=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 3afb8bf709f96407a7e9894ade69f93aea585134 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (19.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/01 job id: 42369379 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/02 job id: 42369380 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/03 job id: 42369381 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/04 job id: 42369382 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/05 job id: 42369384 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/06 job id: 42369385 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/07 job id: 42369386 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/08 job id: 42369387 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/09 job id: 42369388 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/10 job id: 42369389 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/11 job id: 42369390 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/12 job id: 42369391 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/13 job id: 42369392 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/14 job id: 42369393 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/15 job id: 42369395 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/16 job id: 42369396 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/17 job id: 42369397 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/18 job id: 42369398 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/19 job id: 42369400 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/20 job id: 42369401 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/21 job id: 42369402 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/22 job id: 42369403 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/23 job id: 42369404 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/24 job id: 42369405 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/25 job id: 42369407 --- 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/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/26 job id: 42369410 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.042 134.076 134.077 farming_run 1 2.0 133.240 133.242 134.028 134.029 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.455632E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.025 114.889 114.889 qs_energies 1 2.0 0.000 0.000 114.673 114.676 mp2_main 1 3.0 0.000 0.000 112.815 112.818 mp2_gpw_main 1 4.0 0.019 0.025 111.905 111.908 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.884 93.307 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.458 55.881 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.747 47.139 get_2c_integrals 1 6.0 0.000 0.000 36.768 37.254 integrate_v_rspace 273 8.0 0.432 0.446 25.119 30.271 pw_transfer 6555 10.6 0.379 0.390 27.328 27.835 grid_integrate_task_list 273 9.0 20.938 26.580 20.938 26.580 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.008 26.520 fft_wrap_pw1pw2_100 2178 12.4 1.167 1.226 23.569 24.090 compute_2c_integrals 1 7.0 0.002 0.002 19.245 19.246 rpa_ri_compute_en 1 5.0 0.000 0.000 18.911 19.095 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.835 18.945 mp2_eri_2c_integrate_gpw 1 9.0 2.393 2.435 18.832 18.944 cp_fm_cholesky_decompose 12 8.2 17.531 18.038 17.531 18.038 cholesky_decomp 1 7.0 0.000 0.000 16.360 16.867 fft3d_s 5443 13.4 16.113 16.596 16.135 16.616 ao_to_mo_and_store_B_mult_1 272 7.0 10.874 15.612 10.874 15.612 calculate_wavefunction 272 8.0 5.406 5.555 12.538 13.161 rpa_num_int 1 6.0 0.000 0.001 10.667 10.677 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.626 10.649 calc_mat_Q 8 8.0 0.000 0.000 9.394 9.514 contract_S_to_Q 8 9.0 0.000 0.000 8.815 8.932 calc_potential_gpw 544 9.5 0.005 0.006 8.282 8.636 parallel_gemm_fm 14 9.1 0.000 0.000 8.393 8.486 parallel_gemm_fm_cosma 14 10.1 8.393 8.486 8.393 8.486 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.196 8.438 potential_pw2rs 545 10.0 0.108 0.110 7.646 8.316 create_integ_mat 1 6.0 0.014 0.028 7.727 7.737 collocate_single_gaussian 272 10.0 0.040 0.043 7.423 7.649 array2fm 1 7.0 0.000 0.000 6.692 7.086 pw_scatter_s 2720 13.7 4.458 4.670 4.458 4.670 pw_gather_s 2722 13.2 3.898 4.290 3.898 4.290 array2fm_buffer_send 1 8.0 3.009 3.170 3.009 3.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.905711, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.038 395.015 395.017 farming_run 1 2.0 394.056 394.077 394.979 394.982 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.215984E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 23103487. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.050 209.123 209.124 qs_energies 1 2.0 0.000 0.000 208.876 208.906 scf_env_do_scf 1 3.0 0.000 0.000 105.743 105.743 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.905 104.912 rebuild_ks_matrix 4 6.0 0.000 0.000 104.903 104.911 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.066 104.903 104.911 hfx_ks_matrix 4 8.0 0.001 0.001 104.523 104.526 integrate_four_center 4 9.0 0.143 0.456 104.522 104.526 mp2_main 1 3.0 0.000 0.000 102.849 102.879 mp2_gpw_main 1 4.0 0.035 0.056 102.005 102.035 integrate_four_center_main 4 10.0 0.106 0.525 96.563 98.222 integrate_four_center_bin 263 11.0 96.458 98.215 96.458 98.215 init_scf_loop 1 4.0 0.000 0.000 91.272 91.274 mp2_ri_gpw_compute_in 1 5.0 0.093 0.182 74.955 75.928 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.035 54.511 55.499 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 42.305 47.297 integrate_v_rspace 95 8.0 0.398 0.570 28.681 33.477 pw_transfer 2240 10.6 0.146 0.165 29.901 30.363 ao_to_mo_and_store_B_mult_1 91 7.0 10.523 29.523 10.523 29.523 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.907 29.336 grid_integrate_task_list 95 9.0 23.972 28.992 23.972 28.992 mp2_ri_gpw_compute_en 1 5.0 0.067 0.123 26.889 28.584 fft_wrap_pw1pw2_100 730 12.4 1.271 1.425 26.622 27.046 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.908 25.244 25.254 get_2c_integrals 1 6.0 0.005 0.025 20.327 20.361 compute_2c_integrals 1 7.0 0.008 0.027 19.309 19.326 compute_2c_integrals_loop_lm 1 8.0 0.004 0.014 18.890 19.179 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.813 18.886 19.177 fft3d_s 1823 13.4 18.398 18.714 18.412 18.728 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.469 14.470 calculate_wavefunction 91 8.0 2.016 2.055 9.739 9.996 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.584 8.847 9.335 potential_pw2rs 186 10.0 0.033 0.035 8.581 8.960 local_gemm 172 8.0 8.290 8.751 8.290 8.751 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.522 8.174 8.692 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.018 8.268 8.487 calc_potential_gpw 182 9.5 0.002 0.002 7.992 8.254 collocate_single_gaussian 91 10.0 0.027 0.112 7.877 8.030 mp_sendrecv_dm3 2068 8.0 6.197 6.684 6.197 6.684 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.420 6.341 6.420 pw_gather_s 912 13.2 4.884 5.401 4.884 5.401 mp_sync 38 10.4 2.545 5.292 2.545 5.292 pw_scatter_s 910 13.7 3.986 4.282 3.986 4.282 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.989826, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1502.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.055040E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 1361960. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.060 56.006 56.011 qs_mol_dyn_low 1 2.0 0.020 0.074 55.478 55.485 qs_forces 11 3.9 0.002 0.003 55.312 55.333 qs_energies 11 4.9 0.002 0.005 53.698 53.729 scf_env_do_scf 11 5.9 0.000 0.001 47.235 47.235 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.743 44.743 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.546 34.983 qs_scf_new_mos 108 7.5 0.000 0.000 33.527 33.803 qs_scf_loop_do_ot 108 8.5 0.000 0.000 33.526 33.802 ot_scf_mini 108 9.5 0.002 0.002 31.853 32.037 multiply_cannon 2286 13.5 0.179 0.187 26.221 27.822 velocity_verlet 10 3.0 0.001 0.002 27.804 27.805 multiply_cannon_loop 2286 14.5 1.505 1.576 25.318 26.946 ot_mini 108 10.5 0.001 0.001 20.274 20.510 qs_ot_get_derivative 108 11.5 0.001 0.001 17.179 17.363 mp_waitall_1 267858 16.1 9.885 15.819 9.885 15.819 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.206 13.078 multiply_cannon_multrec 54864 15.5 4.132 6.381 7.424 10.779 rebuild_ks_matrix 119 8.3 0.000 0.000 8.664 8.810 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.019 8.663 8.810 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.626 7.759 mp_sum_l 7207 12.9 5.770 7.557 5.770 7.557 qs_ot_get_p 119 10.4 0.001 0.001 6.763 7.069 multiply_cannon_sync_h2d 54864 15.5 5.747 7.031 5.747 7.031 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.892 6.338 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.867 5.971 sum_up_and_integrate 119 10.3 0.012 0.015 5.103 5.116 integrate_v_rspace 119 11.3 0.002 0.002 5.091 5.104 init_scf_run 11 5.9 0.000 0.001 5.027 5.028 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.027 5.028 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.836 5.000 calculate_rho_elec 119 8.7 0.011 0.016 4.836 5.000 dbcsr_mm_accdrv_process 76910 16.1 1.202 1.855 3.216 4.477 rs_pw_transfer 974 11.9 0.012 0.013 3.539 3.817 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.670 3.739 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.012 3.565 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.945 3.179 apply_single 119 13.6 0.000 0.000 2.945 3.179 calculate_dm_sparse 119 9.5 0.000 0.000 2.894 3.030 density_rs2pw 119 9.7 0.004 0.004 2.799 2.950 ot_diis_step 108 11.5 0.006 0.006 2.809 2.810 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.769 2.769 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.592 2.708 calculate_first_density_matrix 1 7.0 0.000 0.004 2.603 2.609 jit_kernel_multiply 13 15.8 1.953 2.536 1.953 2.536 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.472 2.474 init_scf_loop 11 6.9 0.001 0.003 2.467 2.467 potential_pw2rs 119 12.3 0.004 0.004 2.390 2.439 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.416 2.417 cp_fm_redistribute_end 50 14.0 2.185 2.382 2.189 2.384 wfi_extrapolate 11 7.9 0.001 0.001 2.351 2.351 make_m2s 4572 13.5 0.053 0.056 2.277 2.350 cp_fm_diag_elpa_base 50 14.0 0.192 2.321 0.193 2.331 pw_transfer 1439 11.6 0.051 0.055 2.216 2.325 make_images 4572 14.5 0.133 0.139 2.195 2.268 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.141 2.252 mp_sum_d 4129 12.0 1.535 2.209 1.535 2.209 acc_transpose_blocks 54864 15.5 0.227 0.245 1.730 2.160 grid_integrate_task_list 119 12.3 2.004 2.121 2.004 2.121 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.994 2.048 fft3d_ps 1201 14.6 0.359 0.463 1.923 2.029 mp_alltoall_d11v 2130 13.8 1.747 1.929 1.747 1.929 fft_wrap_pw1pw2_140 487 13.2 0.077 0.090 1.628 1.739 mp_waitany 12084 13.8 1.371 1.614 1.371 1.614 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.407 1.426 grid_collocate_task_list 119 9.7 1.291 1.366 1.291 1.366 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.839 1.232 prepare_preconditioner 11 7.9 0.000 0.000 1.187 1.224 make_preconditioner 11 8.9 0.000 0.000 1.187 1.224 make_images_sizes 4572 15.5 0.004 0.005 0.946 1.179 mp_alltoall_i44 4572 16.5 0.942 1.175 0.942 1.175 mp_alltoall_z22v 1201 16.6 1.084 1.166 1.084 1.166 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.121 1.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.011000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.636364, yerr=1.822722 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.428096E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 695101. