=== This is the CP2K Performance-Test === Updating 3a7d65476..6acae23fd Fast-forward arch/CRAY-XC40-gnu.psmp | 6 +++--- arch/CRAY-XC50-gnu.psmp | 6 +++--- arch/Darwin-gnu-arm64.psmp | 6 +++--- arch/Linux-gnu-aarch64.psmp | 6 +++--- arch/Linux-gnu-x86_64.psmp | 4 ++-- arch/Linux-intel-x86_64.psmp | 6 +++--- tools/toolchain/scripts/stage6/install_plumed.sh | 4 ++-- 7 files changed, 19 insertions(+), 19 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 6acae23fd72b559e5a202f07538bd042f9277b92 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (25.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/01 job id: 46829697 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/02 job id: 46829698 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/03 job id: 46829699 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/04 job id: 46829701 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/05 job id: 46829702 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/06 job id: 46829704 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/07 job id: 46829705 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/08 job id: 46829707 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/09 job id: 46829710 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/10 job id: 46829711 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/11 job id: 46829713 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/12 job id: 46829714 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/13 job id: 46829716 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/14 job id: 46829717 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/15 job id: 46829720 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/16 job id: 46829721 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/17 job id: 46829723 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/18 job id: 46829724 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/19 job id: 46829725 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/20 job id: 46829726 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/21 job id: 46829727 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/22 job id: 46829728 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/23 job id: 46829729 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/24 job id: 46829730 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/25 job id: 46829731 --- 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/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/26 job id: 46829732 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/27 job id: 46829733 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.039 133.681 133.682 farming_run 1 2.0 133.172 133.174 133.640 133.643 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.448374E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.026 115.370 115.371 qs_energies 1 2.0 0.000 0.000 115.140 115.142 mp2_main 1 3.0 0.000 0.000 113.047 113.048 mp2_gpw_main 1 4.0 0.019 0.025 112.017 112.019 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.070 93.179 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.315 55.423 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.625 46.566 get_2c_integrals 1 6.0 0.008 0.009 36.842 37.582 integrate_v_rspace 273 8.0 0.439 0.453 25.043 29.634 pw_transfer 6555 10.6 0.381 0.387 27.286 27.430 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.955 26.096 grid_integrate_task_list 273 9.0 20.866 25.898 20.866 25.898 fft_wrap_pw1pw2_100 2178 12.4 1.165 1.222 23.471 23.612 compute_2c_integrals 1 7.0 0.002 0.002 19.146 19.147 rpa_ri_compute_en 1 5.0 0.019 0.020 18.840 18.961 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.763 18.879 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.423 18.760 18.876 cp_fm_cholesky_decompose 12 8.2 17.673 18.447 17.673 18.447 cholesky_decomp 1 7.0 0.000 0.000 16.548 17.333 fft3d_s 5443 13.4 16.087 16.272 16.110 16.296 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.562 10.850 15.562 calculate_wavefunction 272 8.0 5.401 5.482 12.512 13.175 rpa_num_int 1 6.0 0.000 0.000 10.655 10.665 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.642 10.657 calc_mat_Q 8 8.0 0.000 0.000 9.479 9.563 contract_S_to_Q 8 9.0 0.000 0.000 8.899 8.982 calc_potential_gpw 544 9.5 0.005 0.006 8.232 8.599 parallel_gemm_fm 14 9.1 0.000 0.000 8.494 8.579 parallel_gemm_fm_cosma 14 10.1 8.494 8.579 8.494 8.579 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.157 8.372 potential_pw2rs 545 10.0 0.108 0.110 7.632 8.253 create_integ_mat 1 6.0 0.014 0.027 7.981 7.991 collocate_single_gaussian 272 10.0 0.040 0.042 7.413 7.601 array2fm 1 7.0 0.000 0.000 6.963 7.415 pw_scatter_s 2720 13.7 4.452 4.549 4.452 4.549 pw_gather_s 2722 13.2 3.843 4.245 3.843 4.245 array2fm_buffer_send 1 8.0 2.991 3.155 2.991 3.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.016230, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2790.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.040 399.386 399.388 farming_run 1 2.0 398.455 398.467 399.341 399.344 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.217135E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 13903499. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.034 210.131 210.132 qs_energies 1 2.0 0.000 0.000 209.875 209.882 scf_env_do_scf 1 3.0 0.000 0.000 106.084 106.085 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.150 105.160 rebuild_ks_matrix 4 6.0 0.000 0.000 105.149 105.159 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.067 105.149 105.159 hfx_ks_matrix 4 8.0 0.001 0.001 104.753 104.758 integrate_four_center 4 9.0 0.143 0.450 104.752 104.757 mp2_main 1 3.0 0.000 0.000 103.493 103.500 mp2_gpw_main 1 4.0 0.031 0.044 102.491 102.499 integrate_four_center_main 4 10.0 0.127 0.602 96.763 99.012 integrate_four_center_bin 265 11.0 96.636 98.981 96.636 98.981 init_scf_loop 1 4.0 0.000 0.000 91.806 91.807 mp2_ri_gpw_compute_in 1 5.0 0.064 0.068 75.239 76.364 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.549 55.671 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.163 42.272 47.271 integrate_v_rspace 95 8.0 0.399 0.571 28.664 33.499 pw_transfer 2240 10.6 0.144 0.164 30.013 30.604 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.995 29.651 ao_to_mo_and_store_B_mult_1 91 7.0 10.588 29.286 10.588 29.286 grid_integrate_task_list 95 9.0 23.946 28.919 23.946 28.919 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 27.100 28.914 fft_wrap_pw1pw2_100 730 12.4 1.309 1.496 26.665 27.264 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.909 25.305 25.314 get_2c_integrals 1 6.0 0.000 0.000 20.581 20.624 compute_2c_integrals 1 7.0 0.012 0.077 19.552 19.559 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.998 19.315 mp2_eri_2c_integrate_gpw 1 9.0 1.711 1.831 18.997 19.313 fft3d_s 1823 13.4 18.433 18.894 18.447 18.907 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.275 14.275 calculate_wavefunction 91 8.0 2.009 2.037 9.731 9.966 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.588 8.805 9.364 potential_pw2rs 186 10.0 0.034 0.036 8.639 9.168 mp2_ri_gpw_compute_en_comm 22 7.0 0.488 0.511 8.262 8.867 local_gemm 172 8.0 8.248 8.779 8.248 8.779 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.336 8.585 calc_potential_gpw 182 9.5 0.002 0.003 8.010 8.270 collocate_single_gaussian 91 10.0 0.017 0.023 7.946 8.104 mp_sendrecv_dm3 2068 8.0 6.327 6.923 6.327 6.923 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.441 6.355 6.441 pw_gather_s 912 13.2 4.895 5.461 4.895 5.461 mp_sync 37 10.5 2.703 4.995 2.703 4.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.488338, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 448.815104E+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 10329 270. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 52.908 52.910 qs_mol_dyn_low 1 2.0 0.003 0.004 52.673 52.683 qs_forces 11 3.9 0.002 0.002 52.002 52.004 qs_energies 11 4.9 0.001 0.001 50.506 50.519 scf_env_do_scf 11 5.9 0.000 0.001 44.521 44.521 scf_env_do_scf_inner_loop 108 6.5 0.006 0.040 42.401 42.402 dbcsr_multiply_generic 2286 12.5 0.094 0.098 32.169 32.511 qs_scf_new_mos 108 7.5 0.000 0.001 32.161 32.453 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.160 32.453 ot_scf_mini 108 9.5 0.002 0.002 30.573 30.754 multiply_cannon 2286 13.5 0.184 0.192 25.936 27.191 multiply_cannon_loop 2286 14.5 1.478 1.575 25.238 26.528 velocity_verlet 10 3.0 0.002 0.004 25.037 25.044 ot_mini 108 10.5 0.001 0.001 18.464 18.713 qs_ot_get_derivative 108 11.5 0.001 0.002 15.653 15.848 mp_waitall_1 245248 16.5 7.639 14.204 7.639 14.204 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 5.536 12.602 multiply_cannon_multrec 54864 15.5 4.362 6.935 8.275 12.147 rebuild_ks_matrix 119 8.3 0.000 0.000 8.033 8.175 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 8.032 8.175 qs_ot_get_p 119 10.4 0.001 0.002 7.741 8.012 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.056 7.189 multiply_cannon_sync_h2d 54864 15.5 6.123 7.021 6.123 7.021 mp_sum_l 7287 12.8 4.286 5.748 4.286 5.748 dbcsr_mm_accdrv_process 76910 16.1 1.143 1.843 3.834 5.549 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.826 5.257 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.148 5.174 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.791 4.890 init_scf_run 11 5.9 0.000 0.001 4.727 4.727 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.727 4.727 sum_up_and_integrate 119 10.3 0.012 0.014 4.620 4.626 integrate_v_rspace 119 11.3 0.002 0.003 4.608 4.614 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.384 4.384 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.290 4.381 calculate_rho_elec 119 8.7 0.012 0.017 4.290 4.381 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.177 4.178 cp_fm_redistribute_end 50 14.0 2.131 4.152 2.138 4.155 cp_fm_diag_elpa_base 50 14.0 2.011 4.042 2.015 4.051 jit_kernel_multiply 13 15.8 2.630 3.697 2.630 3.697 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.862 3.053 apply_single 119 13.6 0.000 0.000 2.861 3.053 calculate_dm_sparse 119 9.5 0.000 0.001 2.827 2.967 rs_pw_transfer 974 11.9 0.012 0.013 2.810 2.932 ot_diis_step 108 11.5 0.006 0.006 2.626 2.626 calculate_first_density_matrix 1 7.0 0.000 0.000 2.575 2.578 multiply_cannon_metrocomm1 54864 15.5 0.053 0.060 1.401 2.562 density_rs2pw 119 9.7 0.004 0.004 2.252 2.356 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.204 2.267 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.244 2.246 acc_transpose_blocks 54864 15.5 0.226 0.249 1.709 2.194 grid_integrate_task_list 119 12.3 2.065 2.147 2.065 2.147 init_scf_loop 11 6.9 0.000 0.000 2.103 2.104 wfi_extrapolate 11 7.9 0.001 0.001 2.094 2.094 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.987 2.036 potential_pw2rs 119 12.3 0.004 0.004 1.903 1.916 pw_transfer 1439 11.6 0.053 0.058 1.772 1.843 mp_sum_d 4135 12.0 1.176 1.811 1.176 1.811 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.696 1.768 make_m2s 4572 13.5 0.053 0.056 1.635 1.679 make_images 4572 14.5 0.133 0.138 1.552 1.597 mp_waitany 12084 13.8 1.252 1.447 1.252 1.447 grid_collocate_task_list 119 9.7 1.379 1.443 1.379 1.443 mp_alltoall_d11v 2130 13.8 1.258 1.417 1.258 1.417 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.397 1.415 fft3d_ps 1201 14.6 0.380 0.489 1.340 1.413 fft_wrap_pw1pw2_140 487 13.2 0.182 0.201 1.316 1.388 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.910000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 484.433920E+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 10306 303. MP_Sync 54 MP_Alltoall 2060 455234. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.068 41.250 41.255 qs_mol_dyn_low 1 2.0 0.003 0.004 40.821 40.830 qs_forces 11 3.9 0.002 0.004 40.571 40.574 qs_energies 11 4.9 0.002 0.009 38.848 38.854 scf_env_do_scf 11 5.9 0.001 0.002 32.993 32.994 scf_env_do_scf_inner_loop 108 6.5 0.004 0.012 30.472 30.472 dbcsr_multiply_generic 2286 12.5 0.100 0.102 22.275 22.638 qs_scf_new_mos 108 7.5 0.001 0.001 21.028 21.278 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.028 21.278 ot_scf_mini 108 9.5 0.003 0.004 20.135 20.309 velocity_verlet 10 3.0 0.001 0.001 18.730 18.735 multiply_cannon 2286 13.5 0.208 0.218 16.840 18.315 multiply_cannon_loop 2286 14.5 0.902 0.971 15.720 17.328 ot_mini 108 10.5 0.001 0.001 12.737 12.973 mp_waitall_1 200699 16.5 6.219 11.597 6.219 11.597 qs_ot_get_derivative 108 11.5 0.001 0.005 10.276 10.454 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.577 9.888 multiply_cannon_multrec 27432 15.5 1.967 4.608 5.866 8.723 rebuild_ks_matrix 119 8.3 0.000 0.000 7.376 7.532 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.376 7.531 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.502 6.642 dbcsr_mm_accdrv_process 47894 16.0 3.076 5.076 3.830 5.705 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.159 5.320 qs_ot_get_p 119 10.4 0.002 0.011 4.496 4.729 init_scf_run 11 5.9 0.000 0.001 4.552 4.553 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.552 4.552 mp_sum_l 7287 12.8 2.397 4.472 2.397 4.472 sum_up_and_integrate 119 10.3 0.024 0.028 4.359 4.366 integrate_v_rspace 119 11.3 0.002 0.003 4.335 4.344 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.351 4.171 apply_single 119 13.6 0.000 0.000 3.351 4.170 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.973 4.022 calculate_rho_elec 119 8.7 0.021 0.024 3.973 4.021 rs_pw_transfer 974 11.9 0.010 0.011 2.762 3.203 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.034 3.054 calculate_first_density_matrix 1 7.0 0.001 0.003 2.967 2.969 multiply_cannon_sync_h2d 27432 15.5 2.186 2.895 2.186 2.895 make_m2s 4572 13.5 0.052 0.054 2.629 2.858 make_images 4572 14.5 0.201 0.239 2.541 2.767 density_rs2pw 119 9.7 0.004 0.004 2.225 2.711 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.596 2.597 init_scf_loop 11 6.9 0.001 0.003 2.500 2.501 ot_diis_step 108 11.5 0.010 0.011 2.411 2.412 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.197 2.288 acc_transpose_blocks 27432 15.5 0.112 0.117 1.276 2.252 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.230 2.231 cp_fm_redistribute_end 50 14.0 1.131 2.202 1.134 2.205 cp_fm_diag_elpa_base 50 14.0 1.038 2.107 1.066 2.144 calculate_dm_sparse 119 9.5 0.000 0.001 2.038 2.114 pw_transfer 1439 11.6 0.066 0.070 2.019 2.052 potential_pw2rs 119 12.3 0.006 0.006 2.002 2.011 grid_integrate_task_list 119 12.3 1.886 1.980 1.886 1.980 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.927 1.962 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.910 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.698 1.739 make_images_data 4572 15.5 0.045 0.051 1.197 1.652 acc_transpose_blocks_kernels 27432 16.5 0.184 0.273 0.738 1.651 prepare_preconditioner 11 7.9 0.000 0.001 1.558 1.584 make_preconditioner 11 8.9 0.000 0.001 1.558 1.584 jit_kernel_multiply 8 16.1 0.701 1.570 0.701 1.570 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.040 1.548 wfi_extrapolate 11 7.9 0.001 0.006 1.513 1.513 fft_wrap_pw1pw2_140 487 13.2 0.201 0.212 1.480 1.