=== 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: c025c99d5d860bcf6d6a049a55c865b3089da860 ################# 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 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, 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 (26.10.2022) # # \ 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 --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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/01 job id: 43256275 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/02 job id: 43256276 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/03 job id: 43256277 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/04 job id: 43256278 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/05 job id: 43256279 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/06 job id: 43256280 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/07 job id: 43256281 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/08 job id: 43256282 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/09 job id: 43256283 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/10 job id: 43256284 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/11 job id: 43256285 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/12 job id: 43256286 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/13 job id: 43256287 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/14 job id: 43256288 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/15 job id: 43256289 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/16 job id: 43256290 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/17 job id: 43256291 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/18 job id: 43256292 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/19 job id: 43256295 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/20 job id: 43256296 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/21 job id: 43256297 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/22 job id: 43256298 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/23 job id: 43256299 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/24 job id: 43256300 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/25 job id: 43256302 --- 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/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/26 job id: 43256303 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 0.055 0.060 135.104 135.105 farming_run 1 2.0 134.546 134.549 135.040 135.041 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.443647E+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 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.031 115.025 115.026 qs_energies 1 2.0 0.000 0.000 114.762 114.765 mp2_main 1 3.0 0.000 0.000 112.894 112.896 mp2_gpw_main 1 4.0 0.019 0.026 112.053 112.056 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.035 93.531 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 55.432 55.928 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.754 47.194 get_2c_integrals 1 6.0 0.001 0.018 36.786 37.430 integrate_v_rspace 273 8.0 0.438 0.452 25.173 30.369 pw_transfer 6555 10.6 0.374 0.384 27.318 27.948 grid_integrate_task_list 273 9.0 20.984 26.670 20.984 26.670 fft_wrap_pw1pw2 5465 11.4 0.044 0.045 26.007 26.606 fft_wrap_pw1pw2_100 2178 12.4 1.159 1.232 23.572 24.161 compute_2c_integrals 1 7.0 0.009 0.016 19.209 19.210 rpa_ri_compute_en 1 5.0 0.000 0.000 18.911 19.159 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.808 18.910 mp2_eri_2c_integrate_gpw 1 9.0 2.381 2.414 18.805 18.907 cp_fm_cholesky_decompose 12 8.2 17.517 18.165 17.517 18.165 cholesky_decomp 1 7.0 0.000 0.000 16.423 17.064 fft3d_s 5443 13.4 16.162 16.624 16.183 16.645 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.598 10.853 15.598 calculate_wavefunction 272 8.0 5.381 5.533 12.504 13.109 rpa_num_int 1 6.0 0.001 0.009 10.501 10.501 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.471 10.501 calc_mat_Q 8 8.0 0.000 0.000 9.330 9.423 contract_S_to_Q 8 9.0 0.000 0.000 8.751 8.841 calc_potential_gpw 544 9.5 0.005 0.005 8.213 8.571 parallel_gemm_fm 14 9.1 0.000 0.000 8.325 8.417 parallel_gemm_fm_cosma 14 10.1 8.325 8.417 8.325 8.417 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.178 8.407 potential_pw2rs 545 10.0 0.107 0.109 7.689 8.396 create_integ_mat 1 6.0 0.030 0.037 7.818 7.818 collocate_single_gaussian 272 10.0 0.040 0.043 7.429 7.668 array2fm 1 7.0 0.000 0.000 6.576 7.208 pw_scatter_s 2720 13.7 4.437 4.671 4.437 4.671 pw_gather_s 2722 13.2 3.883 4.236 3.883 4.236 array2fm_buffer_send 1 8.0 2.868 3.178 2.868 3.178 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.055904, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 0.029 0.037 399.430 399.432 farming_run 1 2.0 398.096 398.106 399.390 399.394 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.212944E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 5658414. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 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 0.023 0.047 210.367 210.369 qs_energies 1 2.0 0.000 0.000 209.981 209.987 scf_env_do_scf 1 3.0 0.000 0.000 106.390 106.390 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.503 105.510 rebuild_ks_matrix 4 6.0 0.000 0.000 105.502 105.509 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.065 105.502 105.509 hfx_ks_matrix 4 8.0 0.001 0.001 105.130 105.133 integrate_four_center 4 9.0 0.144 0.477 105.129 105.132 mp2_main 1 3.0 0.000 0.000 103.266 103.272 mp2_gpw_main 1 4.0 0.050 0.223 102.405 102.414 integrate_four_center_main 4 10.0 0.098 0.567 96.634 99.245 integrate_four_center_bin 262 11.0 96.536 99.234 96.536 99.234 init_scf_loop 1 4.0 0.000 0.000 92.283 92.284 mp2_ri_gpw_compute_in 1 5.0 0.080 0.440 75.317 76.662 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.105 54.817 56.156 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.163 42.193 47.328 integrate_v_rspace 95 8.0 0.399 0.571 28.559 33.550 pw_transfer 2240 10.6 0.145 0.178 29.923 30.429 ao_to_mo_and_store_B_mult_1 91 7.0 10.936 30.044 10.936 30.044 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.927 29.454 grid_integrate_task_list 95 9.0 23.843 29.040 23.843 29.040 mp2_ri_gpw_compute_en 1 5.0 0.058 0.097 26.860 28.852 fft_wrap_pw1pw2_100 730 12.4 1.280 1.436 26.680 27.292 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.915 24.873 24.885 get_2c_integrals 1 6.0 0.004 0.041 20.377 20.441 compute_2c_integrals 1 7.0 0.007 0.027 19.342 19.372 compute_2c_integrals_loop_lm 1 8.0 0.003 0.015 18.891 19.210 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.849 18.888 19.210 fft3d_s 1823 13.4 18.440 18.911 18.453 18.924 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.105 14.105 calculate_wavefunction 91 8.0 2.012 2.040 9.740 9.992 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.592 8.837 9.502 potential_pw2rs 186 10.0 0.034 0.035 8.649 9.246 local_gemm 172 8.0 8.278 8.921 8.278 8.921 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.252 8.600 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.532 7.814 8.495 collocate_single_gaussian 91 10.0 0.018 0.037 7.900 8.207 calc_potential_gpw 182 9.5 0.002 0.002 7.920 8.150 mp_sendrecv_dm3 2068 8.0 5.837 6.491 5.837 6.491 mp2_ri_gpw_compute_en_ener 172 7.0 6.336 6.409 6.336 6.409 mp_sync 38 10.4 3.162 6.174 3.162 6.174 pw_gather_s 912 13.2 4.932 5.372 4.932 5.372 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.393035, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1498.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 436.690944E+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 271. MP_Sync 580 MP_Alltoall 2083 499316. MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.013 0.035 54.695 54.706 qs_mol_dyn_low 1 2.0 0.003 0.004 54.202 54.309 qs_forces 11 3.9 0.002 0.002 54.010 54.020 qs_energies 11 4.9 0.002 0.002 52.582 52.590 scf_env_do_scf 11 5.9 0.000 0.001 45.285 45.294 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.189 43.197 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.964 36.475 qs_scf_new_mos 108 7.5 0.000 0.001 33.263 33.564 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.262 33.564 ot_scf_mini 108 9.5 0.002 0.002 31.602 31.790 multiply_cannon 2286 13.5 0.182 0.188 28.401 29.818 multiply_cannon_loop 2286 14.5 1.504 1.568 27.722 29.155 velocity_verlet 10 3.0 0.001 0.001 24.843 24.849 ot_mini 108 10.5 0.001 0.001 20.719 20.955 qs_ot_get_derivative 108 11.5 0.001 0.001 17.733 17.923 mp_waitall_1 267858 16.1 9.288 16.264 9.288 16.264 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.596 14.237 multiply_cannon_multrec 54864 15.5 4.182 6.527 9.175 13.024 rebuild_ks_matrix 119 8.3 0.000 0.000 7.846 7.997 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.846 7.997 multiply_cannon_sync_h2d 54864 15.5 5.878 7.449 5.878 7.449 mp_sum_l 7207 12.9 5.660 7.324 5.660 7.324 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.942 7.079 dbcsr_mm_accdrv_process 76910 16.1 1.190 1.838 4.915 6.680 qs_ot_get_p 119 10.4 0.001 0.001 6.297 6.619 init_scf_run 11 5.9 0.000 0.001 6.094 6.096 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.094 6.095 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.995 6.090 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.238 5.689 jit_kernel_multiply 13 15.8 3.661 4.759 3.661 4.759 sum_up_and_integrate 119 10.3 0.012 0.015 4.474 4.483 integrate_v_rspace 119 11.3 0.002 0.002 4.462 4.472 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.936 4.040 calculate_rho_elec 119 8.7 0.011 0.016 3.936 4.040 calculate_first_density_matrix 1 7.0 0.000 0.000 3.981 3.984 calculate_dm_sparse 119 9.5 0.000 0.000 3.800 3.959 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.530 3.602 multiply_cannon_metrocomm1 54864 15.5 0.051 0.058 1.897 3.390 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.889 3.083 apply_single 119 13.6 0.000 0.000 2.889 3.083 rs_pw_transfer 974 11.9 0.011 0.012 2.674 2.770 ot_diis_step 108 11.5 0.006 0.006 2.684 2.685 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.671 2.673 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.641 2.642 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.429 2.496 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.431 2.432 acc_transpose_blocks 54864 15.5 0.234 0.249 1.978 2.422 cp_fm_redistribute_end 50 14.0 2.205 2.407 2.209 2.409 cp_fm_diag_elpa_base 50 14.0 0.198 2.376 0.199 2.385 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.301 2.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.233 2.282 density_rs2pw 119 9.7 0.004 0.004 2.092 2.194 grid_integrate_task_list 119 12.3 2.058 2.156 2.058 2.156 init_scf_loop 11 6.9 0.000 0.000 2.080 2.081 wfi_extrapolate 11 7.9 0.001 0.001 2.055 2.056 mp_sum_d 4125 12.0 1.329 1.921 1.329 1.921 potential_pw2rs 119 12.3 0.004 0.004 1.798 1.816 pw_transfer 1439 11.6 0.051 0.055 1.594 1.664 make_m2s 4572 13.5 0.054 0.056 1.595 1.636 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.520 1.592 make_images 4572 14.5 0.132 0.138 1.511 1.552 mp_alltoall_d11v 2130 13.8 1.251 1.478 1.251 1.478 mp_waitany 12084 13.8 1.240 1.380 1.240 1.380 fft3d_ps 1201 14.6 0.358 0.462 1.298 1.362 grid_collocate_task_list 119 9.7 1.289 1.347 1.289 1.347 acc_transpose_blocks_kernels 54864 16.5 0.252 0.385 0.973 1.267 fft_wrap_pw1pw2_140 487 13.2 0.080 0.092 1.172 1.245 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.749 1.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.706000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=416.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 472.297472E+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 104 MP_Alltoall 2060 981254. