55.889 mp_waitall_1 214728 16.6 29.012 47.026 29.012 47.026 init_scf_loop 11 6.9 0.004 0.008 41.282 41.285 multiply_cannon_metrocomm3 30084 15.6 0.092 0.101 18.232 38.731 multiply_cannon_multrec 30084 15.6 21.410 26.625 30.956 36.636 prepare_preconditioner 11 7.9 0.000 0.001 36.547 36.600 make_preconditioner 11 8.9 0.000 0.001 36.547 36.600 rebuild_ks_matrix 128 8.3 0.001 0.001 35.677 36.332 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.020 35.676 36.331 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.139 35.779 qs_ot_get_derivative 117 11.6 0.001 0.002 32.326 33.078 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.146 32.753 qs_ot_get_p 128 10.4 0.001 0.001 28.740 29.592 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.253 24.192 apply_single 128 13.6 0.001 0.001 22.253 24.192 qs_ot_p2m_diag 83 11.4 0.187 0.216 23.078 23.215 cp_fm_cholesky_invert 11 10.9 22.643 22.662 22.643 22.662 ot_diis_step 117 11.6 0.014 0.015 22.523 22.525 multiply_cannon_sync_h2d 30084 15.6 18.751 22.070 18.751 22.070 cp_dbcsr_syevd 83 12.4 0.006 0.008 21.580 21.582 cp_fm_diag_elpa 83 13.4 0.001 0.002 17.752 17.782 cp_fm_diag_elpa_base 83 14.4 17.462 17.568 17.743 17.767 make_m2s 5014 13.6 0.089 0.095 15.638 17.301 make_images 5014 14.6 1.151 1.341 15.428 17.091 sum_up_and_integrate 128 10.3 0.116 0.132 16.697 16.730 integrate_v_rspace 128 11.3 0.004 0.004 16.581 16.618 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.293 16.323 calculate_rho_elec 128 8.7 0.088 0.104 16.293 16.323 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 13.371 13.889 init_scf_run 11 5.9 0.000 0.001 12.375 12.377 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.375 12.376 density_rs2pw 128 9.7 0.007 0.008 9.590 12.042 rs_pw_transfer 1046 11.9 0.014 0.016 8.940 11.381 make_images_data 5014 15.6 0.063 0.071 9.093 11.144 multiply_cannon_metrocomm4 27577 15.6 0.099 0.114 3.794 11.143 mp_irecv_dv 69486 16.3 3.599 10.745 3.599 10.745 hybrid_alltoall_any 5200 16.5 0.344 1.505 7.653 10.344 pw_transfer 1547 11.6 0.087 0.098 9.700 9.832 dbcsr_mm_accdrv_process 62242 16.2 4.491 5.180 9.035 9.630 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.473 9.609 mp_sum_l 7870 13.0 6.593 9.171 6.593 9.171 wfi_extrapolate 11 7.9 0.001 0.001 9.119 9.119 cp_fm_cholesky_decompose 22 10.9 8.925 8.998 8.925 8.998 fft_wrap_pw1pw2_140 523 13.2 1.330 1.352 7.923 8.073 grid_integrate_task_list 128 12.3 7.122 7.839 7.122 7.839 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.948 7.750 potential_pw2rs 128 12.3 0.015 0.017 6.832 6.923 fft3d_ps 1291 14.7 2.816 2.995 6.789 6.899 calculate_dm_sparse 128 9.5 0.001 0.001 6.672 6.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.352 6.361 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.900 5.996 grid_collocate_task_list 128 9.7 4.700 5.823 4.700 5.823 mp_allgather_i34 2507 14.6 2.357 5.670 2.357 5.670 mp_alltoall_d11v 2415 14.1 4.626 5.634 4.626 5.634 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.382 5.558 mp_waitany 11748 13.9 2.888 5.402 2.888 5.402 mp_sum_d 4465 12.1 3.470 5.048 3.470 5.048 rs_pw_transfer_RS2PW_140 139 11.5 0.358 0.398 2.192 4.647 dbcsr_dot_sd 1318 12.0 0.756 0.834 3.122 4.417 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=218.508000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.090909, yerr=2.314168 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/17/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 2.928533E+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 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 969.334784E+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 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 10977 1007. 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 0.022 0.042 203.831 203.832 qs_mol_dyn_low 1 2.0 0.003 0.004 203.355 203.368 qs_forces 11 3.9 0.004 0.005 203.246 203.250 qs_energies 11 4.9 0.001 0.002 196.362 196.388 scf_env_do_scf 11 5.9 0.001 0.002 179.829 179.830 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 136.665 136.667 velocity_verlet 10 3.0 0.001 0.002 130.804 130.808 qs_scf_new_mos 117 7.6 0.001 0.001 94.371 94.746 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.370 94.745 ot_scf_mini 117 9.6 0.003 0.004 90.078 90.481 dbcsr_multiply_generic 2507 12.6 0.182 0.189 87.075 89.328 multiply_cannon 2507 13.6 0.505 0.531 63.990 72.521 multiply_cannon_loop 2507 14.6 0.855 0.886 60.278 65.868 ot_mini 117 10.6 0.001 0.001 46.971 47.417 init_scf_loop 11 6.9 0.000 0.000 43.055 43.060 mp_waitall_1 170520 16.6 29.029 40.306 29.029 40.306 prepare_preconditioner 11 7.9 0.000 0.000 38.677 38.738 make_preconditioner 11 8.9 0.000 0.001 38.677 38.738 make_full_inverse_cholesky 11 9.9 0.000 0.002 36.103 37.683 rebuild_ks_matrix 128 8.3 0.001 0.001 33.374 34.086 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.374 34.086 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.119 30.784 multiply_cannon_metrocomm3 20056 15.6 0.059 0.064 17.099 30.298 qs_ot_get_p 128 10.4 0.001 0.001 27.698 28.163 qs_ot_get_derivative 117 11.6 0.002 0.002 26.464 26.875 multiply_cannon_multrec 20056 15.6 13.252 16.898 21.867 25.522 qs_ot_p2m_diag 83 11.4 0.266 0.272 22.576 22.678 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.279 21.811 apply_single 128 13.6 0.001 0.001 20.279 21.811 cp_dbcsr_syevd 83 12.4 0.006 0.009 21.259 21.265 