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.450 1.512 fft3d_ps 1201 14.6 0.529 0.585 1.478 1.504 mp_alltoall_d11v 2130 13.8 1.258 1.485 1.258 1.485 grid_collocate_task_list 119 9.7 1.286 1.447 1.286 1.447 jit_kernel_transpose 5 15.5 0.554 1.394 0.554 1.394 mp_allgather_i34 2286 14.5 0.562 1.356 0.562 1.356 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.262 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.179 1.224 mp_sum_d 4135 12.0 0.628 1.085 0.628 1.085 rs_pw_transfer_RS2PW_140 130 11.5 0.146 0.173 0.603 1.047 parallel_gemm_fm 81 9.0 0.000 0.000 1.041 1.046 parallel_gemm_fm_cosma 81 10.0 1.041 1.046 1.041 1.046 qs_energies_init_hamiltonians 11 5.9 0.000 0.003 1.020 1.023 mp_waitany 5720 13.7 0.583 1.015 0.583 1.015 make_basis_sm 11 9.8 0.011 0.041 0.989 0.991 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.932 0.946 multiply_cannon_metrocomm1 27432 15.5 0.031 0.035 0.338 0.905 make_images_sizes 4572 15.5 0.005 0.005 0.612 0.884 mp_alltoall_i44 4572 16.5 0.607 0.879 0.607 0.879 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.606 0.797 0.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.255000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.636364, yerr=2.057038 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 517.963776E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 458078. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.111 0.169 32.672 32.673 qs_mol_dyn_low 1 2.0 0.003 0.003 32.196 32.202 qs_forces 11 3.9 0.002 0.002 32.121 32.123 qs_energies 11 4.9 0.001 0.004 30.511 30.513 scf_env_do_scf 11 5.9 0.001 0.001 25.663 25.663 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.129 23.130 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.326 16.425 velocity_verlet 10 3.0 0.001 0.001 15.315 15.316 qs_scf_new_mos 108 7.5 0.001 0.001 14.844 14.859 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.844 14.858 ot_scf_mini 108 9.5 0.002 0.003 14.125 14.136 multiply_cannon 2286 13.5 0.197 0.206 13.071 13.906 multiply_cannon_loop 2286 14.5 0.633 0.665 12.290 13.158 ot_mini 108 10.5 0.001 0.001 8.690 8.701 qs_ot_get_derivative 108 11.5 0.001 0.001 7.200 7.210 multiply_cannon_multrec 18288 15.5 1.932 2.897 6.768 7.136 rebuild_ks_matrix 119 8.3 0.000 0.000 6.459 6.470 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.459 6.470 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.686 5.697 dbcsr_mm_accdrv_process 38222 16.0 4.163 5.434 4.751 5.612 sum_up_and_integrate 119 10.3 0.031 0.032 4.039 4.046 integrate_v_rspace 119 11.3 0.003 0.003 4.008 4.020 mp_waitall_1 158411 16.6 2.618 3.831 2.618 3.831 init_scf_run 11 5.9 0.000 0.001 3.660 3.660 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.660 3.660 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.508 3.518 calculate_rho_elec 119 8.7 0.031 0.032 3.508 3.518 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.712 3.336 qs_ot_get_p 119 10.4 0.001 0.001 3.272 3.293 rs_pw_transfer 974 11.9 0.009 0.010 2.290 2.538 init_scf_loop 11 6.9 0.001 0.008 2.517 2.519 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.459 2.450 calculate_first_density_matrix 1 7.0 0.000 0.001 2.432 2.434 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.041 2.353 apply_single 119 13.6 0.000 0.000 2.041 2.353 density_rs2pw 119 9.7 0.004 0.004 1.980 2.250 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.205 2.210 grid_integrate_task_list 119 12.3 1.860 1.947 1.860 1.947 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.944 1.944 make_m2s 4572 13.5 0.044 0.045 1.786 1.944 pw_transfer 1439 11.6 0.065 0.069 1.907 1.917 make_images 4572 14.5 0.192 0.205 1.701 1.857 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.814 1.827 calculate_dm_sparse 119 9.5 0.000 0.000 1.812 1.823 potential_pw2rs 119 12.3 0.007 0.008 1.736 1.743 prepare_preconditioner 11 7.9 0.000 0.000 1.718 1.720 make_preconditioner 11 8.9 0.000 0.002 1.717 1.719 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.709 1.710 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.693 1.705 cp_fm_diag_elpa_base 50 14.0 1.669 1.680 1.691 1.703 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.574 1.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.568 1.573 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.541 1.548 multiply_cannon_sync_h2d 18288 15.5 1.407 1.538 1.407 1.538 ot_diis_step 108 11.5 0.011 0.011 1.471 1.471 fft_wrap_pw1pw2_140 487 13.2 0.253 0.260 1.458 1.471 grid_collocate_task_list 119 9.7 1.260 1.417 1.260 1.417 mp_sum_l 7287 12.8 1.083 1.412 1.083 1.412 fft3d_ps 1201 14.6 0.531 0.550 1.302 1.314 acc_transpose_blocks 18288 15.5 0.077 0.079 1.263 1.298 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.184 1.191 wfi_extrapolate 11 7.9 0.001 0.001 1.181 1.181 jit_kernel_multiply 6 16.3 0.537 1.137 0.537 1.137 make_images_data 4572 15.5 0.045 0.048 0.807 0.993 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.970 0.972 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.693 0.909 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.868 0.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.849 0.852 acc_transpose_blocks_kernels 18288 16.5 0.212 0.221 0.817 0.842 mp_waitany 9880 13.7 0.524 0.777 0.524 0.777 mp_alltoall_d11v 2130 13.8 0.659 0.766 0.659 0.766 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.123 0.513 0.758 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.659 0.735 cp_fm_cholesky_invert 11 10.9 0.725 0.728 0.725 0.728 mp_alltoall_z22v 1201 16.6 0.627 0.696 0.627 0.696 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.673000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.272727, yerr=1.813631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 550.080512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.103 35.794 35.795 qs_mol_dyn_low 1 2.0 0.003 0.005 35.374 35.382 qs_forces 11 3.9 0.002 0.002 35.302 35.303 qs_energies 11 4.9 0.003 0.011 33.552 33.557 scf_env_do_scf 11 5.9 0.002 0.007 28.014 28.016 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.620 24.620 dbcsr_multiply_generic 2286 12.5 0.098 0.102 18.084 18.205 velocity_verlet 10 3.0 0.001 0.001 17.827 17.830 qs_scf_new_mos 108 7.5 0.001 0.001 16.061 16.118 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.061 16.117 ot_scf_mini 108 9.5 0.003 0.004 15.127 15.176 multiply_cannon 2286 13.5 0.231 0.272 14.478 14.887 multiply_cannon_loop 2286 14.5 0.941 0.970 13.550 13.901 ot_mini 108 10.5 0.001 0.001 9.127 9.191 multiply_cannon_multrec 27432 15.5 2.337 2.999 8.664 9.055 dbcsr_mm_accdrv_process 47916 15.9 5.370 7.025 6.233 7.417 qs_ot_get_derivative 108 11.5 0.001 0.001 7.321 7.369 rebuild_ks_matrix 119 8.3 0.000 0.000 6.665 6.727 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.664 6.727 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.893 5.949 init_scf_run 11 5.9 0.000 0.001 4.161 4.162 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.161 4.162 sum_up_and_integrate 119 10.3 0.035 0.038 3.898 3.908 integrate_v_rspace 119 11.3 0.003 0.003 3.862 3.873 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.595 3.626 calculate_rho_elec 119 8.7 0.040 0.046 3.594 3.625 qs_ot_get_p 119 10.4 0.001 0.001 3.405 3.484 init_scf_loop 11 6.9 0.001 0.003 3.373 3.376 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.752 3.154 calculate_first_density_matrix 1 7.0 0.001 0.002 2.783 2.785 prepare_preconditioner 11 7.9 0.000 0.000 2.518 2.527 make_preconditioner 11 8.9 0.000 0.003 2.518 2.527 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.125 2.450 make_m2s 4572 13.5 0.054 0.056 2.220 2.328 mp_waitall_1 137007 16.6 1.821 2.301 1.821 2.301 rs_pw_transfer 974 11.9 0.009 0.009 2.039 2.234 make_images 4572 14.5 0.271 0.333 2.113 2.219 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.034 2.179 apply_single 119 13.6 0.000 0.000 2.033 2.178 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.134 2.142 density_rs2pw 119 9.7 0.004 0.004 1.960 2.138 calculate_dm_sparse 119 9.5 0.000 0.000 2.035 2.087 pw_transfer 1439 11.6 0.066 0.070 2.023 2.058 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.930 1.968 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.964 1.966 grid_integrate_task_list 119 12.3 1.868 1.951 1.868 1.951 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.905 1.930 jit_kernel_multiply 10 15.9 0.802 1.920 0.802 1.920 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.816 1.816 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.786 1.800 ot_diis_step 108 11.5 0.012 0.012 1.765 1.765 fft_wrap_pw1pw2_140 487 13.2 0.288 0.300 1.591 1.630 potential_pw2rs 119 12.3 0.008 0.009 1.574 1.578 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.562 1.574 cp_fm_diag_elpa_base 50 14.0 1.527 1.548 1.559 1.571 acc_transpose_blocks 27432 15.5 0.113 0.115 1.484 1.510 fft3d_ps 1201 14.6 0.567 0.621 1.369 1.398 grid_collocate_task_list 119 9.7 1.263 1.393 1.263 1.393 wfi_extrapolate 11 7.9 0.001 0.001 1.324 1.324 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.750 1.297 mp_sum_l 7287 12.8 0.968 1.246 0.968 1.246 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.219 1.229 cp_fm_upper_to_full 72 14.2 0.810 1.153 0.810 1.153 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.118 1.119 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.093 1.111 multiply_cannon_sync_h2d 27432 15.5 0.986 1.044 0.986 1.044 dbcsr_complete_redistribute 329 12.2 0.122 0.145 0.770 1.042 make_images_data 4572 15.5 0.045 0.048 0.882 1.020 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.750 0.968 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.869 acc_transpose_blocks_kernels 27432 16.5 0.271 0.279 0.853 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.861 0.866 mp_alltoall_d11v 2130 13.8 0.753 0.854 0.753 0.854 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.575 0.841 parallel_gemm_fm 81 9.0 0.000 0.000 0.741 0.745 parallel_gemm_fm_cosma 81 10.0 0.741 0.745 0.741 0.745 cp_fm_cholesky_invert 11 10.9 0.735 0.738 0.735 0.738 mp_alltoall_i22 627 13.8 0.431 0.725 0.431 0.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.795000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.090909, yerr=1.676281 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 592.121856E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.036 53.643 53.644 qs_mol_dyn_low 1 2.0 0.003 0.003 53.302 53.310 qs_forces 11 3.9 0.002 0.002 52.954 52.956 qs_energies 11 4.9 0.001 0.001 50.642 50.651 scf_env_do_scf 11 5.9 0.000 0.001 43.537 43.537 scf_env_do_scf_inner_loop 108 6.5 0.003 0.012 36.441 36.443 velocity_verlet 10 3.0 0.001 0.001 30.045 30.066 dbcsr_multiply_generic 2286 12.5 0.095 0.100 21.387 21.872 qs_scf_new_mos 108 7.5 0.001 0.001 21.743 21.832 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.742 21.831 ot_scf_mini 108 9.5 0.002 0.002 20.466 20.516 multiply_cannon 2286 13.5 0.238 0.249 14.409 15.517 multiply_cannon_loop 2286 14.5 0.333 0.345 12.265 13.712 ot_mini 108 10.5 0.001 0.001 11.350 11.443 mp_waitall_1 115863 16.7 8.415 11.423 8.415 11.423 rebuild_ks_matrix 119 8.3 0.000 0.000 10.350 10.491 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 10.349 10.491 qs_ks_update_qs_env 119 7.6 0.001 0.001 9.201 9.326 qs_ot_get_derivative 108 11.5 0.001 0.001 9.278 9.321 qs_rho_update_rho_low 119 7.7 0.001 0.001 7.160 7.194 calculate_rho_elec 119 8.7 0.060 0.061 7.160 7.193 init_scf_loop 11 6.9 0.000 0.000 7.046 7.048 sum_up_and_integrate 119 10.3 0.038 0.042 7.039 7.047 integrate_v_rspace 119 11.3 0.003 0.003 7.002 7.010 multiply_cannon_multrec 9144 15.5 1.685 1.891 6.042 6.269 rs_pw_transfer 974 11.9 0.008 0.008 5.944 6.118 qs_ot_get_p 119 10.4 0.001 0.004 5.842 5.952 make_m2s 4572 13.5 0.035 0.036 4.854 5.780 prepare_preconditioner 11 7.9 0.000 0.000 5.719 5.731 make_preconditioner 11 8.9 0.000 0.000 5.719 5.731 make_images 4572 14.5 0.272 0.308 4.761 5.