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 38.527 38.529 qs_mol_dyn_low 1 2.0 0.013 0.032 38.224 38.232 qs_forces 11 3.9 0.010 0.030 38.158 38.168 qs_energies 11 4.9 0.001 0.002 36.431 36.454 scf_env_do_scf 11 5.9 0.000 0.001 31.334 31.335 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.799 28.799 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.381 21.770 qs_scf_new_mos 108 7.5 0.001 0.001 19.779 20.028 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.778 20.027 ot_scf_mini 108 9.5 0.002 0.003 18.868 19.042 velocity_verlet 10 3.0 0.001 0.002 18.384 18.385 multiply_cannon 2286 13.5 0.206 0.215 16.274 17.863 multiply_cannon_loop 2286 14.5 0.905 0.980 15.130 16.612 ot_mini 108 10.5 0.001 0.001 11.802 12.036 mp_waitall_1 217478 16.2 6.078 11.223 6.078 11.223 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.115 9.564 qs_ot_get_derivative 108 11.5 0.001 0.001 9.344 9.518 multiply_cannon_multrec 27432 15.5 1.975 4.335 5.858 8.621 rebuild_ks_matrix 119 8.3 0.000 0.000 7.186 7.321 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.029 7.186 7.320 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.325 6.446 dbcsr_mm_accdrv_process 47894 16.0 2.935 5.220 3.815 5.653 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.602 4.450 qs_ot_get_p 119 10.4 0.001 0.001 4.159 4.393 sum_up_and_integrate 119 10.3 0.024 0.027 4.203 4.210 integrate_v_rspace 119 11.3 0.002 0.002 4.179 4.187 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.046 4.126 apply_single 119 13.6 0.000 0.000 3.046 4.126 mp_sum_l 7207 12.9 2.064 4.017 2.064 4.017 init_scf_run 11 5.9 0.000 0.001 3.873 3.874 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.873 3.873 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.648 3.674 calculate_rho_elec 119 8.7 0.021 0.024 3.647 3.673 make_m2s 4572 13.5 0.052 0.053 2.592 2.833 multiply_cannon_sync_h2d 27432 15.5 2.190 2.824 2.190 2.824 make_images 4572 14.5 0.199 0.237 2.505 2.744 rs_pw_transfer 974 11.9 0.010 0.011 2.633 2.741 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.701 2.720 init_scf_loop 11 6.9 0.000 0.000 2.511 2.512 ot_diis_step 108 11.5 0.011 0.011 2.403 2.404 calculate_first_density_matrix 1 7.0 0.000 0.000 2.348 2.350 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.259 2.260 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.144 2.236 calculate_dm_sparse 119 9.5 0.000 0.001 2.052 2.129 density_rs2pw 119 9.7 0.004 0.004 2.008 2.097 potential_pw2rs 119 12.3 0.006 0.007 1.937 1.952 jit_kernel_multiply 10 16.0 0.828 1.951 0.828 1.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.922 1.924 grid_integrate_task_list 119 12.3 1.837 1.920 1.837 1.920 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.913 1.913 cp_fm_redistribute_end 50 14.0 1.577 1.888 1.580 1.889 cp_fm_diag_elpa_base 50 14.0 0.299 1.833 0.307 1.863 pw_transfer 1439 11.6 0.063 0.067 1.782 1.810 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.709 1.747 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.693 1.722 prepare_preconditioner 11 7.9 0.000 0.000 1.605 1.632 make_preconditioner 11 8.9 0.000 0.000 1.605 1.632 make_images_data 4572 15.5 0.045 0.052 1.166 1.580 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.499 1.560 acc_transpose_blocks 27432 15.5 0.108 0.112 1.188 1.483 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.011 1.477 wfi_extrapolate 11 7.9 0.001 0.001 1.473 1.473 mp_alltoall_d11v 2130 13.8 1.233 1.434 1.233 1.434 fft3d_ps 1201 14.6 0.497 0.549 1.400 1.426 mp_allgather_i34 2286 14.5 0.590 1.375 0.590 1.375 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.304 1.334 grid_collocate_task_list 119 9.7 1.240 1.313 1.240 1.313 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.262 1.270 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.187 1.234 mp_sum_d 4129 12.0 0.560 1.000 0.560 1.000 qs_energies_init_hamiltonians 11 5.9 0.003 0.009 0.945 0.964 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.917 0.931 acc_transpose_blocks_kernels 27432 16.5 0.183 0.273 0.655 0.864 make_images_sizes 4572 15.5 0.005 0.005 0.583 0.829 mp_alltoall_i44 4572 16.5 0.578 0.825 0.578 0.825 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.606 0.766 0.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.529000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.818182, yerr=1.465865 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.076160E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.125 0.222 34.264 34.269 qs_mol_dyn_low 1 2.0 0.006 0.008 33.112 33.119 qs_forces 11 3.9 0.033 0.040 33.052 33.055 qs_energies 11 4.9 0.012 0.049 31.367 31.401 scf_env_do_scf 11 5.9 0.001 0.002 25.436 25.437 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.843 22.845 dbcsr_multiply_generic 2286 12.5 0.108 0.121 16.642 16.752 velocity_verlet 10 3.0 0.001 0.002 15.183 15.184 qs_scf_new_mos 108 7.5 0.001 0.001 14.788 14.804 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.788 14.803 ot_scf_mini 108 9.5 0.003 0.003 14.062 14.076 multiply_cannon 2286 13.5 0.193 0.200 13.102 13.870 multiply_cannon_loop 2286 14.5 0.640 0.667 12.285 13.073 ot_mini 108 10.5 0.001 0.001 8.790 8.803 qs_ot_get_derivative 108 11.5 0.001 0.001 7.295 7.311 multiply_cannon_multrec 18288 15.5 1.931 2.834 6.776 7.077 rebuild_ks_matrix 119 8.3 0.000 0.000 6.467 6.485 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.467 6.485 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.671 5.687 dbcsr_mm_accdrv_process 38222 16.0 4.033 5.285 4.763 5.557 init_scf_run 11 5.9 0.000 0.001 4.532 4.532 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.532 4.532 mp_waitall_1 169478 16.3 3.064 4.187 3.064 4.187 sum_up_and_integrate 119 10.3 0.030 0.031 4.042 4.063 integrate_v_rspace 119 11.3 0.002 0.003 4.011 4.037 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.727 3.402 calculate_first_density_matrix 1 7.0 0.001 0.002 3.316 3.319 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.284 3.293 calculate_rho_elec 119 8.7 0.030 0.031 3.283 3.292 qs_ot_get_p 119 10.4 0.001 0.002 3.106 3.132 init_scf_loop 11 6.9 0.001 0.003 2.575 2.580 rs_pw_transfer 974 11.9 0.009 0.010 2.380 2.485 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.443 2.429 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.084 2.422 apply_single 119 13.6 0.000 0.000 2.083 2.422 make_m2s 4572 13.5 0.044 0.046 1.941 2.096 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.014 2.020 make_images 4572 14.5 0.190 0.201 1.855 2.009 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.983 1.986 density_rs2pw 119 9.7 0.004 0.004 1.867 1.966 grid_integrate_task_list 119 12.3 1.797 1.875 1.797 1.875 jit_kernel_multiply 10 15.9 0.678 1.867 0.678 1.867 potential_pw2rs 119 12.3 0.007 0.008 1.818 1.845 calculate_dm_sparse 119 9.5 0.000 0.000 1.821 1.829 pw_transfer 1439 11.6 0.063 0.065 1.795 1.818 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.771 1.798 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.751 1.752 prepare_preconditioner 11 7.9 0.000 0.000 1.737 1.740 make_preconditioner 11 8.9 0.000 0.001 1.737 1.740 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.706 1.728 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.588 1.672 mp_sum_l 7207 12.9 1.213 1.628 1.213 1.628 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.571 1.579 multiply_cannon_sync_h2d 18288 15.5 1.385 1.573 1.385 1.573 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.511 1.512 cp_fm_redistribute_end 50 14.0 1.121 1.488 1.122 1.489 ot_diis_step 108 11.5 0.011 0.011 1.475 1.475 cp_fm_diag_elpa_base 50 14.0 0.351 1.425 0.365 1.462 fft3d_ps 1201 14.6 0.506 0.524 1.393 1.415 fft_wrap_pw1pw2_140 487 13.2 0.086 0.091 1.352 1.376 grid_collocate_task_list 119 9.7 1.203 1.301 1.203 1.301 acc_transpose_blocks 18288 15.5 0.075 0.077 1.225 1.245 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.188 1.192 wfi_extrapolate 11 7.9 0.001 0.001 1.167 1.168 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.416 1.142 qs_energies_init_hamiltonians 11 5.9 0.034 0.051 1.056 1.089 make_images_data 4572 15.5 0.045 0.049 0.823 1.004 parallel_gemm_fm 81 9.0 0.000 0.000 0.917 0.920 parallel_gemm_fm_cosma 81 10.0 0.917 0.920 0.917 0.920 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.700 0.903 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.878 0.900 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.866 0.870 mp_alltoall_d11v 2130 13.8 0.730 0.850 0.730 0.850 make_basis_sm 11 9.8 0.000 0.000 0.847 0.849 mp_alltoall_z22v 1201 16.6 0.728 0.799 0.728 0.799 acc_transpose_blocks_kernels 18288 16.5 0.212 0.219 0.784 0.796 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.663 0.747 cp_fm_cholesky_invert 11 10.9 0.739 0.743 0.739 0.743 mp_sum_d 4129 12.0 0.531 0.736 0.531 0.736 dbcsr_complete_redistribute 329 12.2 0.105 0.146 0.542 0.696 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.269000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.818182, yerr=1.898238 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 556.445696E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.100 36.447 36.448 qs_mol_dyn_low 1 2.0 0.003 0.003 35.921 35.929 qs_forces 11 3.9 0.002 0.002 35.863 35.863 qs_energies 11 4.9 0.001 0.003 34.115 34.122 scf_env_do_scf 11 5.9 0.001 0.001 28.986 28.988 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.312 25.313 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.620 18.798 velocity_verlet 10 3.0 0.001 0.002 18.365 18.367 qs_scf_new_mos 108 7.5 0.001 0.001 16.929 16.986 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.929 16.985 ot_scf_mini 108 9.5 0.002 0.003 15.972 16.027 multiply_cannon 2286 13.5 0.229 0.270 14.564 15.129 multiply_cannon_loop 2286 14.5 0.939 0.967 13.530 13.939 ot_mini 108 10.5 0.001 0.001 9.782 9.851 multiply_cannon_multrec 27432 15.5 2.351 3.042 8.608 8.963 qs_ot_get_derivative 108 11.5 0.001 0.001 7.947 8.004 dbcsr_mm_accdrv_process 47916 15.9 5.277 6.929 6.164 7.382 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 5.924 5.980 sum_up_and_integrate 119 10.3 0.035 0.038 3.865 3.873 integrate_v_rspace 119 11.3 0.002 0.003 3.830 3.839 init_scf_run 11 5.9 0.000 0.001 3.742 3.743 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.742 3.742 init_scf_loop 11 6.9 0.001 0.004 3.647 3.648 qs_ot_get_p 119 10.4 0.001 0.001 3.505 3.589 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.440 3.473 calculate_rho_elec 119 8.7 0.040 0.046 3.440 3.473 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.802 3.288 mp_waitall_1 145218 16.4 2.506 3.100 2.506 3.100 prepare_preconditioner 11 7.9 0.000 0.000 2.726 2.735 make_preconditioner 11 8.9 0.000 0.001 2.726 2.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.331 2.657 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.136 2.589 apply_single 119 13.6 0.000 0.000 2.136 2.589 make_m2s 4572 13.5 0.054 0.055 2.462 2.548 make_images 4572 14.5 0.273 0.332 2.354 2.438 rs_pw_transfer 974 11.9 0.009 0.009 2.242 2.355 calculate_first_density_matrix 1 7.0 0.000 0.002 2.303 2.305 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.110 2.122 calculate_dm_sparse 119 9.5 0.000 0.000 2.046 2.103 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.041 2.074 density_rs2pw 119 9.7 0.003 0.004 1.934 2.061 grid_integrate_task_list 119 12.3 1.824 1.913 1.824 1.913 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.881 1.882 pw_transfer 1439 11.6 0.063 0.067 1.781 1.810 ot_diis_step 108 11.5 0.012 0.012 1.796 1.796 mp_sum_l 7207 12.9 1.224 1.782 1.224 1.782 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.781 1.782 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.692 1.723 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.647 1.663 potential_pw2rs 119 12.3 0.009 0.009 1.647 1.653 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.490 1.491 jit_kernel_multiply 8 15.7 0.826 1.483 0.826 1.483 acc_transpose_blocks 27432 15.5 0.113 0.115 1.451 1.471 cp_fm_redistribute_end 50 14.0 0.983 1.458 0.985 1.459 cp_fm_diag_elpa_base 50 14.0 0.450 1.378 0.470 1.417 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.837 1.405 fft3d_ps 1201 14.6 0.530 0.580 1.373 1.398 wfi_extrapolate 11 7.9 0.001 0.001 1.382 1.382 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.317 1.349 grid_collocate_task_list 119 9.7 1.220 1.326 1.220 1.326 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.215 1.228 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.153 1.175 dbcsr_complete_redistribute 329 12.2 0.126 0.179 0.890 1.162 cp_fm_upper_to_full 72 13.5 0.810 1.120 0.810 1.120 qs_energies_init_hamiltonians 11 5.9 0.015 0.022 1.107 1.108 make_images_data 4572 15.5 0.045 0.049 0.948 1.065 multiply_cannon_sync_h2d 27432 15.5 0.976 1.035 0.976 1.035 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.808 0.982 mp_alltoall_d11v 2130 13.8 0.819 0.919 0.819 0.919 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.632 0.901 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.873 cp_fm_cholesky_invert 11 10.9 0.842 0.847 0.842 0.