=== 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: f51da1b24f48180e503e660b45b9ce8da4b1a63d ################# 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 5.2.3, 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 (15.09.2022) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/01 job id: 41884942 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/02 job id: 41884943 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/03 job id: 41884945 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/04 job id: 41884947 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/05 job id: 41884948 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/06 job id: 41884949 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/07 job id: 41884950 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/08 job id: 41884951 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/09 job id: 41884952 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/10 job id: 41884953 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/11 job id: 41884954 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/12 job id: 41884955 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/13 job id: 41884956 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/14 job id: 41884957 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/15 job id: 41884960 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/16 job id: 41884962 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/17 job id: 41884963 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/18 job id: 41884964 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/19 job id: 41884968 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/20 job id: 41884970 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/21 job id: 41884971 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/22 job id: 41884972 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/23 job id: 41884973 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/24 job id: 41884975 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/25 job id: 41884976 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/26 job id: 41884977 --- 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/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 142.393 142.394 farming_run 1 2.0 141.795 141.796 142.361 142.364 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458160E+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 0.027 118.826 118.827 qs_energies 1 2.0 0.000 0.000 118.551 118.553 mp2_main 1 3.0 0.000 0.000 116.338 116.340 mp2_gpw_main 1 4.0 0.020 0.027 115.324 115.326 mp2_ri_gpw_compute_in 1 5.0 0.174 0.181 93.510 93.668 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.406 55.564 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.705 46.627 get_2c_integrals 1 6.0 0.000 0.001 37.411 37.928 integrate_v_rspace 273 8.0 0.438 0.463 25.037 29.653 pw_transfer 6555 10.6 0.377 0.385 27.290 27.725 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.947 26.311 grid_integrate_task_list 273 9.0 20.850 25.903 20.850 25.903 fft_wrap_pw1pw2_100 2178 12.4 1.185 1.253 23.497 23.868 rpa_ri_compute_en 1 5.0 0.009 0.010 21.702 21.770 cp_fm_cholesky_decompose 12 8.2 19.340 21.121 19.340 21.121 compute_2c_integrals 1 7.0 0.002 0.003 19.381 19.382 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.861 19.089 mp2_eri_2c_integrate_gpw 1 9.0 2.399 2.432 18.858 19.087 cholesky_decomp 1 7.0 0.000 0.000 16.873 17.406 fft3d_s 5443 13.4 16.058 16.511 16.080 16.534 ao_to_mo_and_store_B_mult_1 272 7.0 10.860 15.587 10.860 15.587 rpa_num_int 1 6.0 0.000 0.000 13.764 13.774 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 12.210 13.737 calculate_wavefunction 272 8.0 5.498 5.579 12.593 13.257 calc_mat_Q 8 8.0 0.000 0.000 9.662 10.024 contract_S_to_Q 8 9.0 0.000 0.000 9.082 9.442 parallel_gemm_fm 14 9.1 0.000 0.000 8.648 8.936 parallel_gemm_fm_cosma 14 10.1 8.648 8.936 8.648 8.936 calc_potential_gpw 544 9.5 0.005 0.006 8.196 8.664 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.186 8.414 potential_pw2rs 545 10.0 0.108 0.109 7.722 8.353 collocate_single_gaussian 272 10.0 0.041 0.043 7.462 7.728 create_integ_mat 1 6.0 0.014 0.027 7.676 7.686 array2fm 1 7.0 0.000 0.000 6.608 7.106 pw_scatter_s 