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.415 5.559 init_scf_run 11 5.9 0.000 0.001 5.244 5.244 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.244 5.244 density_rs2pw 119 9.7 0.004 0.004 5.051 5.195 dbcsr_mm_accdrv_process 12550 15.8 3.072 4.255 4.253 4.328 pw_transfer 1439 11.6 0.067 0.070 4.156 4.194 potential_pw2rs 119 12.3 0.010 0.011 4.115 4.166 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 4.061 4.097 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 2.920 4.096 make_images_data 4572 15.5 0.039 0.043 2.600 3.905 qs_ot_p2m_diag 50 11.0 0.022 0.023 3.756 3.775 hybrid_alltoall_any 4725 16.4 0.063 0.176 2.224 3.722 fft3d_ps 1201 14.6 0.585 0.595 3.396 3.440 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.987 3.418 cp_fm_cholesky_invert 11 10.9 3.322 3.334 3.322 3.334 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.292 3.293 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.148 3.169 fft_wrap_pw1pw2_140 487 13.2 0.363 0.373 2.844 2.880 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.744 2.768 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.755 2.763 cp_fm_diag_elpa_base 50 14.0 2.707 2.741 2.752 2.758 mp_alltoall_z22v 1201 16.6 2.677 2.749 2.677 2.749 mp_allgather_i34 2286 14.5 1.460 2.720 1.460 2.720 wfi_extrapolate 11 7.9 0.001 0.001 2.678 2.678 make_images_sizes 4572 15.5 0.005 0.005 1.579 2.617 mp_alltoall_i44 4572 16.5 1.574 2.612 1.574 2.612 calculate_dm_sparse 119 9.5 0.000 0.000 2.480 2.555 mp_sum_l 7287 12.8 1.692 2.464 1.692 2.464 calculate_first_density_matrix 1 7.0 0.000 0.000 2.455 2.457 mp_alltoall_d11v 2130 13.8 2.231 2.412 2.231 2.412 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.248 2.252 grid_integrate_task_list 119 12.3 1.992 2.120 1.992 2.120 mp_sendrecv_dv 5355 12.7 1.973 2.023 1.973 2.023 ot_diis_step 108 11.5 0.012 0.013 1.937 1.937 jit_kernel_multiply 10 15.8 1.141 1.918 1.141 1.918 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.722 1.864 apply_single 119 13.6 0.000 0.000 1.722 1.864 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.776 1.836 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.619 1.705 cp_fm_cholesky_decompose 22 10.9 1.674 1.678 1.674 1.678 grid_collocate_task_list 119 9.7 1.382 1.533 1.382 1.533 x_to_yz 595 16.2 0.070 0.074 1.434 1.441 dbcsr_complete_redistribute 329 12.2 0.154 0.168 1.355 1.425 yz_to_x 606 15.1 0.053 0.056 1.366 1.421 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.395 1.418 rs_pw_transfer_RS2PW_50 119 11.7 0.195 0.202 1.238 1.418 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.371 1.373 rs_pw_transfer_PW2RS_50 119 14.3 0.215 0.223 1.175 1.359 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.277 1.294 mp_waitany 5200 13.7 1.015 1.191 1.015 1.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=53.644000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=562.000000, yerr=3.592922 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 762.351616E+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 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.050 41.694 41.694 qs_mol_dyn_low 1 2.0 0.003 0.003 41.370 41.377 qs_forces 11 3.9 0.002 0.002 41.305 41.306 qs_energies 11 4.9 0.001 0.001 39.309 39.313 scf_env_do_scf 11 5.9 0.001 0.001 33.075 33.075 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.287 25.287 velocity_verlet 10 3.0 0.001 0.001 23.217 23.222 dbcsr_multiply_generic 2286 12.5 0.100 0.102 17.320 17.461 qs_scf_new_mos 108 7.5 0.001 0.001 15.558 15.662 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.557 15.661 ot_scf_mini 108 9.5 0.002 0.002 14.485 14.587 multiply_cannon 2286 13.5 0.303 0.317 13.511 14.446 multiply_cannon_loop 2286 14.5 0.342 0.347 12.255 13.180 ot_mini 108 10.5 0.001 0.001 8.668 8.786 multiply_cannon_multrec 9144 15.5 3.363 4.746 8.566 8.644 init_scf_loop 11 6.9 0.000 0.000 7.762 7.763 rebuild_ks_matrix 119 8.3 0.000 0.000 7.293 7.436 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.293 7.435 prepare_preconditioner 11 7.9 0.000 0.000 6.801 6.814 make_preconditioner 11 8.9 0.000 0.000 6.801 6.814 qs_ot_get_derivative 108 11.5 0.001 0.001 6.644 6.744 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.569 6.699 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.343 6.692 dbcsr_mm_accdrv_process 12550 15.8 4.280 5.989 5.079 6.345 cp_fm_upper_to_full 72 14.2 3.166 4.568 3.166 4.568 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.225 4.235 calculate_rho_elec 119 8.7 0.118 0.121 4.224 4.234 init_scf_run 11 5.9 0.000 0.001 4.208 4.208 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.207 4.207 sum_up_and_integrate 119 10.3 0.064 0.066 4.037 4.043 integrate_v_rspace 119 11.3 0.004 0.004 3.972 3.979 qs_ot_get_p 119 10.4 0.001 0.001 3.116 3.258 mp_waitall_1 94719 16.7 2.274 3.235 2.274 3.235 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.396 2.854 dbcsr_complete_redistribute 329 12.2 0.295 0.298 1.994 2.816 calculate_first_density_matrix 1 7.0 0.000 0.000 2.769 2.772 pw_transfer 1439 11.6 0.069 0.070 2.640 2.644 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.541 2.545 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.685 2.499 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.226 2.468 apply_single 119 13.6 0.000 0.000 2.226 2.468 make_m2s 4572 13.5 0.038 0.038 2.247 2.405 mp_alltoall_i22 627 13.8 1.523 2.397 1.523 2.397 make_images 4572 14.5 0.350 0.379 2.127 2.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.453 2.258 calculate_dm_sparse 119 9.5 0.000 0.000 2.161 2.182 fft_wrap_pw1pw2_140 487 13.2 0.617 0.620 2.151 2.155 density_rs2pw 119 9.7 0.004 0.004 2.126 2.148 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.228 2.141 grid_integrate_task_list 119 12.3 2.111 2.128 2.111 2.128 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.028 2.030 ot_diis_step 108 11.5 0.014 0.014 2.000 2.001 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.891 1.891 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.761 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.685 1.744 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.641 1.693 mp_sum_l 7287 12.8 1.018 1.675 1.018 1.675 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.619 1.620 fft3d_ps 1201 14.6 0.603 0.612 1.558 1.563 grid_collocate_task_list 119 9.7 1.506 1.521 1.506 1.521 rs_pw_transfer 974 11.9 0.009 0.009 1.428 1.454 cp_fm_cholesky_invert 11 10.9 1.413 1.416 1.413 1.416 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.148 1.392 potential_pw2rs 119 12.3 0.014 0.014 1.373 1.376 wfi_extrapolate 11 7.9 0.001 0.001 1.371 1.371 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.339 1.340 cp_fm_diag_elpa_base 50 14.0 1.195 1.248 1.338 1.338 make_images_data 4572 15.5 0.043 0.046 1.084 1.284 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.247 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.123 1.141 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.098 1.114 mp_alltoall_d11v 2130 13.8 1.077 1.096 1.077 1.096 multiply_cannon_sync_h2d 9144 15.5 1.040 1.043 1.040 1.043 jit_kernel_multiply 6 15.7 0.772 1.035 0.772 1.035 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.979 1.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.936 0.949 qs_create_task_list 11 7.9 0.001 0.001 0.934 0.944 generate_qs_task_list 11 8.9 0.367 0.388 0.933 0.943 acc_transpose_blocks 9144 15.5 0.039 0.040 0.925 0.936 copy_dbcsr_to_fm 153 11.3 0.002 0.003 0.788 0.861 mp_alltoall_z22v 1201 16.6 0.819 0.836 0.819 0.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.694000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=719.818182, yerr=11.668280 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 497.844224E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 441409. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.075 82.522 82.523 qs_mol_dyn_low 1 2.0 0.003 0.008 82.144 82.153 qs_forces 11 3.9 0.005 0.021 81.605 81.608 qs_energies 11 4.9 0.004 0.023 78.635 78.689 scf_env_do_scf 11 5.9 0.000 0.001 69.411 69.413 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.787 63.788 dbcsr_multiply_generic 2055 12.4 0.106 0.108 50.305 50.667 qs_scf_new_mos 99 7.5 0.000 0.001 46.256 46.398 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.255 46.398 ot_scf_mini 99 9.5 0.002 0.003 43.907 43.987 multiply_cannon 2055 13.4 0.184 0.191 41.612 43.257 velocity_verlet 10 3.0 0.004 0.020 42.526 42.532 multiply_cannon_loop 2055 14.4 1.515 1.555 40.625 42.363 ot_mini 99 10.5 0.001 0.001 25.988 26.066 qs_ot_get_derivative 99 11.5 0.001 0.001 19.272 19.344 multiply_cannon_multrec 49320 15.4 12.578 13.368 17.456 18.483 rebuild_ks_matrix 110 8.3 0.000 0.000 14.799 14.938 qs_ks_build_kohn_sham_matrix 110 9.3 0.018 0.033 14.799 14.938 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.954 13.069 mp_waitall_1 220248 16.4 9.884 11.607 9.884 11.607 multiply_cannon_sync_h2d 49320 15.4 10.488 11.099 10.488 11.099 qs_ot_get_p 110 10.4 0.001 0.001 9.514 9.641 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 8.153 8.664 apply_single 110 13.6 0.000 0.001 8.153 8.663 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.877 8.345 multiply_cannon_metrocomm3 49320 15.4 0.080 0.082 5.897 7.944 sum_up_and_integrate 110 10.3 0.036 0.042 7.274 7.287 integrate_v_rspace 110 11.3 0.003 0.004 7.238 7.260 init_scf_run 11 5.9 0.000 0.001 6.937 6.938 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.937 6.937 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.709 6.827 calculate_rho_elec 110 8.6 0.021 0.026 6.708 6.826 ot_diis_step 99 11.5 0.006 0.007 6.563 6.563 qs_ot_p2m_diag 48 11.0 0.013 0.020 6.485 6.511 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.676 5.676 dbcsr_mm_accdrv_process 87628 16.1 1.902 2.001 4.751 5.647 init_scf_loop 11 6.9 0.001 0.006 5.597 5.598 mp_sum_l 6594 12.7 4.397 5.292 4.397 5.292 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.128 5.198 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.108 5.129 cp_fm_diag_elpa_base 48 14.0 5.096 5.116 5.106 5.128 rs_pw_transfer 902 11.9 0.012 0.013 3.814 4.447 wfi_extrapolate 11 7.9 0.001 0.001 4.082 4.082 density_rs2pw 110 9.6 0.004 0.005 3.430 4.081 make_m2s 4110 13.4 0.060 0.065 3.851 3.989 make_images 4110 14.4 0.177 0.192 3.756 3.893 calculate_dm_sparse 110 9.5 0.001 0.001 3.718 3.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.649 3.654 jit_kernel_multiply 13 15.9 2.574 3.543 2.574 3.543 grid_integrate_task_list 110 12.3 3.334 3.512 3.334 3.512 prepare_preconditioner 11 7.9 0.001 0.010 3.342 3.361 make_preconditioner 11 8.9 0.001 0.007 3.340 3.361 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.302 3.341 pw_transfer 1331 11.6 0.056 0.067 3.155 3.235 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.144 3.192 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.063 3.174 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.111 3.172 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.065 3.148 calculate_first_density_matrix 1 7.0 0.000 0.003 2.761 2.767 fft_wrap_pw1pw2_140 451 13.1 0.452 0.495 2.614 2.704 potential_pw2rs 110 12.3 0.006 0.007 2.681 2.703 mp_waitany 14300 13.8 1.855 2.635 1.855 2.635 mp_alltoall_d11v 2046 13.8 2.067 2.552 2.067 2.552 grid_collocate_task_list 110 9.6 2.172 2.441 2.172 2.441 fft3d_ps 1111 14.6 0.795 0.885 2.260 2.310 acc_transpose_blocks 49320 15.4 0.208 0.217 2.101 2.170 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.979 2.003 mp_sum_d 3889 11.9 1.404 1.981 1.404 1.981 make_images_data 4110 15.4 0.042 0.045 1.783 1.914 cp_fm_cholesky_invert 11 10.9 1.810 1.814 1.810 1.814 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.540 1.798 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.645 1.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.523000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.545455, yerr=2.349609 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 583.442432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2814364. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.038 72.312 72.313 qs_mol_dyn_low 1 2.0 0.003 0.003 71.975 71.984 qs_forces 11 3.9 0.003 0.003 71.901 71.902 qs_energies 11 4.9 0.002 0.007 68.493 68.497 scf_env_do_scf 11 5.9 0.001 0.002 58.941 58.945 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 51.148 51.149 dbcsr_multiply_generic 2055 12.4 0.115 0.120 39.351 39.551 velocity_verlet 10 3.0 0.001 0.002 36.770 36.771 qs_scf_new_mos 99 7.5 0.001 0.001 34.432 34.538 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.432 34.538 multiply_cannon 2055 13.4 0.222 0.243 32.658 33.838 ot_scf_mini 99 9.5 0.003 0.003 32.770 32.883 multiply_cannon_loop 2055 14.4 0.918 0.945 31.390 32.312 ot_mini 99 10.5 0.001 0.001 19.421 19.532 multiply_cannon_multrec 24660 15.4 7.591 9.547 15.355 17.670 rebuild_ks_matrix 110 8.3 0.000 0.000 14.048 14.150 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.047 14.150 qs_ot_get_derivative 99 11.5 0.001 0.001 13.598 13.716 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.324 12.413 mp_waitall_1 176588 16.5 7.849 10.471 7.849 10.471 multiply_cannon_metrocomm3 24660 15.4 0.068 0.073 5.320 8.150 multiply_cannon_sync_h2d 24660 15.4 6.996 8.132 6.996 8.132 dbcsr_mm_accdrv_process 52282 16.1 5.069 7.256 7.605 8.041 init_scf_loop 11 6.9 0.000 0.002 7.755 7.756 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.521 7.174 apply_single 110 13.6 0.000 0.001 6.520 7.174 init_scf_run 11 5.9 0.000 0.001 7.106 7.107 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.106 7.107 sum_up_and_integrate 110 10.3 0.052 0.060 6.894 6.909 qs_ot_get_p 110 10.4 0.001 0.001 6.739 6.863 integrate_v_rspace 110 11.3 0.003 0.003 6.842 6.855 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.423 6.437 calculate_rho_elec 110 8.6 0.040 0.048 6.423 6.436 ot_diis_step 99 11.5 0.010 0.010 5.774 5.774 prepare_preconditioner 11 7.9 0.000 0.000 5.691 5.714 make_preconditioner 11 8.9 0.005 0.035 5.691 5.714 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.270 5.435 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.687 5.413 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.822 4.841 make_m2s 4110 13.4 0.057 0.059 4.266 4.703 make_images 4110 14.4 0.400 0.446 4.158 4.590 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.384 4.384 density_rs2pw 110 9.6 0.005 0.005 3.486 4.030 pw_transfer 1331 11.6 0.067 0.075 3.802 3.942 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.877 3.878 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.694 3.835 jit_kernel_multiply 12 16.1 2.183 3.807 2.183 3.807 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.786 3.798 cp_fm_diag_elpa_base 48 14.0 3.739 3.762 3.784 3.795 rs_pw_transfer 902 11.9 0.012 0.014 3.172 3.747 wfi_extrapolate 11 7.9 0.001 0.001 3.595 3.