ot_diis_step 117 11.6 0.018 0.019 20.292 20.293 cp_fm_cholesky_invert 11 10.9 19.648 19.668 19.648 19.668 cp_fm_diag_elpa 83 13.4 0.001 0.002 17.079 17.100 cp_fm_diag_elpa_base 83 14.4 16.596 16.766 17.072 17.093 make_m2s 5014 13.6 0.082 0.087 15.921 16.968 sum_up_and_integrate 128 10.3 0.133 0.144 16.858 16.890 integrate_v_rspace 128 11.3 0.004 0.004 16.724 16.755 make_images 5014 14.6 1.172 1.261 15.688 16.734 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.408 16.436 calculate_rho_elec 128 8.7 0.130 0.145 16.407 16.436 multiply_cannon_sync_h2d 20056 15.6 14.038 15.711 14.038 15.711 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.083 11.378 init_scf_run 11 5.9 0.000 0.002 11.346 11.347 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.346 11.346 density_rs2pw 128 9.7 0.006 0.008 9.142 11.135 make_images_data 5014 15.6 0.059 0.068 9.581 10.795 rs_pw_transfer 1046 11.9 0.013 0.015 8.439 10.329 pw_transfer 1547 11.6 0.087 0.104 9.743 9.940 hybrid_alltoall_any 5200 16.5 0.435 2.012 8.279 9.921 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.517 9.721 multiply_cannon_metrocomm4 17549 15.6 0.063 0.077 3.511 9.711 cp_fm_cholesky_decompose 22 10.9 9.639 9.709 9.639 9.709 mp_irecv_dv 50230 16.2 3.389 9.457 3.389 9.457 fft_wrap_pw1pw2_140 523 13.2 1.409 1.446 8.097 8.314 dbcsr_mm_accdrv_process 41502 16.2 4.428 5.275 8.097 8.243 wfi_extrapolate 11 7.9 0.001 0.001 8.213 8.213 grid_integrate_task_list 128 12.3 7.259 7.980 7.259 7.980 mp_sum_l 7870 13.0 5.363 7.521 5.363 7.521 cp_fm_upper_to_full 105 14.8 5.760 7.485 5.760 7.485 dbcsr_complete_redistribute 395 12.7 1.164 1.195 5.165 6.959 potential_pw2rs 128 12.3 0.020 0.022 6.742 6.850 fft3d_ps 1291 14.7 2.738 2.966 6.646 6.808 mp_alltoall_d11v 2415 14.1 5.068 6.146 5.068 6.146 calculate_dm_sparse 128 9.5 0.001 0.001 5.924 6.008 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.160 5.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.689 5.694 grid_collocate_task_list 128 9.7 4.910 5.677 4.910 5.677 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.820 5.610 mp_allgather_i34 2507 14.6 2.225 5.286 2.225 5.286 mp_waitany 11748 13.9 2.789 4.876 2.789 4.876 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.615 4.796 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.637 4.704 mp_alltoall_i22 716 14.1 2.349 4.352 2.349 4.352 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.555 4.332 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=203.832000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=911.272727, yerr=21.354544 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 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 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.160188E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922351760 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57632. MP_Allreduce 11104 1087. MP_Sync 88 MP_Alltoall 1724 12509574. 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 0.019 0.033 213.165 213.167 qs_mol_dyn_low 1 2.0 0.003 0.003 212.720 212.731 qs_forces 11 3.9 0.010 0.052 212.607 212.613 qs_energies 11 4.9 0.009 0.036 205.127 205.142 scf_env_do_scf 11 5.9 0.001 0.001 187.147 187.154 velocity_verlet 10 3.0 0.001 0.001 141.719 141.722 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.017 133.018 qs_scf_new_mos 118 7.6 0.001 0.001 91.806 92.122 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.805 92.121 ot_scf_mini 118 9.6 0.003 0.003 87.157 87.452 dbcsr_multiply_generic 2534 12.6 0.190 0.197 83.377 84.549 multiply_cannon 2534 13.6 0.562 0.608 56.436 61.008 multiply_cannon_loop 2534 14.6 1.190 1.221 52.119 55.438 init_scf_loop 11 6.9 0.001 0.006 53.974 53.976 prepare_preconditioner 11 7.9 0.000 0.000 49.684 49.710 make_preconditioner 11 8.9 0.000 0.001 49.684 49.710 make_full_inverse_cholesky 11 9.9 0.000 0.000 43.187 48.218 ot_mini 118 10.6 0.001 0.001 45.549 45.885 rebuild_ks_matrix 129 8.3 0.001 0.001 32.087 32.548 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.087 32.547 mp_waitall_1 149452 16.7 20.172 30.814 20.172 30.814 multiply_cannon_multrec 30408 15.6 14.096 18.681 26.043 30.580 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.981 29.378 qs_ot_get_p 129 10.4 0.001 0.001 27.189 27.506 qs_ot_get_derivative 118 11.6 0.002 0.002 25.448 25.761 make_m2s 5068 13.6 0.098 0.101 21.614 22.896 make_images 5068 14.6 1.951 2.271 21.301 22.579 qs_ot_p2m_diag 84 11.4 0.347 0.395 22.472 22.553 cp_fm_cholesky_invert 11 10.9 21.286 21.302 21.286 21.302 cp_dbcsr_syevd 84 12.4 0.005 0.006 20.974 20.977 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.403 20.060 apply_single 129 13.6 0.001 0.001 19.402 20.059 ot_diis_step 118 11.6 0.018 0.019 19.936 19.938 multiply_cannon_metrocomm3 30408 15.6 0.048 0.050 7.361 18.655 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.122 17.131 cp_fm_diag_elpa_base 84 14.4 16.075 16.370 17.115 17.122 cp_fm_upper_to_full 106 14.8 11.643 17.081 11.643 17.081 sum_up_and_integrate 129 10.3 0.141 0.152 16.528 16.553 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.445 16.466 calculate_rho_elec 129 8.7 0.174 0.191 16.444 16.465 integrate_v_rspace 129 11.3 0.004 0.005 16.387 16.416 make_images_data 5068 15.6 0.063 0.069 11.688 13.441 dbcsr_complete_redistribute 397 12.7 1.497 1.643 9.439 13.356 multiply_cannon_sync_h2d 30408 15.6 11.733 13.061 11.733 13.061 