2720 13.7 4.417 4.599 4.417 4.599 pw_gather_s 2722 13.2 3.907 4.244 3.907 4.244 mp_sync 27 8.8 2.051 4.240 2.051 4.240 compute_Erpa_by_freq_int 8 8.0 0.002 0.003 2.432 3.691 array2fm_buffer_send 1 8.0 2.967 3.080 2.967 3.080 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=115.324342, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 0.030 0.039 406.981 406.982 farming_run 1 2.0 406.262 406.272 406.933 406.937 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222869E+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 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 0.014 0.046 210.615 210.615 qs_energies 1 2.0 0.000 0.000 210.353 210.361 scf_env_do_scf 1 3.0 0.000 0.000 107.238 107.238 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.295 106.304 rebuild_ks_matrix 4 6.0 0.000 0.000 106.294 106.302 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.066 106.294 106.302 hfx_ks_matrix 4 8.0 0.001 0.001 105.863 105.867 integrate_four_center 4 9.0 0.144 0.463 105.863 105.866 mp2_main 1 3.0 0.003 0.027 102.810 102.818 mp2_gpw_main 1 4.0 0.039 0.120 101.912 101.922 integrate_four_center_main 4 10.0 0.144 0.578 96.754 99.725 integrate_four_center_bin 266 11.0 96.609 99.289 96.609 99.289 init_scf_loop 1 4.0 0.000 0.000 92.974 92.974 mp2_ri_gpw_compute_in 1 5.0 0.068 0.092 74.832 75.820 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.456 55.439 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.164 42.012 47.177 integrate_v_rspace 95 8.0 0.398 0.571 28.421 33.421 pw_transfer 2240 10.6 0.147 0.171 29.872 30.291 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.892 29.301 ao_to_mo_and_store_B_mult_1 91 7.0 10.765 29.050 10.765 29.050 grid_integrate_task_list 95 9.0 23.663 28.873 23.663 28.873 mp2_ri_gpw_compute_en 1 5.0 0.060 0.110 26.894 28.542 fft_wrap_pw1pw2_100 730 12.4 1.259 1.446 26.622 27.030 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.847 1.893 25.199 25.208 get_2c_integrals 1 6.0 0.002 0.015 20.262 20.316 compute_2c_integrals 1 7.0 0.003 0.006 19.226 19.239 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.807 19.093 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.866 18.804 19.092 fft3d_s 1823 13.4 18.404 18.725 18.417 18.739 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.262 14.262 calculate_wavefunction 91 8.0 2.036 2.073 9.759 9.979 potential_pw2rs 186 10.0 0.034 0.035 8.656 9.310 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.585 8.766 9.201 mp2_ri_gpw_compute_en_comm 22 7.0 0.513 0.531 8.190 8.682 local_gemm 172 8.0 8.208 8.641 8.208 8.641 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.205 8.576 calc_potential_gpw 182 9.5 0.002 0.002 7.902 8.175 collocate_single_gaussian 91 10.0 0.018 0.029 7.860 8.086 mp_sendrecv_dm3 2068 8.0 6.189 6.662 6.189 6.662 mp2_ri_gpw_compute_en_ener 172 7.0 6.351 6.423 6.351 6.423 mp_sync 38 10.4 3.548 6.327 3.548 6.327 pw_gather_s 912 13.2 4.922 5.475 4.922 5.475 pw_scatter_s 910 13.7 3.937 4.367 3.937 4.367 integrate_four_center_load 4 10.0 0.000 0.000 4.355 4.361 hfx_load_balance 1 11.0 0.000 0.000 4.355 4.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.909475, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 450.002944E+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 592243. 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 0.016 0.033 64.438 64.440 qs_mol_dyn_low 1 2.0 0.003 0.006 64.172 64.180 qs_forces 11 3.9 0.002 0.004 64.063 64.066 qs_energies 11 4.9 0.003 0.004 61.910 61.928 scf_env_do_scf 11 5.9 0.000 0.001 54.889 54.890 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 52.396 52.399 dbcsr_multiply_generic 2286 12.5 0.093 0.097 36.949 37.457 qs_scf_new_mos 108 7.5 0.000 0.001 36.251 36.564 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.250 36.564 ot_scf_mini 108 9.5 0.002 0.002 34.272 34.445 velocity_verlet 10 3.0 0.001 0.002 33.738 33.740 multiply_cannon 2286 13.5 0.181 0.190 27.111 29.115 multiply_cannon_loop 2286 14.5 1.466 1.544 25.906 27.817 ot_mini 108 10.5 0.001 0.001 21.667 21.912 qs_ot_get_derivative 108 11.5 0.001 0.002 18.560 18.733 mp_waitall_1 267858 16.1 12.330 18.102 12.330 18.102 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.342 13.943 multiply_cannon_multrec 54864 15.5 4.181 6.484 7.448 10.798 