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.511 3.552 grid_integrate_task_list 110 12.3 3.255 3.432 3.255 3.432 calculate_first_density_matrix 1 7.0 0.000 0.003 3.420 3.423 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.221 3.283 fft_wrap_pw1pw2_140 451 13.1 0.519 0.540 3.086 3.224 calculate_dm_sparse 110 9.5 0.001 0.001 2.982 3.016 fft3d_ps 1111 14.6 1.117 1.336 2.669 2.835 make_images_data 4110 15.4 0.047 0.051 2.377 2.804 cp_fm_cholesky_invert 11 10.9 2.676 2.684 2.676 2.684 hybrid_alltoall_any 4261 16.3 0.102 0.451 2.056 2.667 potential_pw2rs 110 12.3 0.008 0.009 2.548 2.570 grid_collocate_task_list 110 9.6 2.113 2.561 2.113 2.561 mp_sum_l 6594 12.7 1.805 2.407 1.805 2.407 mp_alltoall_d11v 2046 13.8 1.793 2.118 1.793 2.118 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.961 1.981 qs_energies_init_hamiltonians 11 5.9 0.019 0.027 1.929 1.930 mp_waitany 10164 13.8 1.291 1.913 1.291 1.913 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.883 1.898 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.801 1.852 mp_irecv_dv 57340 16.2 0.672 1.631 0.672 1.631 mp_allgather_i34 2055 14.4 0.599 1.621 0.599 1.621 acc_transpose_blocks 24660 15.4 0.114 0.117 1.577 1.594 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.220 1.017 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.565 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.477 dbcsr_complete_redistribute 325 12.2 0.236 0.321 1.195 1.457 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.313000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.454545, yerr=6.555548 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 657.051648E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.140 61.620 61.622 qs_mol_dyn_low 1 2.0 0.003 0.004 61.001 61.011 qs_forces 11 3.9 0.004 0.014 60.821 60.825 qs_energies 11 4.9 0.003 0.009 57.548 57.554 scf_env_do_scf 11 5.9 0.001 0.004 49.269 49.269 scf_env_do_scf_inner_loop 99 6.5 0.005 0.010 40.764 40.765 velocity_verlet 10 3.0 0.001 0.001 32.616 32.621 dbcsr_multiply_generic 2055 12.4 0.136 0.143 29.154 29.404 qs_scf_new_mos 99 7.5 0.001 0.001 25.499 25.593 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.498 25.593 ot_scf_mini 99 9.5 0.005 0.010 24.289 24.399 multiply_cannon 2055 13.4 0.213 0.230 22.069 23.700 multiply_cannon_loop 2055 14.4 0.615 0.627 20.898 22.638 ot_mini 99 10.5 0.001 0.001 14.363 14.477 rebuild_ks_matrix 110 8.3 0.000 0.000 12.555 12.678 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.554 12.678 multiply_cannon_multrec 16440 15.4 3.999 5.351 9.816 11.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.008 11.112 qs_ot_get_derivative 99 11.5 0.001 0.001 9.974 10.089 mp_waitall_1 139946 16.5 6.710 9.785 6.710 9.785 init_scf_loop 11 6.9 0.005 0.033 8.471 8.472 multiply_cannon_metrocomm3 16440 15.4 0.043 0.046 4.138 7.076 sum_up_and_integrate 110 10.3 0.060 0.061 6.736 6.753 integrate_v_rspace 110 11.3 0.003 0.007 6.675 6.692 dbcsr_mm_accdrv_process 34862 16.1 4.667 6.233 5.670 6.690 prepare_preconditioner 11 7.9 0.000 0.004 6.668 6.684 make_preconditioner 11 8.9 0.008 0.030 6.667 6.684 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.863 6.375 apply_single 110 13.6 0.000 0.000 5.863 6.375 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.011 6.353 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.112 6.120 calculate_rho_elec 110 8.6 0.059 0.059 6.111 6.120 init_scf_run 11 5.9 0.000 0.001 5.700 5.700 scf_env_initial_rho_setup 11 6.9 0.002 0.007 5.699 5.700 qs_ot_get_p 110 10.4 0.001 0.001 5.178 5.296 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.143 4.730 make_m2s 4110 13.4 0.050 0.051 4.074 4.451 density_rs2pw 110 9.6 0.004 0.005 3.120 4.364 ot_diis_step 99 11.5 0.010 0.011 4.361 4.361 make_images 4110 14.4 0.393 0.517 3.960 4.335 multiply_cannon_sync_h2d 16440 15.4 3.718 4.271 3.718 4.271 rs_pw_transfer 902 11.9 0.011 0.011 2.740 3.995 pw_transfer 1331 11.6 0.066 0.073 3.697 3.706 qs_ot_p2m_diag 48 11.0 0.042 0.046 3.621 3.625 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.590 3.601 grid_integrate_task_list 110 12.3 3.277 3.475 3.277 3.475 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.287 3.287 fft_wrap_pw1pw2_140 451 13.1 0.636 0.642 3.104 3.115 mp_sum_l 6594 12.7 2.397 3.088 2.397 3.088 wfi_extrapolate 11 7.9 0.001 0.001 2.961 2.961 hybrid_alltoall_any 4261 16.3 0.106 0.378 2.113 2.882 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.865 2.868 make_images_data 4110 15.4 0.043 0.047 2.337 2.847 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.765 2.776 cp_fm_diag_elpa_base 48 14.0 2.700 2.732 2.763 2.774 calculate_first_density_matrix 1 7.0 0.003 0.009 2.649 2.653 cp_fm_cholesky_invert 11 10.9 2.576 2.582 2.576 2.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.526 2.567 mp_waitany 17072 13.8 1.234 2.533 1.234 2.533 grid_collocate_task_list 110 9.6 2.153 2.532 2.153 2.532 calculate_dm_sparse 110 9.5 0.001 0.001 2.483 2.516 fft3d_ps 1111 14.6 1.095 1.103 2.415 2.424 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.847 2.395 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.289 2.352 potential_pw2rs 110 12.3 0.011 0.011 2.310 2.318 mp_alltoall_d11v 2046 13.8 1.845 2.291 1.845 2.291 mp_irecv_dv 48980 15.7 0.777 2.276 0.777 2.276 jit_kernel_multiply 12 16.4 0.612 2.255 0.612 2.255 rs_pw_transfer_RS2PW_140 121 11.5 0.182 0.211 0.962 2.211 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.141 2.145 dbcsr_complete_redistribute 325 12.2 0.328 0.373 1.412 1.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.742 1.761 cp_fm_upper_to_full 70 14.2 1.386 1.752 1.386 1.752 cp_fm_cholesky_decompose 22 10.9 1.559 1.583 1.559 1.583 mp_allgather_i34 2055 14.4 0.496 1.483 0.496 1.483 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.466 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.453 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.971 1.411 rs_gather_matrices 110 12.3 0.235 0.266 1.003 1.404 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.230 1.320 acc_transpose_blocks 16440 15.4 0.076 0.079 1.261 1.277 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.232 1.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.622000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.818182, yerr=8.166653 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 732.246016E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.063 67.149 67.150 qs_mol_dyn_low 1 2.0 0.003 0.004 66.740 66.771 qs_forces 11 3.9 0.002 0.003 66.636 66.638 qs_energies 11 4.9 0.001 0.006 63.120 63.125 scf_env_do_scf 11 5.9 0.001 0.005 54.187 54.190 scf_env_do_scf_inner_loop 99 6.5 0.010 0.069 42.251 42.252 velocity_verlet 10 3.0 0.001 0.003 37.916 37.920 dbcsr_multiply_generic 2055 12.4 0.114 0.118 29.780 29.929 qs_scf_new_mos 99 7.5 0.001 0.001 26.715 26.831 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.714 26.830 ot_scf_mini 99 9.5 0.003 0.003 25.083 25.189 multiply_cannon 2055 13.4 0.245 0.263 22.577 23.929 multiply_cannon_loop 2055 14.4 0.887 0.906 21.135 21.624 ot_mini 99 10.5 0.001 0.001 14.250 14.382 multiply_cannon_multrec 24660 15.4 4.236 6.867 12.687 13.910 rebuild_ks_matrix 110 8.3 0.000 0.000 12.588 12.693 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.588 12.692 init_scf_loop 11 6.9 0.001 0.002 11.887 11.889 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.061 11.153 qs_ot_get_derivative 99 11.5 0.001 0.001 10.105 10.215 prepare_preconditioner 11 7.9 0.000 0.000 10.029 10.047 make_preconditioner 11 8.9 0.000 0.001 10.029 10.047 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.263 9.699 dbcsr_mm_accdrv_process 52304 16.0 6.854 7.993 8.304 9.230 sum_up_and_integrate 110 10.3 0.068 0.071 6.756 6.772 integrate_v_rspace 110 11.3 0.003 0.003 6.688 6.704 mp_waitall_1 121746 16.5 4.530 6.429 4.530 6.429 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.379 6.387 calculate_rho_elec 110 8.6 0.078 0.081 6.378 6.387 init_scf_run 11 5.9 0.000 0.001 6.112 6.113 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.112 6.112 make_m2s 4110 13.4 0.059 0.062 5.597 5.974 qs_ot_get_p 110 10.4 0.001 0.002 5.771 5.918 make_images 4110 14.4 0.577 0.699 5.457 5.831 cp_fm_upper_to_full 70 14.2 3.338 4.820 3.338 4.820 ot_diis_step 99 11.5 0.011 0.011 4.108 4.108 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.014 4.098 apply_single 110 13.6 0.000 0.000 4.014 4.097 qs_ot_p2m_diag 48 11.0 0.055 0.065 3.948 3.963 pw_transfer 1331 11.6 0.066 0.076 3.922 3.957 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.815 3.853 dbcsr_complete_redistribute 325 12.2 0.416 0.454 2.703 3.840 density_rs2pw 110 9.6 0.004 0.004 3.196 3.718 grid_integrate_task_list 110 12.3 3.382 3.519 3.382 3.519 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.467 3.468 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.342 3.401 fft_wrap_pw1pw2_140 451 13.1 0.667 0.690 3.270 3.311 multiply_cannon_sync_h2d 24660 15.4 3.187 3.306 3.187 3.306 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.373 3.304 make_images_data 4110 15.4 0.046 0.050 2.885 3.295 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.144 3.268 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.410 3.174 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.046 3.105 wfi_extrapolate 11 7.9 0.001 0.001 3.102 3.103 rs_pw_transfer 902 11.9 0.010 0.011 2.546 3.084 calculate_dm_sparse 110 9.5 0.001 0.001 2.944 2.975 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.918 2.926 cp_fm_diag_elpa_base 48 14.0 2.770 2.828 2.916 2.924 calculate_first_density_matrix 1 7.0 0.000 0.002 2.911 2.914 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.881 2.882 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.757 2.866 mp_alltoall_i22 605 13.7 1.652 2.808 1.652 2.808 cp_fm_cholesky_invert 11 10.9 2.786 2.794 2.786 2.794 fft3d_ps 1111 14.6 1.088 1.125 2.577 2.596 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.490 2.526 grid_collocate_task_list 110 9.6 2.242 2.516 2.242 2.516 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.307 2.310 potential_pw2rs 110 12.3 0.012 0.013 2.254 2.266 mp_alltoall_d11v 2046 13.8 1.876 2.228 1.876 2.228 jit_kernel_multiply 10 15.9 1.116 1.851 1.116 1.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.769 1.780 cp_fm_cholesky_decompose 22 10.9 1.725 1.770 1.725 1.770 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.699 1.736 mp_allgather_i34 2055 14.4 0.537 1.702 0.537 1.702 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.701 mp_waitany 13376 13.8 1.139 1.666 1.139 1.666 mp_sum_l 6594 12.7 0.988 1.589 0.988 1.589 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.551 1.566 acc_transpose_blocks 24660 15.4 0.106 0.109 1.539 1.565 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.845 1.524 mp_irecv_dv 62702 16.1 0.744 1.446 0.744 1.446 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.178 0.861 1.408 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.251 1.364 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.150000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.090909, yerr=9.633790 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 838.451200E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.030 59.225 59.226 qs_mol_dyn_low 1 2.0 0.003 0.003 58.945 58.954 qs_forces 11 3.9 0.002 0.003 58.692 58.694 qs_energies 11 4.9 0.001 0.001 54.902 54.908 scf_env_do_scf 11 5.9 0.000 0.001 45.898 45.898 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 37.604 37.605 velocity_verlet 10 3.0 0.001 0.001 32.993 33.007 dbcsr_multiply_generic 2055 12.4 0.104 0.109 23.876 23.987 qs_scf_new_mos 99 7.5 0.001 0.001 21.751 21.788 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.750 21.787 ot_scf_mini 99 9.5 0.002 0.002 20.481 20.505 multiply_cannon 2055 13.4 0.247 0.264 17.824 19.351 multiply_cannon_loop 2055 14.4 0.321 0.333 16.411 16.771 rebuild_ks_matrix 110 8.3 0.000 0.000 12.480 12.504 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.479 12.504 ot_mini 99 10.5 0.001 0.001 11.164 11.178 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.053 11.072 multiply_cannon_multrec 8220 15.4 3.219 4.503 7.475 8.522 mp_waitall_1 103326 16.6 6.667 8.310 6.667 8.310 init_scf_loop 11 6.9 0.000 0.000 8.237 8.238 qs_ot_get_derivative 99 11.5 0.001 0.001 7.355 7.377 sum_up_and_integrate 110 10.3 0.079 0.081 6.949 6.958 integrate_v_rspace 110 11.3 0.003 0.003 6.870 6.878 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.700 6.714 calculate_rho_elec 110 8.6 0.114 0.115 6.700 6.714 prepare_preconditioner 11 7.9 0.000 0.000 6.508 6.510 make_preconditioner 11 8.9 0.000 0.000 6.508 6.510 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.069 6.158 init_scf_run 11 5.9 0.000 0.001 5.802 5.