hybrid_alltoall_any 5255 16.5 0.530 2.241 10.281 12.027 dbcsr_mm_accdrv_process 62840 16.2 7.375 8.347 11.526 12.024 init_scf_run 11 5.9 0.000 0.001 11.755 11.756 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.755 11.756 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.840 11.748 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.895 11.132 pw_transfer 1559 11.6 0.087 0.099 10.228 10.368 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.474 10.318 density_rs2pw 129 9.7 0.006 0.007 8.748 10.269 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.000 10.146 cp_fm_cholesky_decompose 22 10.9 9.584 9.692 9.584 9.692 mp_alltoall_i22 720 14.1 5.668 9.594 5.668 9.594 fft_wrap_pw1pw2_140 527 13.2 1.565 1.610 8.620 8.777 rs_pw_transfer 1054 12.0 0.013 0.015 7.051 8.747 wfi_extrapolate 11 7.9 0.001 0.001 8.545 8.545 grid_integrate_task_list 129 12.3 7.445 8.156 7.445 8.156 multiply_cannon_metrocomm4 25340 15.6 0.084 0.098 2.801 7.865 mp_irecv_dv 76866 16.2 2.650 7.583 2.650 7.583 mp_alltoall_d11v 2429 14.1 5.845 7.084 5.845 7.084 fft3d_ps 1301 14.7 2.861 2.937 6.906 7.012 calculate_dm_sparse 129 9.5 0.001 0.001 6.344 6.424 potential_pw2rs 129 12.3 0.024 0.024 6.046 6.122 grid_collocate_task_list 129 9.7 5.106 5.808 5.106 5.808 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.616 5.674 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.810 4.897 mp_sum_l 7951 13.1 3.392 4.725 3.392 4.725 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.624 4.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.427 4.518 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.511 4.512 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=213.167000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.454545, yerr=22.013895 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 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 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.523077E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+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 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57658. MP_Allreduce 11099 1171. MP_Sync 88 MP_Alltoall 1724 18848021. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.034 199.136 199.137 qs_mol_dyn_low 1 2.0 0.003 0.003 198.626 198.638 qs_forces 11 3.9 0.004 0.004 198.521 198.528 qs_energies 11 4.9 0.001 0.001 190.661 190.669 scf_env_do_scf 11 5.9 0.001 0.001 172.293 172.295 velocity_verlet 10 3.0 0.001 0.001 130.736 130.741 scf_env_do_scf_inner_loop 118 6.6 0.019 0.070 128.502 128.503 qs_scf_new_mos 118 7.6 0.001 0.001 86.141 86.306 qs_scf_loop_do_ot 118 8.6 0.001 0.001 86.140 86.305 ot_scf_mini 118 9.6 0.003 0.004 81.438 81.552 dbcsr_multiply_generic 2535 12.6 0.182 0.192 78.747 79.495 multiply_cannon 2535 13.6 0.588 0.618 56.629 63.819 multiply_cannon_loop 2535 14.6 0.446 0.456 51.791 53.840 init_scf_loop 11 6.9 0.000 0.000 43.621 43.623 ot_mini 118 10.6 0.001 0.001 43.120 43.282 prepare_preconditioner 11 7.9 0.000 0.000 39.352 39.388 make_preconditioner 11 8.9 0.000 0.000 39.352 39.388 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.873 37.419 mp_waitall_1 127164 16.7 29.031 37.026 29.031 37.026 rebuild_ks_matrix 129 8.3 0.001 0.001 32.264 32.399 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 32.264 32.398 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.331 29.447 qs_ot_get_p 129 10.4 0.001 0.001 24.881 25.040 cp_fm_cholesky_invert 11 10.9 23.323 23.331 23.323 23.331 qs_ot_get_derivative 118 11.6 0.002 0.002 22.654 22.768 multiply_cannon_metrocomm3 10140 15.6 0.023 0.024 13.031 21.899 multiply_cannon_multrec 10140 15.6 10.427 13.996 17.937 21.298 qs_ot_p2m_diag 84 11.4 0.501 0.507 20.460 20.494 ot_diis_step 118 11.6 0.020 0.021 20.301 20.302 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.704 20.217 apply_single 129 13.6 0.001 0.001 19.703 20.217 make_m2s 5070 13.6 0.067 0.072 17.389 19.667 make_images 5070 14.6 2.289 2.595 17.079 19.358 cp_dbcsr_syevd 84 12.4 0.005 0.005 19.135 19.139 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.169 17.206 calculate_rho_elec 129 8.7 0.258 0.269 17.168 17.206 sum_up_and_integrate 129 10.3 0.181 0.191 16.878 16.924 integrate_v_rspace 129 11.3 0.004 0.004 16.697 16.753 cp_fm_diag_elpa 84 13.4 0.000 0.000 15.305 15.314 cp_fm_diag_elpa_base 84 14.4 15.040 15.123 15.298 15.308 make_images_data 5070 15.6 0.052 0.061 10.350 13.368 hybrid_alltoall_any 5257 16.5 0.794 3.657 10.123 13.082 multiply_cannon_sync_h2d 10140 15.6 11.646 12.336 11.646 12.336 init_scf_run 11 5.9 0.000 0.002 11.409 11.409 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.408 11.409 pw_transfer 1559 11.6 0.087 0.096 10.683 10.753 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.457 10.535 multiply_cannon_metrocomm1 10140 15.6 0.029 0.029 5.485 10.143 cp_fm_cholesky_decompose 22 10.9 9.845 9.950 9.845 9.950 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.308 9.389 density_rs2pw 129 9.7 0.006 0.006 8.406 9.316 fft_wrap_pw1pw2_140 527 13.2 1.915 1.954 8.999 9.083 wfi_extrapolate 11 7.9 0.001 0.001 8.366 8.366 grid_integrate_task_list 129 12.3 7.749 8.354 7.749 8.354 mp_allgather_i34 2535 14.6 3.024 8.134 3.024 8.134 dbcsr_mm_accdrv_process 20958 16.1 2.906 3.700 7.145 7.833 mp_alltoall_d11v 2429 14.1 6.392 7.737 6.392 7.737 rs_pw_transfer 1054 12.0 0.012 0.013 6.030 7.073 fft3d_ps 1301 14.7 