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.028 0.041 40.393 40.394 qs_mol_dyn_low 1 2.0 0.003 0.005 40.010 40.097 qs_forces 11 3.9 0.002 0.003 39.888 39.903 qs_energies 11 4.9 0.002 0.003 38.222 38.240 scf_env_do_scf 11 5.9 0.001 0.001 31.639 31.639 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.132 29.133 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.604 22.944 qs_scf_new_mos 108 7.5 0.001 0.001 20.172 20.420 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.172 20.420 ot_scf_mini 108 9.5 0.002 0.003 19.275 19.447 multiply_cannon 2286 13.5 0.208 0.216 17.557 19.259 velocity_verlet 10 3.0 0.001 0.001 18.176 18.178 multiply_cannon_loop 2286 14.5 0.902 0.967 16.438 17.871 ot_mini 108 10.5 0.001 0.001 12.151 12.384 mp_waitall_1 217478 16.2 6.313 11.939 6.313 11.939 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.543 10.165 qs_ot_get_derivative 108 11.5 0.001 0.001 9.687 9.862 multiply_cannon_multrec 27432 15.5 1.991 4.608 6.480 9.505 rebuild_ks_matrix 119 8.3 0.000 0.000 7.102 7.234 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.101 7.234 dbcsr_mm_accdrv_process 47894 16.0 3.281 5.987 4.420 6.631 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.283 6.402 init_scf_run 11 5.9 0.000 0.001 5.346 5.347 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.346 5.347 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.766 4.566 qs_ot_get_p 119 10.4 0.001 0.001 4.208 4.437 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.066 4.154 apply_single 119 13.6 0.000 0.000 3.066 4.154 sum_up_and_integrate 119 10.3 0.024 0.028 4.139 4.144 integrate_v_rspace 119 11.3 0.002 0.002 4.115 4.124 mp_sum_l 7207 12.9 2.095 4.038 2.095 4.038 calculate_first_density_matrix 1 7.0 0.001 0.005 3.853 3.856 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.602 3.639 calculate_rho_elec 119 8.7 0.021 0.024 3.602 3.638 multiply_cannon_sync_h2d 27432 15.5 2.178 2.935 2.178 2.935 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.756 2.775 make_m2s 4572 13.5 0.053 0.054 2.506 2.731 make_images 4572 14.5 0.200 0.237 2.417 2.640 rs_pw_transfer 974 11.9 0.009 0.011 2.512 2.573 calculate_dm_sparse 119 9.5 0.000 0.000 2.428 2.500 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.493 2.495 init_scf_loop 11 6.9 0.000 0.002 2.486 2.486 ot_diis_step 108 11.5 0.010 0.011 2.412 2.413 jit_kernel_multiply 9 16.0 1.085 2.395 1.085 2.395 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.309 2.309 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.166 2.259 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.205 2.244 density_rs2pw 119 9.7 0.004 0.004 1.957 2.026 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.978 1.978 acc_transpose_blocks 27432 15.5 0.111 0.115 1.513 1.968 grid_integrate_task_list 119 12.3 1.841 1.958 1.841 1.958 cp_fm_redistribute_end 50 14.0 1.627 1.952 1.631 1.954 cp_fm_diag_elpa_base 50 14.0 0.313 1.895 0.321 1.926 potential_pw2rs 119 12.3 0.006 0.006 1.873 1.883 pw_transfer 1439 11.6 0.063 0.067 1.777 1.807 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.688 1.720 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.638 1.646 prepare_preconditioner 11 7.9 0.000 0.000 1.600 1.626 make_preconditioner 11 8.9 0.000 0.002 1.600 1.626 make_images_data 4572 15.5 0.045 0.052 1.156 1.575 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.499 1.558 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.007 1.475 wfi_extrapolate 11 7.9 0.001 0.001 1.442 1.442 fft3d_ps 1201 14.6 0.500 0.555 1.396 1.426 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.316 1.346 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.974 1.342 grid_collocate_task_list 119 9.7 1.241 1.337 1.241 1.337 mp_allgather_i34 2286 14.5 0.563 1.313 0.563 1.313 mp_alltoall_d11v 2130 13.8 1.205 1.308 1.205 1.308 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.187 1.233 jit_kernel_transpose 5 15.5 0.793 1.073 0.793 1.073 mp_sum_d 4125 12.0 0.594 1.050 0.594 1.050 parallel_gemm_fm 81 9.0 0.000 0.000 0.989 0.994 parallel_gemm_fm_cosma 81 10.0 0.989 0.994 0.989 0.994 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.964 0.980 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.876 0.889 make_basis_sm 11 9.8 0.000 0.000 0.883 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.394000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=449.727273, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 506.372096E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 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 0.136 0.161 34.361 34.363 qs_mol_dyn_low 1 2.0 0.003 0.005 33.685 33.811 qs_forces 11 3.9 0.003 0.003 33.583 33.584 qs_energies 11 4.9 0.005 0.014 31.888 31.892 scf_env_do_scf 11 5.9 0.004 0.009 25.626 25.626 scf_env_do_scf_inner_loop 108 6.5 0.004 0.009 23.056 23.060 dbcsr_multiply_generic 2286 12.5 0.094 0.097 17.391 17.504 qs_scf_new_mos 108 7.5 0.001 0.001 15.164 15.182 qs_scf_loop_do_ot 108 8.5 0.001 0.002 15.163 15.181 velocity_verlet 10 3.0 0.001 0.001 15.176 15.179 multiply_cannon 2286 13.5 0.196 0.203 13.946 14.601 ot_scf_mini 108 9.5 0.003 0.004 14.429 14.446 multiply_cannon_loop 2286 14.5 0.638 0.671 13.158 13.899 ot_mini 108 10.5 0.001 0.001 9.105 9.127 multiply_cannon_multrec 18288 15.5 1.937 2.992 7.147 7.669 qs_ot_get_derivative 108 11.5 0.001 0.001 7.581 7.601 rebuild_ks_matrix 119 8.3 0.000 0.000 6.382 6.402 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.382 6.402 dbcsr_mm_accdrv_process 38222 16.0 4.297 5.936 5.126 5.996 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.546 5.564 init_scf_run 11 5.9 0.001 0.001 4.769 4.770 scf_env_initial_rho_setup 11 6.9 0.041 0.055 4.768 4.770 mp_waitall_1 169478 16.3 3.064 4.368 3.064 4.368 sum_up_and_integrate 119 10.3 0.031 0.032 3.995 4.014 integrate_v_rspace 119 11.3 0.002 0.003 3.964 3.987 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.928 3.662 calculate_first_density_matrix 1 7.0 0.004 0.009 3.536 3.578 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.196 3.206 calculate_rho_elec 119 8.7 0.030 0.031 3.196 3.205 qs_ot_get_p 119 10.4 0.001 0.001 3.160 3.199 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.572 2.770 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.170 2.559 apply_single 119 13.6 0.000 0.000 2.170 2.559 init_scf_loop 11 6.9 0.001 0.003 2.547 2.550 rs_pw_transfer 974 11.9 0.009 0.009 2.291 2.409 calculate_dm_sparse 119 9.5 0.000 0.000 2.254 2.267 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.081 2.086 make_m2s 4572 13.5 0.045 0.046 1.804 1.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.952 1.954 grid_integrate_task_list 119 12.3 1.793 1.920 1.793 1.920 density_rs2pw 119 9.7 0.004 0.004 1.787 1.888 jit_kernel_multiply 8 15.5 0.777 1.875 0.777 1.875 make_images 4572 14.5 0.191 0.203 1.717 1.871 pw_transfer 1439 11.6 0.063 0.065 1.805 1.829 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.819 1.820 potential_pw2rs 119 12.3 0.007 0.008 1.788 1.811 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.793 1.800 prepare_preconditioner 11 7.9 0.000 0.000 1.752 1.754 make_preconditioner 11 8.9 0.001 0.002 1.752 1.754 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.715 1.739 acc_transpose_blocks 18288 15.5 0.077 0.080 1.615 1.716 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.607 1.691 mp_sum_l 7207 12.9 1.260 1.630 1.260 1.630 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.612 1.619 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.594 1.594 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.581 1.590 cp_fm_redistribute_end 50 14.0 1.182 1.573 1.183 1.574 multiply_cannon_sync_h2d 18288 15.5 1.373 1.563 1.373 1.563 cp_fm_diag_elpa_base 50 14.0 0.375 1.520 0.389 1.557 ot_diis_step 108 11.5 0.011 0.011 1.497 1.497 fft3d_ps 1201 14.6 0.507 0.526 1.402 1.427 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.382 1.407 grid_collocate_task_list 119 9.7 1.206 1.290 1.206 1.290 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 1.169 1.264 wfi_extrapolate 11 7.9 0.001 0.001 1.143 1.143 qs_energies_init_hamiltonians 11 5.9 0.008 0.016 1.133 1.141 jit_kernel_transpose 5 15.6 0.960 1.049 0.960 1.049 make_images_data 4572 15.5 0.045 0.048 0.815 0.978 parallel_gemm_fm 81 9.0 0.000 0.000 0.956 0.959 parallel_gemm_fm_cosma 81 10.0 0.956 0.959 0.956 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.920 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.889 0.914 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.694 0.877 make_basis_sm 11 9.8 0.000 0.000 0.821 0.823 mp_alltoall_z22v 1201 16.6 0.740 0.820 0.740 0.820 mp_alltoall_d11v 2130 13.8 0.689 0.786 0.689 0.786 cp_fm_cholesky_invert 11 10.9 0.739 0.743 0.739 0.743 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.739 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.665 0.726 mp_waitany 9880 13.7 0.605 0.704 0.605 0.704 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.123 0.594 0.696 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.363000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=482.000000, yerr=2.296242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 542.384128E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 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 0.082 0.203 37.128 37.132 qs_mol_dyn_low 1 2.0 0.003 0.004 35.986 36.079 qs_forces 11 3.9 0.002 0.002 35.819 35.828 qs_energies 11 4.9 0.026 0.064 34.118 34.134 scf_env_do_scf 11 5.9 0.001 0.002 27.645 27.646 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.092 24.093 dbcsr_multiply_generic 2286 12.5 0.103 0.119 18.317 18.449 velocity_verlet 10 3.0 0.001 0.002 17.547 17.553 qs_scf_new_mos 108 7.5 0.001 0.001 16.036 16.085 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.035 16.084 ot_scf_mini 108 9.5 0.003 0.003 15.101 15.150 multiply_cannon 2286 13.5 0.227 0.278 14.633 15.083 multiply_cannon_loop 2286 14.5 0.942 0.968 13.690 14.069 ot_mini 108 10.5 0.001 0.001 9.229 9.292 multiply_cannon_multrec 27432 15.5 2.356 3.033 8.878 9.218 dbcsr_mm_accdrv_process 47916 15.9 5.529 7.273 6.427 7.677 qs_ot_get_derivative 108 11.5 0.001 0.002 7.422 7.473 rebuild_ks_matrix 119 8.3 0.000 0.000 6.558 6.606 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.028 6.557 6.606 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.838 5.880 init_scf_run 11 5.9 0.000 0.001 4.389 4.390 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.389 4.390 sum_up_and_integrate 119 10.3 0.035 0.038 3.719 3.730 integrate_v_rspace 119 11.3 0.002 0.003 3.684 3.695 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.540 3.575 calculate_rho_elec 119 8.7 0.040 0.046 3.540 3.574 init_scf_loop 11 6.9 0.004 0.011 3.534 3.535 qs_ot_get_p 119 10.4 0.001 0.001 3.266 3.335 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.679 3.159 calculate_first_density_matrix 1 7.0 0.001 0.002 2.743 2.745 rs_pw_transfer 974 11.9 0.008 0.009 2.453 2.734 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.128 2.594 apply_single 119 13.6 0.000 0.000 2.128 2.594 prepare_preconditioner 11 7.9 0.000 0.000 2.559 2.566 make_preconditioner 11 8.9 0.008 0.023 2.559 2.566 mp_waitall_1 145218 16.4 1.962 2.558 1.962 2.558 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.160 2.492 make_m2s 4572 13.5 0.058 0.078 2.221 2.336 calculate_dm_sparse 119 9.5 0.000 0.000 2.258 2.310 make_images 4572 14.5 0.275 0.333 2.110 2.227 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.006 2.015 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.912 1.939 grid_integrate_task_list 119 12.3 1.828 1.912 1.828 1.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.899 1.901 jit_kernel_multiply 9 16.0 0.836 1.897 0.836 1.897 density_rs2pw 119 9.7 0.004 0.004 1.772 1.876 pw_transfer 1439 11.6 0.063 0.066 1.744 1.773 ot_diis_step 108 11.5 0.012 0.012 1.767 1.767 qs_energies_init_hamiltonians 11 5.9 0.017 0.111 1.731 1.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.714 1.727 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.691 1.692 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.655 1.687 mp_sum_l 7207 12.9 1.052 1.556 1.052 1.556 potential_pw2rs 119 12.3 0.008 0.009 1.502 1.508 acc_transpose_blocks 27432 15.5 0.112 0.114 1.481 1.500 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.443 1.453 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.444 1.446 cp_fm_redistribute_end 50 14.0 0.950 1.421 0.952 1.422 cp_fm_diag_elpa_base 50 14.0 0.447 1.361 0.467 1.403 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.355 1.390 grid_collocate_task_list 119 9.7 1.247 1.376 1.247 1.376 fft3d_ps 1201 14.6 0.536 0.589 1.337 1.359 wfi_extrapolate 11 7.9 0.001 0.001 1.295 1.295 mp_waitany 7280 13.7 0.995 1.294 0.995 1.294 rs_pw_transfer_RS2PW_140 130 11.5 0.108 0.113 1.001 1.283 qs_env_update_s_mstruct 11 6.9 0.005 0.028 1.105 1.268 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.707 1.200 mp_alltoall_d11v 2130 13.8 0.885 1.126 0.885 1.126 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.106 1.125 cp_fm_upper_to_full 72 13.5 0.804 1.112 0.804 1.112 dbcsr_complete_redistribute 329 12.2 0.128 0.158 0.788 1.059 multiply_cannon_sync_h2d 27432 15.5 0.987 1.037 0.987 1.037 make_images_data 4572 15.5 0.045 0.048 0.851 0.969 build_core_hamiltonian_matrix 11 6.9 0.036 0.166 0.516 0.903 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.732 0.894 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.792 0.870 acc_transpose_blocks_kernels 27432 16.5 0.268 0.277 0.850 0.867 copy_fm_to_dbcsr 176 11.2 0.001 0.003 0.585 0.849 mp_sum_d 4123 12.0 0.625 0.832 0.625 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.813 calculate_rho_core 11 7.9 0.057 0.167 0.651 0.802 cp_fm_cholesky_invert 11 10.9 0.760 0.763 0.760 0.