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.802 5.802 qs_ot_get_p 110 10.4 0.001 0.001 5.287 5.317 dbcsr_mm_accdrv_process 17442 15.9 2.871 3.957 4.125 5.050 make_m2s 4110 13.4 0.039 0.040 4.525 4.846 make_images 4110 14.4 0.637 0.695 4.395 4.714 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.073 4.397 pw_transfer 1331 11.6 0.067 0.074 4.260 4.280 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.152 4.175 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.849 4.078 apply_single 110 13.6 0.000 0.000 3.848 4.078 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.784 3.790 ot_diis_step 99 11.5 0.012 0.012 3.783 3.783 density_rs2pw 110 9.6 0.004 0.004 3.335 3.714 grid_integrate_task_list 110 12.3 3.454 3.672 3.454 3.672 fft_wrap_pw1pw2_140 451 13.1 0.833 0.845 3.530 3.558 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.444 3.445 cp_fm_cholesky_invert 11 10.9 3.161 3.165 3.161 3.165 make_images_data 4110 15.4 0.038 0.044 2.591 3.101 multiply_cannon_sync_h2d 8220 15.4 2.911 3.061 2.911 3.061 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.509 2.990 wfi_extrapolate 11 7.9 0.001 0.001 2.908 2.908 rs_pw_transfer 902 11.9 0.010 0.010 2.496 2.889 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.872 2.880 cp_fm_diag_elpa_base 48 14.0 2.811 2.836 2.870 2.878 calculate_first_density_matrix 1 7.0 0.000 0.000 2.795 2.796 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.711 2.713 fft3d_ps 1111 14.6 1.153 1.181 2.696 2.711 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.615 2.616 calculate_dm_sparse 110 9.5 0.001 0.001 2.560 2.600 grid_collocate_task_list 110 9.6 2.336 2.598 2.336 2.598 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.235 2.405 potential_pw2rs 110 12.3 0.015 0.015 2.296 2.300 mp_alltoall_d11v 2046 13.8 1.916 2.224 1.916 2.224 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.181 2.189 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.990 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.927 1.941 cp_fm_cholesky_decompose 22 10.9 1.850 1.864 1.850 1.864 mp_allgather_i34 2055 14.4 0.598 1.771 0.598 1.771 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.760 1.764 dbcsr_complete_redistribute 325 12.2 0.562 0.587 1.599 1.703 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.550 1.678 mp_waitany 9240 13.8 1.160 1.571 1.160 1.571 jit_kernel_multiply 9 15.6 0.942 1.541 0.942 1.541 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.506 1.519 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.068 1.502 mp_alltoall_z22v 1111 16.6 1.308 1.347 1.308 1.347 qs_create_task_list 11 7.9 0.001 0.001 1.238 1.339 generate_qs_task_list 11 8.9 0.375 0.441 1.237 1.338 rs_gather_matrices 110 12.3 0.327 0.366 1.019 1.336 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.188 0.859 1.259 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.237 1.253 acc_transpose_blocks 8220 15.4 0.036 0.037 0.982 1.248 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.481 1.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.226000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=774.818182, yerr=15.182635 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 1.427046E+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 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.051 87.586 87.586 qs_mol_dyn_low 1 2.0 0.003 0.003 87.135 87.145 qs_forces 11 3.9 0.003 0.003 87.037 87.037 qs_energies 11 4.9 0.001 0.001 82.806 82.808 scf_env_do_scf 11 5.9 0.001 0.001 72.140 72.140 velocity_verlet 10 3.0 0.001 0.001 55.611 55.621 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 44.066 44.068 dbcsr_multiply_generic 2055 12.4 0.137 0.148 28.265 28.349 init_scf_loop 11 6.9 0.000 0.000 28.003 28.007 prepare_preconditioner 11 7.9 0.000 0.000 25.975 25.981 make_preconditioner 11 8.9 0.000 0.000 25.975 25.981 qs_scf_new_mos 99 7.5 0.001 0.001 25.735 25.777 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.735 25.776 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.195 25.440 ot_scf_mini 99 9.5 0.002 0.002 23.987 24.012 multiply_cannon 2055 13.4 0.341 0.375 21.358 22.141 multiply_cannon_loop 2055 14.4 0.340 0.342 19.604 19.923 cp_fm_upper_to_full 70 14.2 12.527 18.022 12.527 18.022 rebuild_ks_matrix 110 8.3 0.000 0.001 14.183 14.215 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.183 14.214 ot_mini 99 10.5 0.001 0.001 13.360 13.387 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.804 12.834 dbcsr_complete_redistribute 325 12.2 1.050 1.086 7.383 10.607 multiply_cannon_multrec 8220 15.4 4.370 4.576 9.592 9.737 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.361 9.582 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.765 8.951 qs_ot_get_derivative 99 11.5 0.001 0.001 8.900 8.927 mp_alltoall_i22 605 13.7 5.374 8.608 5.374 8.608 mp_waitall_1 84994 16.7 7.326 8.198 7.326 8.198 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.793 7.826 calculate_rho_elec 110 8.6 0.225 0.226 7.793 7.825 sum_up_and_integrate 110 10.3 0.150 0.152 7.415 7.433 integrate_v_rspace 110 11.3 0.004 0.004 7.265 7.281 init_scf_run 11 5.9 0.000 0.001 6.382 6.383 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.382 6.383 make_m2s 4110 13.4 0.043 0.043 5.218 5.703 qs_ot_get_p 110 10.4 0.001 0.001 5.493 5.520 make_images 4110 14.4 0.882 0.934 5.032 5.516 dbcsr_mm_accdrv_process 11614 15.7 3.303 3.683 5.078 5.328 cp_fm_cholesky_invert 11 10.9 5.179 5.184 5.179 5.184 pw_transfer 1331 11.6 0.076 0.076 5.070 5.077 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.591 5.065 apply_single 110 13.6 0.000 0.000 4.591 5.065 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.610 4.971 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.952 4.959 ot_diis_step 99 11.5 0.015 0.015 4.445 4.445 fft_wrap_pw1pw2_140 451 13.1 1.336 1.343 4.323 4.329 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.990 3.998 multiply_cannon_sync_h2d 8220 15.4 3.942 3.950 3.942 3.950 grid_integrate_task_list 110 12.3 3.788 3.850 3.788 3.850 density_rs2pw 110 9.6 0.004 0.005 3.766 3.790 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.672 3.673 hybrid_alltoall_any 4261 16.3 0.260 0.564 2.867 3.612 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.551 3.551 make_images_data 4110 15.4 0.042 0.044 2.834 3.458 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.881 3.333 wfi_extrapolate 11 7.9 0.001 0.001 3.329 3.329 calculate_dm_sparse 110 9.5 0.001 0.001 3.153 3.180 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.964 2.964 cp_fm_diag_elpa_base 48 14.0 2.426 2.627 2.962 2.962 calculate_first_density_matrix 1 7.0 0.000 0.000 2.929 2.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.896 2.898 fft3d_ps 1111 14.6 1.303 1.309 2.881 2.886 grid_collocate_task_list 110 9.6 2.682 2.694 2.682 2.694 potential_pw2rs 110 12.3 0.021 0.021 2.396 2.399 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.376 2.390 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.345 2.363 rs_pw_transfer 902 11.9 0.011 0.011 2.286 2.304 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.198 2.271 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.074 2.213 mp_alltoall_d11v 2046 13.8 1.987 2.085 1.987 2.085 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.980 1.982 cp_fm_cholesky_decompose 22 10.9 1.944 1.970 1.944 1.970 qs_create_task_list 11 7.9 0.001 0.001 1.896 1.945 generate_qs_task_list 11 8.9 0.731 0.784 1.895 1.944 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.774 1.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.586000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1267.454545, yerr=71.756360 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 624.500736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57694. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 1888386. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.029 204.939 204.940 qs_mol_dyn_low 1 2.0 0.003 0.004 204.490 204.516 qs_forces 11 3.9 0.004 0.005 204.397 204.400 qs_energies 11 4.9 0.001 0.001 198.727 198.738 scf_env_do_scf 11 5.9 0.001 0.001 181.691 181.696 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 161.174 161.176 dbcsr_multiply_generic 2485 12.5 0.182 0.186 122.045 122.549 velocity_verlet 10 3.0 0.001 0.001 122.459 122.460 qs_scf_new_mos 116 7.6 0.001 0.001 121.036 121.231 qs_scf_loop_do_ot 116 8.6 0.001 0.001 121.035 121.231 ot_scf_mini 116 9.6 0.003 0.003 114.555 114.697 multiply_cannon 2485 13.5 0.238 0.249 99.745 101.180 multiply_cannon_loop 2485 14.5 2.025 2.093 97.579 99.015 ot_mini 116 10.6 0.001 0.001 63.718 63.899 multiply_cannon_multrec 59640 15.5 33.348 35.356 41.409 43.282 qs_ot_get_derivative 116 11.6 0.001 0.001 39.172 39.341 rebuild_ks_matrix 127 8.3 0.001 0.001 33.800 33.982 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.016 33.799 33.982 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.262 30.450 mp_waitall_1 264810 16.5 27.224 30.305 27.224 30.305 multiply_cannon_sync_h2d 59640 15.5 27.636 29.572 27.636 29.572 qs_ot_get_p 127 10.4 0.001 0.001 29.008 29.274 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.150 25.029 apply_single 127 13.6 0.001 0.001 24.150 25.029 ot_diis_step 116 11.6 0.007 0.008 24.333 24.334 qs_ot_p2m_diag 82 11.4 0.078 0.090 22.501 22.657 init_scf_loop 11 6.9 0.000 0.001 20.443 20.445 cp_dbcsr_syevd 82 12.4 0.005 0.005 20.043 20.043 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.156 18.253 multiply_cannon_metrocomm3 59640 15.5 0.113 0.118 15.057 17.209 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.078 17.111 cp_fm_diag_elpa_base 82 14.4 17.006 17.036 17.074 17.107 prepare_preconditioner 11 7.9 0.000 0.000 15.800 15.852 make_preconditioner 11 8.9 0.000 0.000 15.799 15.852 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.001 15.175 sum_up_and_integrate 127 10.3 0.088 0.107 14.460 14.474 integrate_v_rspace 127 11.3 0.004 0.005 14.371 14.393 make_m2s 4970 13.5 0.103 0.113 13.612 14.022 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.880 13.994 calculate_rho_elec 127 8.7 0.046 0.064 13.880 13.993 make_images 4970 14.5 0.397 0.415 13.432 13.851 init_scf_run 11 5.9 0.000 0.001 12.801 12.801 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.800 12.801 density_rs2pw 127 9.7 0.006 0.007 7.115 10.501 rs_pw_transfer 1038 11.9 0.017 0.019 5.799 9.182 wfi_extrapolate 11 7.9 0.001 0.001 9.143 9.143 cp_fm_cholesky_invert 11 10.9 8.964 8.971 8.964 8.971 calculate_dm_sparse 127 9.5 0.001 0.001 8.300 8.391 mp_sum_l 7884 12.9 7.416 8.371 7.416 8.371 dbcsr_mm_accdrv_process 123452 16.2 3.020 3.223 7.617 8.218 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.873 7.996 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.887 7.995 pw_transfer 1535 11.6 0.075 0.091 7.772 7.951 multiply_cannon_metrocomm1 59640 15.5 0.089 0.092 5.716 7.816 grid_integrate_task_list 127 12.3 7.186 7.751 7.186 7.751 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 7.571 7.747 make_images_data 4970 15.5 0.065 0.071 6.699 7.587 hybrid_alltoall_any 5155 16.4 0.289 2.252 5.866 7.026 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.997 7.008 fft_wrap_pw1pw2_140 519 13.2 1.261 1.316 6.618 6.791 mp_alltoall_d11v 2401 14.1 4.404 6.568 4.404 6.568 mp_waitany 15900 13.9 2.716 6.168 2.716 6.168 grid_collocate_task_list 127 9.7 4.688 5.930 4.688 5.930 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.833 5.909 fft3d_ps 1281 14.7 2.131 2.768 5.357 5.679 rs_pw_transfer_RS2PW_140 138 11.5 0.277 0.292 2.141 5.541 potential_pw2rs 127 12.3 0.009 0.010 4.612 4.629 cp_fm_cholesky_decompose 22 10.9 4.599 4.615 4.599 4.615 rs_gather_matrices 127 12.3 0.364 0.410 2.415 4.532 mp_sum_d 4454 12.1 3.374 4.144 3.374 4.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.940000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.454545, yerr=4.997520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 834.572288E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57909. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 4978482. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.074 191.734 191.736 qs_mol_dyn_low 1 2.0 0.003 0.004 191.197 191.208 qs_forces 11 3.9 0.004 0.004 191.094 191.096 qs_energies 11 4.9 0.001 0.001 184.199 184.213 scf_env_do_scf 11 5.9 0.001 0.001 167.498 167.508 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 134.776 134.778 velocity_verlet 10 3.0 0.001 0.002 120.756 120.758 dbcsr_multiply_generic 2527 12.6 0.189 0.194 97.727 98.895 qs_scf_new_mos 118 7.6 0.001 0.001 94.063 94.557 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.062 94.556 ot_scf_mini 118 9.6 0.004 0.004 89.213 89.755 multiply_cannon 2527 13.6 0.482 0.540 77.638 81.738 multiply_cannon_loop 2527 14.6 1.252 1.297 74.692 77.152 ot_mini 118 10.6 0.001 0.001 49.813 50.439 mp_waitall_1 216438 16.6 24.478 37.582 24.478 37.582 multiply_cannon_multrec 30324 15.6 22.309 26.643 31.991 36.535 rebuild_ks_matrix 129 8.3 0.001 0.001 33.848 34.428 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.848 34.428 init_scf_loop 11 6.9 0.000 0.000 32.633 32.634 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.400 30.937 multiply_cannon_metrocomm3 30324 15.6 0.092 0.097 15.653 28.535 qs_ot_get_derivative 118 11.6 0.001 0.002 27.735 28.284 prepare_preconditioner 11 7.9 0.000 0.000 28.223 28.277 make_preconditioner 11 8.9 0.000 0.000 28.223 28.277 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.917 27.453 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.170 23.326 apply_single 129 13.6 0.001 0.001 22.170 23.326 multiply_cannon_sync_h2d 30324 15.6 19.560 22.303 19.560 22.303 ot_diis_step 118 11.6 0.014 0.015 21.903 21.906 qs_ot_get_p 129 10.4 0.001 0.002 21.083 21.731 cp_fm_cholesky_invert 11 10.9 16.480 16.492 16.480 16.492 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.258 16.295 make_m2s 5054 13.6 0.089 0.094 14.407 15.613 make_images 5054 14.6 1.164 1.342 14.198 15.406 sum_up_and_integrate 129 10.3 0.117 0.133 15.086 15.121 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.110 15.110 integrate_v_rspace 129 11.3 0.004 0.004 14.969 15.007 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.316 14.349 calculate_rho_elec 129 8.7 0.088 0.105 14.315 14.348 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.962 11.992 cp_fm_diag_elpa_base 83 14.4 11.706 11.766 11.956 11.989 init_scf_run 11 5.9 0.000 0.001 11.762 11.763 scf_env_initial_rho_setup 11 6.9 0.016 0.017 11.762 11.763 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.812 11.233 multiply_cannon_metrocomm4 27797 15.6 0.105 0.120 3.740 10.476 make_images_data 5054 15.6 0.065 0.072 8.771 10.300 mp_irecv_dv 70031 16.3 3.537 10.084 3.537 10.084 density_rs2pw 129 9.7 0.007 0.007 7.517 10.038 dbcsr_mm_accdrv_process 62734 16.2 4.580 5.376 9.135 9.758 hybrid_alltoall_any 5240 16.5 0.347 1.541 7.472 9.738 pw_transfer 1559 11.6 0.087 0.099 8.862 8.918 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.635 8.695 wfi_extrapolate 11 7.9 0.001 0.001 8.391 8.391 rs_pw_transfer 1054 12.0 0.014 0.017 5.750 8.286 grid_integrate_task_list 129 12.3 7.463 7.880 7.463 7.880 fft_wrap_pw1pw2_140 527 13.2 1.344 1.369 7.637 7.721 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.286 7.047 cp_fm_cholesky_decompose 22 10.9 6.859 6.946 6.859 6.946 calculate_dm_sparse 129 9.5 0.001 0.001 6.531 6.674 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.226 6.237 mp_sum_l 8010 12.9 3.922 6.175 3.922 6.175 grid_collocate_task_list 129 9.7 4.865 6.063 4.865 6.063 fft3d_ps 1301 14.7 2.840 3.010 5.925 5.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.328 5.502 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.366 5.435 mp_waitany 11836 13.9 2.517 5.115 2.517 5.115 potential_pw2rs 129 12.3 0.015 0.018 4.975 4.993 mp_alltoall_d11v 2423 14.1 4.205 4.800 4.205 4.800 rs_pw_transfer_RS2PW_140 140 11.5 0.356 0.385 2.125 4.649 mp_allgather_i34 2527 14.6 1.516 4.546 1.516 4.546 dbcsr_complete_redistribute 395 12.7 0.820 0.905 3.164 4.020 mp_sum_d 4499 12.1 2.620 3.930 2.620 3.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.839 3.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.736000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.090909, yerr=1.928473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 950.640640E+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 4024 57903. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 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 0.025 0.041 183.783 183.785 qs_mol_dyn_low 1 2.0 0.003 0.012 183.026 183.047 qs_forces 11 3.9 0.004 0.004 182.618 182.622 qs_energies 11 4.9 0.002 0.009 175.813 175.823 scf_env_do_scf 11 5.9 0.001 0.001 159.275 159.276 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 123.082 123.083 velocity_verlet 10 3.0 0.001 0.001 116.664 116.676 dbcsr_multiply_generic 2527 12.6 0.181 0.185 83.697 84.926 qs_scf_new_mos 118 7.6 0.001 0.001 83.090 83.391 qs_scf_loop_do_ot 118 8.6 0.001 0.001 83.090 83.390 ot_scf_mini 118 9.6 0.003 0.004 78.777 79.125 multiply_cannon 2527 13.6 0.508 0.525 62.995 67.725 multiply_cannon_loop 2527 14.6 0.862 0.888 59.802 62.968 ot_mini 118 10.6 0.001 0.001 43.912 44.274 mp_waitall_1 171878 16.6 26.599 36.627 26.599 36.627 init_scf_loop 11 6.9 0.000 0.000 36.087 36.088 rebuild_ks_matrix 129 8.3 0.001 0.001 31.887 32.343 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 31.886 32.343 prepare_preconditioner 11 7.9 0.000 0.000 31.995 32.033 make_preconditioner 11 8.9 0.000 0.000 31.995 32.033 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.580 31.023 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.694 29.107 multiply_cannon_metrocomm3 20216 15.6 0.059 0.062 16.247 26.710 multiply_cannon_multrec 20216 15.6 13.403 16.853 22.123 25.535 qs_ot_get_derivative 118 11.6 0.002 0.002 23.569 23.909 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.316 21.487 apply_single 129 13.6 0.001 0.001 20.315 21.487 qs_ot_get_p 129 10.4 0.001 0.001 19.877 20.316 ot_diis_step 118 11.6 0.018 0.018 20.200 20.201 make_m2s 5054 13.6 0.081 0.087 15.314 16.106 multiply_cannon_sync_h2d 20216 15.6 14.219 15.890 14.219 15.890 make_images 5054 14.6 1.195 1.291 15.081 15.875 qs_ot_p2m_diag 83 11.4 0.266 0.274 15.412 15.432 sum_up_and_integrate 129 10.3 0.135 0.146 15.369 15.396 integrate_v_rspace 129 11.3 0.004 0.005 15.234 15.261 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.150 15.196 calculate_rho_elec 129 8.7 0.131 0.146 15.149 15.196 cp_fm_cholesky_invert 11 10.9 15.056 15.065 15.056 15.065 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.340 14.341 init_scf_run 11 5.9 0.000 0.001 11.308 11.308 scf_env_initial_rho_setup 11 6.9 0.004 0.004 11.308 11.308 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.095 11.111 cp_fm_diag_elpa_base 83 14.4 10.675 10.833 11.091 11.108 make_images_data 5054 15.6 0.060 0.067 9.427 10.683 hybrid_alltoall_any 5240 16.5 0.438 1.991 8.141 10.113 density_rs2pw 129 9.7 0.006 0.006 7.703 9.874 multiply_cannon_metrocomm4 17689 15.6 0.062 0.074 3.516 9.594 pw_transfer 1559 11.6 0.087 0.105 9.476 9.577 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.275 9.530 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.248 9.355 mp_irecv_dv 50625 16.2 3.393 9.337 3.393 9.337 dbcsr_mm_accdrv_process 41830 16.2 4.465 4.956 8.184 8.304 fft_wrap_pw1pw2_140 527 13.2 1.421 1.445 8.115 8.230 grid_integrate_task_list 129 12.3 7.536 7.990 7.536 7.990 rs_pw_transfer 1054 12.0 0.014 0.015 5.600 7.802 cp_fm_cholesky_decompose 22 10.9 7.745 7.782 7.745 7.782 wfi_extrapolate 11 7.9 0.001 0.001 7.635 7.637 cp_fm_upper_to_full 105 14.8 5.724 7.249 5.724 7.249 dbcsr_complete_redistribute 395 12.7 1.199 1.225 4.773 6.518 fft3d_ps 1301 14.7 2.767 2.992 6.354 6.433 calculate_dm_sparse 129 9.5 0.001 0.001 5.975 6.071 grid_collocate_task_list 129 9.7 5.060 6.006 5.060 6.006 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.594 5.599 mp_alltoall_d11v 2423 14.1 4.825 5.491 4.825 5.491 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.803 5.479 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.541 5.282 mp_sum_l 8010 12.9 3.562 5.088 3.562 5.088 potential_pw2rs 129 12.3 0.020 0.023 5.022 5.038 mp_waitany 11836 13.9 2.520 4.806 2.520 4.806 mp_allgather_i34 2527 14.6 1.709 4.787 1.709 4.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.594 4.728 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.223 4.266 rs_pw_transfer_RS2PW_140 140 11.5 0.337 0.361 2.037 4.231 transfer_fm_to_dbcsr 11 9.9 0.020 0.029 2.395 4.081 qs_energies_init_hamiltonians 11 5.9 0.020 0.026 3.901 3.903 mp_alltoall_i22 718 14.1 2.010 3.839 2.010 3.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.712 3.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=183.785000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=903.545455, yerr=10.924232 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.151889E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.092 190.208 190.210 qs_mol_dyn_low 1 2.0 0.003 0.004 189.549 189.563 qs_forces 11 3.9 0.004 0.007 188.712 188.717 qs_energies 11 4.9 0.008 0.037 181.355 181.370 scf_env_do_scf 11 5.9 0.001 0.002 164.019 164.029 velocity_verlet 10 3.0 0.001 0.002 126.241 126.274 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.511 117.513 dbcsr_multiply_generic 2507 12.6 0.186 0.191 79.431 80.150 qs_scf_new_mos 117 7.6 0.001 0.001 79.517 79.837 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.516 79.837 ot_scf_mini 117 9.6 0.004 0.005 75.027 75.329 multiply_cannon 2507 13.6 0.559 0.605 54.716 58.007 multiply_cannon_loop 2507 14.6 1.180 1.212 50.976 52.653 init_scf_loop 11 6.9 0.002 0.011 46.382 46.384 ot_mini 117 10.6 0.001 0.001 42.306 42.624 prepare_preconditioner 11 7.9 0.000 0.000 42.060 42.089 make_preconditioner 11 8.9 0.000 0.002 42.060 42.089 make_full_inverse_cholesky 11 9.9 0.013 0.027 35.800 40.714 rebuild_ks_matrix 128 8.3 0.001 0.001 30.436 30.721 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 30.435 30.720 multiply_cannon_multrec 30084 15.6 14.056 19.219 26.011 30.632 mp_waitall_1 147882 16.7 17.453 28.003 17.453 28.003 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.308 27.562 qs_ot_get_derivative 117 11.6 0.004 0.025 22.567 22.871 make_m2s 5014 13.6 0.098 0.120 20.353 21.576 make_images 5014 14.6 1.948 2.269 20.048 21.275 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.167 19.639 apply_single 128 13.6 0.001 0.001 19.166 19.639 ot_diis_step 117 11.6 0.018 0.019 19.623 19.624 qs_ot_get_p 128 10.4 0.002 0.005 19.035 19.404 cp_fm_upper_to_full 105 14.8 11.385 16.706 11.385 16.706 cp_fm_cholesky_invert 11 10.9 16.126 16.135 16.126 16.135 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.516 15.651 sum_up_and_integrate 128 10.3 0.140 0.153 15.174 15.209 integrate_v_rspace 128 11.3 0.004 0.004 15.033 15.073 qs_ot_p2m_diag 83 11.4 0.343 0.391 14.898 14.950 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.808 14.851 calculate_rho_elec 128 8.7 0.173 0.189 14.807 14.850 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.622 13.623 multiply_cannon_sync_h2d 30084 15.6 11.705 12.755 11.705 12.755 dbcsr_complete_redistribute 395 12.7 1.514 1.638 8.910 12.629 make_images_data 5014 15.6 0.063 0.067 11.009 12.606 hybrid_alltoall_any 5200 16.5 0.528 2.213 9.865 12.099 dbcsr_mm_accdrv_process 62264 16.2 7.379 8.390 11.530 12.035 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.516 11.265 init_scf_run 11 5.9 0.000 0.001 11.160 11.162 scf_env_initial_rho_setup 11 6.9 0.002 0.004 11.160 11.161 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.554 10.571 cp_fm_diag_elpa_base 83 14.4 9.589 9.910 10.547 10.562 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.237 9.915 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.325 9.552 pw_transfer 1547 11.6 0.087 0.102 9.249 9.323 mp_alltoall_i22 716 14.1 5.468 9.191 5.468 9.191 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.021 9.105 density_rs2pw 128 9.7 0.006 0.006 7.048 8.419 grid_integrate_task_list 128 12.3 7.750 8.229 7.750 8.229 fft_wrap_pw1pw2_140 523 13.2 1.559 1.601 8.025 8.112 wfi_extrapolate 11 7.9 0.001 0.001 7.856 7.856 cp_fm_cholesky_decompose 22 10.9 7.529 7.629 7.529 7.629 multiply_cannon_metrocomm4 25070 15.6 0.080 0.092 2.789 7.079 mp_irecv_dv 76098 16.2 2.640 6.806 2.640 6.806 rs_pw_transfer 1046 11.9 0.013 0.015 4.853 6.313 calculate_dm_sparse 128 9.5 0.001 0.001 6.180 6.272 mp_alltoall_d11v 2415 14.1 5.058 6.036 5.058 6.036 fft3d_ps 1291 14.7 2.834 2.902 5.949 5.993 grid_collocate_task_list 128 9.7 5.204 5.900 5.204 5.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.573 5.625 potential_pw2rs 128 12.3 0.024 0.024 4.653 4.679 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 4.569 4.570 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.446 4.545 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.350 4.433 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.193 4.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.210000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.272727, yerr=22.328502 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.530225E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3989 58394. MP_Allreduce 11050 1168. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 184.978 184.979 qs_mol_dyn_low 1 2.0 0.003 0.004 184.476 184.497 qs_forces 11 3.9 0.004 0.004 184.362 184.366 qs_energies 11 4.9 0.001 0.001 176.631 176.640 scf_env_do_scf 11 5.9 0.001 0.001 158.777 158.782 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 120.577 120.579 velocity_verlet 10 3.0 0.001 0.001 118.507 118.512 qs_scf_new_mos 116 7.6 0.001 0.001 80.994 81.112 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.993 81.112 ot_scf_mini 116 9.6 0.003 0.004 76.532 76.642 dbcsr_multiply_generic 2485 12.5 0.184 0.196 75.265 75.823 multiply_cannon 2485 13.5 0.581 0.617 54.877 59.103 multiply_cannon_loop 2485 14.5 0.440 0.453 50.221 51.391 ot_mini 116 10.6 0.001 0.001 40.592 40.733 init_scf_loop 11 6.9 0.000 0.000 38.039 38.041 mp_waitall_1 124680 16.7 26.946 34.223 26.946 34.223 prepare_preconditioner 11 7.9 0.000 0.000 33.948 33.971 make_preconditioner 11 8.9 0.000 0.000 33.948 33.971 make_full_inverse_cholesky 11 9.9 0.015 0.026 31.714 32.008 rebuild_ks_matrix 127 8.3 0.001 0.001 30.663 30.811 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.663 30.810 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.757 27.887 qs_ot_get_p 127 10.4 0.001 0.001 22.927 23.103 multiply_cannon_multrec 9940 15.5 10.371 15.967 17.661 21.729 qs_ot_get_derivative 116 11.6 0.001 0.002 20.550 20.656 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.772 20.359 ot_diis_step 116 11.6 0.020 0.020 19.956 19.959 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.460 19.798 apply_single 127 13.6 0.001 0.001 19.460 19.798 cp_fm_cholesky_invert 11 10.9 19.695 19.703 19.695 19.703 qs_ot_p2m_diag 82 11.4 0.490 0.495 18.743 18.762 make_m2s 4970 13.5 0.066 0.071 16.509 18.683 make_images 4970 14.5 2.272 2.750 16.206 18.383 cp_dbcsr_syevd 82 12.4 0.005 0.005 17.547 17.550 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.845 15.873 calculate_rho_elec 127 8.7 0.254 0.265 15.845 15.872 sum_up_and_integrate 127 10.3 0.178 0.188 15.709 15.758 integrate_v_rspace 127 11.3 0.004 0.004 15.530 15.588 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.290 14.299 cp_fm_diag_elpa_base 82 14.4 14.044 14.125 14.283 14.293 make_images_data 4970 15.5 0.051 0.059 10.029 12.491 hybrid_alltoall_any 5155 16.4 0.834 3.795 9.841 12.227 multiply_cannon_sync_h2d 9940 15.5 11.430 12.016 11.430 12.016 init_scf_run 11 5.9 0.000 0.001 10.864 10.864 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.863 10.864 pw_transfer 1535 11.6 0.086 0.096 9.932 9.962 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 9.707 9.744 fft_wrap_pw1pw2_140 519 13.2 1.900 1.930 8.487 8.524 cp_fm_cholesky_decompose 22 10.9 8.401 8.498 8.401 8.498 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.346 8.427 grid_integrate_task_list 127 12.3 7.976 8.195 7.976 8.195 density_rs2pw 127 9.7 0.006 0.006 7.245 8.167 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.735 7.877 wfi_extrapolate 11 7.9 0.001 0.001 7.756 7.756 dbcsr_mm_accdrv_process 20590 16.1 2.644 3.424 6.925 7.561 mp_allgather_i34 2485 14.5 2.876 7.045 2.876 7.045 fft3d_ps 1281 14.7 2.748 2.830 6.211 6.252 calculate_dm_sparse 127 9.5 0.001 0.001 6.128 6.227 grid_collocate_task_list 127 9.7 5.467 6.138 5.467 6.138 mp_alltoall_d11v 2401 14.1 5.615 6.068 5.615 6.068 dbcsr_complete_redistribute 393 12.7 2.115 2.153 5.543 5.942 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.861 5.583 mp_irecv_dv 28618 15.9 1.827 5.494 1.827 5.494 rs_pw_transfer 1038 11.9 0.012 0.013 4.522 5.482 qs_energies_init_hamiltonians 11 5.9 0.004 0.005 5.389 5.391 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.315 5.338 potential_pw2rs 127 12.3 0.026 0.026 4.795 4.815 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.186 4.213 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.600 3.930 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.560 3.867 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.802 3.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.811 3.828 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.747 3.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=184.979000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1403.727273, yerr=58.557029 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.143148E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3981 59443. MP_Allreduce 11030 1505. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.051 306.568 306.569 qs_mol_dyn_low 1 2.0 0.003 0.003 305.942 305.960 qs_forces 11 3.9 0.004 0.004 305.840 305.844 qs_energies 11 4.9 0.001 0.001 296.247 296.256 scf_env_do_scf 11 5.9 0.001 0.001 273.270 273.284 velocity_verlet 10 3.0 0.001 0.001 220.045 220.053 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 145.233 145.236 init_scf_loop 11 6.9 0.000 0.000 127.763 127.769 prepare_preconditioner 11 7.9 0.000 0.000 122.572 122.592 make_preconditioner 11 8.9 0.000 0.000 122.572 122.592 make_full_inverse_cholesky 11 9.9 0.037 0.039 98.030 119.671 qs_scf_new_mos 116 7.6 0.001 0.001 93.656 93.742 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.656 93.741 ot_scf_mini 116 9.6 0.004 0.004 88.818 88.888 dbcsr_multiply_generic 2485 12.5 0.210 0.222 84.409 85.115 cp_fm_upper_to_full 104 14.8 53.627 77.123 53.627 77.123 multiply_cannon 2485 13.5 0.696 0.733 58.736 59.537 multiply_cannon_loop 2485 14.5 0.466 0.475 55.095 56.835 ot_mini 116 10.6 0.001 0.001 45.154 45.226 dbcsr_complete_redistribute 393 12.7 3.995 4.039 30.455 43.497 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.943 40.010 rebuild_ks_matrix 127 8.3 0.001 0.001 39.014 39.098 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 39.014 39.098 transfer_fm_to_dbcsr 11 9.9 0.032 0.035 24.495 37.402 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.824 35.904 mp_alltoall_i22 712 14.1 22.233 35.473 22.233 35.473 cp_fm_cholesky_invert 11 10.9 34.843 34.849 34.843 34.849 mp_waitall_1 102768 16.8 29.308 32.782 29.308 32.782 qs_ot_get_p 127 10.4 0.001 0.001 27.951 28.020 qs_ot_get_derivative 116 11.6 0.002 0.002 24.809 24.880 qs_ot_p2m_diag 82 11.4 0.868 0.874 23.640 23.674 cp_dbcsr_syevd 82 12.4 0.006 0.006 21.868 21.869 make_m2s 4970 13.5 0.075 0.078 20.788 21.757 make_images 4970 14.5 3.752 3.963 20.313 21.284 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.953 20.965 calculate_rho_elec 127 8.7 0.477 0.477 20.953 20.964 multiply_cannon_metrocomm3 9940 15.5 0.023 0.026 19.122 20.701 ot_diis_step 116 11.6 0.022 0.022 20.302 20.303 sum_up_and_integrate 127 10.3 0.319 0.320 20.160 20.245 integrate_v_rspace 127 11.3 0.004 0.004 19.841 19.926 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.509 19.776 apply_single 127 13.6 0.001 0.001 19.509 19.776 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.606 18.607 cp_fm_diag_elpa_base 82 14.4 14.166 15.720 18.602 18.602 multiply_cannon_multrec 9940 15.5 10.467 12.215 17.856 17.949 multiply_cannon_sync_h2d 9940 15.5 15.532 15.546 15.532 15.546 hybrid_alltoall_any 5155 16.4 1.290 3.014 11.306 13.190 make_images_data 4970 15.5 0.058 0.062 11.314 13.143 init_scf_run 11 5.9 0.000 0.001 12.711 12.712 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.711 12.712 pw_transfer 1535 11.6 0.093 0.094 11.881 11.894 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.645 11.653 fft_wrap_pw1pw2_140 519 13.2 3.053 3.110 10.212 10.223 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.675 9.741 mp_alltoall_d11v 2401 14.1 8.670 9.639 8.670 9.639 wfi_extrapolate 11 7.9 0.001 0.001 9.444 9.444 cp_fm_cholesky_decompose 22 10.9 9.385 9.396 9.385 9.396 grid_integrate_task_list 127 12.3 8.789 8.965 8.789 8.965 dbcsr_mm_accdrv_process 20590 16.0 3.865 5.675 7.152 8.945 density_rs2pw 127 9.7 0.005 0.006 8.333 8.477 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.096 8.098 fft3d_ps 1281 14.7 2.823 2.835 6.757 6.780 calculate_dm_sparse 127 9.5 0.001 0.001 6.693 6.759 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.477 6.547 grid_collocate_task_list 127 9.7 6.432 6.490 6.432 6.490 rs_scatter_matrices 138 9.7 3.753 4.677 6.182 6.406 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.216 6.303 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=306.569000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2838.363636, yerr=169.084311 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 1.256129E+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 3139 6114. MP_Sync 4 MP_Alltoall 54 22321977. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.118 85.648 85.650 qs_energies 1 2.0 0.002 0.013 85.006 85.052 ls_scf 1 3.0 0.005 0.037 84.038 84.056 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.556 72.725 multiply_cannon 111 7.7 0.017 0.020 55.862 57.061 multiply_cannon_loop 111 8.7 0.208 0.224 52.439 53.676 ls_scf_main 1 4.0 0.006 0.057 52.344 52.351 density_matrix_trs4 2 5.0 0.002 0.004 46.674 46.763 ls_scf_init_scf 1 4.0 0.001 0.005 28.650 28.676 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.337 27.382 mp_waitall_1 11031 10.9 22.238 25.120 22.238 25.120 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.040 25.058 multiply_cannon_multrec 2664 9.7 8.179 9.014 15.613 17.410 multiply_cannon_sync_h2d 2664 9.7 13.664 15.449 13.664 15.449 make_m2s 222 7.7 0.008 0.011 13.014 13.447 make_images 222 8.7 0.099 0.109 12.992 13.428 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.580 12.838 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.369 8.822 make_images_data 222 9.7 0.004 0.005 7.622 8.246 hybrid_alltoall_any 227 10.6 0.215 1.837 6.586 8.168 dbcsr_mm_accdrv_process 4760 10.4 0.514 0.641 7.051 8.023 dbcsr_mm_accdrv_process_sort 4760 11.4 6.339 7.259 6.339 7.259 calculate_norms 4752 9.8 5.554 6.174 5.554 6.174 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.052 5.248 mp_sum_l 887 5.1 3.184 4.673 3.184 4.673 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.028 3.584 mp_irecv_dv 6231 10.9 2.012 3.553 2.012 3.553 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.332 3.394 make_images_sizes 222 9.7 0.000 0.000 0.753 3.323 mp_alltoall_i44 222 10.7 0.753 3.323 0.753 3.323 arnoldi_extremal 4 6.8 0.000 0.000 3.246 3.275 arnoldi_normal_ev 4 7.8 0.001 0.003 3.246 3.274 build_subspace 16 8.4 0.009 0.012 3.128 3.134 ls_scf_post 1 4.0 0.001 0.009 3.039 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.849 2.893 dbcsr_special_finalize 555 9.7 0.005 0.006 2.287 2.680 dbcsr_merge_single_wm 555 10.7 0.462 0.595 2.279 2.672 make_images_pack 222 9.7 2.205 2.624 2.207 2.625 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.313 2.558 dbcsr_matrix_vector_mult_local 304 10.0 2.060 2.459 2.062 2.460 dbcsr_sort_data 658 11.4 2.076 2.436 2.076 2.436 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.354 2.432 buffer_matrices_ensure_size 222 8.7 1.768 2.125 1.768 2.125 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.844 1.874 rebuild_ks_matrix 3 7.3 0.000 0.000 1.835 1.864 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.010 1.835 1.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.650000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1127.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 2.076750E+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 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.075 91.071 91.073 qs_energies 1 2.0 0.000 0.000 90.551 90.554 ls_scf 1 3.0 0.000 0.001 89.205 89.209 dbcsr_multiply_generic 111 6.7 0.015 0.015 74.981 75.398 multiply_cannon 111 7.7 0.029 0.043 53.320 56.632 ls_scf_main 1 4.0 0.000 0.003 54.821 54.830 multiply_cannon_loop 111 8.7 0.116 0.123 50.064 52.915 density_matrix_trs4 2 5.0 0.002 0.004 49.020 49.205 ls_scf_init_scf 1 4.0 0.001 0.005 30.868 30.872 mp_waitall_1 9105 10.9 21.088 30.797 21.088 30.797 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.620 29.736 multiply_cannon_multrec 1332 9.7 13.098 16.987 22.421 27.462 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.004 27.101 27.112 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.803 21.589 make_m2s 222 7.7 0.006 0.007 15.218 15.824 make_images 222 8.7 1.578 1.922 15.187 15.795 dbcsr_mm_accdrv_process 4041 10.4 0.292 0.454 8.925 10.460 dbcsr_mm_accdrv_process_sort 4041 11.4 8.518 10.031 8.518 10.031 make_images_data 222 9.7 0.004 0.004 8.706 9.559 hybrid_alltoall_any 227 10.6 0.523 2.430 8.214 9.281 mp_sum_l 887 5.1 5.125 8.399 5.125 8.399 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.233 7.533 mp_irecv_dv 3311 11.0 3.213 7.463 3.213 7.463 calculate_norms 2376 9.8 5.999 6.821 5.999 6.821 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.949 6.701 multiply_cannon_sync_h2d 1332 9.7 4.800 5.969 4.800 5.969 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.058 5.239 arnoldi_extremal 4 6.8 0.000 0.000 4.629 4.643 arnoldi_normal_ev 4 7.8 0.003 0.007 4.629 4.643 build_subspace 16 8.4 0.014 0.021 4.370 4.374 ls_scf_post 1 4.0 0.000 0.001 3.517 3.521 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.124 3.361 ls_scf_store_result 1 5.0 0.000 0.000 3.212 3.341 dbcsr_matrix_vector_mult_local 304 10.0 2.734 3.216 2.737 3.218 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.662 2.756 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.215 2.744 make_images_pack 222 9.7 2.026 2.402 2.028 2.404 mp_allgather_i34 111 8.7 0.960 2.314 0.960 2.314 dbcsr_sort_data 436 11.2 1.862 2.103 1.862 2.103 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.904 1.908 rebuild_ks_matrix 3 7.3 0.000 0.000 1.892 1.896 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.013 1.892 1.896 dbcsr_data_new 4174 10.1 1.614 1.836 1.614 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.073000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1678.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 2.704323E+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 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.112 94.992 94.994 qs_energies 1 2.0 0.000 0.000 94.130 94.133 ls_scf 1 3.0 0.000 0.001 92.639 92.644 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.212 77.487 ls_scf_main 1 4.0 0.000 0.002 58.087 58.092 multiply_cannon 111 7.7 0.047 0.103 52.928 57.157 multiply_cannon_loop 111 8.7 0.100 0.106 49.373 53.137 density_matrix_trs4 2 5.0 0.002 0.004 52.088 52.212 mp_waitall_1 7281 11.0 24.026 34.189 24.026 34.189 ls_scf_init_scf 1 4.0 0.000 0.002 31.002 31.005 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.747 29.825 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.020 27.272 27.286 multiply_cannon_multrec 888 9.7 12.618 15.833 21.248 25.119 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.151 23.006 make_m2s 222 7.7 0.006 0.007 17.355 18.588 make_images 222 8.7 1.969 2.278 17.316 18.551 hybrid_alltoall_any 227 10.6 0.622 2.876 9.599 11.164 make_images_data 222 9.7 0.003 0.004 9.927 10.925 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.414 8.165 9.412 mp_sum_l 887 5.1 5.385 9.200 5.385 9.200 dbcsr_mm_accdrv_process_sort 3754 11.4 7.792 9.000 7.792 9.000 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.210 7.290 multiply_cannon_sync_h2d 888 9.7 6.089 7.146 6.089 7.146 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.574 7.013 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.421 6.729 mp_irecv_dv 2335 11.1 2.405 6.687 2.405 6.687 arnoldi_extremal 4 6.8 0.000 0.000 5.092 5.109 arnoldi_normal_ev 4 7.8 0.013 0.031 5.092 5.109 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.892 5.081 build_subspace 16 8.4 0.014 0.020 4.767 4.779 calculate_norms 1584 9.8 4.261 4.659 4.261 4.659 mp_allgather_i34 111 8.7 1.393 3.938 1.393 3.938 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.416 3.745 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.582 3.015 3.584 ls_scf_post 1 4.0 0.001 0.017 3.550 3.557 ls_scf_store_result 1 5.0 0.000 0.000 3.289 3.375 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.889 3.009 make_images_sizes 222 9.7 0.000 0.000 1.093 2.216 mp_alltoall_i44 222 10.7 1.093 2.215 1.093 2.215 dbcsr_sort_data 325 11.1 1.900 2.154 1.900 2.154 make_images_pack 222 9.7 1.816 2.125 1.819 2.128 dbcsr_data_release 9322 10.9 1.302 2.023 1.302 2.023 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.969 1.971 rebuild_ks_matrix 3 7.3 0.000 0.000 1.950 1.952 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.950 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.994000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2164.