rebuild_ks_matrix 119 8.3 0.000 0.000 10.574 10.712 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.015 10.573 10.711 qs_ks_update_qs_env 119 7.6 0.001 0.001 9.059 9.183 qs_rho_update_rho_low 119 7.7 0.000 0.001 8.239 8.339 calculate_rho_elec 119 8.7 0.011 0.016 8.238 8.339 mp_sum_l 7207 12.9 6.369 8.267 6.369 8.267 qs_ot_get_p 119 10.4 0.001 0.001 7.561 7.927 multiply_cannon_sync_h2d 54864 15.5 5.777 7.478 5.777 7.478 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.975 7.391 sum_up_and_integrate 119 10.3 0.013 0.016 6.448 6.456 integrate_v_rspace 119 11.3 0.002 0.003 6.435 6.445 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.784 5.895 rs_pw_transfer 974 11.9 0.012 0.013 5.494 5.869 density_rs2pw 119 9.7 0.003 0.004 5.220 5.334 init_scf_run 11 5.9 0.000 0.001 4.990 4.991 scf_env_initial_rho_setup 11 6.9 0.003 0.008 4.990 4.990 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 2.469 4.569 pw_transfer 1439 11.6 0.051 0.056 4.291 4.547 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 4.216 4.475 dbcsr_mm_accdrv_process 76910 16.1 1.145 1.807 3.192 4.473 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.314 4.392 fft3d_ps 1201 14.6 0.361 0.463 3.997 4.252 fft_wrap_pw1pw2_140 487 13.2 0.077 0.091 3.541 3.798 potential_pw2rs 119 12.3 0.004 0.004 3.438 3.576 calculate_dm_sparse 119 9.5 0.000 0.000 3.244 3.434 mp_alltoall_d11v 2130 13.8 3.212 3.421 3.212 3.421 make_m2s 4572 13.5 0.053 0.057 3.163 3.376 make_images 4572 14.5 0.133 0.139 3.081 3.292 cp_dbcsr_syevd 50 12.0 0.002 0.003 3.282 3.283 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.008 3.231 apply_single 119 13.6 0.000 0.000 3.008 3.230 mp_alltoall_z22v 1201 16.6 2.994 3.220 2.994 3.220 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.951 3.041 ot_diis_step 108 11.5 0.006 0.009 2.812 2.813 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.610 2.613 jit_kernel_multiply 13 15.8 1.986 2.585 1.986 2.585 cp_fm_redistribute_end 50 14.0 2.360 2.579 2.382 2.582 calculate_first_density_matrix 1 7.0 0.000 0.001 2.531 2.545 init_scf_loop 11 6.9 0.000 0.000 2.461 2.467 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.386 2.391 wfi_extrapolate 11 7.9 0.001 0.006 2.372 2.374 cp_fm_diag_elpa_base 50 14.0 0.195 2.357 0.196 2.366 acc_transpose_blocks 54864 15.5 0.221 0.243 1.717 2.171 mp_sum_d 4129 12.0 1.516 2.138 1.516 2.138 grid_integrate_task_list 119 12.3 1.994 2.106 1.994 2.106 make_images_sizes 4572 15.5 0.004 0.005 1.632 2.064 yz_to_x 249 15.8 0.025 0.033 1.846 2.063 mp_alltoall_i44 4572 16.5 1.628 2.059 1.628 2.059 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.009 2.053 rs_scatter_matrices 130 9.7 0.038 0.048 1.790 1.900 mp_waitany 12084 13.8 1.701 1.890 1.701 1.890 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.633 1.671 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.564 1.583 grid_collocate_task_list 119 9.7 1.289 1.371 1.289 1.371 make_images_data 4572 15.5 0.038 0.042 0.865 1.320 rs_pw_transfer_RS2PW_20 119 11.7 0.055 0.061 1.081 1.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=64.440000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 486.174720E+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 1682609. 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 0.020 0.041 45.846 45.846 qs_mol_dyn_low 1 2.0 0.003 0.006 45.565 45.599 qs_forces 11 3.9 0.002 0.006 45.478 45.480 qs_energies 11 4.9 0.002 0.006 43.460 43.464 scf_env_do_scf 11 5.9 0.000 0.001 37.501 37.502 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 34.346 34.347 dbcsr_multiply_generic 2286 12.5 0.101 0.103 23.955 24.447 qs_scf_new_mos 108 7.5 0.001 0.001 23.152 23.384 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.151 23.383 velocity_verlet 10 3.0 0.001 0.002 22.311 22.313 ot_scf_mini 108 9.5 0.002 0.003 22.101 22.267 multiply_cannon 2286 13.5 0.207 0.217 17.359 19.818 multiply_cannon_loop 2286 14.5 0.901 0.978 15.974 17.939 ot_mini 108 10.5 0.001 0.003 13.376 13.629 mp_waitall_1 217478 16.2 8.644 13.350 8.644 13.350 qs_ot_get_derivative 108 11.5 0.001 0.001 10.680 10.861 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.884 10.335 multiply_cannon_multrec 27432 15.5 1.928 4.459 5.852 8.673 rebuild_ks_matrix 119 8.3 0.000 0.000 8.456 8.671 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.018 