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.132000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=514.636364, yerr=3.867645 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 596.430848E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 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 0.013 0.029 29.650 29.650 qs_mol_dyn_low 1 2.0 0.003 0.004 29.402 29.413 qs_forces 11 3.9 0.002 0.002 29.324 29.325 qs_energies 11 4.9 0.001 0.001 27.639 27.641 scf_env_do_scf 11 5.9 0.001 0.001 21.097 21.097 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.642 18.642 velocity_verlet 10 3.0 0.001 0.001 14.020 14.024 dbcsr_multiply_generic 2286 12.5 0.093 0.096 13.717 13.854 multiply_cannon 2286 13.5 0.230 0.241 11.203 11.670 qs_scf_new_mos 108 7.5 0.001 0.001 11.168 11.189 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.168 11.188 multiply_cannon_loop 2286 14.5 0.330 0.341 10.315 10.604 ot_scf_mini 108 9.5 0.002 0.002 10.491 10.514 multiply_cannon_multrec 9144 15.5 1.605 1.823 6.614 6.938 ot_mini 108 10.5 0.001 0.001 6.055 6.085 rebuild_ks_matrix 119 8.3 0.000 0.000 5.735 5.755 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.734 5.755 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.110 5.129 init_scf_run 11 5.9 0.000 0.001 5.118 5.118 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.118 5.118 dbcsr_mm_accdrv_process 12550 15.8 3.935 4.928 4.909 5.094 qs_ot_get_derivative 108 11.5 0.001 0.001 4.767 4.792 calculate_first_density_matrix 1 7.0 0.000 0.000 4.012 4.013 sum_up_and_integrate 119 10.3 0.037 0.041 3.508 3.513 integrate_v_rspace 119 11.3 0.002 0.003 3.471 3.476 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.161 3.171 calculate_rho_elec 119 8.7 0.059 0.061 3.161 3.170 qs_ot_get_p 119 10.4 0.001 0.001 2.668 2.715 jit_kernel_multiply 10 16.1 0.935 2.512 0.935 2.512 init_scf_loop 11 6.9 0.000 0.000 2.434 2.435 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.432 2.433 calculate_dm_sparse 119 9.5 0.000 0.000 2.258 2.291 mp_waitall_1 121218 16.5 1.807 2.284 1.807 2.284 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.171 2.179 grid_integrate_task_list 119 12.3 1.847 1.938 1.847 1.938 make_m2s 4572 13.5 0.035 0.036 1.702 1.862 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.769 1.771 make_images 4572 14.5 0.269 0.305 1.611 1.770 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.693 1.729 prepare_preconditioner 11 7.9 0.000 0.000 1.711 1.716 make_preconditioner 11 8.9 0.000 0.000 1.711 1.716 pw_transfer 1439 11.6 0.063 0.066 1.695 1.702 rs_pw_transfer 974 11.9 0.008 0.008 1.590 1.693 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.674 1.680 density_rs2pw 119 9.7 0.003 0.004 1.575 1.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.601 1.626 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.605 1.613 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.573 1.573 acc_transpose_blocks 9144 15.5 0.039 0.040 1.526 1.552 grid_collocate_task_list 119 9.7 1.272 1.355 1.272 1.355 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.309 1.318 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.318 1.318 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.283 1.306 potential_pw2rs 119 12.3 0.010 0.010 1.295 1.298 cp_fm_redistribute_end 50 14.0 0.653 1.297 0.654 1.297 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.175 1.294 apply_single 119 13.6 0.000 0.000 1.175 1.294 cp_fm_diag_elpa_base 50 14.0 0.599 1.229 0.643 1.285 fft3d_ps 1201 14.6 0.539 0.550 1.273 1.281 ot_diis_step 108 11.5 0.012 0.013 1.276 1.276 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.213 1.214 jit_kernel_transpose 5 15.6 1.166 1.189 1.166 1.189 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.120 1.131 wfi_extrapolate 11 7.9 0.001 0.001 1.061 1.061 make_images_data 4572 15.5 0.039 0.042 0.795 0.987 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.763 0.985 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.868 0.917 mp_alltoall_d11v 2130 13.8 0.761 0.896 0.761 0.896 cp_fm_cholesky_invert 11 10.9 0.869 0.872 0.869 0.872 multiply_cannon_sync_h2d 9144 15.5 0.712 0.802 0.712 0.802 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.465 0.780 make_basis_sm 11 9.8 0.000 0.000 0.756 0.757 parallel_gemm_fm 81 9.0 0.000 0.000 0.750 0.751 parallel_gemm_fm_cosma 81 10.0 0.750 0.751 0.750 0.751 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.745 0.747 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.729 0.737 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.669 0.719 mp_alltoall_z22v 1201 16.6 0.608 0.642 0.608 0.642 dbcsr_complete_redistribute 329 12.2 0.178 0.274 0.601 0.626 mp_allgather_i34 2286 14.5 0.222 0.624 0.222 0.624 multiply_cannon_metrocomm1 9144 15.5 0.022 0.022 0.392 0.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.650000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=560.909091, yerr=9.922010 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 723.562496E+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_ISendRecv 4998 189067. MP_Wait 8898 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 0.079 0.119 43.846 43.847 qs_mol_dyn_low 1 2.0 0.003 0.004 42.561 43.160 qs_forces 11 3.9 0.002 0.002 42.280 42.280 qs_energies 11 4.9 0.001 0.001 40.361 40.365 scf_env_do_scf 11 5.9 0.001 0.001 32.931 32.931 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.098 25.099 velocity_verlet 10 3.0 0.001 0.001 22.845 22.867 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.466 19.723 multiply_cannon 2286 13.5 0.301 0.309 15.140 16.522 qs_scf_new_mos 108 7.5 0.001 0.001 16.330 16.432 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.329 16.431 ot_scf_mini 108 9.5 0.002 0.002 15.253 15.355 multiply_cannon_loop 2286 14.5 0.344 0.351 13.890 15.258 multiply_cannon_multrec 9144 15.5 3.332 4.751 9.156 9.598 ot_mini 108 10.5 0.001 0.001 9.427 9.546 init_scf_loop 11 6.9 0.000 0.000 7.806 7.808 qs_ot_get_derivative 108 11.5 0.001 0.001 7.392 7.493 dbcsr_mm_accdrv_process 12550 15.8 4.577 6.351 5.702 7.325 prepare_preconditioner 11 7.9 0.000 0.000 6.894 6.907 make_preconditioner 11 8.9 0.000 0.000 6.894 6.907 rebuild_ks_matrix 119 8.3 0.000 0.000 6.756 6.900 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.756 6.899 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.410 6.784 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.098 6.229 init_scf_run 11 5.9 0.000 0.001 5.417 5.417 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.416 5.416 cp_fm_upper_to_full 72 14.2 3.208 4.635 3.208 4.635 calculate_first_density_matrix 1 7.0 0.000 0.000 4.025 4.027 mp_waitall_1 97218 16.6 2.763 3.840 2.763 3.840 sum_up_and_integrate 119 10.3 0.064 0.065 3.622 3.630 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.620 3.628 calculate_rho_elec 119 8.7 0.118 0.121 3.619 3.627 integrate_v_rspace 119 11.3 0.003 0.003 3.558 3.565 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.736 3.549 qs_ot_get_p 119 10.4 0.001 0.001 3.088 3.227 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.593 3.082 apply_single 119 13.6 0.000 0.000 2.593 3.081 dbcsr_complete_redistribute 329 12.2 0.291 0.298 2.024 2.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.734 2.734 calculate_dm_sparse 119 9.5 0.000 0.000 2.712 2.732 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.426 2.577 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.711 2.563 make_m2s 4572 13.5 0.039 0.039 2.377 2.546 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.470 2.514 make_images 4572 14.5 0.352 0.386 2.255 2.423 mp_sum_l 7207 12.9 1.433 2.385 1.433 2.385 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.479 2.324 mp_alltoall_i22 627 13.8 1.453 2.289 1.453 2.289 jit_kernel_multiply 8 15.3 1.098 2.113 1.098 2.113 grid_integrate_task_list 119 12.3 2.001 2.029 2.001 2.029 ot_diis_step 108 11.5 0.014 0.014 2.008 2.008 pw_transfer 1439 11.6 0.066 0.066 1.997 1.999 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.902 1.904 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.864 1.865 acc_transpose_blocks 9144 15.5 0.038 0.039 1.665 1.820 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.782 1.791 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.750 1.751 density_rs2pw 119 9.7 0.003 0.004 1.727 1.746 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.670 1.722 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.599 1.604 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.589 1.589 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 1.412 1.568 fft3d_ps 1201 14.6 0.566 0.576 1.535 1.537 jit_kernel_transpose 5 15.6 1.294 1.452 1.294 1.452 grid_collocate_task_list 119 9.7 1.443 1.451 1.443 1.451 cp_fm_cholesky_invert 11 10.9 1.439 1.442 1.439 1.442 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.183 1.435 rs_pw_transfer 974 11.9 0.009 0.009 1.328 1.366 wfi_extrapolate 11 7.9 0.001 0.001 1.334 1.334 make_images_data 4572 15.5 0.042 0.046 1.117 1.327 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.301 1.301 cp_fm_diag_elpa_base 50 14.0 1.159 1.211 1.300 1.300 mp_alltoall_d11v 2130 13.8 1.156 1.186 1.156 1.186 potential_pw2rs 119 12.3 0.014 0.014 1.173 1.176 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.145 1.165 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.087 1.102 multiply_cannon_sync_h2d 9144 15.5 1.038 1.041 1.038 1.041 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.978 1.014 qs_create_task_list 11 7.9 0.000 0.001 0.944 0.960 generate_qs_task_list 11 8.9 0.373 0.395 0.943 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.867 0.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.847000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=683.727273, yerr=8.655959 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 486.154240E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 2831633. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 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 83.017 83.018 qs_mol_dyn_low 1 2.0 0.003 0.004 82.675 82.703 qs_forces 11 3.9 0.003 0.003 82.383 82.383 qs_energies 11 4.9 0.001 0.002 79.593 79.611 scf_env_do_scf 11 5.9 0.000 0.001 70.134 70.136 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.530 64.530 dbcsr_multiply_generic 2055 12.4 0.108 0.111 52.615 52.914 qs_scf_new_mos 99 7.5 0.000 0.001 47.776 47.894 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.775 47.894 ot_scf_mini 99 9.5 0.002 0.002 45.360 45.490 multiply_cannon 2055 13.4 0.176 0.183 43.118 44.277 multiply_cannon_loop 2055 14.4 1.557 1.597 41.851 43.333 velocity_verlet 10 3.0 0.001 0.001 42.032 42.035 ot_mini 99 10.5 0.001 0.001 27.659 27.798 qs_ot_get_derivative 99 11.5 0.001 0.001 20.832 20.956 multiply_cannon_multrec 49320 15.4 12.100 13.011 17.395 18.055 rebuild_ks_matrix 110 8.3 0.000 0.001 14.248 14.391 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.247 14.391 mp_waitall_1 241148 16.1 11.559 12.690 11.559 12.690 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.534 12.658 multiply_cannon_sync_h2d 49320 15.4 9.913 10.637 9.913 10.637 qs_ot_get_p 110 10.4 0.001 0.001 9.186 9.298 multiply_cannon_metrocomm3 49320 15.4 0.078 0.085 6.859 8.242 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.148 7.685 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.159 7.639 apply_single 110 13.6 0.000 0.001 7.159 7.639 init_scf_run 11 5.9 0.000 0.001 7.440 7.441 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.440 7.441 sum_up_and_integrate 110 10.3 0.036 0.043 6.899 6.915 integrate_v_rspace 110 11.3 0.002 0.003 6.863 6.887 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.570 6.638 ot_diis_step 99 11.5 0.005 0.006 6.520 6.520 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.086 6.209 calculate_rho_elec 110 8.6 0.020 0.024 6.085 6.209 mp_sum_l 6514 12.8 5.212 6.068 5.212 6.068 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.007 6.033 dbcsr_mm_accdrv_process 87628 16.1 2.097 2.199 5.175 5.621 init_scf_loop 11 6.9 0.000 0.000 5.577 5.578 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.046 5.046 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.582 4.583 cp_fm_redistribute_end 48 14.0 3.985 4.556 3.990 4.557 cp_fm_diag_elpa_base 48 14.0 0.561 4.492 0.564 4.518 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.115 4.119 calculate_dm_sparse 110 9.5 0.000 0.001 3.977 4.075 wfi_extrapolate 11 7.9 0.001 0.001 4.002 4.002 make_m2s 4110 13.4 0.061 0.066 3.823 3.953 make_images 4110 14.4 0.176 0.190 3.727 3.859 rs_pw_transfer 902 11.9 0.011 0.013 3.606 3.841 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.706 3.755 multiply_cannon_metrocomm1 49320 15.4 0.058 0.061 2.740 3.739 prepare_preconditioner 11 7.9 0.000 0.000 3.418 3.440 make_preconditioner 11 8.9 0.000 0.000 3.418 3.440 grid_integrate_task_list 110 12.3 3.259 3.392 3.259 3.392 calculate_first_density_matrix 1 7.0 0.000 0.000 3.351 3.358 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.306 3.357 density_rs2pw 110 9.6 0.004 0.004 3.150 3.306 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.201 3.241 jit_kernel_multiply 13 15.9 2.791 2.941 2.791 2.941 pw_transfer 1331 11.6 0.053 0.063 2.769 2.827 acc_transpose_blocks 49320 15.4 0.225 0.237 2.317 2.762 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.682 2.743 mp_alltoall_d11v 2046 13.8 2.097 2.551 2.097 2.551 potential_pw2rs 110 12.3 0.006 0.007 2.512 2.531 fft_wrap_pw1pw2_140 451 13.1 0.169 0.190 2.277 2.336 fft3d_ps 1111 14.6 0.753 0.838 2.220 2.278 grid_collocate_task_list 110 9.6 2.086 2.232 2.086 2.232 mp_waitany 14300 13.8 1.814 2.223 1.814 2.223 mp_sum_d 3879 11.9 1.458 2.011 1.458 2.011 make_images_data 4110 15.4 0.042 0.045 1.745 1.883 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.852 1.873 cp_fm_cholesky_invert 11 10.9 1.864 1.868 1.864 1.868 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.832 1.860 hybrid_alltoall_any 4261 16.3 0.083 0.480 1.536 1.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.018000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=462.090909, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 576.339968E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3493132. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 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 0.055 0.097 70.883 70.884 qs_mol_dyn_low 1 2.0 0.003 0.004 70.251 70.339 qs_forces 11 3.9 0.002 0.002 69.368 69.369 qs_energies 11 4.9 0.003 0.008 66.081 66.085 scf_env_do_scf 11 5.9 0.001 0.002 56.141 56.144 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.516 48.518 dbcsr_multiply_generic 2055 12.4 0.165 0.171 38.760 38.947 velocity_verlet 10 3.0 0.001 0.001 36.433 36.449 multiply_cannon 2055 13.4 0.222 0.246 32.038 33.271 qs_scf_new_mos 99 7.5 0.001 0.001 32.798 32.969 qs_scf_loop_do_ot 99 8.5 0.001 0.002 32.797 32.968 multiply_cannon_loop 2055 14.4 0.926 0.947 30.816 31.774 ot_scf_mini 99 9.5 0.003 0.004 31.134 31.292 ot_mini 99 10.5 0.001 0.001 18.430 18.597 multiply_cannon_multrec 24660 15.4 7.621 9.286 14.579 16.173 rebuild_ks_matrix 110 8.3 0.000 0.001 13.525 13.626 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.525 13.625 qs_ot_get_derivative 99 11.5 0.001 0.001 12.583 12.742 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.920 12.005 mp_waitall_1 186928 16.3 7.895 10.663 7.895 10.663 multiply_cannon_metrocomm3 24660 15.4 0.068 0.070 5.255 8.302 multiply_cannon_sync_h2d 24660 15.4 7.014 7.997 7.014 7.997 init_scf_loop 11 6.9 0.001 0.004 7.591 7.595 init_scf_run 11 5.9 0.000 0.001 7.554 7.554 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.553 7.554 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.577 7.285 apply_single 110 13.6 0.000 0.001 6.577 7.284 dbcsr_mm_accdrv_process 52282 16.1 5.279 6.066 6.799 7.097 sum_up_and_integrate 110 10.3 0.052 0.060 6.457 6.469 integrate_v_rspace 110 11.3 0.002 0.002 6.405 6.418 qs_ot_get_p 110 10.4 0.001 0.001 6.136 6.331 ot_diis_step 99 11.5 0.010 0.010 5.799 5.799 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.666 5.678 calculate_rho_elec 110 8.6 0.039 0.047 5.666 5.677 prepare_preconditioner 11 7.9 0.000 0.000 5.543 5.569 make_preconditioner 11 8.9 0.000 0.001 5.543 5.568 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.695 5.428 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.103 5.266 make_m2s 4110 13.4 0.057 0.059 4.208 4.595 make_images 4110 14.4 0.401 0.447 4.099 4.484 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.223 4.245 calculate_first_density_matrix 1 7.0 0.001 0.003 4.005 4.007 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.910 3.911 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.787 3.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.499 3.531 wfi_extrapolate 11 7.9 0.001 0.001 3.462 3.463 pw_transfer 1331 11.6 0.065 0.069 3.249 3.381 calculate_dm_sparse 110 9.5 0.001 0.001 3.328 3.358 grid_integrate_task_list 110 12.3 3.165 3.283 3.165 3.283 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.143 3.274 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.182 3.267 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.257 3.259 cp_fm_redistribute_end 48 14.0 2.423 3.229 2.425 3.229 cp_fm_diag_elpa_base 48 14.0 0.769 3.120 0.801 3.205 density_rs2pw 110 9.6 0.004 0.004 2.968 3.119 rs_pw_transfer 902 11.9 0.012 0.013 2.850 3.028 fft_wrap_pw1pw2_140 451 13.1 0.201 0.219 2.665 2.796 make_images_data 4110 15.4 0.047 0.050 2.388 2.779 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.084 2.701 fft3d_ps 1111 14.6 1.074 1.268 2.500 2.612 cp_fm_cholesky_invert 11 10.9 2.556 2.563 2.556 2.563 mp_sum_l 6514 12.8 1.840 2.511 1.840 2.511 potential_pw2rs 110 12.3 0.008 0.009 2.329 2.341 grid_collocate_task_list 110 9.6 2.122 2.287 2.122 2.287 jit_kernel_multiply 8 16.6 1.166 2.124 1.166 2.124 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.961 1.981 mp_alltoall_d11v 2046 13.8 1.759 1.933 1.759 1.933 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.896 1.907 qs_energies_init_hamiltonians 11 5.9 0.005 0.009 1.874 1.875 acc_transpose_blocks 24660 15.4 0.111 0.116 1.839 1.865 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.770 1.786 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.771 1.768 mp_allgather_i34 2055 14.4 0.552 1.597 0.552 1.597 mp_irecv_dv 57340 16.2 0.647 1.556 0.647 1.556 dbcsr_complete_redistribute 325 12.2 0.302 0.397 1.248 1.511 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.474 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.884000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=543.727273, yerr=8.258229 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 645.980160E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 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 0.061 0.084 65.277 65.278 qs_mol_dyn_low 1 2.0 0.003 0.005 64.059 64.150 qs_forces 11 3.9 0.002 0.002 62.974 62.987 qs_energies 11 4.9 0.002 0.005 58.852 58.867 scf_env_do_scf 11 5.9 0.001 0.002 49.633 49.633 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.211 41.212 velocity_verlet 10 3.0 0.001 0.001 32.939 32.972 dbcsr_multiply_generic 2055 12.4 0.109 0.111 31.369 31.577 qs_scf_new_mos 99 7.5 0.001 0.001 26.678 26.784 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.678 26.784 multiply_cannon 2055 13.4 0.212 0.225 24.629 25.830 ot_scf_mini 99 9.5 0.003 0.003 25.466 25.587 multiply_cannon_loop 2055 14.4 0.616 0.638 23.424 24.976 ot_mini 99 10.5 0.001 0.001 15.531 15.651 rebuild_ks_matrix 110 8.3 0.000 0.000 13.070 13.202 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.070 13.202 multiply_cannon_multrec 16440 15.4 4.002 5.159 11.065 12.115 mp_waitall_1 146766 16.3 7.824 11.718 7.824 11.718 qs_ot_get_derivative 99 11.5 0.001 0.001 11.150 11.270 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.657 10.779 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.971 8.534 init_scf_loop 11 6.9 0.000 0.002 8.388 8.391 sum_up_and_integrate 110 10.3 0.060 0.061 7.183 7.323 integrate_v_rspace 110 11.3 0.002 0.002 7.123 7.262 dbcsr_mm_accdrv_process 34862 16.1 5.842 6.697 6.916 7.141 prepare_preconditioner 11 7.9 0.000 0.000 6.619 6.634 make_preconditioner 11 8.9 0.000 0.001 6.619 6.634 init_scf_run 11 5.9 0.000 0.001 6.620 6.620 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.619 6.620 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.950 6.287 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.492 6.146 apply_single 110 13.6 0.000 0.000 5.492 6.146 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.610 5.618 calculate_rho_elec 110 8.6 0.058 0.058 5.610 5.618 qs_ot_get_p 110 10.4 0.001 0.001 5.144 5.274 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.001 5.231 make_m2s 4110 13.4 0.050 0.052 4.137 4.548 make_images 4110 14.4 0.395 0.511 4.021 4.432 multiply_cannon_sync_h2d 16440 15.4 3.715 4.415 3.715 4.415 ot_diis_step 99 11.5 0.010 0.011 4.353 4.353 pw_transfer 1331 11.6 0.064 0.070 4.018 4.146 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.912 4.042 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 3.482 3.614 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.587 3.589 calculate_first_density_matrix 1 7.0 0.000 0.001 3.486 3.488 grid_integrate_task_list 110 12.3 3.181 3.382 3.181 3.382 fft3d_ps 1111 14.6 1.066 1.077 3.227 3.360 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.255 3.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.222 3.224 potential_pw2rs 110 12.3 0.011 0.011 3.009 3.143 density_rs2pw 110 9.6 0.004 0.004 2.861 3.119 calculate_dm_sparse 110 9.5 0.001 0.001 3.066 3.098 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.897 2.935 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.186 2.933 rs_pw_transfer 902 11.9 0.010 0.011 2.677 2.927 make_images_data 4110 15.4 0.043 0.047 2.411 2.919 wfi_extrapolate 11 7.9 0.001 0.001 2.906 2.906 mp_sum_l 6514 12.8 2.000 2.854 2.000 2.854 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.754 2.756 cp_fm_redistribute_end 48 14.0 1.712 2.734 1.714 2.734 cp_fm_diag_elpa_base 48 14.0 0.959 2.609 1.016 2.712 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.612 2.625 cp_fm_cholesky_invert 11 10.9 2.554 2.559 2.554 2.559 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.282 2.352 grid_collocate_task_list 110 9.6 2.195 2.346 2.195 2.346 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.841 2.257 mp_irecv_dv 48980 15.7 0.772 2.139 0.772 2.139 mp_alltoall_d11v 2046 13.8 1.809 2.131 1.809 2.131 mp_alltoall_z22v 1111 16.6 1.948 2.096 1.948 2.096 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.035 2.047 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.961 1.974 jit_kernel_multiply 6 16.7 0.683 1.934 0.683 1.934 dbcsr_complete_redistribute 325 12.2 0.329 0.358 1.413 1.850 cp_fm_upper_to_full 70 13.6 1.366 1.825 1.366 1.825 acc_transpose_blocks 16440 15.4 0.073 0.076 1.602 1.759 yz_to_x 561 15.1 0.099 0.102 1.522 1.657 mp_allgather_i34 2055 14.4 0.527 1.580 0.527 1.580 cp_fm_cholesky_decompose 22 10.9 1.555 1.577 1.555 1.577 mp_waitany 17072 13.8 1.245 1.514 1.245 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.358 1.454 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.974 1.411 acc_transpose_blocks_kernels 16440 16.4 0.208 0.211 1.174 1.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.278000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=612.181818, yerr=8.850456 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 718.344192E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 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 0.038 0.092 67.852 67.853 qs_mol_dyn_low 1 2.0 0.003 0.004 67.454 67.479 qs_forces 11 3.9 0.002 0.002 66.661 66.661 qs_energies 11 4.9 0.002 0.004 63.329 63.333 scf_env_do_scf 11 5.9 0.001 0.001 53.195 53.199 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.593 41.594 velocity_verlet 10 3.0 0.005 0.011 36.961 36.993 dbcsr_multiply_generic 2055 12.4 0.115 0.118 31.444 31.654 qs_scf_new_mos 99 7.5 0.001 0.001 27.439 27.544 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.438 27.543 ot_scf_mini 99 9.5 0.003 0.003 25.824 25.929 multiply_cannon 2055 13.4 0.244 0.260 24.606 25.347 multiply_cannon_loop 2055 14.4 0.887 0.905 23.194 23.773 ot_mini 99 10.5 0.001 0.001 15.369 15.493 multiply_cannon_multrec 24660 15.4 4.313 7.322 13.822 15.205 rebuild_ks_matrix 110 8.3 0.000 0.000 11.693 11.806 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 11.692 11.806 init_scf_loop 11 6.9 0.001 0.003 11.563 11.563 qs_ot_get_derivative 99 11.5 0.096 0.382 11.256 11.365 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.352 10.454 dbcsr_mm_accdrv_process 52304 16.0 7.693 9.571 9.362 10.365 prepare_preconditioner 11 7.9 0.000 0.000 9.842 9.859 make_preconditioner 11 8.9 0.000 0.001 9.842 9.859 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.093 9.536 init_scf_run 11 5.9 0.000 0.001 6.931 6.931 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.930 6.931 mp_waitall_1 126806 16.4 4.835 6.920 4.835 6.920 sum_up_and_integrate 110 10.3 0.068 0.071 6.166 6.177 integrate_v_rspace 110 11.3 0.002 0.003 6.098 6.110 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.704 5.715 calculate_rho_elec 110 8.6 0.077 0.081 5.703 5.714 qs_ot_get_p 110 10.4 0.001 0.001 5.500 5.654 make_m2s 4110 13.4 0.060 0.063 5.210 5.593 make_images 4110 14.4 0.577 0.704 5.069 5.449 cp_fm_upper_to_full 70 13.8 3.228 4.598 3.228 4.598 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.037 4.135 apply_single 110 13.6 0.000 0.000 4.037 4.135 ot_diis_step 99 11.5 0.011 0.012 4.076 4.076 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.787 3.881 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.753 3.767 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.768 3.767 dbcsr_complete_redistribute 325 12.2 0.421 0.471 2.622 3.739 calculate_first_density_matrix 1 7.0 0.000 0.002 3.673 3.676 calculate_dm_sparse 110 9.5 0.001 0.001 3.507 3.536 grid_integrate_task_list 110 12.3 3.269 3.426 3.269 3.426 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.416 3.416 multiply_cannon_sync_h2d 24660 15.4 3.195 3.332 3.195 3.332 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.280 3.280 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.096 3.201 rs_pw_transfer 902 11.9 0.010 0.011 2.880 3.151 pw_transfer 1331 11.6 0.064 0.071 3.116 3.138 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.064 3.093 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.985 3.041 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.011 3.036 density_rs2pw 110 9.6 0.004 0.004 2.866 3.026 hybrid_alltoall_any 4261 16.3 0.120 0.461 2.254 2.993 make_images_data 4110 15.4 0.046 0.049 2.586 2.953 jit_kernel_multiply 10 15.7 1.335 2.913 1.335 2.913 wfi_extrapolate 11 7.9 0.001 0.001 2.912 2.912 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.740 2.825 mp_alltoall_i22 605 13.7 1.624 2.784 1.624 2.784 cp_fm_cholesky_invert 11 10.9 2.736 2.744 2.736 2.744 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.738 2.740 cp_fm_redistribute_end 48 14.0 1.362 2.712 1.363 2.713 cp_fm_diag_elpa_base 48 14.0 1.270 2.587 1.347 2.696 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.647 2.648 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 2.554 2.585 grid_collocate_task_list 110 9.6 2.255 2.493 2.255 2.493 fft3d_ps 1111 14.6 1.061 1.105 2.315 2.334 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.126 2.141 acc_transpose_blocks 24660 15.4 0.105 0.107 1.944 2.050 potential_pw2rs 110 12.3 0.012 0.013 1.996 2.003 mp_waitany 13376 13.8 1.599 1.916 1.599 1.916 mp_alltoall_d11v 2046 13.8 1.700 1.896 1.700 1.896 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.687 1.721 cp_fm_cholesky_decompose 22 10.9 1.650 1.704 1.650 1.704 qs_env_update_s_mstruct 11 6.9 0.005 0.014 1.607 1.702 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.699 mp_sum_l 6514 12.8 1.003 1.620 1.003 1.620 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.174 1.358 1.620 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.828 