2.783 2.852 6.923 7.002 calculate_dm_sparse 129 9.5 0.001 0.001 6.396 6.537 dbcsr_complete_redistribute 397 12.7 2.097 2.148 5.981 6.409 grid_collocate_task_list 129 9.7 5.440 6.109 5.440 6.109 potential_pw2rs 129 12.3 0.026 0.027 5.810 5.890 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.365 5.371 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.307 5.309 multiply_cannon_metrocomm4 7605 15.6 0.024 0.026 1.905 4.360 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.206 4.308 mp_irecv_dv 29150 15.9 1.870 4.282 1.870 4.282 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.918 4.273 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.127 4.179 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.046 4.152 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.902 3.990 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=199.137000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1428.636364, yerr=54.839058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1399806074880 0.0% 0.0% 100.0% flops 32 x 32 x 32 1906965479424 0.0% 0.0% 100.0% flops 22 x 9 x 32 1943685697536 0.0% 0.0% 100.0% flops 9 x 22 x 32 1949313282048 0.0% 0.0% 100.0% flops 22 x 22 x 32 2694948675584 0.0% 0.0% 100.0% flops 32 x 32 x 9 4338990710784 0.0% 0.0% 100.0% flops 32 x 32 x 22 5303210868736 0.0% 0.0% 100.0% flops 9 x 32 x 32 5347334946816 0.0% 0.0% 100.0% flops 22 x 32 x 32 6535631601664 0.0% 0.0% 100.0% flops 9 x 32 x 9 11360521838592 0.0% 0.0% 100.0% flops 22 x 32 x 9 14909203279872 0.0% 0.0% 100.0% flops 9 x 32 x 22 14909203279872 0.0% 0.0% 100.0% flops 22 x 32 x 22 19481731219456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.080547E+12 0.0% 0.0% 100.0% flops max/rank 11.516856E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6655093856 0.0% 0.0% 100.0% number of processed stacks 1931936 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.8 marketing flops 142.438295E+12 ------------------------------------------------------------------------------- # multiplications 2464 max memory usage/rank 3.030540E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 98560 MPI messages size (bytes): total size 1.118682E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.350266E+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 44368 34435235840 4194304 < size <= 16777216 43616 373376942080 16777216 < size 9960 710867104896 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4016 58941. MP_Allreduce 11037 1514. MP_Sync 87 MP_Alltoall 1688 36934280. MP_SendRecv 1764 218624. MP_ISendRecv 1764 218624. MP_Wait 9654 MP_ISend 6312 1080635. MP_IRecv 6312 1080635. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.045 322.630 322.632 qs_mol_dyn_low 1 2.0 0.003 0.003 322.005 322.018 qs_forces 11 3.9 0.004 0.005 321.897 321.900 qs_energies 11 4.9 0.001 0.001 312.039 312.057 scf_env_do_scf 11 5.9 0.001 0.001 288.434 288.442 velocity_verlet 10 3.0 0.001 0.001 229.889 229.899 scf_env_do_scf_inner_loop 115 6.6 0.006 0.014 154.297 154.299 init_scf_loop 11 6.9 0.000 0.000 133.852 133.855 prepare_preconditioner 11 7.9 0.000 0.000 128.378 128.407 make_preconditioner 11 8.9 0.000 0.000 128.378 128.407 make_full_inverse_cholesky 11 9.9 0.000 0.000 104.545 125.346 qs_scf_new_mos 115 7.6 0.001 0.001 99.915 100.033 qs_scf_loop_do_ot 115 8.6 0.001 0.001 99.914 100.032 ot_scf_mini 115 9.6 0.004 0.004 95.063 95.236 dbcsr_multiply_generic 2464 12.5 0.208 0.219 85.373 85.746 cp_fm_upper_to_full 105 14.8 52.268 74.862 52.268 74.862 multiply_cannon 2464 13.5 0.703 0.743 59.873 62.113 multiply_cannon_loop 2464 14.5 0.460 0.470 56.226 57.259 ot_mini 115 10.6 0.001 0.001 46.011 46.258 dbcsr_complete_redistribute 395 12.7 3.996 4.032 30.001 42.544 rebuild_ks_matrix 126 8.3 0.001 0.001 40.798 41.085 qs_ks_build_kohn_sham_matrix 126 9.3 0.017 0.018 40.797 41.085 cp_fm_cholesky_invert 11 10.9 40.415 40.423 40.415 40.423 copy_fm_to_dbcsr 209 11.7 0.002 0.002 26.349 38.850 qs_ks_update_qs_env 126 7.6 0.001 0.001 37.465 37.713 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.787 36.139 mp_alltoall_i22 712 14.1 21.449 33.986 21.449 33.986 qs_ot_get_p 126 10.4 0.001 0.001 33.283 33.562 mp_waitall_1 101902 16.8 30.710 32.183 30.710 32.183 qs_ot_p2m_diag 83 11.4 0.877 0.882 28.672 28.689 cp_dbcsr_syevd 83 12.4 0.006 0.006 26.793 26.797 qs_ot_get_derivative 115 11.6 0.002 0.002 25.552 25.724 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.989 22.990 cp_fm_diag_elpa_base 83 14.4 18.590 20.134 22.985 22.986 qs_rho_update_rho_low 126 7.7 0.001 0.001 22.435 22.464 calculate_rho_elec 126 8.7 0.473 0.474 22.435 22.463 multiply_cannon_metrocomm3 9856 15.5 0.023 0.024 20.607 21.902 sum_up_and_integrate 126 10.3 0.318 0.320 21.468 21.545 integrate_v_rspace 126 11.3 0.004 0.004 21.149 21.230 make_m2s 4928 13.5 0.076 0.078 19.990 20.685 ot_diis_step 115 11.6 0.022 0.022 20.349 20.349 make_images 4928 14.5 3.709 3.820 19.517 20.216 apply_preconditioner_dbcsr 126 12.6 0.000 0.000 19.508 19.886 apply_single 126 13.6 0.001 0.001 19.508 19.886 multiply_cannon_multrec 9856 15.5 10.342 12.166 17.655 17.722 multiply_cannon_sync_h2d 9856 15.5 15.428 15.443 15.428 15.443 pw_transfer 1523 11.6 0.092 0.092 13.533 13.575 fft_wrap_pw1pw2 1271 12.7 0.011 0.011 13.301 13.344 init_scf_run 11 5.9 0.000 0.001 13.331 13.332 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.331 13.331 make_images_data 4928 15.5 0.058 0.063 10.764 12.264 hybrid_alltoall_any 5114 16.4 1.282 3.010 10.877 12.229 fft_wrap_pw1pw2_140 515 13.2 3.032 3.080 11.634 11.680 cp_fm_cholesky_decompose 22 10.9 11.574 11.583 11.574 11.583 mp_alltoall_d11v 2399 14.1 10.315 11.081 10.315 11.081 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.170 10.308 wfi_extrapolate 11 7.9 0.001 0.001 10.104 10.104 density_rs2pw 126 9.7 0.005 0.006 9.565 9.690 dbcsr_mm_accdrv_process 20422 16.0 3.835 5.643 7.076 8.869 grid_integrate_task_list 126 12.3 8.458 8.623 8.458 8.623 fft3d_ps 1271 14.7 2.807 2.815 8.444 8.486 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.112 8.113 rs_scatter_matrices 137 9.7 3.659 4.469 6.608 6.818 calculate_dm_sparse 126 9.5 0.001 0.001 6.727 6.