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.839 acc_transpose_blocks_kernels 27432 16.5 0.271 0.280 0.823 0.831 mp_alltoall_z22v 1201 16.6 0.726 0.751 0.726 0.751 mp_alltoall_i22 627 13.8 0.437 0.735 0.437 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.448000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.272727, yerr=4.069642 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 598.585344E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.088 29.532 29.532 qs_mol_dyn_low 1 2.0 0.009 0.011 28.902 28.910 qs_forces 11 3.9 0.002 0.005 28.828 28.833 qs_energies 11 4.9 0.001 0.001 27.130 27.138 scf_env_do_scf 11 5.9 0.001 0.017 21.595 21.595 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.980 18.981 velocity_verlet 10 3.0 0.017 0.017 14.391 14.394 dbcsr_multiply_generic 2286 12.5 0.093 0.095 12.611 12.730 qs_scf_new_mos 108 7.5 0.001 0.001 11.344 11.371 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.343 11.370 ot_scf_mini 108 9.5 0.002 0.002 10.657 10.682 multiply_cannon 2286 13.5 0.230 0.239 9.990 10.325 multiply_cannon_loop 2286 14.5 0.332 0.342 9.062 9.264 multiply_cannon_multrec 9144 15.5 1.672 1.902 6.022 6.258 ot_mini 108 10.5 0.001 0.001 6.015 6.044 rebuild_ks_matrix 119 8.3 0.000 0.000 5.821 5.844 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.821 5.844 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.191 5.211 qs_ot_get_derivative 108 11.5 0.001 0.001 4.708 4.731 dbcsr_mm_accdrv_process 12550 15.8 3.253 4.183 4.251 4.323 init_scf_run 11 5.9 0.000 0.001 3.924 3.924 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.924 3.924 sum_up_and_integrate 119 10.3 0.037 0.041 3.565 3.569 integrate_v_rspace 119 11.3 0.003 0.003 3.527 3.532 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.273 3.280 calculate_rho_elec 119 8.7 0.059 0.061 3.272 3.279 qs_ot_get_p 119 10.4 0.001 0.001 2.859 2.899 calculate_first_density_matrix 1 7.0 0.000 0.000 2.788 2.788 init_scf_loop 11 6.9 0.000 0.000 2.579 2.581 mp_waitall_1 121218 16.5 1.832 2.344 1.832 2.344 make_m2s 4572 13.5 0.034 0.035 1.815 1.959 grid_integrate_task_list 119 12.3 1.847 1.939 1.847 1.939 jit_kernel_multiply 10 15.9 0.960 1.932 0.960 1.932 make_images 4572 14.5 0.267 0.302 1.726 1.869 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.857 1.861 calculate_dm_sparse 119 9.5 0.000 0.000 1.822 1.842 prepare_preconditioner 11 7.9 0.000 0.000 1.818 1.821 make_preconditioner 11 8.9 0.000 0.000 1.818 1.821 rs_pw_transfer 974 11.9 0.008 0.008 1.688 1.807 density_rs2pw 119 9.7 0.004 0.004 1.673 1.783 pw_transfer 1439 11.6 0.063 0.066 1.744 1.753 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.701 1.725 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.718 1.720 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.660 1.662 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.653 1.662 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.602 1.614 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.481 1.489 qs_energies_init_hamiltonians 11 5.9 0.092 0.129 1.381 1.387 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.365 1.366 grid_collocate_task_list 119 9.7 1.271 1.360 1.271 1.360 potential_pw2rs 119 12.3 0.010 0.010 1.344 1.347 cp_fm_redistribute_end 50 14.0 0.680 1.341 0.681 1.342 fft3d_ps 1201 14.6 0.539 0.549 1.321 1.331 fft_wrap_pw1pw2_140 487 13.2 0.082 0.089 1.321 1.329 cp_fm_diag_elpa_base 50 14.0 0.616 1.268 0.659 1.323 ot_diis_step 108 11.5 0.013 0.013 1.293 1.293 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.232 1.237 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.198 1.210 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.165 1.183 apply_single 119 13.6 0.000 0.000 1.165 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.091 1.091 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.825 1.056 make_images_data 4572 15.5 0.039 0.043 0.859 1.039 acc_transpose_blocks 9144 15.5 0.039 0.040 1.011 1.017 mp_alltoall_d11v 2130 13.8 0.840 0.963 0.840 0.963 cp_fm_cholesky_invert 11 10.9 0.920 0.923 0.920 0.923 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.918 multiply_cannon_sync_h2d 9144 15.5 0.713 0.780 0.713 0.780 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.770 0.772 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.707 0.757 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.755 parallel_gemm_fm 81 9.0 0.000 0.000 0.750 0.751 parallel_gemm_fm_cosma 81 10.0 0.750 0.751 0.750 0.751 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.740 0.749 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.385 0.725 make_basis_sm 11 9.8 0.000 0.000 0.713 0.714 mp_allgather_i34 2286 14.5 0.260 0.702 0.260 0.702 mp_alltoall_z22v 1201 16.6 0.656 0.695 0.656 0.695 jit_kernel_transpose 5 15.6 0.652 0.655 0.652 0.655 dbcsr_complete_redistribute 329 12.2 0.201 0.245 0.617 0.645 qs_create_task_list 11 7.9 0.036 0.036 0.581 0.605 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.532000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.454545, yerr=4.075729 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 755.159040E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.093 41.460 41.461 qs_mol_dyn_low 1 2.0 0.054 0.085 40.851 40.858 qs_forces 11 3.9 0.003 0.006 40.742 40.753 qs_energies 11 4.9 0.001 0.001 38.790 38.803 scf_env_do_scf 11 5.9 0.001 0.001 32.928 32.928 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.980 24.981 velocity_verlet 10 3.0 0.002 0.002 22.988 22.993 dbcsr_multiply_generic 2286 12.5 0.101 0.101 17.655 17.722 qs_scf_new_mos 108 7.5 0.001 0.001 16.065 16.159 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.065 16.158 ot_scf_mini 108 9.5 0.002 0.002 14.985 15.080 multiply_cannon 2286 13.5 0.308 0.318 13.711 14.533 multiply_cannon_loop 2286 14.5 0.343 0.350 12.407 13.245 ot_mini 108 10.5 0.001 0.001 9.023 9.135 multiply_cannon_multrec 9144 15.5 3.433 4.804 8.567 8.637 init_scf_loop 11 6.9 0.000 0.000 7.919 7.922 qs_ot_get_derivative 108 11.5 0.001 0.001 6.993 7.089 rebuild_ks_matrix 119 8.3 0.000 0.000 6.904 7.041 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.903 7.041 prepare_preconditioner 11 7.9 0.000 0.000 6.941 6.954 make_preconditioner 11 8.9 0.000 0.000 6.940 6.954 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.484 6.831 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.247 6.371 dbcsr_mm_accdrv_process 12550 15.8 4.094 5.611 5.013 6.350 cp_fm_upper_to_full 72 14.2 3.188 4.588 3.188 4.588 init_scf_run 11 5.9 0.000 0.001 3.817 3.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.817 3.817 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.694 3.702 calculate_rho_elec 119 8.7 0.118 0.121 3.694 3.702 sum_up_and_integrate 119 10.3 0.064 0.066 3.690 3.696 integrate_v_rspace 119 11.3 0.003 0.003 3.625 3.632 mp_waitall_1 97218 16.6 2.600 3.495 2.600 3.495 qs_ot_get_p 119 10.4 0.001 0.001 3.203 3.341 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.446 2.897 dbcsr_complete_redistribute 329 12.2 0.288 0.296 2.007 2.843 make_m2s 4572 13.5 0.037 0.038 2.331 2.551 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.688 2.504 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.223 2.457 apply_single 119 13.6 0.000 0.000 2.223 2.457 make_images 4572 14.5 0.351 0.381 2.212 2.430 calculate_first_density_matrix 1 7.0 0.000 0.000 2.392 2.395 mp_alltoall_i22 627 13.8 1.493 2.339 1.493 2.339 calculate_dm_sparse 119 9.5 0.000 0.000 2.246 2.264 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.452 2.261 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.327 2.140 grid_integrate_task_list 119 12.3 2.021 2.047 2.021 2.047 pw_transfer 1439 11.6 0.066 0.067 2.041 2.045 ot_diis_step 108 11.5 0.014 0.015 2.006 2.007 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.945 1.950 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.934 1.936 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.928 1.929 density_rs2pw 119 9.7 0.003 0.003 1.783 1.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.776 1.787 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.697 1.748 mp_sum_l 7207 12.9 1.069 1.740 1.069 1.740 jit_kernel_multiply 10 15.5 0.893 1.724 0.893 1.724 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.661 1.661 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.598 1.624 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.618 1.624 fft3d_ps 1201 14.6 0.567 0.578 1.580 1.583 cp_fm_cholesky_invert 11 10.9 1.505 1.509 1.505 1.509 grid_collocate_task_list 119 9.7 1.457 1.469 1.457 1.469 rs_pw_transfer 974 11.9 0.009 0.009 1.418 1.447 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.169 1.442 make_images_data 4572 15.5 0.043 0.046 1.113 1.374 wfi_extrapolate 11 7.9 0.001 0.001 1.369 1.369 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.359 1.360 cp_fm_diag_elpa_base 50 14.0 1.213 1.268 1.357 1.357 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.315 1.323 potential_pw2rs 119 12.3 0.014 0.015 1.221 1.225 mp_alltoall_d11v 2130 13.8 1.169 1.189 1.169 1.189 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.150 1.168 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.113 1.139 multiply_cannon_sync_h2d 9144 15.5 1.040 1.045 1.040 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.990 1.042 acc_transpose_blocks 9144 15.5 0.039 0.039 0.973 0.979 qs_create_task_list 11 7.9 0.001 0.001 0.941 0.951 generate_qs_task_list 11 8.9 0.372 0.391 0.941 0.950 mp_alltoall_z22v 1201 16.6 0.877 0.894 0.877 0.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.869 0.882 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.787 0.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.461000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=710.727273, yerr=11.378536 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 501.358592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1441588. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 83.387 83.388 qs_mol_dyn_low 1 2.0 0.003 0.004 82.942 82.952 qs_forces 11 3.9 0.038 0.052 82.703 82.703 qs_energies 11 4.9 0.001 0.002 79.873 79.917 scf_env_do_scf 11 5.9 0.000 0.001 70.988 70.991 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.436 65.437 dbcsr_multiply_generic 2055 12.4 0.105 0.108 52.490 52.710 qs_scf_new_mos 99 7.5 0.000 0.001 48.525 48.651 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.524 48.650 ot_scf_mini 99 9.5 0.002 0.002 46.061 46.136 multiply_cannon 2055 13.4 0.174 0.178 42.650 43.358 velocity_verlet 10 3.0 0.011 0.016 42.634 42.637 multiply_cannon_loop 2055 14.4 1.571 1.611 41.653 42.307 ot_mini 99 10.5 0.001 0.001 27.960 28.037 qs_ot_get_derivative 99 11.5 0.001 0.001 21.117 21.196 multiply_cannon_multrec 49320 15.4 12.026 12.713 17.090 17.825 rebuild_ks_matrix 110 8.3 0.000 0.001 14.385 14.478 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.384 14.477 mp_waitall_1 241148 16.1 11.983 12.821 11.983 12.821 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.649 12.725 multiply_cannon_sync_h2d 49320 15.4 9.884 10.503 9.884 10.503 qs_ot_get_p 110 10.4 0.001 0.001 9.439 9.536 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.145 8.175 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.228 7.775 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.160 7.676 apply_single 110 13.6 0.000 0.000 7.160 7.675 sum_up_and_integrate 110 10.3 0.036 0.043 6.998 7.012 integrate_v_rspace 110 11.3 0.002 0.003 6.962 6.984 init_scf_run 11 5.9 0.000 0.001 6.868 6.869 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.868 6.869 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.686 6.722 ot_diis_step 99 11.5 0.006 0.006 6.526 6.527 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.223 6.355 calculate_rho_elec 110 8.6 0.020 0.024 6.223 6.355 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.115 6.143 mp_sum_l 6514 12.8 5.482 6.143 5.482 6.143 init_scf_loop 11 6.9 0.000 0.000 5.522 5.523 dbcsr_mm_accdrv_process 87628 16.1 2.125 2.234 4.944 5.252 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.079 5.080 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.576 4.577 cp_fm_redistribute_end 48 14.0 3.984 4.548 3.988 4.549 cp_fm_diag_elpa_base 48 14.0 0.555 4.447 0.558 4.473 make_m2s 4110 13.4 0.061 0.065 3.910 4.035 wfi_extrapolate 11 7.9 0.001 0.001 4.015 4.016 rs_pw_transfer 902 11.9 0.012 0.013 3.773 3.983 make_images 4110 14.4 0.178 0.192 3.815 3.944 calculate_dm_sparse 110 9.5 0.000 0.001 3.829 3.938 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.712 3.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.610 3.613 density_rs2pw 110 9.6 0.004 0.004 3.290 3.488 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.419 3.466 grid_integrate_task_list 110 12.3 3.263 3.399 3.263 3.399 prepare_preconditioner 11 7.9 0.000 0.000 3.376 3.390 make_preconditioner 11 8.9 0.000 0.000 3.376 3.390 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.218 3.250 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.158 3.204 pw_transfer 1331 11.6 0.054 0.061 2.854 2.917 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.767 2.832 calculate_first_density_matrix 1 7.0 0.000 0.000 2.765 2.771 potential_pw2rs 110 12.3 0.006 0.007 2.589 2.616 jit_kernel_multiply 13 15.9 2.536 2.548 2.536 2.548 mp_alltoall_d11v 2046 13.8 2.128 2.543 2.128 2.543 fft_wrap_pw1pw2_140 451 13.1 0.168 0.187 2.297 2.362 fft3d_ps 1111 14.6 0.749 0.836 2.305 2.359 acc_transpose_blocks 49320 15.4 0.216 0.226 2.173 2.238 grid_collocate_task_list 110 9.6 2.090 2.199 2.090 2.199 mp_waitany 14300 13.8 1.837 2.153 1.837 2.153 mp_sum_d 3883 11.9 1.466 1.960 1.466 1.960 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.875 1.896 make_images_data 4110 15.4 0.043 0.046 1.764 1.885 cp_fm_cholesky_invert 11 10.9 1.832 1.836 1.832 1.836 hybrid_alltoall_any 4261 16.3 0.081 0.482 1.534 1.789 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.388000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.377581 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 591.151104E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1808072. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.192 0.241 70.939 70.939 qs_mol_dyn_low 1 2.0 0.004 0.004 70.230 70.261 qs_forces 11 3.9 0.002 0.003 70.159 70.161 qs_energies 11 4.9 0.001 0.001 66.811 66.816 scf_env_do_scf 11 5.9 0.000 0.001 57.905 57.908 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.712 49.713 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.006 38.158 velocity_verlet 10 3.0 0.001 0.002 37.202 37.203 qs_scf_new_mos 99 7.5 0.001 0.001 33.376 33.503 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.375 33.503 multiply_cannon 2055 13.4 0.220 0.243 31.016 32.155 ot_scf_mini 99 9.5 0.003 0.003 31.698 31.820 multiply_cannon_loop 2055 14.4 0.932 0.955 29.627 30.525 ot_mini 99 10.5 0.001 0.001 18.643 18.772 multiply_cannon_multrec 24660 15.4 7.648 9.192 13.869 15.267 rebuild_ks_matrix 110 8.3 0.000 0.001 13.906 13.993 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.906 13.992 qs_ot_get_derivative 99 11.5 0.001 0.001 12.833 12.953 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.277 12.352 mp_waitall_1 186928 16.3 8.248 10.888 8.248 10.888 multiply_cannon_sync_h2d 24660 15.4 7.006 8.295 7.006 8.295 init_scf_loop 11 6.9 0.000 0.000 8.153 8.154 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.077 7.499 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.535 7.292 apply_single 110 13.6 0.000 0.001 6.535 7.292 sum_up_and_integrate 110 10.3 0.052 0.059 6.852 6.861 integrate_v_rspace 110 11.3 0.002 0.003 6.799 6.813 qs_ot_get_p 110 10.4 0.001 0.001 6.380 6.530 init_scf_run 11 5.9 0.000 0.001 6.440 6.440 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.439 6.440 dbcsr_mm_accdrv_process 52282 16.1 4.730 5.407 6.062 6.395 prepare_preconditioner 11 7.9 0.000 0.000 6.103 6.133 make_preconditioner 11 8.9 0.000 0.000 6.103 6.133 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.050 6.058 calculate_rho_elec 110 8.6 0.039 0.047 6.049 6.058 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.663 5.826 ot_diis_step 99 11.5 0.010 0.010 5.762 5.763 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.744 5.477 make_m2s 4110 13.4 0.056 0.059 4.510 4.950 make_images 4110 14.4 0.396 0.439 4.402 4.838 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.285 4.306 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.824 3.824 pw_transfer 1331 11.6 0.065 0.070 3.504 3.656 wfi_extrapolate 11 7.9 0.001 0.001 3.636 3.636 rs_pw_transfer 902 11.9 0.012 0.013 3.380 3.550 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.397 3.548 density_rs2pw 110 9.6 0.004 0.004 3.359 3.524 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.366 3.433 grid_integrate_task_list 110 12.3 3.135 3.325 3.135 3.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.275 3.277 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.162 3.164 cp_fm_redistribute_end 48 14.0 2.347 3.119 2.350 3.120 cp_fm_diag_elpa_base 48 14.0 0.734 2.983 0.765 3.066 calculate_dm_sparse 110 9.5 0.001 0.001 2.982 3.009 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.890 2.947 fft3d_ps 1111 14.6 1.071 1.267 2.758 2.893 make_images_data 4110 15.4 0.048 0.051 2.420 2.864 cp_fm_cholesky_invert 11 10.9 2.843 2.851 2.843 2.851 fft_wrap_pw1pw2_140 451 13.1 0.197 0.217 2.670 2.815 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.126 2.786 potential_pw2rs 110 12.3 0.008 0.009 2.706 2.742 calculate_first_density_matrix 1 7.0 0.000 0.000 2.711 2.714 mp_sum_l 6514 12.8 1.897 2.608 1.897 2.608 grid_collocate_task_list 110 9.6 2.121 2.271 2.121 2.271 mp_alltoall_d11v 2046 13.8 1.908 2.137 1.908 2.137 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.996 2.016 qs_energies_init_hamiltonians 11 5.9 0.020 0.033 1.947 1.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.795 1.808 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.784 1.716 mp_allgather_i34 2055 14.4 0.718 1.698 0.718 1.698 jit_kernel_multiply 8 16.2 0.980 1.617 0.980 1.617 mp_irecv_dv 57340 16.2 0.661 1.602 0.661 1.602 cp_fm_cholesky_decompose 22 10.9 1.559 1.565 1.559 1.565 dbcsr_complete_redistribute 325 12.2 0.234 0.302 1.272 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.549 acc_transpose_blocks 24660 15.4 0.107 0.110 1.516 1.538 mp_alltoall_z22v 1111 16.6 1.396 1.493 1.396 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.374 1.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.939000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.272727, yerr=8.708313 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 658.571264E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.084 60.845 60.846 qs_mol_dyn_low 1 2.0 0.003 0.003 60.335 60.344 qs_forces 11 3.9 0.036 0.053 60.270 60.271 qs_energies 11 4.9 0.001 0.003 57.066 57.103 scf_env_do_scf 11 5.9 0.001 0.001 48.917 48.917 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.106 40.107 velocity_verlet 10 3.0 0.001 0.002 32.778 32.780 dbcsr_multiply_generic 2055 12.4 0.108 0.112 28.551 28.776 qs_scf_new_mos 99 7.5 0.001 0.001 25.201 25.289 qs_scf_loop_do_ot 99 8.5 0.001 0.002 25.200 25.289 ot_scf_mini 99 9.5 0.003 0.004 23.970 24.081 multiply_cannon 2055 13.4 0.210 0.220 22.106 23.295 multiply_cannon_loop 2055 14.4 0.621 0.637 20.861 21.874 ot_mini 99 10.5 0.001 0.001 13.799 13.913 rebuild_ks_matrix 110 8.3 0.000 0.000 12.502 12.609 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.501 12.608 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.067 11.162 multiply_cannon_multrec 16440 15.4 3.951 5.269 9.704 10.917 mp_waitall_1 146766 16.3 7.289 10.504 7.289 10.504 qs_ot_get_derivative 99 11.5 0.001 0.001 9.247 9.359 init_scf_loop 11 6.9 0.000 0.002 8.775 8.777 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.187 7.282 prepare_preconditioner 11 7.9 0.000 0.000 6.962 6.976 make_preconditioner 11 8.9 0.000 0.001 6.962 6.976 sum_up_and_integrate 110 10.3 0.060 0.061 6.720 6.736 integrate_v_rspace 110 11.3 0.003 0.003 6.660 6.677 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.305 6.664 dbcsr_mm_accdrv_process 34862 16.1 4.665 5.367 5.606 5.762 init_scf_run 11 5.9 0.000 0.001 5.728 5.729 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.728 5.728 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.638 5.648 calculate_rho_elec 110 8.6 0.058 0.058 5.638 5.648 qs_ot_get_p 110 10.4 0.001 0.001 5.404 5.543 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.019 5.452 apply_single 110 13.6 0.000 0.000 5.019 5.451 make_m2s 4110 13.4 0.050 0.051 4.330 4.707 make_images 4110 14.4 0.390 0.512 4.215 4.595 ot_diis_step 99 11.5 0.011 0.011 4.524 4.525 multiply_cannon_sync_h2d 16440 15.4 3.722 4.384 3.722 4.384 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.147 3.778 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.677 3.680 grid_integrate_task_list 110 12.3 3.198 3.400 3.198 3.400 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.328 3.328 pw_transfer 1331 11.6 0.064 0.070 3.284 3.294 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.178 3.190 density_rs2pw 110 9.6 0.004 0.004 2.913 3.098 rs_pw_transfer 902 11.9 0.010 0.011 2.819 2.992 make_images_data 4110 15.4 0.044 0.049 2.484 2.982 wfi_extrapolate 11 7.9 0.001 0.001 2.972 2.972 hybrid_alltoall_any 4261 16.3 0.105 0.371 2.217 2.937 cp_fm_cholesky_invert 11 10.9 2.776 2.783 2.776 2.783 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.775 2.779 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.756 2.758 cp_fm_redistribute_end 48 14.0 1.726 2.747 1.728 2.747 cp_fm_diag_elpa_base 48 14.0 0.955 2.597 1.013 2.705 fft_wrap_pw1pw2_140 451 13.1 0.208 0.212 2.670 2.682 calculate_first_density_matrix 1 7.0 0.000 0.003 2.673 2.674 calculate_dm_sparse 110 9.5 0.001 0.001 2.510 2.538 fft3d_ps 1111 14.6 1.063 1.075 2.493 2.509 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.430 2.469 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.398 2.457 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.857 2.384 grid_collocate_task_list 110 9.6 2.174 2.375 2.174 2.375 potential_pw2rs 110 12.3 0.011 0.011 2.344 2.361 mp_alltoall_d11v 2046 13.8 2.015 2.344 2.015 2.344 mp_irecv_dv 48980 15.7 0.788 2.261 0.788 2.261 mp_sum_l 6514 12.8 1.546 2.206 1.546 2.206 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.974 2.008 dbcsr_complete_redistribute 325 12.2 0.321 0.346 1.421 1.888 cp_fm_upper_to_full 70 13.6 1.390 1.829 1.390 1.829 cp_fm_cholesky_decompose 22 10.9 1.691 1.730 1.691 1.730 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.647 mp_allgather_i34 2055 14.4 0.571 1.578 0.571 1.578 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.463 1.474 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.364 1.469 jit_kernel_multiply 8 16.3 0.552 1.469 0.552 1.469 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.968 1.427 rs_gather_matrices 110 12.3 0.138 0.150 1.064 1.381 mp_waitany 17072 13.8 1.151 1.295 1.151 1.295 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.261 1.271 mp_alltoall_z22v 1111 16.6 1.217 1.245 1.217 1.245 acc_transpose_blocks 16440 15.4 0.073 0.076 1.216 1.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.846000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.181818, yerr=7.952337 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 734.347264E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.041 67.333 67.334 qs_mol_dyn_low 1 2.0 0.003 0.003 66.667 66.677 qs_forces 11 3.9 0.002 0.002 66.589 66.590 qs_energies 11 4.9 0.005 0.030 63.191 63.196 scf_env_do_scf 11 5.9 0.001 0.001 54.772 54.775 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.277 42.278 velocity_verlet 10 3.0 0.001 0.002 38.086 38.088 dbcsr_multiply_generic 2055 12.4 0.116 0.119 29.908 30.126 qs_scf_new_mos 99 7.5 0.001 0.001 27.476 27.585 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.476 27.585 ot_scf_mini 99 9.5 0.003 0.003 25.829 25.947 multiply_cannon 2055 13.4 0.242 0.261 22.529 23.719 multiply_cannon_loop 2055 14.4 0.883 0.907 20.997 21.648 ot_mini 99 10.5 0.001 0.001 14.482 14.616 multiply_cannon_multrec 24660 15.4 4.163 6.800 12.600 13.845 init_scf_loop 11 6.9 0.001 0.003 12.442 12.443 rebuild_ks_matrix 110 8.3 0.000 0.000 12.123 12.219 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.123 12.219 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.747 10.830 prepare_preconditioner 11 7.9 0.000 0.000 10.675 10.691 make_preconditioner 11 8.9 0.000 0.002 10.675 10.691 qs_ot_get_derivative 99 11.5 0.001 0.001 10.268 10.396 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.862 10.353 dbcsr_mm_accdrv_process 52304 16.0 6.883 8.737 8.292 9.229 mp_waitall_1 126806 16.4 4.911 6.897 4.911 6.897 sum_up_and_integrate 110 10.3 0.068 0.071 6.500 6.513 integrate_v_rspace 110 11.3 0.003 0.003 6.432 6.445 qs_ot_get_p 110 10.4 0.001 0.001 6.259 6.423 make_m2s 4110 13.4 0.059 0.062 5.654 5.965 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.876 5.885 calculate_rho_elec 110 8.6 0.077 0.081 5.875 5.884 make_images 4110 14.4 0.575 0.690 5.513 5.820 init_scf_run 11 5.9 0.000 0.001 5.594 5.595 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.594 5.595 cp_fm_upper_to_full 70 13.8 3.397 4.813 3.397 4.813 qs_ot_p2m_diag 48 11.0 0.054 0.063 4.276 4.292 ot_diis_step 99 11.5 0.011 0.012 4.171 4.172 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.063 4.169 apply_single 110 13.6 0.000 0.000 4.063 4.168 dbcsr_complete_redistribute 325 12.2 0.463 0.508 2.884 4.057 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.772 3.774 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.521 3.462 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.402 3.452 grid_integrate_task_list 110 12.3 3.283 3.439 3.283 3.439 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.271 3.437 pw_transfer 1331 11.6 0.065 0.072 3.398 3.430 multiply_cannon_sync_h2d 24660 15.4 3.176 3.377 3.176 3.377 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.292 3.327 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.175 3.250 make_images_data 4110 15.4 0.046 0.050 2.878 3.194 density_rs2pw 110 9.6 0.004 0.004 3.020 3.185 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.445 3.147 cp_fm_cholesky_invert 11 10.9 3.123 3.134 3.123 3.134 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.128 3.131 cp_fm_redistribute_end 48 14.0 1.560 3.089 1.561 3.090 wfi_extrapolate 11 7.9 0.001 0.001 3.082 3.082 cp_fm_diag_elpa_base 48 14.0 1.443 2.942 1.524 3.055 calculate_dm_sparse 110 9.5 0.001 0.001 2.936 2.970 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.804 2.955 mp_alltoall_i22 605 13.7 1.708 2.953 1.708 2.953 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.870 2.872 rs_pw_transfer 902 11.9 0.010 0.010 2.692 2.867 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.731 2.768 fft3d_ps 1111 14.6 1.064 1.105 2.596 2.