1.604 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.586 1.597 mp_irecv_dv 62702 16.1 0.731 1.525 0.731 1.525 acc_transpose_blocks_kernels 24660 16.4 0.303 0.309 1.299 1.397 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.853000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=679.636364, yerr=9.603718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 810.606592E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 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 0.021 0.047 58.781 58.782 qs_mol_dyn_low 1 2.0 0.006 0.056 57.838 58.116 qs_forces 11 3.9 0.043 0.056 57.090 57.094 qs_energies 11 4.9 0.009 0.009 53.442 53.458 scf_env_do_scf 11 5.9 0.001 0.001 43.370 43.371 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.568 35.569 velocity_verlet 10 3.0 0.001 0.001 31.346 31.390 dbcsr_multiply_generic 2055 12.4 0.105 0.106 25.413 25.581 qs_scf_new_mos 99 7.5 0.001 0.001 21.321 21.393 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.321 21.393 multiply_cannon 2055 13.4 0.248 0.264 19.919 21.197 ot_scf_mini 99 9.5 0.002 0.002 20.090 20.151 multiply_cannon_loop 2055 14.4 0.321 0.332 18.588 19.022 rebuild_ks_matrix 110 8.3 0.000 0.000 11.427 11.472 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.427 11.472 ot_mini 99 10.5 0.001 0.001 11.276 11.332 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.180 10.220 multiply_cannon_multrec 8220 15.4 3.354 5.017 8.789 10.043 mp_waitall_1 106626 16.5 6.303 7.890 6.303 7.890 init_scf_loop 11 6.9 0.000 0.000 7.755 7.756 qs_ot_get_derivative 99 11.5 0.001 0.001 7.390 7.450 init_scf_run 11 5.9 0.000 0.001 6.835 6.835 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.834 6.834 dbcsr_mm_accdrv_process 17442 15.9 3.469 4.631 5.305 6.401 sum_up_and_integrate 110 10.3 0.079 0.081 6.158 6.169 prepare_preconditioner 11 7.9 0.000 0.000 6.126 6.131 make_preconditioner 11 8.9 0.000 0.000 6.126 6.131 integrate_v_rspace 110 11.3 0.003 0.003 6.079 6.089 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.712 5.784 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.540 5.554 calculate_rho_elec 110 8.6 0.115 0.115 5.540 5.553 qs_ot_get_p 110 10.4 0.001 0.001 4.907 4.955 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.194 4.603 make_m2s 4110 13.4 0.039 0.041 4.131 4.381 make_images 4110 14.4 0.637 0.691 4.000 4.248 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.821 4.085 apply_single 110 13.6 0.000 0.000 3.821 4.084 calculate_first_density_matrix 1 7.0 0.000 0.000 4.055 4.057 ot_diis_step 99 11.5 0.012 0.012 3.860 3.861 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.546 3.550 grid_integrate_task_list 110 12.3 3.366 3.530 3.366 3.530 calculate_dm_sparse 110 9.5 0.001 0.001 3.371 3.418 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.383 3.384 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.242 3.243 pw_transfer 1331 11.6 0.064 0.070 3.219 3.235 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.113 3.132 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.025 3.040 multiply_cannon_sync_h2d 8220 15.4 2.907 3.026 2.907 3.026 cp_fm_cholesky_invert 11 10.9 2.939 2.943 2.939 2.943 make_images_data 4110 15.4 0.038 0.044 2.349 2.737 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.721 2.721 density_rs2pw 110 9.6 0.004 0.004 2.600 2.705 fft_wrap_pw1pw2_140 451 13.1 0.211 0.214 2.675 2.701 cp_fm_redistribute_end 48 14.0 0.687 2.694 0.692 2.694 wfi_extrapolate 11 7.9 0.001 0.001 2.692 2.692 cp_fm_diag_elpa_base 48 14.0 1.823 2.506 1.995 2.664 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.654 2.664 hybrid_alltoall_any 4261 16.3 0.200 0.859 2.283 2.628 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.458 2.554 grid_collocate_task_list 110 9.6 2.327 2.470 2.327 2.470 fft3d_ps 1111 14.6 1.115 1.177 2.373 2.394 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.341 2.353 rs_pw_transfer 902 11.9 0.010 0.010 2.146 2.265 acc_transpose_blocks 8220 15.4 0.036 0.038 1.967 2.246 acc_transpose_blocks_kernels 8220 16.4 0.109 0.112 1.733 2.009 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.991 jit_kernel_multiply 7 15.8 1.524 1.974 1.524 1.974 potential_pw2rs 110 12.3 0.015 0.015 1.907 1.911 jit_kernel_transpose 5 15.6 1.624 1.898 1.624 1.898 mp_alltoall_d11v 2046 13.8 1.632 1.865 1.632 1.865 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.788 1.818 cp_fm_cholesky_decompose 22 10.9 1.726 1.742 1.726 1.742 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.501 1.607 mp_allgather_i34 2055 14.4 0.521 1.589 0.521 1.589 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.578 1.584 dbcsr_complete_redistribute 325 12.2 0.562 0.589 1.454 1.539 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.830 1.444 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.479 1.354 qs_create_task_list 11 7.9 0.001 0.001 1.220 1.317 generate_qs_task_list 11 8.9 0.379 0.446 1.219 1.316 mp_irecv_dv 24056 15.7 0.455 1.315 0.455 1.315 mp_waitany 9240 13.8 1.060 1.193 1.060 1.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.782000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=765.363636, yerr=10.191554 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 1.361867E+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_ISendRecv 4620 360267. MP_Wait 7524 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 0.049 0.074 91.146 91.146 qs_mol_dyn_low 1 2.0 0.003 0.004 88.483 90.044 qs_forces 11 3.9 0.002 0.002 87.565 87.566 qs_energies 11 4.9 0.001 0.001 83.534 83.536 scf_env_do_scf 11 5.9 0.001 0.001 72.344 72.344 velocity_verlet 10 3.0 0.001 0.001 54.997 55.082 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.263 44.265 dbcsr_multiply_generic 2055 12.4 0.119 0.121 31.411 31.457 init_scf_loop 11 6.9 0.000 0.000 28.008 28.011 qs_scf_new_mos 99 7.5 0.001 0.001 27.724 27.753 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.723 27.752 prepare_preconditioner 11 7.9 0.000 0.000 26.058 26.063 make_preconditioner 11 8.9 0.000 0.000 26.058 26.063 ot_scf_mini 99 9.5 0.002 0.002 25.960 25.984 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.342 25.524 multiply_cannon 2055 13.4 0.348 0.383 23.943 24.672 multiply_cannon_loop 2055 14.4 0.341 0.344 22.165 22.953 cp_fm_upper_to_full 70 14.2 12.605 18.025 12.605 18.025 ot_mini 99 10.5 0.001 0.001 15.189 15.207 rebuild_ks_matrix 110 8.3 0.000 0.001 13.138 13.179 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.138 13.179 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.950 11.987 multiply_cannon_multrec 8220 15.4 4.746 4.934 10.797 11.002 qs_ot_get_derivative 99 11.5 0.001 0.001 10.597 10.621 dbcsr_complete_redistribute 325 12.2 1.030 1.066 7.295 10.448 mp_waitall_1 87304 16.6 8.379 9.519 8.379 9.519 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.261 9.413 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.700 8.829 mp_alltoall_i22 605 13.7 5.312 8.466 5.312 8.466 init_scf_run 11 5.9 0.000 0.001 6.915 6.915 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.914 6.914 sum_up_and_integrate 110 10.3 0.151 0.153 6.600 6.615 integrate_v_rspace 110 11.3 0.004 0.004 6.449 6.464 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.415 6.449 calculate_rho_elec 110 8.6 0.227 0.227 6.415 6.448 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.252 6.206 dbcsr_mm_accdrv_process 11614 15.7 3.529 3.929 5.909 6.176 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.072 5.900 apply_single 110 13.6 0.000 0.000 5.072 5.900 make_m2s 4110 13.4 0.044 0.045 5.373 5.877 make_images 4110 14.4 0.876 0.920 5.184 5.688 qs_ot_get_p 110 10.4 0.001 0.001 5.548 5.572 cp_fm_cholesky_invert 11 10.9 5.222 5.227 5.222 5.227 ot_diis_step 99 11.5 0.015 0.015 4.567 4.567 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.696 4.516 calculate_dm_sparse 110 9.5 0.001 0.001 4.063 4.082 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.022 4.028 multiply_cannon_sync_h2d 8220 15.4 3.949 3.955 3.949 3.955 hybrid_alltoall_any 4261 16.3 0.256 0.554 3.014 3.779 pw_transfer 1331 11.6 0.072 0.073 3.724 3.730 grid_integrate_task_list 110 12.3 3.661 3.720 3.661 3.720 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.682 3.683 make_images_data 4110 15.4 0.041 0.044 3.016 3.657 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.609 3.614 calculate_first_density_matrix 1 7.0 0.000 0.000 3.589 3.591 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.584 3.585 wfi_extrapolate 11 7.9 0.001 0.001 3.220 3.220 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.153 3.157 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.134 3.141 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.000 3.000 cp_fm_diag_elpa_base 48 14.0 2.477 2.662 2.998 2.998 density_rs2pw 110 9.6 0.004 0.004 2.921 2.939 fft3d_ps 1111 14.6 1.265 1.277 2.818 2.824 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.646 2.655 grid_collocate_task_list 110 9.6 2.630 2.648 2.630 2.648 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.592 2.610 jit_kernel_multiply 10 15.3 2.177 2.532 2.177 2.532 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.364 2.376 md_write_output 11 3.9 0.240 1.919 0.285 2.265 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.179 2.227 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.159 rs_pw_transfer 902 11.9 0.010 0.010 2.080 2.109 mp_alltoall_d11v 2046 13.8 2.026 2.075 2.026 2.075 potential_pw2rs 110 12.3 0.021 0.021 2.035 2.042 md_output 10 3.0 0.000 0.000 0.254 2.007 cp_fm_cholesky_decompose 22 10.9 1.971 1.997 1.971 1.997 mp_sum_l 6514 12.8 1.225 1.977 1.225 1.977 qs_create_task_list 11 7.9 0.000 0.001 1.896 1.939 generate_qs_task_list 11 8.9 0.741 0.792 1.895 1.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.146000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1213.363636, yerr=59.556349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 614.445056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 3052320. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 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 0.039 0.069 206.022 206.023 qs_mol_dyn_low 1 2.0 0.003 0.005 205.066 205.099 qs_forces 11 3.9 0.003 0.003 204.941 204.942 qs_energies 11 4.9 0.002 0.005 199.362 199.384 scf_env_do_scf 11 5.9 0.001 0.001 181.700 181.703 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 160.926 160.929 dbcsr_multiply_generic 2507 12.6 0.177 0.184 125.669 126.186 qs_scf_new_mos 117 7.6 0.001 0.001 122.095 122.329 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.094 122.329 velocity_verlet 10 3.0 0.004 0.028 121.914 121.915 ot_scf_mini 117 9.6 0.003 0.005 115.544 115.756 multiply_cannon 2507 13.6 0.235 0.240 102.167 103.878 multiply_cannon_loop 2507 14.6 2.128 2.196 100.008 101.686 ot_mini 117 10.6 0.001 0.001 66.557 66.753 multiply_cannon_multrec 60168 15.6 33.043 35.024 42.076 43.618 qs_ot_get_derivative 117 11.6 0.001 0.004 41.937 42.137 rebuild_ks_matrix 128 8.3 0.001 0.001 33.317 33.721 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.316 33.720 mp_waitall_1 291448 16.2 29.244 32.034 29.244 32.034 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.905 30.230 multiply_cannon_sync_h2d 60168 15.6 27.215 28.695 27.215 28.695 qs_ot_get_p 128 10.4 0.001 0.002 27.168 27.431 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.200 24.999 apply_single 128 13.6 0.001 0.001 24.199 24.998 ot_diis_step 117 11.6 0.007 0.008 24.237 24.238 init_scf_loop 11 6.9 0.002 0.008 20.691 20.700 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.513 20.600 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.021 19.178 multiply_cannon_metrocomm3 60168 15.6 0.116 0.122 16.018 18.256 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.948 17.949 prepare_preconditioner 11 7.9 0.000 0.000 16.050 16.097 make_preconditioner 11 8.9 0.000 0.001 16.050 16.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.280 15.445 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.142 15.148 cp_fm_redistribute_end 83 14.4 11.932 15.070 11.945 15.076 cp_fm_diag_elpa_base 83 14.4 3.083 14.846 3.116 14.970 make_m2s 5014 13.6 0.104 0.110 14.105 14.574 make_images 5014 14.6 0.398 0.417 13.924 14.403 sum_up_and_integrate 128 10.3 0.089 0.107 14.021 14.050 integrate_v_rspace 128 11.3 0.003 0.004 13.932 13.963 init_scf_run 11 5.9 0.000 0.001 13.026 13.027 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.026 13.027 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.551 12.636 calculate_rho_elec 128 8.7 0.045 0.063 12.551 12.635 mp_sum_l 7870 13.0 8.168 9.535 8.168 9.535 dbcsr_mm_accdrv_process 124484 16.2 3.324 3.476 8.601 9.203 cp_fm_cholesky_invert 11 10.9 9.180 9.188 9.180 9.188 wfi_extrapolate 11 7.9 0.001 0.001 9.030 9.030 calculate_dm_sparse 128 9.5 0.001 0.001 8.771 8.851 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.543 8.741 multiply_cannon_metrocomm1 60168 15.6 0.086 0.089 6.302 8.049 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.765 7.877 make_images_data 5014 15.6 0.067 0.074 6.892 7.821 grid_integrate_task_list 128 12.3 7.105 7.575 7.105 7.575 hybrid_alltoall_any 5200 16.5 0.293 2.272 6.045 7.361 pw_transfer 1547 11.6 0.074 0.106 6.869 7.139 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.007 7.017 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.667 6.914 density_rs2pw 128 9.7 0.006 0.007 6.350 6.867 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.150 6.280 rs_pw_transfer 1046 11.9 0.016 0.018 5.571 6.197 fft_wrap_pw1pw2_140 523 13.2 0.447 0.508 5.783 5.965 fft3d_ps 1291 14.7 2.091 2.564 5.458 5.644 mp_alltoall_d11v 2415 14.1 4.237 5.594 4.237 5.594 grid_collocate_task_list 128 9.7 4.729 5.138 4.729 5.138 cp_fm_cholesky_decompose 22 10.9 4.666 4.681 4.666 4.681 potential_pw2rs 128 12.3 0.009 0.010 4.484 4.528 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.023000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=581.272727, yerr=7.398905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 814.297088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57173. MP_Allreduce 11257 946. MP_Sync 168 MP_Alltoall 1955 6058387. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.110 0.183 192.929 192.930 qs_mol_dyn_low 1 2.0 0.003 0.004 191.723 191.748 qs_forces 11 3.9 0.003 0.003 189.814 189.815 qs_energies 11 4.9 0.008 0.029 183.120 183.130 scf_env_do_scf 11 5.9 0.001 0.001 166.247 166.257 scf_env_do_scf_inner_loop 116 6.6 0.009 0.026 133.243 133.245 velocity_verlet 10 3.0 0.007 0.032 119.596 119.635 dbcsr_multiply_generic 2485 12.5 0.190 0.221 98.245 99.476 qs_scf_new_mos 116 7.6 0.001 0.001 94.854 95.338 qs_scf_loop_do_ot 116 8.6 0.001 0.002 94.853 95.337 ot_scf_mini 116 9.6 0.004 0.005 90.093 90.709 multiply_cannon 2485 13.5 0.474 0.525 78.221 82.398 multiply_cannon_loop 2485 14.5 1.240 1.276 74.940 77.289 ot_mini 116 10.6 0.001 0.001 50.600 51.146 mp_waitall_1 224796 16.4 25.165 38.136 25.165 38.136 multiply_cannon_multrec 29820 15.5 21.944 26.748 32.049 37.075 rebuild_ks_matrix 127 8.3 0.001 0.001 32.433 32.919 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 32.432 32.919 init_scf_loop 11 6.9 0.002 0.014 32.913 32.914 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.199 29.660 qs_ot_get_derivative 116 11.6 0.003 0.012 28.952 29.561 multiply_cannon_metrocomm3 29820 15.5 0.091 0.096 16.152 28.834 prepare_preconditioner 11 7.9 0.000 0.000 28.519 28.571 make_preconditioner 11 8.9 0.000 0.001 28.519 28.571 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.224 27.764 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.087 23.358 apply_single 127 13.6 0.001 0.001 22.087 23.358 qs_ot_get_p 127 10.4 0.001 0.001 21.488 22.079 multiply_cannon_sync_h2d 29820 15.5 19.123 21.558 19.123 21.558 ot_diis_step 116 11.6 0.014 0.015 21.478 21.480 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.767 16.799 cp_fm_cholesky_invert 11 10.9 16.724 16.736 16.724 16.736 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.649 15.651 make_m2s 4970 13.5 0.088 0.094 14.101 15.620 make_images 4970 14.5 1.154 1.344 13.892 15.411 sum_up_and_integrate 127 10.3 0.115 0.134 14.169 14.204 integrate_v_rspace 127 11.3 0.003 0.003 14.053 14.094 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.775 12.815 calculate_rho_elec 127 8.7 0.087 0.104 12.775 12.815 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.614 12.620 cp_fm_redistribute_end 82 14.4 7.332 12.548 7.343 12.550 cp_fm_diag_elpa_base 82 14.4 4.975 12.103 5.190 12.468 init_scf_run 11 5.9 0.001 0.003 11.952 11.954 scf_env_initial_rho_setup 11 6.9 0.001 0.004 11.951 11.954 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.795 11.220 multiply_cannon_metrocomm4 27335 15.5 0.096 0.112 3.719 10.586 make_images_data 4970 15.5 0.064 0.071 8.570 10.450 mp_irecv_dv 68888 16.3 3.526 10.206 3.526 10.206 dbcsr_mm_accdrv_process 61726 16.2 4.780 5.556 9.567 10.177 hybrid_alltoall_any 5155 16.4 0.341 1.485 7.311 9.693 wfi_extrapolate 11 7.9 0.001 0.001 8.297 8.297 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.112 8.146 pw_transfer 1535 11.6 0.084 0.106 7.641 7.694 grid_integrate_task_list 127 12.3 7.113 7.500 7.113 7.500 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.419 7.472 cp_fm_cholesky_decompose 22 10.9 6.956 7.047 6.956 7.047 density_rs2pw 127 9.7 0.006 0.006 6.608 6.961 calculate_dm_sparse 127 9.5 0.001 0.001 6.750 6.896 fft_wrap_pw1pw2_140 519 13.2 0.464 0.507 6.509 6.567 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.302 6.314 mp_sum_l 7804 13.0 4.199 5.994 4.199 5.994 rs_pw_transfer 1038 11.9 0.014 0.016 5.489 5.869 fft3d_ps 1281 14.7 2.756 2.904 5.804 5.847 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.410 5.524 grid_collocate_task_list 127 9.7 4.871 5.307 4.871 5.307 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.230 5.290 mp_allgather_i34 2485 14.5 1.875 5.006 1.875 5.006 potential_pw2rs 127 12.3 0.015 0.017 4.704 4.727 mp_alltoall_d11v 2401 14.1 4.073 4.599 4.073 4.599 dbcsr_complete_redistribute 393 12.7 0.789 0.921 3.134 3.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.930000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=776.636364, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 949.813248E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56898. MP_Allreduce 11306 983. MP_Sync 170 MP_Alltoall 1724 9394172. MP_ISendRecv 15996 75008. MP_Wait 29984 MP_comm_split 83 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.056 179.467 179.469 qs_mol_dyn_low 1 2.0 0.003 0.003 178.698 178.710 qs_forces 11 3.9 0.011 0.022 178.572 178.575 qs_energies 11 4.9 0.002 0.005 172.076 172.095 scf_env_do_scf 11 5.9 0.001 0.001 155.429 155.430 scf_env_do_scf_inner_loop 118 6.6 0.004 0.017 120.332 120.333 velocity_verlet 10 3.0 0.001 0.001 113.529 113.532 dbcsr_multiply_generic 2527 12.6 0.182 0.186 84.229 85.527 qs_scf_new_mos 118 7.6 0.001 0.001 83.346 83.665 qs_scf_loop_do_ot 118 8.6 0.001 0.002 83.345 83.664 ot_scf_mini 118 9.6 0.004 0.004 79.178 79.576 multiply_cannon 2527 13.6 0.503 0.531 63.791 67.989 multiply_cannon_loop 2527 14.6 0.863 0.889 60.522 63.513 ot_mini 118 10.6 0.001 0.001 43.919 44.320 mp_waitall_1 179876 16.5 25.985 35.707 25.985 35.707 init_scf_loop 11 6.9 0.000 0.001 34.995 34.996 prepare_preconditioner 11 7.9 0.000 0.000 30.998 31.055 make_preconditioner 11 8.9 0.000 0.001 30.998 31.055 rebuild_ks_matrix 129 8.3 0.001 0.001 30.290 30.743 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.289 30.743 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.635 30.008 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.383 27.803 multiply_cannon_multrec 20216 15.6 13.561 16.632 23.100 26.020 multiply_cannon_metrocomm3 20216 15.6 0.060 0.064 15.615 25.274 qs_ot_get_derivative 118 11.6 0.001 0.002 24.023 24.421 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.281 21.349 apply_single 129 13.6 0.001 0.001 20.281 21.349 qs_ot_get_p 129 10.4 0.001 0.001 20.532 20.985 ot_diis_step 118 11.6 0.018 0.018 19.794 19.795 make_m2s 5054 13.6 0.081 0.087 15.107 16.318 qs_ot_p2m_diag 83 11.4 0.266 0.274 16.242 16.253 multiply_cannon_sync_h2d 20216 15.6 14.460 16.244 14.460 16.244 make_images 5054 14.6 1.195 1.275 14.872 16.084 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.245 15.246 cp_fm_cholesky_invert 11 10.9 14.451 14.460 14.451 14.460 sum_up_and_integrate 129 10.3 0.135 0.147 14.207 14.231 integrate_v_rspace 129 11.3 0.003 0.004 14.071 14.100 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.096 13.130 calculate_rho_elec 129 8.7 0.133 0.148 13.095 13.130 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.086 12.087 cp_fm_redistribute_end 83 14.4 4.537 12.022 4.551 12.024 cp_fm_diag_elpa_base 83 14.4 7.044 11.467 7.452 11.938 init_scf_run 11 5.9 0.000 0.001 11.543 11.544 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.543 11.543 make_images_data 5054 15.6 0.060 0.067 9.387 11.146 hybrid_alltoall_any 5240 16.5 0.433 1.970 8.077 10.269 multiply_cannon_metrocomm4 17689 15.6 0.060 0.069 3.451 9.270 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.945 9.261 dbcsr_mm_accdrv_process 41830 16.2 5.125 6.238 8.989 9.234 mp_irecv_dv 50625 16.2 3.331 9.027 3.331 9.027 pw_transfer 1559 11.6 0.085 0.104 7.796 7.906 grid_integrate_task_list 129 12.3 7.376 7.757 7.376 7.757 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.572 7.685 cp_fm_upper_to_full 105 14.5 5.770 7.485 5.770 7.485 cp_fm_cholesky_decompose 22 10.9 7.368 7.452 7.368 7.452 wfi_extrapolate 11 7.9 0.001 0.001 7.390 7.390 density_rs2pw 129 9.7 0.006 0.006 6.485 6.787 fft_wrap_pw1pw2_140 527 13.2 0.481 0.529 6.663 6.784 calculate_dm_sparse 129 9.5 0.001 0.001 6.383 6.488 dbcsr_complete_redistribute 395 12.7 1.171 1.199 4.638 6.447 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 5.262 6.241 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.910 5.916 fft3d_ps 1301 14.7 2.714 2.920 5.831 5.896 grid_collocate_task_list 129 9.7 5.115 5.503 5.115 5.503 rs_pw_transfer 1054 12.0 0.013 0.014 5.104 5.483 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.455 5.262 mp_allgather_i34 2527 14.6 1.792 5.188 1.792 5.188 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.990 5.138 mp_sum_l 7930 13.1 3.506 4.958 3.506 4.958 mp_alltoall_d11v 2423 14.1 4.334 4.902 4.334 4.902 potential_pw2rs 129 12.3 0.020 0.022 4.506 4.522 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.343 4.133 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.015 4.048 mp_alltoall_i22 718 14.1 1.953 3.977 1.953 3.977 calculate_first_density_matrix 1 7.0 0.003 0.026 3.966 3.969 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.820 3.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.469000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.272727, yerr=17.868667 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.156264E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57136. MP_Allreduce 11261 1067. MP_Sync 168 MP_Alltoall 1700 12496371. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.289 0.325 187.322 187.323 qs_mol_dyn_low 1 2.0 0.003 0.004 186.214 186.227 qs_forces 11 3.9 0.003 0.003 186.099 186.110 qs_energies 11 4.9 0.010 0.019 179.093 179.103 scf_env_do_scf 11 5.9 0.001 0.001 161.461 161.472 velocity_verlet 10 3.0 0.002 0.012 121.939 121.941 scf_env_do_scf_inner_loop 116 6.6 0.005 0.019 115.313 115.314 dbcsr_multiply_generic 2485 12.5 0.188 0.191 79.820 80.413 qs_scf_new_mos 116 7.6 0.001 0.001 79.953 80.225 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.952 80.224 ot_scf_mini 116 9.6 0.003 0.004 75.479 75.780 multiply_cannon 2485 13.5 0.554 0.587 55.345 58.663 multiply_cannon_loop 2485 14.5 1.175 1.205 51.902 53.474 init_scf_loop 11 6.9 0.001 0.003 46.022 46.023 ot_mini 116 10.6 0.001 0.001 42.485 42.782 prepare_preconditioner 11 7.9 0.000 0.000 41.993 42.016 make_preconditioner 11 8.9 0.000 0.002 41.992 42.016 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.714 40.657 multiply_cannon_multrec 29820 15.5 14.154 19.398 26.752 31.392 rebuild_ks_matrix 127 8.3 0.001 0.001 28.732 29.024 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.731 29.023 mp_waitall_1 152434 16.5 17.605 27.295 17.605 27.295 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.960 26.217 qs_ot_get_derivative 116 11.6 0.004 0.025 23.211 23.521 make_m2s 4970 13.5 0.095 0.099 20.185 21.123 make_images 4970 14.5 1.952 2.235 19.881 20.822 qs_ot_get_p 127 10.4 0.001 0.001 19.303 19.654 ot_diis_step 116 11.6 0.018 0.021 19.136 19.148 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.663 19.120 apply_single 127 13.6 0.001 0.001 18.663 19.120 cp_fm_upper_to_full 104 14.7 10.926 16.160 10.926 16.160 cp_fm_cholesky_invert 11 10.9 16.121 16.130 16.121 16.130 qs_ot_p2m_diag 82 11.4 0.339 0.386 15.242 15.294 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.538 15.076 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.993 13.993 sum_up_and_integrate 127 10.3 0.139 0.150 13.908 13.935 integrate_v_rspace 127 11.3 0.003 0.004 13.768 13.800 multiply_cannon_sync_h2d 29820 15.5 11.744 13.085 11.744 13.085 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.957 12.979 calculate_rho_elec 127 8.7 0.174 0.189 12.957 12.978 dbcsr_mm_accdrv_process 61748 16.2 7.795 8.818 12.178 12.908 dbcsr_complete_redistribute 393 12.7 1.554 1.690 8.949 12.675 make_images_data 4970 15.5 0.063 0.067 10.885 12.373 init_scf_run 11 5.9 0.000 0.001 11.547 11.549 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.547 11.548 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.574 11.281 hybrid_alltoall_any 5155 16.4 0.521 2.197 9.619 11.098 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.967 10.969 cp_fm_redistribute_end 82 14.4 1.865 10.901 1.878 10.906 cp_fm_diag_elpa_base 82 14.4 8.430 10.325 8.996 10.805 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.256 9.846 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.220 9.422 mp_alltoall_i22 712 14.1 5.522 9.199 5.522 9.199 pw_transfer 1535 11.6 0.084 0.097 7.779 7.849 grid_integrate_task_list 127 12.3 7.496 7.781 7.496 7.781 cp_fm_cholesky_decompose 22 10.9 7.579 7.684 7.579 7.684 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.556 7.629 wfi_extrapolate 11 7.9 0.001 0.001 7.561 7.561 multiply_cannon_metrocomm4 24850 15.5 0.075 0.086 2.681 7.091 mp_irecv_dv 75445 16.2 2.538 6.820 2.538 6.820 fft_wrap_pw1pw2_140 519 13.2 0.478 0.484 6.677 6.759 density_rs2pw 127 9.7 0.005 0.006 6.158 6.701 calculate_dm_sparse 127 9.5 0.001 0.001 6.594 6.671 fft3d_ps 1281 14.7 2.776 2.842 5.803 5.