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.673 6.722 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.491 6.654 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=322.632000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2723.818182, yerr=159.539575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 1.262170E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 31004404. 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 0.016 0.038 84.774 84.778 qs_energies 1 2.0 0.000 0.000 84.345 84.359 ls_scf 1 3.0 0.000 0.000 83.437 83.450 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.548 72.720 multiply_cannon 111 7.7 0.017 0.020 55.901 57.090 multiply_cannon_loop 111 8.7 0.202 0.219 52.477 53.769 ls_scf_main 1 4.0 0.000 0.000 52.152 52.164 density_matrix_trs4 2 5.0 0.002 0.003 46.699 46.807 ls_scf_init_scf 1 4.0 0.000 0.000 28.234 28.235 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.149 27.200 mp_waitall_1 11031 10.9 22.436 25.829 22.436 25.829 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.013 25.029 multiply_cannon_multrec 2664 9.7 8.164 8.954 15.455 17.180 multiply_cannon_sync_h2d 2664 9.7 13.608 15.285 13.608 15.285 make_m2s 222 7.7 0.008 0.011 13.019 13.502 make_images 222 8.7 0.098 0.110 12.997 13.483 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.664 12.770 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.543 9.485 make_images_data 222 9.7 0.004 0.005 7.564 8.089 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.617 6.913 7.802 hybrid_alltoall_any 227 10.6 0.216 1.831 6.481 7.623 dbcsr_mm_accdrv_process_sort 4760 11.4 6.206 7.062 6.206 7.062 calculate_norms 4752 9.8 5.519 6.110 5.519 6.110 apply_matrix_preconditioner 6 5.3 0.011 0.025 5.040 5.210 mp_sum_l 807 5.4 3.093 4.614 3.093 4.614 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.086 3.824 mp_irecv_dv 6231 10.9 2.068 3.793 2.068 3.793 make_images_sizes 222 9.7 0.000 0.000 0.752 3.486 mp_alltoall_i44 222 10.7 0.751 3.485 0.751 3.485 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.301 3.470 arnoldi_extremal 4 6.8 0.000 0.000 3.204 3.229 arnoldi_normal_ev 4 7.8 0.001 0.002 3.204 3.229 build_subspace 16 8.4 0.009 0.012 3.108 3.111 ls_scf_post 1 4.0 0.000 0.000 3.050 3.064 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.352 2.782 dbcsr_merge_single_wm 555 10.7 0.476 0.609 2.344 2.774 make_images_pack 222 9.7 2.206 2.621 2.208 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.325 2.559 dbcsr_sort_data 658 11.4 2.123 2.474 2.123 2.474 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.455 2.073 2.456 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.243 2.291 buffer_matrices_ensure_size 222 8.7 1.803 2.137 1.803 2.137 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.764 1.765 rebuild_ks_matrix 3 7.3 0.000 0.000 1.755 1.756 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.755 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.778000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 2.098704E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_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 0.031 0.068 90.886 90.887 qs_energies 1 2.0 0.000 0.000 90.348 90.352 ls_scf 1 3.0 0.000 0.000 89.031 89.034 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.061 75.422 multiply_cannon 111 7.7 0.028 0.042 53.248 56.534 ls_scf_main 1 4.0 0.000 0.000 54.741 54.744 multiply_cannon_loop 111 8.7 0.116 0.122 49.952 52.837 density_matrix_trs4 2 5.0 0.002 0.003 48.998 49.170 ls_scf_init_scf 1 4.0 0.000 0.000 30.764 30.765 mp_waitall_1 9105 10.9 21.084 30.528 21.084 30.528 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.521 29.627 multiply_cannon_multrec 1332 9.7 13.117 17.273 22.293 27.612 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.128 27.147 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.771 21.312 make_m2s 222 7.7 0.006 0.007 15.213 15.883 make_images 222 8.7 1.570 1.962 15.182 15.854 dbcsr_mm_accdrv_process 4041 10.4 0.262 0.449 8.782 10.361 dbcsr_mm_accdrv_process_sort 4041 11.4 8.376 9.912 8.376 9.912 make_images_data 222 9.7 0.004 0.004 8.701 9.673 hybrid_alltoall_any 227 10.6 0.521 2.481 8.165 9.027 mp_sum_l 807 5.4 5.294 8.211 5.294 8.211 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.200 7.734 mp_irecv_dv 3311 11.0 3.180 7.680 3.180 7.680 calculate_norms 2376 9.8 6.008 6.870 6.008 6.870 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.127 6.506 multiply_cannon_sync_h2d 1332 9.7 4.849 6.034 4.849 6.034 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.023 5.274 arnoldi_extremal 4 6.8 0.000 0.000 4.609 4.634 