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.413 2.453 calculate_first_density_matrix 1 7.0 0.000 0.002 2.423 2.427 grid_collocate_task_list 110 9.6 2.219 2.374 2.219 2.374 potential_pw2rs 110 12.3 0.012 0.013 2.263 2.281 qs_energies_init_hamiltonians 11 5.9 0.025 0.028 2.273 2.275 mp_alltoall_d11v 2046 13.8 1.904 2.155 1.904 2.155 jit_kernel_multiply 11 15.6 1.080 1.980 1.080 1.980 cp_fm_cholesky_decompose 22 10.9 1.877 1.920 1.877 1.920 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.755 1.784 mp_allgather_i34 2055 14.4 0.639 1.753 0.639 1.753 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.603 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.624 1.636 mp_sum_l 6514 12.8 1.039 1.571 1.039 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.546 acc_transpose_blocks 24660 15.4 0.106 0.107 1.500 1.516 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.852 1.422 mp_alltoall_z22v 1111 16.6 1.316 1.383 1.316 1.383 mp_irecv_dv 62702 16.1 0.753 1.353 0.753 1.353 mp_waitany 13376 13.8 1.094 1.349 1.094 1.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.334000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.090909, yerr=11.040871 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 836.612096E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.190 0.225 56.931 56.932 qs_mol_dyn_low 1 2.0 0.003 0.003 56.135 56.197 qs_forces 11 3.9 0.020 0.021 56.066 56.066 qs_energies 11 4.9 0.001 0.001 52.407 52.410 scf_env_do_scf 11 5.9 0.000 0.001 43.838 43.838 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.537 35.538 velocity_verlet 10 3.0 0.002 0.002 31.512 31.515 dbcsr_multiply_generic 2055 12.4 0.104 0.107 23.180 23.320 qs_scf_new_mos 99 7.5 0.001 0.001 21.270 21.352 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.269 21.351 ot_scf_mini 99 9.5 0.002 0.002 20.017 20.096 multiply_cannon 2055 13.4 0.252 0.273 17.487 18.806 multiply_cannon_loop 2055 14.4 0.324 0.338 16.133 16.384 rebuild_ks_matrix 110 8.3 0.000 0.000 11.611 11.659 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.611 11.659 ot_mini 99 10.5 0.001 0.001 10.666 10.742 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.353 10.396 multiply_cannon_multrec 8220 15.4 3.246 4.821 7.457 8.601 mp_waitall_1 106626 16.5 6.534 8.518 6.534 8.518 init_scf_loop 11 6.9 0.000 0.000 8.252 8.254 qs_ot_get_derivative 99 11.5 0.001 0.001 6.822 6.905 prepare_preconditioner 11 7.9 0.000 0.000 6.548 6.558 make_preconditioner 11 8.9 0.000 0.000 6.548 6.558 sum_up_and_integrate 110 10.3 0.079 0.081 6.270 6.286 integrate_v_rspace 110 11.3 0.003 0.003 6.191 6.205 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.114 6.188 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.720 5.732 calculate_rho_elec 110 8.6 0.115 0.115 5.719 5.731 qs_ot_get_p 110 10.4 0.001 0.001 5.363 5.443 init_scf_run 11 5.9 0.000 0.001 5.344 5.344 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.344 5.344 dbcsr_mm_accdrv_process 17442 15.9 2.838 3.944 4.080 4.971 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.146 4.790 make_m2s 4110 13.4 0.038 0.040 4.326 4.597 make_images 4110 14.4 0.636 0.688 4.196 4.466 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.898 3.902 ot_diis_step 99 11.5 0.012 0.014 3.819 3.819 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.734 3.773 apply_single 110 13.6 0.000 0.000 3.734 3.773 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.566 3.567 grid_integrate_task_list 110 12.3 3.378 3.557 3.378 3.557 pw_transfer 1331 11.6 0.065 0.069 3.337 3.341 cp_fm_cholesky_invert 11 10.9 3.247 3.251 3.247 3.251 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.230 3.239 multiply_cannon_sync_h2d 8220 15.4 2.881 3.025 2.881 3.025 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.993 2.994 cp_fm_redistribute_end 48 14.0 0.769 2.959 0.775 2.959 cp_fm_diag_elpa_base 48 14.0 1.993 2.746 2.174 2.906 make_images_data 4110 15.4 0.038 0.044 2.463 2.898 density_rs2pw 110 9.6 0.004 0.004 2.772 2.856 wfi_extrapolate 11 7.9 0.001 0.001 2.799 2.799 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.767 2.767 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.735 2.747 hybrid_alltoall_any 4261 16.3 0.199 0.857 2.367 2.729 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.536 2.538 calculate_dm_sparse 110 9.5 0.001 0.001 2.487 2.517 fft3d_ps 1111 14.6 1.111 1.122 2.489 2.500 grid_collocate_task_list 110 9.6 2.324 2.494 2.324 2.494 calculate_first_density_matrix 1 7.0 0.000 0.000 2.457 2.459 rs_pw_transfer 902 11.9 0.010 0.010 2.325 2.454 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.160 2.170 potential_pw2rs 110 12.3 0.015 0.016 1.994 2.000 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.995 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.958 1.993 mp_alltoall_d11v 2046 13.8 1.713 1.943 1.713 1.943 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.886 1.930 cp_fm_cholesky_decompose 22 10.9 1.828 1.849 1.828 1.849 jit_kernel_multiply 10 15.8 0.936 1.734 0.936 1.734 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.557 1.707 mp_allgather_i34 2055 14.4 0.539 1.660 0.539 1.660 dbcsr_complete_redistribute 325 12.2 0.553 0.576 1.535 1.630 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.593 1.599 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.860 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.440 1.449 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.477 1.379 mp_irecv_dv 24056 15.7 0.453 1.336 0.453 1.336 qs_create_task_list 11 7.9 0.001 0.001 1.226 1.323 generate_qs_task_list 11 8.9 0.378 0.447 1.226 1.322 mp_waitany 9240 13.8 1.103 1.256 1.103 1.256 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.204 1.256 mp_alltoall_z22v 1111 16.6 1.144 1.157 1.144 1.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.932000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.454545, yerr=13.693818 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 1.315090E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.159 0.162 87.072 87.073 qs_mol_dyn_low 1 2.0 0.034 0.050 86.171 86.179 qs_forces 11 3.9 0.002 0.003 85.739 85.756 qs_energies 11 4.9 0.001 0.001 81.687 81.705 scf_env_do_scf 11 5.9 0.000 0.001 71.851 71.851 velocity_verlet 10 3.0 0.002 0.002 55.678 55.726 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.730 42.732 init_scf_loop 11 6.9 0.000 0.000 29.047 29.053 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.487 28.564 prepare_preconditioner 11 7.9 0.000 0.000 27.102 27.108 make_preconditioner 11 8.9 0.000 0.000 27.102 27.108 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.076 26.579 qs_scf_new_mos 99 7.5 0.001 0.001 26.350 26.398 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.350 26.397 ot_scf_mini 99 9.5 0.002 0.002 24.577 24.611 multiply_cannon 2055 13.4 0.350 0.382 21.545 22.272 multiply_cannon_loop 2055 14.4 0.342 0.344 19.755 20.091 cp_fm_upper_to_full 70 14.2 12.972 18.724 12.972 18.724 ot_mini 99 10.5 0.001 0.001 13.549 13.585 rebuild_ks_matrix 110 8.3 0.000 0.001 13.130 13.172 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.129 13.172 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.945 11.984 dbcsr_complete_redistribute 325 12.2 1.024 1.044 7.608 11.093 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.585 10.064 multiply_cannon_multrec 8220 15.4 4.362 4.592 9.585 9.699 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.010 9.448 mp_alltoall_i22 605 13.7 5.636 9.133 5.636 9.133 qs_ot_get_derivative 99 11.5 0.001 0.001 9.004 9.040 mp_waitall_1 87304 16.6 7.799 8.679 7.799 8.679 sum_up_and_integrate 110 10.3 0.151 0.152 6.635 6.648 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.484 6.520 calculate_rho_elec 110 8.6 0.227 0.227 6.483 6.519 integrate_v_rspace 110 11.3 0.004 0.004 6.484 6.497 qs_ot_get_p 110 10.4 0.001 0.001 5.830 5.884 make_m2s 4110 13.4 0.043 0.043 5.193 5.700 init_scf_run 11 5.9 0.000 0.001 5.587 5.587 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.586 5.587 cp_fm_cholesky_invert 11 10.9 5.529 5.533 5.529 5.533 make_images 4110 14.4 0.879 0.927 5.007 5.513 dbcsr_mm_accdrv_process 11614 15.7 3.301 3.947 5.079 5.329 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.815 5.163 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.659 5.127 apply_single 110 13.6 0.000 0.000 4.659 5.126 ot_diis_step 99 11.5 0.015 0.015 4.522 4.523 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.277 4.283 multiply_cannon_sync_h2d 8220 15.4 3.947 3.955 3.947 3.955 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.831 3.832 pw_transfer 1331 11.6 0.073 0.074 3.766 3.771 grid_integrate_task_list 110 12.3 3.665 3.716 3.665 3.716 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.650 3.668 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.650 3.655 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.868 3.535 make_images_data 4110 15.4 0.042 0.045 2.814 3.486 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.883 3.331 wfi_extrapolate 11 7.9 0.001 0.001 3.228 3.228 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.207 3.207 cp_fm_diag_elpa_base 48 14.0 2.660 2.866 3.205 3.206 calculate_dm_sparse 110 9.5 0.001 0.001 3.161 3.185 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.154 3.163 density_rs2pw 110 9.6 0.004 0.004 2.984 3.002 fft3d_ps 1111 14.6 1.265 1.274 2.859 2.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.839 2.844 grid_collocate_task_list 110 9.6 2.627 2.653 2.627 2.653 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.392 2.414 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.339 2.351 calculate_first_density_matrix 1 7.0 0.000 0.000 2.252 2.254 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.174 2.238 rs_pw_transfer 902 11.9 0.010 0.011 2.143 2.219 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.187 cp_fm_cholesky_decompose 22 10.9 2.087 2.106 2.087 2.106 mp_alltoall_d11v 2046 13.8 2.030 2.071 2.030 2.071 potential_pw2rs 110 12.3 0.021 0.021 2.063 2.067 qs_create_task_list 11 7.9 0.000 0.000 1.894 1.940 generate_qs_task_list 11 8.9 0.738 0.790 1.894 1.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.796 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.748 1.772 jit_kernel_multiply 10 15.4 1.578 1.759 1.578 1.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.073000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1184.363636, yerr=56.253117 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.758848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 3352577. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.212 207.636 207.639 qs_mol_dyn_low 1 2.0 0.030 0.068 206.545 206.559 qs_forces 11 3.9 0.023 0.055 206.409 206.437 qs_energies 11 4.9 0.008 0.023 200.874 200.927 scf_env_do_scf 11 5.9 0.001 0.021 183.562 183.566 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.487 162.489 dbcsr_multiply_generic 2507 12.6 0.175 0.179 124.864 125.374 velocity_verlet 10 3.0 0.006 0.007 124.312 124.313 qs_scf_new_mos 117 7.6 0.001 0.001 123.343 123.551 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.342 123.551 ot_scf_mini 117 9.6 0.003 0.004 116.748 116.901 multiply_cannon 2507 13.6 0.239 0.247 101.110 103.556 multiply_cannon_loop 2507 14.6 2.128 2.204 98.765 100.733 ot_mini 117 10.6 0.001 0.001 66.448 66.636 multiply_cannon_multrec 60168 15.6 33.049 35.006 41.442 43.208 qs_ot_get_derivative 117 11.6 0.001 0.001 41.648 41.813 rebuild_ks_matrix 128 8.3 0.001 0.001 33.409 33.663 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.019 33.408 33.662 mp_waitall_1 291448 16.2 29.241 32.476 29.241 32.476 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.074 30.325 multiply_cannon_sync_h2d 60168 15.6 27.175 28.979 27.175 28.979 qs_ot_get_p 128 10.4 0.001 0.001 28.404 28.630 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.193 24.827 apply_single 128 13.6 0.001 0.001 24.192 24.827 ot_diis_step 117 11.6 0.007 0.008 24.448 24.449 qs_ot_p2m_diag 83 11.4 0.077 0.091 21.542 21.621 init_scf_loop 11 6.9 0.001 0.004 20.975 20.976 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.332 19.490 multiply_cannon_metrocomm3 60168 15.6 0.110 0.115 15.884 19.474 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.849 18.850 prepare_preconditioner 11 7.9 0.000 0.000 16.366 16.416 make_preconditioner 11 8.9 0.000 0.003 16.366 16.416 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.813 15.819 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.589 15.769 cp_fm_redistribute_end 83 14.4 12.474 15.722 12.487 15.725 cp_fm_diag_elpa_base 83 14.4 3.188 15.420 3.224 15.539 make_m2s 5014 13.6 0.103 0.110 13.929 14.256 sum_up_and_integrate 128 10.3 0.090 0.110 14.083 14.101 make_images 5014 14.6 0.403 0.422 13.749 14.087 integrate_v_rspace 128 11.3 0.003 0.004 13.993 14.011 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.926 13.032 calculate_rho_elec 128 8.7 0.045 0.063 12.925 13.031 init_scf_run 11 5.9 0.000 0.001 12.686 12.687 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.686 12.687 mp_sum_l 7870 13.0 8.599 9.746 8.599 9.746 cp_fm_cholesky_invert 11 10.9 9.440 9.449 9.440 9.449 wfi_extrapolate 11 7.9 0.001 0.001 9.072 9.072 calculate_dm_sparse 128 9.5 0.001 0.001 8.394 8.477 dbcsr_mm_accdrv_process 124484 16.2 3.364 3.527 7.961 8.457 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.328 8.440 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.805 7.921 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 6.178 7.911 make_images_data 5014 15.6 0.070 0.076 6.816 7.823 grid_integrate_task_list 128 12.3 7.090 7.607 7.090 7.607 pw_transfer 1547 11.6 0.075 0.107 6.985 7.269 density_rs2pw 128 9.7 0.006 0.007 6.670 7.259 hybrid_alltoall_any 5200 16.5 0.290 2.266 5.940 7.129 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.780 7.042 rs_pw_transfer 1046 11.9 0.017 0.019 6.057 6.776 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 6.677 6.686 mp_alltoall_d11v 2415 14.1 4.321 5.912 4.321 5.912 fft_wrap_pw1pw2_140 523 13.2 0.441 0.509 5.704 5.903 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.774 5.838 fft3d_ps 1291 14.7 2.092 2.567 5.577 5.772 grid_collocate_task_list 128 9.7 4.727 5.093 4.727 5.093 cp_fm_cholesky_decompose 22 10.9 4.704 4.718 4.704 4.718 potential_pw2rs 128 12.3 0.010 0.011 4.564 4.606 mp_sum_d 4459 12.1 3.766 4.449 3.766 4.