8.455 8.671 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.461 7.641 dbcsr_mm_accdrv_process 47894 16.0 2.970 5.236 3.856 5.720 qs_ot_get_p 119 10.4 0.001 0.002 5.359 5.609 sum_up_and_integrate 119 10.3 0.024 0.029 5.300 5.306 integrate_v_rspace 119 11.3 0.002 0.004 5.275 5.283 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.982 5.034 calculate_rho_elec 119 8.7 0.021 0.024 4.981 5.034 mp_sum_l 7207 12.9 2.854 5.017 2.854 5.017 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.079 5.009 init_scf_run 11 5.9 0.000 0.001 4.616 4.617 scf_env_initial_rho_setup 11 6.9 0.003 0.018 4.616 4.617 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.235 4.397 apply_single 119 13.6 0.000 0.000 3.235 4.397 rs_pw_transfer 974 11.9 0.010 0.011 4.213 4.394 make_m2s 4572 13.5 0.052 0.053 3.291 3.608 make_images 4572 14.5 0.201 0.239 3.204 3.518 qs_ot_p2m_diag 50 11.0 0.009 0.016 3.296 3.339 density_rs2pw 119 9.7 0.004 0.004 3.164 3.242 init_scf_loop 11 6.9 0.000 0.001 3.135 3.136 potential_pw2rs 119 12.3 0.006 0.006 2.949 3.003 multiply_cannon_sync_h2d 27432 15.5 2.122 2.733 2.122 2.733 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.701 2.703 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.582 2.673 ot_diis_step 108 11.5 0.011 0.015 2.613 2.615 pw_transfer 1439 11.6 0.063 0.067 2.461 2.555 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.371 2.468 calculate_first_density_matrix 1 7.0 0.000 0.001 2.401 2.403 calculate_dm_sparse 119 9.5 0.000 0.001 2.207 2.289 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.265 2.270 fft3d_ps 1201 14.6 0.498 0.549 2.079 2.169 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.156 2.156 wfi_extrapolate 11 7.9 0.002 0.011 2.134 2.134 cp_fm_redistribute_end 50 14.0 1.761 2.109 1.768 2.110 prepare_preconditioner 11 7.9 0.000 0.000 2.046 2.085 make_preconditioner 11 8.9 0.000 0.000 2.046 2.085 cp_fm_diag_elpa_base 50 14.0 0.332 2.019 0.340 2.057 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.935 2.007 make_images_data 4572 15.5 0.045 0.051 1.378 1.996 jit_kernel_multiply 10 16.1 0.834 1.930 0.834 1.930 mp_alltoall_d11v 2130 13.8 1.781 1.924 1.781 1.924 grid_integrate_task_list 119 12.3 1.819 1.913 1.819 1.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.807 1.889 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.207 1.712 mp_allgather_i34 2286 14.5 0.828 1.705 0.828 1.705 fft_wrap_pw1pw2_140 487 13.2 0.076 0.081 1.600 1.691 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.513 1.554 acc_transpose_blocks 27432 15.5 0.106 0.112 1.205 1.516 make_images_sizes 4572 15.5 0.005 0.005 1.099 1.454 mp_alltoall_i44 4572 16.5 1.094 1.449 1.094 1.449 mp_alltoall_z22v 1201 16.6 1.196 1.357 1.196 1.357 grid_collocate_task_list 119 9.7 1.243 1.345 1.243 1.345 mp_sum_d 4129 12.0 0.788 1.289 0.788 1.289 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.265 1.274 rs_pw_transfer_PW2RS_50 119 14.3 0.586 0.605 1.108 1.198 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.063 1.099 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.036 1.038 dbcsr_complete_redistribute 329 12.2 0.083 0.101 0.943 1.015 multiply_cannon_metrocomm1 27432 15.5 0.031 0.035 0.439 0.933 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.516 0.920 acc_transpose_blocks_kernels 27432 16.5 0.181 0.267 0.679 0.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.846000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.636364, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 517.677056E+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 0.025 0.051 37.696 37.698 qs_mol_dyn_low 1 2.0 0.003 0.006 37.404 37.413 qs_forces 11 3.9 0.004 0.018 37.333 37.334 qs_energies 11 4.9 0.007 0.049 35.593 35.597 scf_env_do_scf 11 5.9 0.001 0.001 30.354 30.354 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.554 26.561 dbcsr_multiply_generic 2286 12.5 0.095 0.099 19.028 19.201 qs_scf_new_mos 108 7.5 0.001 0.001 17.774 17.806 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.773 17.806 velocity_verlet 10 3.0 0.001 0.002 17.666 17.668 ot_scf_mini 108 9.5 0.003 0.005 16.817 16.856 multiply_cannon 2286 13.5 0.193 0.202 14.345 15.188 multiply_cannon_loop 2286 14.5 0.638 0.669 13.312 14.391 ot_mini 108 10.5 0.001 0.002 10.327 