arnoldi_normal_ev 4 7.8 0.001 0.004 4.609 4.634 build_subspace 16 8.4 0.014 0.021 4.348 4.351 ls_scf_post 1 4.0 0.000 0.000 3.526 3.530 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.128 3.368 ls_scf_store_result 1 5.0 0.000 0.000 3.228 3.350 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.221 2.739 3.223 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.621 2.730 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.242 2.702 make_images_pack 222 9.7 2.020 2.411 2.022 2.413 mp_allgather_i34 111 8.7 0.994 2.298 0.994 2.298 dbcsr_sort_data 436 11.2 1.830 2.062 1.830 2.062 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.914 1.917 rebuild_ks_matrix 3 7.3 0.000 0.000 1.902 1.904 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.010 1.902 1.904 dbcsr_data_new 4174 10.1 1.612 1.857 1.612 1.857 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.887000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1758.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 2.801947E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_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 0.066 0.154 94.317 94.320 qs_energies 1 2.0 0.000 0.000 93.531 93.585 ls_scf 1 3.0 0.001 0.007 92.035 92.092 dbcsr_multiply_generic 111 6.7 0.017 0.024 76.528 76.915 ls_scf_main 1 4.0 0.001 0.013 57.366 57.372 multiply_cannon 111 7.7 0.044 0.104 52.907 57.158 multiply_cannon_loop 111 8.7 0.100 0.105 49.265 53.190 density_matrix_trs4 2 5.0 0.003 0.024 51.343 51.554 mp_waitall_1 7281 11.0 23.871 34.009 23.871 34.009 ls_scf_init_scf 1 4.0 0.000 0.002 31.059 31.062 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.745 29.817 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.334 27.342 multiply_cannon_multrec 888 9.7 12.515 15.163 21.019 24.278 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.203 23.321 make_m2s 222 7.7 0.006 0.007 16.804 18.114 make_images 222 8.7 1.966 2.286 16.766 18.074 make_images_data 222 9.7 0.003 0.004 9.661 10.911 hybrid_alltoall_any 227 10.6 0.621 2.875 9.286 10.557 dbcsr_mm_accdrv_process 3754 10.4 0.254 0.415 8.027 9.258 dbcsr_mm_accdrv_process_sort 3754 11.4 7.656 8.844 7.656 8.844 mp_sum_l 807 5.4 5.306 8.675 5.306 8.675 multiply_cannon_sync_h2d 888 9.7 6.026 7.135 6.026 7.135 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.441 7.083 mp_irecv_dv 2335 11.1 2.426 7.045 2.426 7.045 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.997 6.700 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.630 6.331 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.974 5.208 arnoldi_extremal 4 6.8 0.000 0.000 5.084 5.097 arnoldi_normal_ev 4 7.8 0.001 0.005 5.084 5.097 build_subspace 16 8.4 0.014 0.020 4.770 4.781 calculate_norms 1584 9.8 4.315 4.699 4.315 4.699 mp_allgather_i34 111 8.7 1.482 3.826 1.482 3.826 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.427 3.752 ls_scf_post 1 4.0 0.004 0.028 3.608 3.665 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.585 3.017 3.587 ls_scf_store_result 1 5.0 0.000 0.000 3.314 3.423 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.861 2.992 make_images_sizes 222 9.7 0.000 0.000 0.846 2.136 mp_alltoall_i44 222 10.7 0.846 2.136 0.846 2.136 make_images_pack 222 9.7 1.812 2.120 1.815 2.123 dbcsr_sort_data 325 11.1 1.879 2.120 1.879 2.120 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.997 1.999 rebuild_ks_matrix 3 7.3 0.000 0.000 1.979 1.981 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.020 1.979 1.981 dbcsr_data_release 9322 10.9 1.297 1.976 1.297 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.320000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2169.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 3.346665E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_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 0.066 0.122 96.696 96.723 qs_energies 1 2.0 0.000 0.000 96.016 96.024 ls_scf 1 3.0 0.000 0.001 94.353 94.361 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.108 78.390 ls_scf_main 1 4.0 0.000 0.002 58.657 58.660 multiply_cannon 111 7.7 0.047 0.120 51.569 56.002 density_matrix_trs4 2 5.0 0.002 0.004 52.577 52.682 multiply_cannon_loop 111 8.7 0.115 0.124 46.486 48.858 ls_scf_init_scf 1 4.0 0.000 0.002 32.464 32.466 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.255 31.317 mp_waitall_1 6369 11.0 22.605 28.849 22.605 28.849 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.798 28.812 multiply_cannon_multrec 1332 9.7 14.241 17.221 22.077 24.895 make_m2s 222 7.7 0.007 0.008 21.021 22.451 make_images 222 8.7 3.125 3.604 20.971 22.403 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.222 17.211 make_images_data 222 9.7 0.004 0.004 11.722 13.335 hybrid_alltoall_any 227 10.6 0.797 3.739 11.102 12.893 dbcsr_mm_accdrv_process 3641 10.4 0.230 0.406 7.471 8.974 dbcsr_mm_accdrv_process_sort 3641 11.4 7.096 8.556 7.096 8.556 mp_sum_l 807 5.4 3.949 7.417 3.949 7.417 multiply_cannon_sync_h2d 1332 9.7 5.497 6.302 5.497 6.302 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.052 6.017 mp_irecv_dv 3229 10.9 2.029 5.931 2.029 5.931 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.050 5.845 arnoldi_extremal 4 6.8 0.000 0.000 5.227 5.239 arnoldi_normal_ev 4 7.8 0.001 0.004 5.227 5.239 build_subspace 16 8.4 0.014 0.021 4.882 4.892 mp_allgather_i34 111 8.7 2.213 4.724 2.213 4.724 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.547 4.696 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.489 4.578 calculate_norms 2376 9.8 4.209 4.532 4.209 4.532 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.572 3.890 dbcsr_matrix_vector_mult_local 304 10.0 3.174 3.669 3.176 3.670 dbcsr_sort_data 658 11.4 3.042 3.263 3.042 3.263 ls_scf_post 1 4.0 0.000 0.001 3.232 3.238 dbcsr_special_finalize 555 9.7 0.006 0.008 2.790 3.087 dbcsr_merge_single_wm 555 10.7 0.531 0.652 2.781 3.079 ls_scf_store_result 1 5.0 0.000 0.000 2.968 3.035 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.973 3.017 dbcsr_data_release 10477 10.7 1.581 2.414 1.581 2.414 dbcsr_finalize 304 7.8 0.049 0.061 1.790 1.976 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.934 1.