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.614 5.657 grid_collocate_task_list 127 9.7 5.173 5.529 5.173 5.529 mp_alltoall_d11v 2401 14.1 4.655 5.151 4.655 5.151 rs_pw_transfer 1038 11.9 0.013 0.014 4.515 5.111 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.940 5.039 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.462 4.548 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.522 4.523 potential_pw2rs 127 12.3 0.022 0.023 4.204 4.221 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.146 4.217 calculate_first_density_matrix 1 7.0 0.001 0.003 3.798 3.802 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.323000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.545455, yerr=24.780357 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 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 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.501782E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516877E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592388592 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096701. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 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 0.025 0.056 174.577 174.577 qs_mol_dyn_low 1 2.0 0.003 0.004 174.139 174.180 qs_forces 11 3.9 0.003 0.003 173.499 173.502 qs_energies 11 4.9 0.001 0.002 166.219 166.226 scf_env_do_scf 11 5.9 0.001 0.001 147.432 147.445 velocity_verlet 10 3.0 0.001 0.001 113.362 113.399 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.651 111.652 dbcsr_multiply_generic 2534 12.6 0.184 0.189 76.925 77.527 qs_scf_new_mos 118 7.6 0.001 0.001 75.706 75.819 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.705 75.818 ot_scf_mini 118 9.6 0.004 0.004 71.225 71.348 multiply_cannon 2534 13.6 0.591 0.617 57.093 61.621 multiply_cannon_loop 2534 14.6 0.452 0.466 52.277 53.443 ot_mini 118 10.6 0.001 0.001 40.866 40.972 init_scf_loop 11 6.9 0.000 0.000 35.629 35.631 mp_waitall_1 130986 16.6 26.583 33.768 26.583 33.768 prepare_preconditioner 11 7.9 0.000 0.000 31.742 31.771 make_preconditioner 11 8.9 0.000 0.000 31.742 31.771 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.661 29.916 rebuild_ks_matrix 129 8.3 0.001 0.001 28.687 28.834 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.687 28.833 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.201 26.334 multiply_cannon_multrec 10136 15.6 10.617 16.155 18.789 22.920 qs_ot_get_derivative 118 11.6 0.002 0.002 20.532 20.651 multiply_cannon_metrocomm3 10136 15.6 0.023 0.024 12.912 20.344 ot_diis_step 118 11.6 0.020 0.021 20.266 20.266 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.794 20.196 apply_single 129 13.6 0.001 0.001 19.794 20.196 make_m2s 5068 13.6 0.067 0.072 16.184 18.721 make_images 5068 14.6 2.352 2.720 15.875 18.403 cp_fm_cholesky_invert 11 10.9 18.235 18.240 18.235 18.240 qs_ot_get_p 129 10.4 0.001 0.001 17.228 17.372 sum_up_and_integrate 129 10.3 0.181 0.190 14.060 14.103 integrate_v_rspace 129 11.3 0.004 0.005 13.878 13.930 qs_ot_p2m_diag 84 11.4 0.502 0.508 13.492 13.508 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.099 13.153 calculate_rho_elec 129 8.7 0.260 0.271 13.099 13.153 make_images_data 5068 15.6 0.053 0.061 9.794 12.417 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.358 12.359 multiply_cannon_sync_h2d 10136 15.6 11.656 12.328 11.656 12.328 hybrid_alltoall_any 5255 16.5 0.824 3.652 9.665 12.045 init_scf_run 11 5.9 0.000 0.001 11.688 11.688 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.688 11.688 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.359 9.360 cp_fm_diag_elpa_base 84 14.4 9.105 9.182 9.352 9.352 dbcsr_mm_accdrv_process 20954 16.1 2.991 3.953 7.801 8.544 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.024 8.114 grid_integrate_task_list 129 12.3 7.817 8.110 7.817 8.110 cp_fm_cholesky_decompose 22 10.9 7.884 7.993 7.884 7.993 mp_allgather_i34 2534 14.6 3.022 7.905 3.022 7.905 pw_transfer 1559 11.6 0.083 0.091 7.775 7.799 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.554 7.582 multiply_cannon_metrocomm1 10136 15.6 0.028 0.029 4.477 7.574 wfi_extrapolate 11 7.9 0.001 0.001 7.303 7.303 fft_wrap_pw1pw2_140 527 13.2 0.505 0.528 6.674 6.706 calculate_dm_sparse 129 9.5 0.001 0.001 6.603 6.678 density_rs2pw 129 9.7 0.005 0.006 5.888 6.137 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.961 5.983 grid_collocate_task_list 129 9.7 5.525 5.749 5.525 5.749 fft3d_ps 1301 14.7 2.742 2.812 5.699 5.735 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.578 5.579 dbcsr_complete_redistribute 397 12.7 2.167 2.252 5.187 5.578 multiply_cannon_metrocomm4 7602 15.6 0.023 0.026 1.905 5.454 mp_irecv_dv 29142 15.9 1.870 5.371 1.870 5.371 mp_alltoall_d11v 2429 14.1 4.647 5.230 4.647 5.230 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.789 4.836 rs_pw_transfer 1054 12.0 0.012 0.013 4.026 4.322 calculate_first_density_matrix 1 7.0 0.000 0.000 4.191 4.194 potential_pw2rs 129 12.3 0.026 0.026 4.033 4.048 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.803 3.842 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.546 3.834 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.786 3.827 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.406 3.722 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.525 3.635 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.577000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1403.363636, yerr=49.348985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1348710248448 0.0% 0.0% 100.0% flops 32 x 32 x 32 1816771166208 0.0% 0.0% 100.0% flops 22 x 9 x 32 1872750477312 0.0% 0.0% 100.0% flops 9 x 22 x 32 1878180581376 0.0% 0.0% 100.0% flops 22 x 22 x 32 2596603592704 0.0% 0.0% 100.0% flops 32 x 32 x 9 4145717182464 0.0% 0.0% 100.0% flops 32 x 32 x 22 5066987667456 0.0% 0.0% 100.0% flops 9 x 32 x 32 5105743036416 0.0% 0.0% 100.0% flops 22 x 32 x 32 6240352600064 0.0% 0.0% 100.0% flops 9 x 32 x 9 10939554164736 0.0% 0.0% 100.0% flops 22 x 32 x 9 14359317147648 0.0% 0.0% 100.0% flops 9 x 32 x 22 14359317147648 0.0% 0.0% 100.0% flops 22 x 32 x 22 18766081540096 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 88.496087E+12 0.0% 0.0% 100.0% flops max/rank 11.068543E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6402989536 0.0% 0.0% 100.0% number of processed stacks 1851840 0.0% 0.0% 100.0% average stack size 0.0 0.0 3457.6 marketing flops 137.086766E+12 ------------------------------------------------------------------------------- # multiplications 2357 max memory usage/rank 3.016389E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 94280 MPI messages size (bytes): total size 1.074917E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.401329E+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 42288 32816234496 4194304 < size <= 16777216 41776 357438586880 16777216 < size 9600 684659414064 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3856 61299. MP_Allreduce 10640 1528. MP_Sync 81 MP_Alltoall 1628 36829524. MP_ISendRecv 3388 218624. MP_Wait 10978 MP_ISend 6072 1081461. MP_IRecv 6072 1081461. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.070 285.909 285.910 qs_mol_dyn_low 1 2.0 0.003 0.004 284.682 284.696 qs_forces 11 3.9 0.003 0.003 284.488 284.489 qs_energies 11 4.9 0.001 0.001 275.891 275.902 scf_env_do_scf 11 5.9 0.001 0.001 253.018 253.036 velocity_verlet 10 3.0 0.001 0.001 201.030 201.050 scf_env_do_scf_inner_loop 110 6.5 0.003 0.008 128.334 128.336 init_scf_loop 11 6.9 0.000 0.000 124.429 124.430 prepare_preconditioner 11 7.9 0.000 0.000 119.712 119.746 make_preconditioner 11 8.9 0.000 0.000 119.712 119.746 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.477 116.866 qs_scf_new_mos 110 7.5 0.001 0.001 89.002 89.067 qs_scf_loop_do_ot 110 8.5 0.001 0.001 89.001 89.066 ot_scf_mini 110 9.5 0.003 0.003 84.486 84.525 dbcsr_multiply_generic 2357 12.5 0.201 0.208 83.933 84.498 cp_fm_upper_to_full 99 14.7 52.494 75.544 52.494 75.544 multiply_cannon 2357 13.5 0.664 0.713 59.451 60.377 multiply_cannon_loop 2357 14.5 0.445 0.458 55.764 57.072 ot_mini 110 10.5 0.001 0.001 45.762 45.790 dbcsr_complete_redistribute 383 12.6 3.989 4.032 29.870 42.741 copy_fm_to_dbcsr 203 11.6 0.001 0.002 26.496 39.401 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.190 36.944 mp_alltoall_i22 690 14.0 21.979 35.112 21.979 35.112 cp_fm_cholesky_invert 11 10.9 33.484 33.490 33.484 33.490 mp_waitall_1 99186 16.7 28.443 31.849 28.443 31.849 rebuild_ks_matrix 121 8.3 0.001 0.001 31.725 31.790 qs_ks_build_kohn_sham_matrix 121 9.3 0.016 0.017 31.724 31.790 qs_ks_update_qs_env 121 7.6 0.001 0.001 29.380 29.442 qs_ot_get_derivative 110 11.5 0.001 0.001 26.285 26.324 qs_ot_get_p 121 10.4 0.001 0.001 24.039 24.094 make_m2s 4714 13.5 0.073 0.075 19.373 20.543 qs_ot_p2m_diag 77 11.4 0.813 0.818 20.235 20.263 make_images 4714 14.5 3.569 3.705 18.919 20.093 multiply_cannon_metrocomm3 9428 15.5 0.022 0.023 18.817 20.011 apply_preconditioner_dbcsr 121 12.6 0.000 0.000 19.639 19.989 apply_single 121 13.6 0.001 0.001 19.639 19.989 ot_diis_step 110 11.5 0.020 0.021 19.441 19.441 multiply_cannon_multrec 9428 15.5 10.415 12.511 19.092 19.365 cp_dbcsr_syevd 77 12.4 0.005 0.005 18.637 18.638 cp_fm_diag_elpa 77 13.4 0.000 0.000 15.687 15.688 cp_fm_diag_elpa_base 77 14.4 11.701 13.117 15.684 15.684 sum_up_and_integrate 121 10.3 0.304 0.306 14.882 14.967 multiply_cannon_sync_h2d 9428 15.5 14.808 14.822 14.808 14.822 integrate_v_rspace 121 11.3 0.003 0.004 14.578 14.663 qs_rho_update_rho_low 121 7.7 0.001 0.001 14.114 14.129 calculate_rho_elec 121 8.7 0.456 0.457 14.113 14.128 init_scf_run 11 5.9 0.000 0.001 13.086 13.086 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.085 13.086 hybrid_alltoall_any 4894 16.4 1.229 2.867 10.740 12.644 make_images_data 4714 15.5 0.056 0.059 10.602 12.534 dbcsr_mm_accdrv_process 19574 16.0 4.560 6.585 8.455 10.750 cp_fm_cholesky_decompose 22 10.9 8.900 8.944 8.900 8.944 wfi_extrapolate 11 7.9 0.001 0.001 8.739 8.739 qs_ot_get_derivative_diag 71 12.3 0.002 0.002 8.702 8.727 grid_integrate_task_list 121 12.3 8.107 8.290 8.107 8.290 pw_transfer 1463 11.6 0.085 0.085 7.871 7.883 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.826 7.827 fft_wrap_pw1pw2 1221 12.7 0.010 0.010 7.650 7.659 calculate_dm_sparse 121 9.5 0.001 0.001 6.978 7.030 qs_ot_get_derivative_taylor 39 13.0 0.001 0.001 5.906 7.025 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.899 6.975 mp_alltoall_d11v 2323 14.0 6.700 6.837 6.700 6.837 fft_wrap_pw1pw2_140 495 13.2 0.509 0.511 6.774 6.781 copy_dbcsr_to_fm 180 11.7 0.004 0.004 6.025 6.155 grid_collocate_task_list 121 9.7 5.994 6.036 5.994 6.036 fft3d_ps 1221 14.7 2.615 2.629 5.798 5.809 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=285.910000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2667.272727, yerr=174.819955 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 1.245409E+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 MP_ISendRecv 570 19200. MP_Wait 1302 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 0.083 0.239 85.657 85.667 qs_energies 1 2.0 0.000 0.000 84.864 84.897 ls_scf 1 3.0 0.001 0.008 83.542 83.576 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.369 72.558 multiply_cannon 111 7.7 0.017 0.020 55.814 57.346 multiply_cannon_loop 111 8.7 0.213 0.227 52.400 53.915 ls_scf_main 1 4.0 0.000 0.008 51.965 51.966 density_matrix_trs4 2 5.0 0.002 0.003 46.595 46.685 ls_scf_init_scf 1 4.0 0.000 0.003 28.544 28.546 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.154 27.208 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.014 25.027 mp_waitall_1 11316 10.9 21.918 24.613 21.918 24.613 multiply_cannon_multrec 2664 9.7 8.221 8.955 15.609 17.314 multiply_cannon_sync_h2d 2664 9.7 13.897 15.392 13.897 15.392 make_m2s 222 7.7 0.009 0.011 12.994 13.532 make_images 222 8.7 0.099 0.108 12.972 13.512 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.348 11.722 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.325 9.091 make_images_data 222 9.7 0.004 0.005 7.571 8.159 dbcsr_mm_accdrv_process 4760 10.4 0.519 0.632 7.007 8.046 hybrid_alltoall_any 227 10.6 0.221 1.841 6.570 8.004 dbcsr_mm_accdrv_process_sort 4760 11.4 6.280 7.226 6.280 7.226 calculate_norms 4752 9.8 5.554 6.152 5.554 6.152 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.935 5.085 mp_sum_l 807 5.4 3.147 4.811 3.147 4.811 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.050 3.639 make_images_sizes 222 9.7 0.000 0.000 0.710 3.635 mp_alltoall_i44 222 10.7 0.710 3.635 0.710 3.635 mp_irecv_dv 6231 10.9 2.033 3.610 2.033 3.610 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.247 3.528 arnoldi_extremal 4 6.8 0.000 0.000 3.212 3.236 arnoldi_normal_ev 4 7.8 0.001 0.003 3.211 3.236 build_subspace 16 8.4 0.009 0.012 3.113 3.115 ls_scf_post 1 4.0 0.005 0.040 3.033 3.068 ls_scf_store_result 1 5.0 0.000 0.000 2.839 2.891 dbcsr_special_finalize 555 9.7 0.005 0.006 2.357 2.757 dbcsr_merge_single_wm 555 10.7 0.459 0.592 2.349 2.749 make_images_pack 222 9.7 2.209 2.629 2.211 2.631 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.307 2.550 dbcsr_sort_data 658 11.4 2.146 2.507 2.146 2.507 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.458 2.065 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.187 2.284 buffer_matrices_ensure_size 222 8.7 1.752 2.074 1.752 2.074 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.776 1.777 rebuild_ks_matrix 3 7.3 0.000 0.000 1.767 1.768 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.025 1.767 1.768 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.667000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1126.