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.639000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.181818, yerr=8.451123 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 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 830.693376E+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 1473826487232 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56823. MP_Allreduce 11322 944. MP_Sync 170 MP_Alltoall 1983 5090107. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 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 0.024 0.066 191.190 191.198 qs_mol_dyn_low 1 2.0 0.013 0.052 190.577 190.589 qs_forces 11 3.9 0.003 0.005 189.707 189.715 qs_energies 11 4.9 0.001 0.002 183.002 183.020 scf_env_do_scf 11 5.9 0.001 0.002 166.515 166.525 scf_env_do_scf_inner_loop 118 6.6 0.005 0.024 133.618 133.621 velocity_verlet 10 3.0 0.002 0.002 120.616 120.633 dbcsr_multiply_generic 2527 12.6 0.186 0.191 98.015 99.134 qs_scf_new_mos 118 7.6 0.001 0.001 94.749 95.195 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.748 95.194 ot_scf_mini 118 9.6 0.004 0.004 89.915 90.480 multiply_cannon 2527 13.6 0.478 0.530 77.873 82.010 multiply_cannon_loop 2527 14.6 1.270 1.308 74.565 77.223 ot_mini 118 10.6 0.001 0.001 50.075 50.579 mp_waitall_1 228564 16.4 25.125 39.690 25.125 39.690 multiply_cannon_multrec 30324 15.6 22.065 26.684 31.762 36.828 rebuild_ks_matrix 129 8.3 0.001 0.001 32.920 33.473 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.920 33.473 init_scf_loop 11 6.9 0.000 0.000 32.806 32.808 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.692 30.195 multiply_cannon_metrocomm3 30324 15.6 0.096 0.102 15.888 28.772 qs_ot_get_derivative 118 11.6 0.001 0.002 28.019 28.574 prepare_preconditioner 11 7.9 0.000 0.000 28.431 28.492 make_preconditioner 11 8.9 0.000 0.000 28.431 28.491 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.146 27.690 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.118 23.277 apply_single 129 13.6 0.001 0.001 22.118 23.276 multiply_cannon_sync_h2d 30324 15.6 19.302 22.238 19.302 22.238 qs_ot_get_p 129 10.4 0.001 0.001 21.521 22.116 ot_diis_step 118 11.6 0.014 0.015 21.879 21.882 qs_ot_p2m_diag 83 11.4 0.188 0.218 16.651 16.686 cp_fm_cholesky_invert 11 10.9 16.635 16.648 16.635 16.648 make_m2s 5054 13.6 0.089 0.095 14.313 15.960 make_images 5054 14.6 1.159 1.350 14.106 15.751 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.484 15.502 sum_up_and_integrate 129 10.3 0.117 0.133 14.452 14.480 integrate_v_rspace 129 11.3 0.003 0.004 14.334 14.367 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.029 13.068 calculate_rho_elec 129 8.7 0.088 0.106 13.028 13.068 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.371 12.383 cp_fm_redistribute_end 83 14.4 7.252 12.315 7.266 12.318 cp_fm_diag_elpa_base 83 14.4 4.816 11.840 5.031 12.199 init_scf_run 11 5.9 0.000 0.001 11.647 11.648 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.646 11.648 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.949 11.364 multiply_cannon_metrocomm4 27797 15.6 0.097 0.111 3.829 10.948 make_images_data 5054 15.6 0.068 0.074 8.685 10.619 mp_irecv_dv 70031 16.3 3.634 10.562 3.634 10.562 hybrid_alltoall_any 5240 16.5 0.343 1.500 7.379 9.879 dbcsr_mm_accdrv_process 62734 16.2 4.619 5.490 9.153 9.707 wfi_extrapolate 11 7.9 0.001 0.001 8.338 8.338 pw_transfer 1559 11.6 0.085 0.101 7.792 7.861 grid_integrate_task_list 129 12.3 7.234 7.647 7.234 7.647 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.566 7.626 density_rs2pw 129 9.7 0.006 0.006 6.765 7.202 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.338 7.105 cp_fm_cholesky_decompose 22 10.9 6.919 7.006 6.919 7.006 calculate_dm_sparse 129 9.5 0.001 0.001 6.579 6.738 fft_wrap_pw1pw2_140 527 13.2 0.474 0.523 6.601 6.663 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.177 6.185 rs_pw_transfer 1054 12.0 0.015 0.017 5.637 6.149 fft3d_ps 1301 14.7 2.792 2.948 5.926 5.957 mp_sum_l 7930 13.1 4.102 5.895 4.102 5.895 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.323 5.450 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.349 5.412 grid_collocate_task_list 129 9.7 4.948 5.332 4.948 5.332 mp_allgather_i34 2527 14.6 1.875 5.030 1.875 5.030 potential_pw2rs 129 12.3 0.015 0.018 4.813 4.828 mp_alltoall_d11v 2423 14.1 4.138 4.664 4.138 4.664 mp_sum_d 4496 12.2 2.668 4.011 2.668 4.011 dbcsr_complete_redistribute 395 12.7 0.766 0.838 3.119 3.975 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.198000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.818182, yerr=6.671761 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 938.672128E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.089 175.366 175.367 qs_mol_dyn_low 1 2.0 0.003 0.003 174.514 174.527 qs_forces 11 3.9 0.003 0.004 174.413 174.419 qs_energies 11 4.9 0.002 0.011 167.900 167.912 scf_env_do_scf 11 5.9 0.001 0.002 152.470 152.470 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.701 117.702 velocity_verlet 10 3.0 0.001 0.002 111.973 111.975 dbcsr_multiply_generic 2507 12.6 0.181 0.185 81.542 82.710 qs_scf_new_mos 117 7.6 0.001 0.001 80.790 81.082 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.789 81.081 ot_scf_mini 117 9.6 0.003 0.004 76.655 77.006 multiply_cannon 2507 13.6 0.503 0.522 61.870 66.906 multiply_cannon_loop 2507 14.6 0.864 0.896 58.594 61.444 ot_mini 117 10.6 0.001 0.001 42.445 42.779 init_scf_loop 11 6.9 0.000 0.000 34.669 34.671 mp_waitall_1 178456 16.5 25.082 34.197 25.082 34.197 prepare_preconditioner 11 7.9 0.000 0.000 30.679 30.728 make_preconditioner 11 8.9 0.000 0.000 30.679 30.728 rebuild_ks_matrix 128 8.3 0.001 0.001 30.049 30.491 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.049 30.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.265 29.710 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.131 27.531 multiply_cannon_multrec 20056 15.6 13.549 16.590 22.211 25.340 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.060 24.437 qs_ot_get_derivative 117 11.6 0.001 0.002 22.804 23.154 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.800 20.769 apply_single 128 13.6 0.001 0.001 19.800 20.768 qs_ot_get_p 128 10.4 0.001 0.001 19.570 20.004 ot_diis_step 117 11.6 0.017 0.018 19.543 19.543 make_m2s 5014 13.6 0.080 0.086 14.624 15.865 multiply_cannon_sync_h2d 20056 15.6 14.301 15.752 14.301 15.752 make_images 5014 14.6 1.165 1.265 14.393 15.634 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.291 15.299 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.298 14.300 cp_fm_cholesky_invert 11 10.9 14.274 14.283 14.274 14.283 sum_up_and_integrate 128 10.3 0.134 0.145 14.081 14.104 integrate_v_rspace 128 11.3 0.004 0.013 13.946 13.974 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.953 12.975 calculate_rho_elec 128 8.7 0.132 0.146 12.952 12.974 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.312 11.313 cp_fm_redistribute_end 83 14.4 4.266 11.257 4.280 11.259 cp_fm_diag_elpa_base 83 14.4 6.554 10.675 6.960 11.159 make_images_data 5014 15.6 0.061 0.069 9.004 10.714 init_scf_run 11 5.9 0.000 0.001 10.308 10.309 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.308 10.309 hybrid_alltoall_any 5200 16.5 0.434 1.980 7.912 9.928 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.940 9.207 multiply_cannon_metrocomm4 17549 15.6 0.063 0.071 3.417 9.086 mp_irecv_dv 50230 16.2 3.293 8.841 3.293 8.841 dbcsr_mm_accdrv_process 41502 16.2 4.463 5.038 8.108 8.245 pw_transfer 1547 11.6 0.084 0.103 7.731 7.844 grid_integrate_task_list 128 12.3 7.296 7.659 7.296 7.659 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.509 7.625 wfi_extrapolate 11 7.9 0.001 0.001 7.356 7.356 cp_fm_upper_to_full 105 14.5 5.682 7.330 5.682 7.330 cp_fm_cholesky_decompose 22 10.9 7.268 7.304 7.268 7.304 density_rs2pw 128 9.7 0.006 0.006 6.397 6.737 fft_wrap_pw1pw2_140 523 13.2 0.479 0.526 6.613 6.730 dbcsr_complete_redistribute 395 12.7 1.166 1.197 4.492 6.215 fft3d_ps 1291 14.7 2.696 2.900 5.775 5.849 calculate_dm_sparse 128 9.5 0.001 0.001 5.726 5.826 grid_collocate_task_list 128 9.7 5.071 5.588 5.071 5.588 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.411 5.415 rs_pw_transfer 1046 11.9 0.013 0.014 4.965 5.380 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.550 5.279 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.325 5.051 mp_alltoall_d11v 2415 14.1 4.175 4.698 4.175 4.698 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.542 4.656 mp_allgather_i34 2507 14.6 1.798 4.646 1.798 4.646 mp_sum_l 7870 13.0 3.233 4.594 3.233 4.594 potential_pw2rs 128 12.3 0.020 0.022 4.427 4.445 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.394 4.086 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.973 3.997 mp_alltoall_i22 716 14.1 2.008 3.895 2.008 3.895 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.804 3.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.367000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=887.181818, yerr=10.760350 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/18/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 4.320337E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.153466E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57301. MP_Allreduce 11231 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.119 187.643 187.644 qs_mol_dyn_low 1 2.0 0.003 0.003 186.837 186.850 qs_forces 11 3.9 0.003 0.004 186.728 186.735 qs_energies 11 4.9 0.004 0.007 179.694 179.708 scf_env_do_scf 11 5.9 0.001 0.002 162.983 162.994 velocity_verlet 10 3.0 0.001 0.002 123.586 123.588 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.045 116.046 qs_scf_new_mos 116 7.6 0.001 0.001 80.882 81.192 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.882 81.191 dbcsr_multiply_generic 2485 12.5 0.188 0.193 78.830 79.388 ot_scf_mini 116 9.6 0.003 0.004 76.407 76.719 multiply_cannon 2485 13.5 0.551 0.589 54.223 56.477 multiply_cannon_loop 2485 14.5 1.171 1.201 50.524 51.771 init_scf_loop 11 6.9 0.000 0.001 46.812 46.813 prepare_preconditioner 11 7.9 0.000 0.000 42.703 42.734 make_preconditioner 11 8.9 0.000 0.001 42.703 42.734 ot_mini 116 10.6 0.001 0.001 42.089 42.415 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.284 41.345 multiply_cannon_multrec 29820 15.5 14.116 18.868 25.962 30.474 rebuild_ks_matrix 127 8.3 0.001 0.001 28.818 29.041 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.817 29.040 mp_waitall_1 152434 16.5 17.327 27.261 17.327 27.261 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.024 26.223 qs_ot_get_derivative 116 11.6 0.001 0.002 22.612 22.926 make_m2s 4970 13.5 0.094 0.099 20.203 21.200 qs_ot_get_p 127 10.4 0.001 0.001 20.597 20.910 make_images 4970 14.5 1.954 2.297 19.899 20.896 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.951 19.338 apply_single 127 13.6 0.001 0.001 18.951 19.338 ot_diis_step 116 11.6 0.017 0.018 19.333 19.335 cp_fm_upper_to_full 104 14.7 11.183 16.564 11.183 16.564 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.497 16.550 cp_fm_cholesky_invert 11 10.9 16.346 16.356 16.346 16.356 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.216 15.217 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.224 15.217 sum_up_and_integrate 127 10.3 0.140 0.154 13.998 14.023 integrate_v_rspace 127 11.3 0.003 0.004 13.858 13.890 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.016 13.071 calculate_rho_elec 127 8.7 0.174 0.189 13.015 13.070 dbcsr_complete_redistribute 393 12.7 1.514 1.625 9.072 12.876 multiply_cannon_sync_h2d 29820 15.5 11.705 12.875 11.705 12.875 make_images_data 4970 15.5 0.064 0.068 10.868 12.565 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.093 12.096 cp_fm_redistribute_end 82 14.4 2.077 12.002 2.095 12.008 dbcsr_mm_accdrv_process 61748 16.2 7.306 8.309 11.427 11.893 cp_fm_diag_elpa_base 82 14.4 9.296 11.388 9.878 11.865 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.675 11.476 hybrid_alltoall_any 5155 16.4 0.521 2.176 9.804 11.398 init_scf_run 11 5.9 0.000 0.001 10.722 10.