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.723000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2768.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 4.724052E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_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 0.057 0.065 100.497 100.498 qs_energies 1 2.0 0.000 0.000 99.616 99.621 ls_scf 1 3.0 0.000 0.000 97.674 97.684 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.487 78.686 ls_scf_main 1 4.0 0.000 0.000 62.835 62.836 multiply_cannon 111 7.7 0.098 0.212 55.615 61.043 density_matrix_trs4 2 5.0 0.002 0.003 55.461 55.582 multiply_cannon_loop 111 8.7 0.070 0.085 51.171 52.706 mp_waitall_1 5436 11.0 27.093 31.530 27.093 31.530 ls_scf_init_scf 1 4.0 0.000 0.000 31.191 31.196 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.876 29.915 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.674 27.679 multiply_cannon_multrec 444 9.7 14.004 16.476 21.025 22.715 make_m2s 222 7.7 0.004 0.005 18.126 20.553 make_images 222 8.7 3.710 4.408 18.065 20.494 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.372 17.165 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.280 13.851 make_images_data 222 9.7 0.003 0.004 10.231 12.724 hybrid_alltoall_any 227 10.6 0.786 3.760 9.903 12.643 dbcsr_mm_accdrv_process 3003 10.4 0.193 0.368 6.723 7.849 dbcsr_mm_accdrv_process_sort 3003 11.4 6.402 7.497 6.402 7.497 multiply_cannon_sync_h2d 444 9.7 6.580 7.482 6.580 7.482 mp_allgather_i34 111 8.7 2.672 7.073 2.672 7.073 arnoldi_extremal 4 6.8 0.000 0.000 6.221 6.228 arnoldi_normal_ev 4 7.8 0.001 0.004 6.220 6.228 build_subspace 16 8.4 0.015 0.019 5.833 5.845 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.650 4.787 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.303 4.478 dbcsr_matrix_vector_mult_local 304 10.0 3.670 4.158 3.672 4.160 mp_sum_l 807 5.4 2.818 4.022 2.818 4.022 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.719 3.865 make_images_sizes 222 9.7 0.000 0.000 1.274 3.820 mp_alltoall_i44 222 10.7 1.274 3.820 1.274 3.820 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.701 3.667 calculate_norms 792 9.8 3.548 3.659 3.548 3.659 ls_scf_post 1 4.0 0.000 0.000 3.647 3.652 mp_irecv_dv 1241 11.2 1.682 3.633 1.682 3.633 ls_scf_store_result 1 5.0 0.000 0.000 3.422 3.474 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.884 3.026 dbcsr_finalize 304 7.8 0.062 0.078 2.190 2.269 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.175 2.177 rebuild_ks_matrix 3 7.3 0.000 0.000 2.143 2.145 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.143 2.145 dbcsr_merge_all 275 8.9 0.474 0.525 2.042 2.106 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.498000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3681.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 8.666419E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_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 0.078 0.098 110.529 110.531 qs_energies 1 2.0 0.000 0.000 109.098 109.112 ls_scf 1 3.0 0.000 0.000 106.174 106.187 dbcsr_multiply_generic 111 6.7 0.023 0.027 79.279 79.387 ls_scf_main 1 4.0 0.000 0.000 66.356 66.357 density_matrix_trs4 2 5.0 0.002 0.003 56.943 57.004 multiply_cannon 111 7.7 0.164 0.237 51.718 53.447 multiply_cannon_loop 111 8.7 0.067 0.069 48.280 49.222 ls_scf_init_scf 1 4.0 0.000 0.000 36.121 36.122 ls_scf_init_matrix_S 1 5.0 0.000 0.001 34.501 34.519 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.674 31.681 mp_waitall_1 4527 11.1 22.722 26.282 22.722 26.282 make_m2s 222 7.7 0.005 0.005 23.927 24.893 make_images 222 8.7 4.577 4.958 23.821 24.786 multiply_cannon_multrec 444 9.7 17.875 18.690 22.457 23.286 hybrid_alltoall_any 227 10.6 1.661 3.618 12.929 15.797 make_images_data 222 9.7 0.003 0.003 13.221 15.576 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.886 13.108 multiply_cannon_sync_h2d 444 9.7 8.855 8.896 8.855 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.655 7.667 arnoldi_normal_ev 4 7.8 0.002 0.008 7.655 7.667 build_subspace 16 8.4 0.025 0.036 7.098 7.109 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.610 5.638 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.326 5.421 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.083 5.319 dbcsr_matrix_vector_mult_local 304 10.0 4.994 5.300 4.997 5.303 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.323 4.414 4.532 dbcsr_mm_accdrv_process_sort 1814 11.4 4.115 4.245 4.115 4.245 ls_scf_post 1 4.0 0.000 0.000 3.697 3.710 make_images_sizes 222 9.7 0.000 0.000 1.394 3.601 mp_alltoall_i44 222 10.7 1.394 3.601 1.394 3.601 mp_allgather_i34 111 8.7 1.090 