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 2.102505E+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_ISendRecv 282 57600. MP_Wait 828 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 0.035 0.043 91.915 91.917 qs_energies 1 2.0 0.000 0.000 91.289 91.293 ls_scf 1 3.0 0.000 0.000 89.782 89.786 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.540 75.907 multiply_cannon 111 7.7 0.028 0.042 53.758 57.145 ls_scf_main 1 4.0 0.000 0.000 54.704 54.710 multiply_cannon_loop 111 8.7 0.116 0.122 50.526 53.300 density_matrix_trs4 2 5.0 0.002 0.003 49.109 49.276 ls_scf_init_scf 1 4.0 0.000 0.001 31.558 31.559 mp_waitall_1 9246 10.9 21.363 30.453 21.363 30.453 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.996 30.091 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.549 27.565 multiply_cannon_multrec 1332 9.7 13.112 16.674 22.429 27.136 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.947 21.119 make_m2s 222 7.7 0.006 0.007 15.294 16.004 make_images 222 8.7 1.581 1.978 15.263 15.975 dbcsr_mm_accdrv_process 4041 10.4 0.296 0.486 8.917 10.540 dbcsr_mm_accdrv_process_sort 4041 11.4 8.483 10.054 8.483 10.054 make_images_data 222 9.7 0.004 0.004 8.788 9.769 hybrid_alltoall_any 227 10.6 0.540 2.511 8.251 9.148 mp_sum_l 807 5.4 5.187 8.255 5.187 8.255 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.218 7.778 mp_irecv_dv 3311 11.0 3.198 7.727 3.198 7.727 calculate_norms 2376 9.8 6.003 6.715 6.003 6.715 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.976 6.627 multiply_cannon_sync_h2d 1332 9.7 4.772 6.164 4.772 6.164 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.015 5.188 arnoldi_extremal 4 6.8 0.000 0.000 4.576 4.601 arnoldi_normal_ev 4 7.8 0.001 0.007 4.575 4.601 build_subspace 16 8.4 0.014 0.021 4.324 4.326 ls_scf_post 1 4.0 0.000 0.000 3.521 3.525 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.117 3.349 ls_scf_store_result 1 5.0 0.000 0.000 3.198 3.345 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.222 2.746 3.224 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.259 2.746 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.494 2.566 make_images_pack 222 9.7 2.026 2.431 2.028 2.433 mp_allgather_i34 111 8.7 0.949 2.400 0.949 2.400 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.247 2.249 rebuild_ks_matrix 3 7.3 0.000 0.000 2.233 2.236 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.233 2.236 dbcsr_sort_data 436 11.2 1.825 2.051 1.825 2.051 dbcsr_data_new 4174 10.1 1.602 1.856 1.602 1.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.917000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 2.685374E+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_ISendRecv 186 57600. MP_Wait 732 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 0.037 0.054 94.291 94.292 qs_energies 1 2.0 0.000 0.000 93.647 93.650 ls_scf 1 3.0 0.000 0.000 92.245 92.247 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.584 76.860 ls_scf_main 1 4.0 0.000 0.000 57.434 57.438 multiply_cannon 111 7.7 0.041 0.125 52.936 56.891 multiply_cannon_loop 111 8.7 0.100 0.105 49.388 53.035 density_matrix_trs4 2 5.0 0.002 0.003 51.484 51.697 mp_waitall_1 7374 11.0 23.681 33.937 23.681 33.937 ls_scf_init_scf 1 4.0 0.000 0.001 31.229 31.231 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.058 30.116 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.543 27.556 multiply_cannon_multrec 888 9.7 12.692 15.415 21.306 24.631 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.050 23.196 make_m2s 222 7.7 0.006 0.006 16.867 18.151 make_images 222 8.7 1.981 2.271 16.829 18.110 hybrid_alltoall_any 227 10.6 0.618 2.846 9.361 10.685 make_images_data 222 9.7 0.003 0.004 9.613 10.675 dbcsr_mm_accdrv_process 3754 10.4 0.256 0.456 8.132 9.378 dbcsr_mm_accdrv_process_sort 3754 11.4 7.743 8.922 7.743 8.922 mp_sum_l 807 5.4 5.210 8.878 5.210 8.878 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.451 7.200 mp_irecv_dv 2335 11.1 2.435 7.158 2.435 7.158 multiply_cannon_sync_h2d 888 9.7 6.050 7.130 6.050 7.130 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.000 7.077 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.634 6.650 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.908 5.117 arnoldi_extremal 4 6.8 0.000 0.000 5.084 5.095 arnoldi_normal_ev 4 7.8 0.001 0.005 5.084 5.095 build_subspace 16 8.4 0.014 0.020 4.775 4.781 calculate_norms 1584 9.8 4.271 4.657 4.271 4.657 mp_allgather_i34 111 8.7 1.399 3.791 1.399 3.791 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.438 3.771 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.610 3.029 3.612 ls_scf_post 1 4.0 0.000 0.000 3.583 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.307 3.409 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.729 2.817 make_images_pack 222 9.7 1.817 2.135 1.820 2.138 dbcsr_sort_data 325 11.1 1.898 2.135 1.898 2.135 make_images_sizes 222 9.7 0.000 0.000 0.906 2.090 mp_alltoall_i44 222 10.7 0.906 2.090 0.906 2.090 dbcsr_data_release 9322 10.9 1.316 1.965 1.316 1.965 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.900 1.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.292000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2151.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 3.331920E+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_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.102 0.513 97.351 97.402 qs_energies 1 2.0 0.000 0.000 96.212 96.225 ls_scf 1 3.0 0.001 0.009 94.272 94.286 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.918 78.172 ls_scf_main 1 4.0 0.001 0.010 58.340 58.343 multiply_cannon 111 7.7 0.054 0.148 51.440 55.820 density_matrix_trs4 2 5.0 0.002 0.003 52.360 52.478 multiply_cannon_loop 111 8.7 0.114 0.126 46.396 49.239 ls_scf_init_scf 1 4.0 0.001 0.009 32.723 32.725 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.450 31.523 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.706 28.721 mp_waitall_1 6438 11.0 22.251 28.180 22.251 28.180 multiply_cannon_multrec 1332 9.7 14.240 17.118 22.187 24.854 make_m2s 222 7.7 0.007 0.008 20.852 22.266 make_images 222 8.7 3.143 3.603 20.802 22.218 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.064 16.962 make_images_data 222 9.7 0.004 0.004 11.578 13.218 hybrid_alltoall_any 227 10.6 0.795 3.744 10.940 12.937 dbcsr_mm_accdrv_process 3641 10.4 0.216 0.439 7.593 9.099 dbcsr_mm_accdrv_process_sort 3641 11.4 7.188 8.682 7.188 8.682 mp_sum_l 807 5.4 4.063 7.564 4.063 7.564 multiply_cannon_sync_h2d 1332 9.7 5.516 6.228 5.516 6.228 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.154 6.068 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.069 5.947 mp_irecv_dv 3229 10.9 2.045 5.894 2.045 5.894 arnoldi_extremal 4 6.8 0.000 0.000 5.208 5.225 arnoldi_normal_ev 4 7.8 0.001 0.004 5.208 5.225 build_subspace 16 8.4 0.014 0.021 4.867 4.876 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.435 4.874 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.567 4.763 calculate_norms 2376 9.8 4.183 4.519 4.183 4.519 mp_allgather_i34 111 8.7 2.171 4.492 2.171 4.492 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.587 3.890 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.702 3.205 3.703 dbcsr_sort_data 658 11.4 3.067 3.422 3.067 3.422 ls_scf_post 1 4.0 0.002 0.013 3.208 3.221 dbcsr_special_finalize 555 9.7 0.006 0.007 2.816 3.167 dbcsr_merge_single_wm 555 10.7 0.534 0.641 2.808 3.159 ls_scf_store_result 1 5.0 0.000 0.000 2.945 3.004 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.876 2.933 dbcsr_data_release 10477 10.7 1.586 2.452 1.586 2.452 dbcsr_finalize 304 7.8 0.049 0.061 1.795 1.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.402000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2684.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 4.612022E+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_ISendRecv 90 115200. MP_Wait 573 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 0.062 0.100 98.750 98.751 qs_energies 1 2.0 0.000 0.000 97.669 97.674 ls_scf 1 3.0 0.000 0.000 95.747 95.754 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.227 77.422 ls_scf_main 1 4.0 0.000 0.000 61.491 61.491 multiply_cannon 111 7.7 0.077 0.164 54.920 60.355 density_matrix_trs4 2 5.0 0.002 0.003 54.590 54.677 multiply_cannon_loop 111 8.7 0.069 0.079 50.341 52.002 mp_waitall_1 5481 11.0 25.773 30.734 25.773 30.734 ls_scf_init_scf 1 4.0 0.000 0.001 30.665 30.671 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.497 29.540 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.321 27.327 multiply_cannon_multrec 444 9.7 14.046 16.432 21.175 24.626 make_m2s 222 7.7 0.004 0.005 17.422 20.052 make_images 222 8.7 3.723 4.396 17.361 19.992 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.957 16.014 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.906 15.165 make_images_data 222 9.7 0.003 0.004 9.712 12.241 hybrid_alltoall_any 227 10.6 0.791 3.756 9.396 11.922 multiply_cannon_sync_h2d 444 9.7 6.538 9.008 6.538 9.008 dbcsr_mm_accdrv_process 3003 10.4 0.189 0.476 6.840 7.961 dbcsr_mm_accdrv_process_sort 3003 11.4 6.475 7.582 6.475 7.582 mp_allgather_i34 111 8.7 2.818 7.022 2.818 7.022 arnoldi_extremal 4 6.8 0.000 0.000 5.812 5.819 arnoldi_normal_ev 4 7.8 0.004 0.008 5.812 5.819 mp_sum_l 807 5.4 2.929 5.605 2.929 5.605 build_subspace 16 8.4 0.015 0.019 5.429 5.439 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.548 4.696 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.194 4.407 dbcsr_matrix_vector_mult_local 304 10.0 3.732 4.208 3.734 4.210 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.566 4.201 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.020 4.179 mp_irecv_dv 1241 11.2 1.546 4.174 1.546 4.174 calculate_norms 792 9.8 3.560 3.723 3.560 3.723 ls_scf_post 1 4.0 0.000 0.000 3.592 3.597 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.337 3.424 ls_scf_store_result 1 5.0 0.000 0.000 3.359 3.419 make_images_sizes 222 9.7 0.000 0.000 1.065 3.254 mp_alltoall_i44 222 10.7 1.065 3.253 1.065 3.253 dbcsr_finalize 304 7.8 0.062 0.078 2.196 2.253 dbcsr_merge_all 275 8.9 0.473 0.516 2.045 2.090 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.751000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3593.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c025c99d5d860bcf6d6a049a55c865b3089da860_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 8.742158E+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_ISendRecv 84 732600. MP_Wait 309 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 0.107 0.228 108.275 108.313 qs_energies 1 2.0 0.000 0.000 106.567 106.580 ls_scf 1 3.0 0.000 0.000 103.677 103.690 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.696 77.843 ls_scf_main 1 4.0 0.000 0.000 65.734 65.735 density_matrix_trs4 2 5.0 0.002 0.003 56.923 56.975 multiply_cannon 111 7.7 0.106 0.160 50.113 51.983 multiply_cannon_loop 111 8.7 0.068 0.072 46.629 47.850 ls_scf_init_scf 1 4.0 0.001 0.001 34.262 34.263 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.832 32.854 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.999 30.011 mp_waitall_1 4569 11.1 22.182 25.999 22.182 25.999 make_m2s 222 7.7 0.005 0.005 23.883 24.904 make_images 222 8.7 4.583 4.984 23.777 24.796 multiply_cannon_multrec 444 9.7 17.858 18.593 22.499 23.200 make_images_data 222 9.7 0.003 0.004 13.126 15.589 hybrid_alltoall_any 227 10.6 1.659 3.627 12.919 15.534 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.471 10.984 multiply_cannon_sync_h2d 444 9.7 8.843 8.887 8.843 8.887 arnoldi_extremal 4 6.8 0.000 0.000 7.453 7.468 arnoldi_normal_ev 4 7.8 0.003 0.009 7.453 7.468 build_subspace 16 8.4 0.026 0.036 6.908 6.919 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.536 5.707 dbcsr_matrix_vector_mult_local 304 10.0 5.114 5.454 5.116 5.456 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.019 5.267 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.753 4.840 dbcsr_mm_accdrv_process 1814 10.4 0.167 0.337 4.480 4.617 dbcsr_mm_accdrv_process_sort 1814 11.4 4.165 4.296 4.165 4.296 ls_scf_post 1 4.0 0.000 0.000 3.681 3.693 mp_allgather_i34 111 8.7 1.150 3.632 1.150 3.632 make_images_sizes 222 9.7 0.000 0.000 1.436 3.516 mp_alltoall_i44 222 10.7 1.436 3.515 1.436 3.515 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.418 calculate_norms 792 9.8 3.243 3.277 3.243 3.277 dbcsr_finalize 304 7.8 0.082 0.090 3.080 3.131 dbcsr_merge_all 275 8.9 0.891 0.923 2.865 2.908 dbcsr_complete_redistribute 5 7.6 1.436 1.462 2.757 2.862 dbcsr_data_release 12724 10.6 2.330 2.860 2.330 2.860 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.860 2.860 matrix_ls_to_qs 2 6.0 0.000 0.000 2.398 2.511 dbcsr_sort_data 325 11.1 2.441 2.501 2.441 2.501 dbcsr_new_transposed 4 7.5 0.244 0.253 2.286 2.302 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.255 2.257 dbcsr_frobenius_norm 74 6.6 2.058 2.136 2.204 2.248 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.211 dbcsr_add_anytype 103 7.2 0.860 0.894 2.132 2.211 rebuild_ks_matrix 3 7.3 0.000 0.000 2.190 2.191 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.189 2.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.313000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6857.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c025c99d5d860bcf6d6a049a55c865b3089da860 Summary: empty Status: OK