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.721 10.723 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.397 10.097 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.332 9.559 mp_alltoall_i22 712 14.1 5.627 9.405 5.627 9.405 grid_integrate_task_list 127 12.3 7.453 7.890 7.453 7.890 pw_transfer 1535 11.6 0.084 0.098 7.776 7.853 cp_fm_cholesky_decompose 22 10.9 7.697 7.791 7.697 7.791 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.552 7.636 wfi_extrapolate 11 7.9 0.001 0.001 7.587 7.587 multiply_cannon_metrocomm4 24850 15.5 0.074 0.084 2.682 7.357 mp_irecv_dv 75445 16.2 2.540 7.091 2.540 7.091 fft_wrap_pw1pw2_140 519 13.2 0.477 0.485 6.653 6.754 density_rs2pw 127 9.7 0.006 0.006 6.209 6.611 calculate_dm_sparse 127 9.5 0.001 0.001 6.178 6.258 fft3d_ps 1281 14.7 2.775 2.848 5.792 5.846 grid_collocate_task_list 127 9.7 5.165 5.715 5.165 5.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.419 5.496 mp_alltoall_d11v 2401 14.1 4.780 5.480 4.780 5.480 rs_pw_transfer 1038 11.9 0.013 0.015 4.627 4.971 qs_energies_init_hamiltonians 11 5.9 0.015 0.029 4.491 4.492 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.366 4.455 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.336 4.404 potential_pw2rs 127 12.3 0.022 0.023 4.282 4.305 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.194 4.259 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.644000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.727273, yerr=24.140087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.523577E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.067 173.705 173.706 qs_mol_dyn_low 1 2.0 0.003 0.003 173.149 173.161 qs_forces 11 3.9 0.003 0.004 173.041 173.044 qs_energies 11 4.9 0.001 0.002 165.734 165.742 scf_env_do_scf 11 5.9 0.001 0.001 148.681 148.694 velocity_verlet 10 3.0 0.015 0.019 113.030 113.034 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 112.666 112.668 qs_scf_new_mos 117 7.6 0.001 0.001 76.945 77.013 qs_scf_loop_do_ot 117 8.6 0.001 0.001 76.944 77.012 dbcsr_multiply_generic 2507 12.6 0.182 0.188 74.062 74.430 ot_scf_mini 117 9.6 0.004 0.004 72.542 72.578 multiply_cannon 2507 13.6 0.583 0.615 54.326 58.313 multiply_cannon_loop 2507 14.6 0.452 0.486 49.723 50.442 ot_mini 117 10.6 0.001 0.001 39.687 39.725 init_scf_loop 11 6.9 0.000 0.000 35.856 35.858 mp_waitall_1 129618 16.6 26.152 33.525 26.152 33.525 prepare_preconditioner 11 7.9 0.000 0.000 31.993 32.021 make_preconditioner 11 8.9 0.000 0.000 31.993 32.021 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.974 30.254 rebuild_ks_matrix 128 8.3 0.001 0.001 28.606 28.651 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.606 28.650 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.132 26.168 multiply_cannon_multrec 10028 15.6 10.441 14.507 17.718 21.063 qs_ot_get_p 128 10.4 0.001 0.001 19.966 20.028 qs_ot_get_derivative 117 11.6 0.002 0.002 19.797 19.846 ot_diis_step 117 11.6 0.019 0.020 19.822 19.823 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.431 19.613 apply_single 128 13.6 0.001 0.001 19.431 19.612 multiply_cannon_metrocomm3 10028 15.6 0.022 0.024 12.347 19.054 cp_fm_cholesky_invert 11 10.9 18.386 18.392 18.386 18.392 make_m2s 5014 13.6 0.065 0.070 16.086 18.342 make_images 5014 14.6 2.316 2.829 15.783 18.039 qs_ot_p2m_diag 83 11.4 0.495 0.502 16.119 16.139 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.943 14.945 sum_up_and_integrate 128 10.3 0.180 0.191 14.153 14.196 integrate_v_rspace 128 11.3 0.004 0.004 13.972 14.026 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.243 13.286 calculate_rho_elec 128 8.7 0.258 0.269 13.242 13.285 make_images_data 5014 15.6 0.053 0.061 9.710 12.196 multiply_cannon_sync_h2d 10028 15.6 11.566 12.153 11.566 12.153 hybrid_alltoall_any 5200 16.5 0.837 3.775 9.502 12.064 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.741 11.746 cp_fm_diag_elpa_base 83 14.4 11.494 11.568 11.734 11.739 init_scf_run 11 5.9 0.000 0.001 10.256 10.256 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.255 10.256 grid_integrate_task_list 128 12.3 7.701 8.192 7.701 8.192 cp_fm_cholesky_decompose 22 10.9 8.061 8.181 8.061 8.181 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.025 8.058 pw_transfer 1547 11.6 0.084 0.092 7.831 7.852 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.610 7.640 dbcsr_mm_accdrv_process 20762 16.1 2.584 3.428 6.911 7.533 wfi_extrapolate 11 7.9 0.001 0.001 7.336 7.336 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.461 7.300 mp_allgather_i34 2507 14.6 2.809 6.866 2.809 6.866 fft_wrap_pw1pw2_140 523 13.2 0.502 0.521 6.644 6.684 density_rs2pw 128 9.7 0.005 0.006 6.058 6.264 calculate_dm_sparse 128 9.5 0.001 0.001 6.098 6.188 fft3d_ps 1291 14.7 2.723 2.807 5.771 5.797 grid_collocate_task_list 128 9.7 5.495 5.728 5.495 5.728 dbcsr_complete_redistribute 395 12.7 2.108 2.168 5.201 5.569 mp_alltoall_d11v 2415 14.1 4.770 5.567 4.770 5.567 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.234 5.235 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.146 5.157 rs_pw_transfer 1046 11.9 0.013 0.013 4.258 4.498 multiply_cannon_metrocomm4 7521 15.6 0.024 0.026 1.856 4.458 mp_irecv_dv 28860 15.9 1.820 4.385 1.820 4.385 potential_pw2rs 128 12.3 0.026 0.028 4.178 4.195 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.046 4.080 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.559 3.869 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.718 3.745 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.601 3.671 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.347 3.661 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.466 3.479 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.706000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1424.636364, yerr=50.492615 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.932748E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 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 0.051 0.076 290.185 290.185 qs_mol_dyn_low 1 2.0 0.003 0.004 289.432 289.444 qs_forces 11 3.9 0.066 0.067 289.339 289.342 qs_energies 11 4.9 0.002 0.002 280.722 280.736 scf_env_do_scf 11 5.9 0.001 0.002 258.952 258.963 velocity_verlet 10 3.0 0.002 0.002 209.438 209.445 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.215 132.216 init_scf_loop 11 6.9 0.000 0.000 126.472 126.474 prepare_preconditioner 11 7.9 0.000 0.000 121.707 121.741 make_preconditioner 11 8.9 0.000 0.000 121.707 121.741 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.765 118.863 qs_scf_new_mos 116 7.6 0.001 0.001 90.454 90.554 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.453 90.553 ot_scf_mini 116 9.6 0.004 0.004 85.710 85.747 dbcsr_multiply_generic 2485 12.5 0.212 0.219 82.183 82.749 cp_fm_upper_to_full 104 14.8 54.208 78.189 54.208 78.189 multiply_cannon 2485 13.5 0.711 0.796 58.002 58.734 multiply_cannon_loop 2485 14.5 0.468 0.473 54.390 55.779 dbcsr_complete_redistribute 393 12.7 4.006 4.068 30.710 44.205 ot_mini 116 10.6 0.001 0.001 43.693 43.730 copy_fm_to_dbcsr 208 11.6 0.001 0.002 27.264 40.737 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.897 38.271 mp_alltoall_i22 712 14.1 22.701 36.419 22.701 36.419 cp_fm_cholesky_invert 11 10.9 33.437 33.444 33.437 33.444 rebuild_ks_matrix 127 8.3 0.001 0.001 33.340 33.384 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.339 33.384 mp_waitall_1 104546 16.7 27.852 31.729 27.852 31.729 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.104 31.149 qs_ot_get_p 127 10.4 0.001 0.001 26.780 26.798 qs_ot_get_derivative 116 11.6 0.002 0.002 23.988 24.028 qs_ot_p2m_diag 82 11.4 0.868 0.873 22.625 22.651 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.912 20.913 make_m2s 4970 13.5 0.075 0.077 19.712 20.666 make_images 4970 14.5 3.737 3.826 19.240 20.195 ot_diis_step 116 11.6 0.022 0.023 19.679 19.679 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.433 19.547 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.904 19.097 apply_single 127 13.6 0.001 0.001 18.904 19.097 multiply_cannon_multrec 9940 15.5 10.431 12.171 17.871 17.976 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.712 17.714 cp_fm_diag_elpa_base 82 14.4 13.256 14.933 17.707 17.709 sum_up_and_integrate 127 10.3 0.318 0.321 15.657 15.756 multiply_cannon_sync_h2d 9940 15.5 15.533 15.549 15.533 15.549 integrate_v_rspace 127 11.3 0.004 0.004 15.338 15.437 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.929 14.942 calculate_rho_elec 127 8.7 0.479 0.480 14.928 14.942 hybrid_alltoall_any 5155 16.4 1.293 3.019 10.626 12.447 make_images_data 4970 15.5 0.060 0.065 10.478 12.167 init_scf_run 11 5.9 0.000 0.001 11.836 11.836 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.835 11.836 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.388 9.421 dbcsr_mm_accdrv_process 20590 16.0 3.800 5.659 7.204 8.992 cp_fm_cholesky_decompose 22 10.9 8.924 8.951 8.924 8.951 wfi_extrapolate 11 7.9 0.001 0.001 8.781 8.782 grid_integrate_task_list 127 12.3 8.495 8.665 8.495 8.665 pw_transfer 1535 11.6 0.090 0.091 8.364 8.370 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.132 8.137 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 7.918 7.919 fft_wrap_pw1pw2_140 519 13.2 0.535 0.538 7.184 7.193 mp_alltoall_d11v 2401 14.1 6.984 7.120 6.984 7.120 calculate_dm_sparse 127 9.5 0.001 0.001 6.589 6.674 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.420 6.496 grid_collocate_task_list 127 9.7 6.287 6.321 6.287 6.321 fft3d_ps 1281 14.7 2.734 2.742 6.191 6.200 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.117 6.200 density_rs2pw 127 9.7 0.005 0.005 6.030 6.061 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.185000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2593.818182, yerr=173.547907 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 1.261318E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.100 85.061 85.075 qs_energies 1 2.0 0.000 0.000 84.289 84.299 ls_scf 1 3.0 0.000 0.000 83.335 83.346 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.459 72.604 multiply_cannon 111 7.7 0.017 0.020 55.738 57.112 multiply_cannon_loop 111 8.7 0.212 0.225 52.289 53.855 ls_scf_main 1 4.0 0.000 0.000 52.154 52.154 density_matrix_trs4 2 5.0 0.002 0.003 46.703 46.777 ls_scf_init_scf 1 4.0 0.000 0.000 28.139 28.140 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.045 27.107 mp_waitall_1 11316 10.9 22.551 25.396 22.551 25.396 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.967 24.988 multiply_cannon_multrec 2664 9.7 8.113 9.050 15.390 17.260 multiply_cannon_sync_h2d 2664 9.7 13.523 15.752 13.523 15.752 make_m2s 222 7.7 0.009 0.011 13.043 13.598 make_images 222 8.7 0.099 0.108 13.020 13.577 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.716 12.519 make_images_data 222 9.7 0.004 0.005 7.636 8.119 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.516 8.080 hybrid_alltoall_any 227 10.6 0.215 1.835 6.540 7.937 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.619 6.897 7.827 dbcsr_mm_accdrv_process_sort 4760 11.4 6.188 7.057 6.188 7.057 calculate_norms 4752 9.8 5.497 6.058 5.497 6.058 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.042 5.179 mp_sum_l 807 5.4 3.172 4.677 3.172 4.677 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.361 3.569 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.051 3.279 mp_irecv_dv 6231 10.9 2.034 3.256 2.034 3.256 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.236 arnoldi_normal_ev 4 7.8 0.001 0.003 3.209 3.236 make_images_sizes 222 9.7 0.000 0.000 0.647 3.206 mp_alltoall_i44 222 10.7 0.647 3.206 0.647 3.206 build_subspace 16 8.4 0.009 0.012 3.115 3.117 ls_scf_post 1 4.0 0.000 0.000 3.042 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.843 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.406 2.822 dbcsr_merge_single_wm 555 10.7 0.458 0.591 2.398 2.813 make_images_pack 222 9.7 2.207 2.626 2.209 2.628 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.309 2.564 dbcsr_sort_data 658 11.4 2.198 2.535 2.198 2.535 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.467 2.065 2.469 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.188 2.259 buffer_matrices_ensure_size 222 8.7 1.748 2.072 1.748 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.743 1.744 rebuild_ks_matrix 3 7.3 0.000 0.000 1.734 1.735 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.734 1.