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 3.301446E+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 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.110 97.286 97.288 qs_energies 1 2.0 0.000 0.000 96.476 96.483 ls_scf 1 3.0 0.001 0.002 94.818 94.826 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.544 78.795 ls_scf_main 1 4.0 0.000 0.002 58.952 58.954 multiply_cannon 111 7.7 0.063 0.163 51.430 56.155 density_matrix_trs4 2 5.0 0.002 0.005 52.789 52.904 multiply_cannon_loop 111 8.7 0.114 0.125 46.347 49.994 ls_scf_init_scf 1 4.0 0.000 0.002 32.640 32.643 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.393 31.464 mp_waitall_1 6369 11.0 22.516 29.204 22.516 29.204 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.933 28.946 multiply_cannon_multrec 1332 9.7 14.180 17.848 22.145 25.002 make_m2s 222 7.7 0.006 0.008 21.244 22.684 make_images 222 8.7 3.132 3.587 21.194 22.636 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.948 17.298 make_images_data 222 9.7 0.004 0.004 11.901 13.631 hybrid_alltoall_any 227 10.6 0.796 3.792 11.218 12.949 dbcsr_mm_accdrv_process 3641 10.4 0.221 0.398 7.601 9.161 dbcsr_mm_accdrv_process_sort 3641 11.4 7.226 8.736 7.226 8.736 mp_sum_l 887 5.1 4.278 7.559 4.278 7.559 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.093 5.997 multiply_cannon_sync_h2d 1332 9.7 5.495 5.969 5.495 5.969 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.354 5.959 mp_irecv_dv 3229 10.9 2.070 5.921 2.070 5.921 arnoldi_extremal 4 6.8 0.000 0.000 5.181 5.197 arnoldi_normal_ev 4 7.8 0.001 0.005 5.181 5.196 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.499 5.050 build_subspace 16 8.4 0.014 0.021 4.831 4.837 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.516 4.730 mp_allgather_i34 111 8.7 2.204 4.637 2.204 4.637 calculate_norms 2376 9.8 4.217 4.631 4.217 4.631 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.574 3.859 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.680 3.186 3.682 dbcsr_sort_data 658 11.4 3.076 3.428 3.076 3.428 ls_scf_post 1 4.0 0.000 0.001 3.226 3.232 dbcsr_special_finalize 555 9.7 0.006 0.007 2.804 3.160 dbcsr_merge_single_wm 555 10.7 0.531 0.653 2.795 3.152 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.068 3.132 ls_scf_store_result 1 5.0 0.000 0.000 2.971 3.030 dbcsr_data_release 10477 10.7 1.600 2.442 1.600 2.442 dbcsr_finalize 304 7.8 0.050 0.061 1.820 2.020 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.980 1.982 rebuild_ks_matrix 3 7.3 0.000 0.000 1.957 1.959 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 1.957 1.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.288000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2703.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 4.668502E+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 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.060 99.156 99.157 qs_energies 1 2.0 0.000 0.000 98.354 98.358 ls_scf 1 3.0 0.000 0.000 96.282 96.292 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.680 77.853 ls_scf_main 1 4.0 0.000 0.000 61.909 61.910 multiply_cannon 111 7.7 0.084 0.171 55.241 60.089 density_matrix_trs4 2 5.0 0.002 0.003 54.693 54.781 multiply_cannon_loop 111 8.7 0.070 0.085 50.686 52.137 mp_waitall_1 5436 11.0 26.206 32.436 26.206 32.436 ls_scf_init_scf 1 4.0 0.000 0.000 30.764 30.768 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.504 29.533 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.296 27.304 multiply_cannon_multrec 444 9.7 13.997 16.526 21.125 23.421 make_m2s 222 7.7 0.004 0.005 17.624 20.172 make_images 222 8.7 3.721 4.425 17.563 20.111 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.772 15.998 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.373 15.398 make_images_data 222 9.7 0.003 0.004 9.862 12.322 hybrid_alltoall_any 227 10.6 0.788 3.761 9.687 12.153 dbcsr_mm_accdrv_process 3003 10.4 0.165 0.344 6.841 7.991 multiply_cannon_sync_h2d 444 9.7 6.571 7.817 6.571 7.817 dbcsr_mm_accdrv_process_sort 3003 11.4 6.524 7.644 6.524 7.644 mp_allgather_i34 111 8.7 2.789 6.962 2.789 6.962 arnoldi_extremal 4 6.8 0.000 0.000 5.876 5.885 arnoldi_normal_ev 4 7.8 0.002 0.005 5.876 5.885 build_subspace 16 8.4 0.015 0.019 5.468 5.480 mp_sum_l 887 5.1 2.861 4.854 2.861 4.854 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.626 4.763 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.196 4.391 dbcsr_matrix_vector_mult_local 304 10.0 3.695 4.179 3.697 4.182 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.647 3.940 mp_irecv_dv 1241 11.2 1.627 3.921 1.627 3.921 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.598 3.699 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.969 3.687 calculate_norms 792 9.8 3.541 3.670 3.541 3.670 ls_scf_post 1 4.0 0.000 0.000 3.609 3.613 make_images_sizes 222 9.7 0.000 0.000 1.117 3.523 mp_alltoall_i44 222 10.7 1.117 3.523 1.117 3.523 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.434 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.291 dbcsr_merge_all 275 8.9 0.472 0.525 2.048 2.128 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.103 2.104 rebuild_ks_matrix 3 7.3 0.000 0.000 2.070 2.071 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.070 2.071 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.056 2.056 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.157000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3676.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_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 8.712958E+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 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.100 109.193 109.194 qs_energies 1 2.0 0.000 0.000 107.747 107.759 ls_scf 1 3.0 0.000 0.000 104.790 104.802 dbcsr_multiply_generic 111 6.7 0.023 0.028 78.178 78.286 ls_scf_main 1 4.0 0.000 0.000 66.103 66.106 density_matrix_trs4 2 5.0 0.002 0.003 56.821 56.877 multiply_cannon 111 7.7 0.188 0.334 50.480 52.361 multiply_cannon_loop 111 8.7 0.067 0.071 46.924 47.873 ls_scf_init_scf 1 4.0 0.000 0.000 34.999 35.000 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.392 33.410 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.520 30.532 mp_waitall_1 4527 11.1 22.793 26.705 22.793 26.705 make_m2s 222 7.7 0.005 0.005 24.075 25.134 make_images 222 8.7 4.597 4.984 23.969 25.027 multiply_cannon_multrec 444 9.7 17.807 18.546 22.472 22.995 hybrid_alltoall_any 227 10.6 1.666 3.654 13.037 15.886 make_images_data 222 9.7 0.003 0.003 13.228 15.813 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.974 11.435 multiply_cannon_sync_h2d 444 9.7 8.852 8.883 8.852 8.883 arnoldi_extremal 4 6.8 0.000 0.000 7.492 7.516 arnoldi_normal_ev 4 7.8 0.003 0.009 7.492 7.516 build_subspace 16 8.4 0.026 0.036 6.915 6.924 dbcsr_matrix_vector_mult 304 9.0 0.017 0.032 5.483 5.621 dbcsr_matrix_vector_mult_local 304 10.0 5.050 5.350 5.053 5.352 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.220 5.310 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.054 5.289 dbcsr_mm_accdrv_process 1814 10.4 0.186 0.332 4.500 4.645 dbcsr_mm_accdrv_process_sort 1814 11.4 4.198 4.330 4.198 4.330 ls_scf_post 1 4.0 0.000 0.000 3.687 3.699 make_images_sizes 222 9.7 0.000 0.000 1.498 3.633 mp_alltoall_i44 222 10.7 1.498 3.633 1.498 3.633 mp_allgather_i34 111 8.7 1.181 3.619 1.181 3.619 ls_scf_store_result 1 5.0 0.000 0.000 3.425 3.433 calculate_norms 792 9.8 3.238 3.271 3.238 3.271 dbcsr_finalize 304 7.8 0.082 0.090 3.093 3.172 dbcsr_merge_all 275 8.9 0.890 0.916 2.879 2.953 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.926 2.927 dbcsr_data_release 12724 10.6 2.339 2.891 2.339 2.891 dbcsr_complete_redistribute 5 7.6 1.440 1.469 2.773 2.872 matrix_ls_to_qs 2 6.0 0.000 0.000 2.430 2.548 dbcsr_sort_data 325 11.1 2.450 2.512 2.450 2.512 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.495 2.496 rebuild_ks_matrix 3 7.3 0.000 0.000 2.427 2.428 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.427 2.428 dbcsr_new_transposed 4 7.5 0.254 0.272 2.286 2.301 dbcsr_frobenius_norm 74 6.6 2.057 2.139 2.199 2.248 dbcsr_add_d 103 6.2 0.000 0.000 2.129 2.201 dbcsr_add_anytype 103 7.2 0.859 0.891 2.129 2.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.194000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6acae23fd72b559e5a202f07538bd042f9277b92_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 582.144000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 655984749. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.223 244.424 244.425 qs_mol_dyn_low 1 2.0 0.006 0.059 243.094 243.132 qs_forces 5 3.8 0.007 0.064 242.873 242.877 qs_energies 5 4.8 0.003 0.052 239.657 239.681 scf_env_do_scf 5 5.8 0.002 0.041 221.651 221.654 scf_env_do_scf_inner_loop 105 6.6 0.007 0.064 193.412 193.428 qs_scf_new_mos 105 7.6 0.000 0.001 149.710 149.891 qs_scf_loop_do_ot 105 8.6 0.001 0.001 149.710 149.891 ot_scf_mini 105 9.6 0.003 0.019 139.955 140.116 dbcsr_multiply_generic 1445 12.2 0.126 0.134 134.867 135.368 multiply_cannon 1445 13.2 0.275 0.290 115.316 117.791 multiply_cannon_loop 1445 14.2 2.379 2.513 113.381 114.922 velocity_verlet 4 3.0 0.004 0.059 108.632 108.634 ot_mini 105 10.6 0.002 0.032 59.407 59.564 qs_ot_get_p 112 10.4 0.001 0.002 50.795 51.094 multiply_cannon_multrec 69360 15.2 31.756 37.129 39.496 44.971 mp_waitall_1 488190 16.1 34.103 40.944 34.103 40.944 qs_ot_p2m_diag 40 11.0 0.020 0.031 39.616 39.752 qs_ot_get_derivative 55 11.6 0.001 0.001 37.801 37.949 cp_dbcsr_syevd 40 12.0 0.002 0.003 36.351 36.352 multiply_cannon_sync_h2d 69360 15.2 30.620 34.995 30.620 34.995 multiply_cannon_metrocomm3 69360 15.2 0.196 0.209 24.627 32.101 cp_fm_syevd 40 13.0 0.000 0.002 30.949 31.105 rebuild_ks_matrix 110 8.4 0.000 0.000 30.884 31.086 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.062 30.883 31.086 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.329 28.509 init_scf_loop 7 6.6 0.003 0.055 28.196 28.200 cp_fm_redistribute_end 40 14.0 13.069 26.077 13.076 26.080 cp_fm_syevd_base 40 14.0 12.993 26.005 12.993 26.005 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.011 23.204 apply_single 62 13.6 0.000 0.000 23.011 23.203 prepare_preconditioner 7 7.6 0.000 0.001 22.864 22.901 make_preconditioner 7 8.6 0.003 0.066 22.864 22.901 ot_new_cg_direction 55 11.6 0.001 0.002 20.960 20.961 qs_rho_update_rho_low 110 7.6 0.001 0.017 19.255 19.601 calculate_rho_elec 110 8.6 0.030 0.038 19.255 19.600 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.492 15.560 init_scf_run 5 5.8 0.000 0.001 15.046 15.047 scf_env_initial_rho_setup 5 6.8 0.002 0.044 15.046 15.047 rs_pw_transfer 690 11.5 0.011 0.013 13.399 14.688 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.551 14.681 density_rs2pw 110 9.6 0.006 0.012 13.048 14.308 pw_transfer 1645 12.4 0.083 0.110 13.425 13.705 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.568 13.665 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.282 13.561 mp_sum_l 4764 12.2 12.001 12.849 12.001 12.849 fft_wrap_pw1pw2_240 915 15.0 1.180 1.279 11.392 11.685 calculate_dm_sparse 110 9.5 0.000 0.001 11.376 11.579 qs_vxc_create 110 10.4 0.007 0.046 11.250 11.310 cp_fm_cholesky_invert 7 10.6 10.621 10.630 10.621 10.630 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.243 10.308 calculate_first_density_matrix 1 7.0 0.000 0.005 9.981 10.004 check_diag 80 13.5 8.586 8.854 9.837 9.979 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 9.939 9.961 fft3d_pb 915 16.0 2.375 2.669 9.489 9.831 acc_transpose_blocks 69360 15.2 0.362 0.381 9.280 9.630 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.516 9.583 sum_up_and_integrate 60 10.3 0.028 0.037 9.536 9.561 integrate_v_rspace 60 11.3 0.002 0.004 9.508 9.534 dbcsr_mm_accdrv_process 154766 15.8 4.032 4.214 7.608 8.428 xc_rho_set_and_dset_create 110 12.4 0.077 0.125 7.878 8.137 make_m2s 2890 13.2 0.079 0.087 7.129 7.747 acc_transpose_blocks_kernels 69360 16.2 0.860 0.906 7.392 7.653 make_images 2890 14.2 0.242 0.260 7.022 7.640 xc_vxc_pw_create 60 11.3 0.039 0.053 7.536 7.592 make_full_single_inverse 7 9.6 0.001 0.038 7.076 7.114 multiply_cannon_metrocomm1 69360 15.2 0.090 0.097 4.176 7.093 xc_pw_derive 510 13.4 0.005 0.007 6.928 7.028 mp_alltoall_z22v 2340 17.7 6.607 6.987 6.607 6.987 jit_kernel_transpose 5 15.0 6.532 6.768 6.532 6.768 mp_waitany 7680 13.5 4.674 6.161 4.674 6.161 potential_pw2rs 60 12.3 0.003 0.003 5.547 5.623 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=244.425000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=554.000000, yerr=2.449490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 6acae23fd72b559e5a202f07538bd042f9277b92 Summary: empty Status: OK