3.466 1.090 3.466 ls_scf_store_result 1 5.0 0.000 0.000 3.450 3.458 calculate_norms 792 9.8 3.239 3.269 3.239 3.269 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.129 dbcsr_merge_all 275 8.9 0.890 0.914 2.862 2.910 dbcsr_complete_redistribute 5 7.6 1.426 1.467 2.805 2.907 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.894 2.894 dbcsr_data_release 12724 10.6 2.328 2.859 2.328 2.859 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.593 2.594 matrix_ls_to_qs 2 6.0 0.000 0.000 2.455 2.578 rebuild_ks_matrix 3 7.3 0.000 0.000 2.526 2.528 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.526 2.528 dbcsr_sort_data 325 11.1 2.443 2.506 2.443 2.506 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.430 2.436 dbcsr_new_transposed 4 7.5 0.246 0.258 2.297 2.310 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.192 2.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.531000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6884.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0165da4645489888fae55c9cbb9ccbdcebf313e1_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 587.997184E+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 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 1155319307. 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 0.066 0.129 314.097 314.133 qs_mol_dyn_low 1 2.0 0.005 0.032 312.986 313.019 qs_forces 5 3.8 0.006 0.048 312.837 312.873 qs_energies 5 4.8 0.005 0.064 309.029 309.110 scf_env_do_scf 5 5.8 0.001 0.020 293.911 293.949 scf_env_do_scf_inner_loop 105 6.6 0.002 0.008 253.138 253.172 qs_scf_new_mos 105 7.6 0.000 0.001 198.535 198.978 qs_scf_loop_do_ot 105 8.6 0.001 0.001 198.534 198.977 ot_scf_mini 105 9.6 0.003 0.004 187.837 188.070 dbcsr_multiply_generic 1445 12.2 0.127 0.160 139.766 141.352 velocity_verlet 4 3.0 0.001 0.001 127.634 127.642 multiply_cannon 1445 13.2 0.282 0.293 116.728 121.046 multiply_cannon_loop 1445 14.2 2.366 2.497 113.477 118.113 qs_ot_get_p 112 10.4 0.001 0.002 92.111 92.572 qs_ot_p2m_diag 40 11.0 0.020 0.038 78.874 78.950 cp_dbcsr_syevd 40 12.0 0.002 0.002 75.251 75.262 cp_fm_syevd 40 13.0 0.000 0.000 68.335 68.524 ot_mini 105 10.6 0.001 0.004 64.418 64.594 cp_fm_redistribute_end 40 14.0 31.288 62.433 31.299 62.443 cp_fm_syevd_base 40 14.0 31.122 62.275 31.122 62.275 mp_waitall_1 488190 16.1 43.678 56.745 43.678 56.745 multiply_cannon_metrocomm3 69360 15.2 0.192 0.207 28.892 44.515 multiply_cannon_multrec 69360 15.2 30.805 36.778 38.580 44.297 qs_ot_get_derivative 55 11.6 0.001 0.001 42.513 42.723 init_scf_loop 7 6.6 0.001 0.016 40.712 40.719 rebuild_ks_matrix 110 8.4 0.000 0.000 35.321 35.646 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.015 35.320 35.646 prepare_preconditioner 7 7.6 0.000 0.002 34.724 34.757 make_preconditioner 7 8.6 0.000 0.004 34.724 34.757 multiply_cannon_sync_h2d 69360 15.2 29.128 33.969 29.128 33.969 qs_ks_update_qs_env 112 7.6 0.001 0.001 32.332 32.628 qs_rho_update_rho_low 110 7.6 0.001 0.002 26.628 27.047 calculate_rho_elec 110 8.6 0.030 0.033 26.628 27.047 make_full_inverse_cholesky 7 9.6 0.000 0.000 25.545 25.611 rs_pw_transfer 690 11.5 0.011 0.012 23.528 24.801 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.157 23.383 apply_single 62 13.6 0.000 0.000 23.157 23.383 density_rs2pw 110 9.6 0.006 0.007 19.873 21.292 ot_new_cg_direction 55 11.6 0.001 0.005 21.173 21.176 cp_fm_cholesky_invert 7 10.6 17.437 17.475 17.437 17.475 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.947 16.292 multiply_cannon_metrocomm1 69360 15.2 0.091 0.098 5.644 15.855 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.352 15.460 mp_sum_l 4684 12.4 13.273 14.895 13.273 14.895 pw_transfer 1645 12.4 0.085 0.114 13.972 14.312 fft_wrap_pw1pw2 1425 13.5 0.013 0.018 13.827 14.172 sum_up_and_integrate 60 10.3 0.028 0.031 13.851 13.864 integrate_v_rspace 60 11.3 0.002 0.005 13.823 13.837 calculate_dm_sparse 110 9.5 0.000 0.001 12.402 12.728 check_diag 80 13.5 8.585 8.857 11.882 12.047 init_scf_run 5 5.8 0.000 0.001 12.011 12.013 scf_env_initial_rho_setup 5 6.8 0.001 0.002 12.011 12.013 fft_wrap_pw1pw2_240 915 15.0 1.214 1.320 11.407 11.724 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.328 11.412 qs_vxc_create 110 10.4 0.002 0.008 11.229 11.304 fft3d_pb 915 16.0 2.387 2.701 9.439 9.951 make_m2s 2890 13.2 0.077 0.083 9.367 9.947 make_images 2890 14.2 0.244 0.263 9.263 9.843 potential_pw2rs 60 12.3 0.003 0.003 9.508 9.570 rs_pw_transfer_RS2PW_30 110 11.6 1.549 1.646 9.017 9.545 mp_sendrecv_dv 168740 12.6 9.194 9.363 9.194 9.363 make_full_single_inverse 7 9.6 0.001 0.003 8.694 8.727 dbcsr_mm_accdrv_process 154766 15.8 3.971 4.229 7.654 8.495 xc_rho_set_and_dset_create 110 12.4 0.076 0.100 7.658 7.925 xc_vxc_pw_create 60 11.3 0.039 0.071 7.627 7.702 cp_fm_cholesky_decompose 14 10.2 7.502 7.511 7.502 7.511 mp_alltoall_z22v 2340 17.7 6.615 7.273 6.615 7.273 xc_pw_derive 510 13.4 0.005 0.007 6.855 6.929 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 6.755 6.770 acc_transpose_blocks 69360 15.2 0.353 0.375 5.990 6.338 calculate_first_density_matrix 1 7.0 0.000 0.005 6.284 6.294 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=314.133000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.600000, yerr=2.332381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 0165da4645489888fae55c9cbb9ccbdcebf313e1 Summary: empty Status: OK