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.075000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 2.089828E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.126 91.868 91.894 qs_energies 1 2.0 0.000 0.000 90.864 90.876 ls_scf 1 3.0 0.000 0.001 89.440 89.451 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.412 75.774 multiply_cannon 111 7.7 0.028 0.040 53.488 57.108 ls_scf_main 1 4.0 0.000 0.000 55.085 55.096 multiply_cannon_loop 111 8.7 0.116 0.123 50.147 53.245 density_matrix_trs4 2 5.0 0.002 0.003 49.404 49.595 ls_scf_init_scf 1 4.0 0.000 0.002 30.778 30.779 mp_waitall_1 9246 10.9 21.391 30.374 21.391 30.374 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.546 29.658 multiply_cannon_multrec 1332 9.7 13.163 17.309 22.355 27.590 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.127 27.148 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.974 21.023 make_m2s 222 7.7 0.006 0.008 15.276 15.979 make_images 222 8.7 1.571 1.918 15.246 15.951 dbcsr_mm_accdrv_process 4041 10.4 0.295 0.454 8.793 10.376 dbcsr_mm_accdrv_process_sort 4041 11.4 8.384 9.922 8.384 9.922 make_images_data 222 9.7 0.004 0.004 8.813 9.752 hybrid_alltoall_any 227 10.6 0.522 2.425 8.207 9.132 mp_sum_l 807 5.4 5.402 8.284 5.402 8.284 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.250 7.718 mp_irecv_dv 3311 11.0 3.230 7.652 3.230 7.652 calculate_norms 2376 9.8 5.985 6.739 5.985 6.739 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.088 6.443 multiply_cannon_sync_h2d 1332 9.7 4.786 5.812 4.786 5.812 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.127 5.369 arnoldi_extremal 4 6.8 0.000 0.000 4.695 4.720 arnoldi_normal_ev 4 7.8 0.001 0.004 4.695 4.720 build_subspace 16 8.4 0.014 0.021 4.436 4.439 ls_scf_post 1 4.0 0.001 0.006 3.577 3.588 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.183 3.407 ls_scf_store_result 1 5.0 0.000 0.000 3.260 3.386 dbcsr_matrix_vector_mult_local 304 10.0 2.772 3.257 2.774 3.258 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.228 2.811 ls_scf_dm_to_ks 2 5.0 0.000 0.005 2.520 2.603 mp_allgather_i34 111 8.7 1.045 2.533 1.045 2.533 make_images_pack 222 9.7 2.019 2.438 2.021 2.440 dbcsr_sort_data 436 11.2 1.796 2.021 1.796 2.021 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.889 rebuild_ks_matrix 3 7.3 0.000 0.000 1.874 1.876 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.011 1.873 1.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.894000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1699.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 2.710307E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.131 94.678 94.679 qs_energies 1 2.0 0.000 0.000 93.704 93.707 ls_scf 1 3.0 0.000 0.000 92.316 92.319 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.049 77.319 ls_scf_main 1 4.0 0.000 0.000 57.987 57.992 multiply_cannon 111 7.7 0.038 0.080 53.037 57.106 multiply_cannon_loop 111 8.7 0.100 0.110 49.433 53.383 density_matrix_trs4 2 5.0 0.002 0.003 52.000 52.216 mp_waitall_1 7374 11.0 24.243 34.042 24.243 34.042 ls_scf_init_scf 1 4.0 0.000 0.000 30.745 30.747 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.637 29.695 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.249 27.266 multiply_cannon_multrec 888 9.7 12.558 15.253 21.070 24.377 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.315 23.049 make_m2s 222 7.7 0.006 0.006 17.148 18.395 make_images 222 8.7 1.967 2.280 17.110 18.358 make_images_data 222 9.7 0.003 0.004 9.802 10.833 hybrid_alltoall_any 227 10.6 0.618 2.829 9.458 10.779 dbcsr_mm_accdrv_process 3754 10.4 0.241 0.415 8.031 9.241 dbcsr_mm_accdrv_process_sort 3754 11.4 7.660 8.826 7.660 8.826 mp_sum_l 807 5.4 5.394 8.699 5.394 8.699 multiply_cannon_sync_h2d 888 9.7 5.995 7.692 5.995 7.692 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.462 7.055 mp_irecv_dv 2335 11.1 2.446 7.010 2.446 7.010 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.058 6.817 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.737 6.763 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.063 5.257 arnoldi_extremal 4 6.8 0.000 0.000 5.221 5.237 arnoldi_normal_ev 4 7.8 0.001 0.005 5.220 5.237 build_subspace 16 8.4 0.014 0.020 4.917 4.924 calculate_norms 1584 9.8 4.240 4.571 4.240 4.571 mp_allgather_i34 111 8.7 1.436 3.960 1.436 3.960 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.493 3.811 dbcsr_matrix_vector_mult_local 304 10.0 3.041 3.599 3.043 3.601 ls_scf_post 1 4.0 0.000 0.000 3.584 3.588 ls_scf_store_result 1 5.0 0.000 0.000 3.323 3.408 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.769 2.872 make_images_sizes 222 9.7 0.000 0.000 1.033 2.222 mp_alltoall_i44 222 10.7 1.033 2.221 1.033 2.221 dbcsr_sort_data 325 11.1 1.884 2.121 1.884 2.121 make_images_pack 222 9.7 1.812 2.117 1.815 2.119 dbcsr_data_release 9322 10.9 1.303 1.938 1.303 1.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.679000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2179.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 3.340837E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.113 0.234 97.885 97.888 qs_energies 1 2.0 0.000 0.000 96.601 96.615 ls_scf 1 3.0 0.000 0.001 94.953 94.967 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.757 79.024 ls_scf_main 1 4.0 0.000 0.000 59.066 59.067 multiply_cannon 111 7.7 0.055 0.117 51.600 56.324 density_matrix_trs4 2 5.0 0.002 0.003 53.072 53.173 multiply_cannon_loop 111 8.7 0.114 0.128 46.582 49.881 ls_scf_init_scf 1 4.0 0.009 0.023 32.632 32.633 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.424 31.513 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.924 28.936 mp_waitall_1 6438 11.0 23.007 28.606 23.007 28.606 multiply_cannon_multrec 1332 9.7 14.100 17.153 21.908 24.520 make_m2s 222 7.7 0.007 0.008 21.242 22.631 make_images 222 8.7 3.130 3.585 21.192 22.583 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.331 16.964 make_images_data 222 9.7 0.004 0.004 11.902 13.565 hybrid_alltoall_any 227 10.6 0.798 3.771 11.297 13.092 dbcsr_mm_accdrv_process 3641 10.4 0.196 0.407 7.455 8.971 dbcsr_mm_accdrv_process_sort 3641 11.4 7.082 8.553 7.082 8.553 mp_sum_l 807 5.4 4.397 8.289 4.397 8.289 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.335 6.542 multiply_cannon_sync_h2d 1332 9.7 5.545 6.261 5.545 6.261 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.083 5.956 mp_irecv_dv 3229 10.9 2.060 5.889 2.060 5.889 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.602 5.435 arnoldi_extremal 4 6.8 0.000 0.000 5.299 5.312 arnoldi_normal_ev 4 7.8 0.001 0.004 5.299 5.312 build_subspace 16 8.4 0.015 0.021 4.955 4.963 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.524 4.739 calculate_norms 2376 9.8 4.177 4.554 4.177 4.554 mp_allgather_i34 111 8.7 2.136 4.383 2.136 4.383 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.615 3.915 dbcsr_matrix_vector_mult_local 304 10.0 3.186 3.690 3.188 3.691 dbcsr_sort_data 658 11.4 3.061 3.371 3.061 3.371 ls_scf_post 1 4.0 0.002 0.013 3.255 3.269 dbcsr_special_finalize 555 9.7 0.006 0.007 2.818 3.258 dbcsr_merge_single_wm 555 10.7 0.536 0.664 2.810 3.250 ls_scf_store_result 1 5.0 0.000 0.000 3.002 3.049 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.886 2.943 dbcsr_data_release 10477 10.7 1.560 2.355 1.560 2.355 dbcsr_finalize 304 7.8 0.049 0.061 1.784 2.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.888000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 4.615889E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.092 0.104 99.860 99.861 qs_energies 1 2.0 0.000 0.000 98.442 98.446 ls_scf 1 3.0 0.000 0.000 96.484 96.488 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.251 78.455 ls_scf_main 1 4.0 0.000 0.000 62.295 62.296 multiply_cannon 111 7.7 0.102 0.173 55.679 61.228 density_matrix_trs4 2 5.0 0.002 0.003 55.272 55.368 multiply_cannon_loop 111 8.7 0.069 0.077 51.164 52.931 mp_waitall_1 5481 11.0 27.023 32.216 27.023 32.216 ls_scf_init_scf 1 4.0 0.000 0.000 30.582 30.585 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.420 29.460 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.253 27.260 multiply_cannon_multrec 444 9.7 14.097 16.470 21.097 24.442 make_m2s 222 7.7 0.004 0.005 17.789 20.287 make_images 222 8.7 3.718 4.392 17.727 20.226 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 12.052 16.931 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.660 15.294 make_images_data 222 9.7 0.003 0.004 10.107 12.502 hybrid_alltoall_any 227 10.6 0.788 3.776 9.817 12.249 multiply_cannon_sync_h2d 444 9.7 6.572 8.805 6.572 8.805 dbcsr_mm_accdrv_process 3003 10.4 0.166 0.343 6.703 7.842 dbcsr_mm_accdrv_process_sort 3003 11.4 6.387 7.493 6.387 7.493 mp_allgather_i34 111 8.7 2.744 6.969 2.744 6.969 arnoldi_extremal 4 6.8 0.000 0.000 5.821 5.829 arnoldi_normal_ev 4 7.8 0.002 0.005 5.821 5.829 build_subspace 16 8.4 0.015 0.019 5.436 5.444 mp_sum_l 807 5.4 2.829 4.859 2.829 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.597 4.751 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.605 4.564 mp_irecv_dv 1241 11.2 1.586 4.541 1.586 4.541 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.154 4.372 dbcsr_matrix_vector_mult_local 304 10.0 3.651 4.110 3.653 4.112 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.962 3.804 calculate_norms 792 9.8 3.525 3.677 3.525 3.677 ls_scf_post 1 4.0 0.000 0.000 3.608 3.612 make_images_sizes 222 9.7 0.000 0.000 1.047 3.601 mp_alltoall_i44 222 10.7 1.047 3.601 1.047 3.601 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.398 3.507 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.439 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.266 dbcsr_merge_all 275 8.9 0.473 0.524 2.047 2.100 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.861000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3661.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3afb8bf709f96407a7e9894ade69f93aea585134_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 8.759861E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.181 0.222 108.363 108.367 qs_energies 1 2.0 0.000 0.000 106.445 106.457 ls_scf 1 3.0 0.000 0.000 103.512 103.524 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.515 77.624 ls_scf_main 1 4.0 0.000 0.000 65.791 65.792 density_matrix_trs4 2 5.0 0.002 0.003 56.890 56.950 multiply_cannon 111 7.7 0.150 0.246 49.904 52.109 multiply_cannon_loop 111 8.7 0.067 0.070 46.379 47.199 ls_scf_init_scf 1 4.0 0.000 0.000 34.030 34.031 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.692 32.704 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.857 29.869 mp_waitall_1 4569 11.1 22.259 26.079 22.259 26.079 make_m2s 222 7.7 0.005 0.005 24.025 25.091 make_images 222 8.7 4.575 4.982 23.919 24.983 multiply_cannon_multrec 444 9.7 17.934 18.548 22.521 23.075 hybrid_alltoall_any 227 10.6 1.657 3.613 13.074 15.783 make_images_data 222 9.7 0.003 0.003 13.255 15.764 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.402 10.961 multiply_cannon_sync_h2d 444 9.7 8.846 8.902 8.846 8.902 arnoldi_extremal 4 6.8 0.000 0.000 7.419 7.431 arnoldi_normal_ev 4 7.8 0.003 0.009 7.419 7.431 build_subspace 16 8.4 0.026 0.035 6.841 6.850 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.415 5.574 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.045 5.290 dbcsr_matrix_vector_mult_local 304 10.0 4.960 5.243 4.962 5.246 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.848 4.929 dbcsr_mm_accdrv_process 1814 10.4 0.232 0.313 4.418 4.551 dbcsr_mm_accdrv_process_sort 1814 11.4 4.117 4.253 4.117 4.253 mp_allgather_i34 111 8.7 1.180 3.717 1.180 3.717 ls_scf_post 1 4.0 0.000 0.000 3.691 3.703 make_images_sizes 222 9.7 0.000 0.000 1.435 3.561 mp_alltoall_i44 222 10.7 1.435 3.561 1.435 3.561 ls_scf_store_result 1 5.0 0.000 0.000 3.419 3.428 calculate_norms 792 9.8 3.241 3.283 3.241 3.283 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.134 dbcsr_merge_all 275 8.9 0.890 0.923 2.864 2.915 dbcsr_complete_redistribute 5 7.6 1.443 1.484 2.783 2.912 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.902 2.902 dbcsr_data_release 12724 10.6 2.346 2.887 2.346 2.887 matrix_ls_to_qs 2 6.0 0.000 0.000 2.425 2.569 dbcsr_sort_data 325 11.1 2.436 2.499 2.436 2.499 dbcsr_new_transposed 4 7.5 0.269 0.298 2.375 2.390 dbcsr_frobenius_norm 74 6.6 2.056 2.132 2.188 2.230 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.213 2.214 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.207 dbcsr_add_anytype 103 7.2 0.860 0.892 2.132 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.367000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6873.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 3afb8bf709f96407a7e9894ade69f93aea585134 Summary: empty Status: OK