10.358 qs_ot_get_derivative 108 11.5 0.001 0.002 8.724 8.750 multiply_cannon_multrec 18288 15.5 1.996 3.174 6.872 7.369 rebuild_ks_matrix 119 8.3 0.000 0.000 6.638 6.660 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.638 6.660 mp_waitall_1 169478 16.3 4.802 6.587 4.802 6.587 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.750 5.771 dbcsr_mm_accdrv_process 38222 16.0 4.030 5.510 4.793 5.604 sum_up_and_integrate 119 10.3 0.030 0.031 4.226 4.235 integrate_v_rspace 119 11.3 0.002 0.003 4.196 4.209 qs_ot_get_p 119 10.4 0.001 0.001 4.038 4.075 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.913 3.924 calculate_rho_elec 119 8.7 0.030 0.031 3.912 3.923 init_scf_loop 11 6.9 0.000 0.002 3.756 3.766 init_scf_run 11 5.9 0.000 0.001 3.764 3.764 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.763 3.764 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.976 3.700 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.929 3.635 rs_pw_transfer 974 11.9 0.009 0.010 2.967 3.099 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.397 2.825 apply_single 119 13.6 0.000 0.000 2.397 2.825 make_m2s 4572 13.5 0.044 0.046 2.442 2.706 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.607 2.655 make_images 4572 14.5 0.192 0.205 2.357 2.618 prepare_preconditioner 11 7.9 0.000 0.000 2.521 2.527 make_preconditioner 11 8.9 0.000 0.001 2.521 2.527 density_rs2pw 119 9.7 0.004 0.004 2.420 2.522 calculate_first_density_matrix 1 7.0 0.000 0.001 2.410 2.411 mp_sum_l 7207 12.9 1.859 2.262 1.859 2.262 calculate_dm_sparse 119 9.5 0.000 0.000 2.154 2.172 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.025 2.025 potential_pw2rs 119 12.3 0.007 0.008 1.971 1.990 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.892 1.983 pw_transfer 1439 11.6 0.063 0.066 1.916 1.946 jit_kernel_multiply 10 16.1 0.712 1.920 0.712 1.920 grid_integrate_task_list 119 12.3 1.807 1.899 1.807 1.899 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.826 1.857 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.783 1.785 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.749 1.750 cp_fm_redistribute_end 50 14.0 1.294 1.719 1.296 1.720 cp_fm_diag_elpa_base 50 14.0 0.406 1.648 0.421 1.686 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.652 1.664 multiply_cannon_sync_h2d 18288 15.5 1.371 1.609 1.371 1.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.590 1.602 mp_alltoall_d11v 2130 13.8 1.411 1.570 1.411 1.570 ot_diis_step 108 11.5 0.012 0.015 1.556 1.557 fft3d_ps 1201 14.6 0.507 0.524 1.514 1.546 make_images_data 4572 15.5 0.044 0.049 1.119 1.447 dbcsr_complete_redistribute 329 12.2 0.093 0.100 1.309 1.426 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.377 1.408 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.303 1.323 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.302 1.310 wfi_extrapolate 11 7.9 0.001 0.002 1.304 1.304 grid_collocate_task_list 119 9.7 1.205 1.290 1.205 1.290 acc_transpose_blocks 18288 15.5 0.074 0.077 1.262 1.289 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.793 1.279 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.956 1.246 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.984 1.097 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 1.028 1.030 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.958 0.960 mp_allgather_i34 2286 14.5 0.495 0.959 0.495 0.959 mp_alltoall_z22v 1201 16.6 0.832 0.906 0.832 0.906 cp_fm_cholesky_invert 11 10.9 0.886 0.892 0.886 0.892 make_images_sizes 4572 15.5 0.005 0.005 0.634 0.861 mp_alltoall_i44 4572 16.5 0.629 0.856 0.629 0.856 acc_transpose_blocks_kernels 18288 16.5 0.207 0.217 0.826 0.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=37.698000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=493.181818, yerr=0.833196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 552.472576E+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 0.030 0.053 41.324 41.325 qs_mol_dyn_low 1 2.0 0.003 0.006 40.999 41.007 qs_forces 11 3.9 0.002 0.005 40.930 40.932 qs_energies 11 4.9 0.002 0.003 39.208 39.216 scf_env_do_scf 11 5.9 0.001 0.004 33.970 33.971 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 28.372 28.397 velocity_verlet 10 3.0 0.001 0.002 20.360 20.363 dbcsr_multiply_generic 2286 12.5 0.101 0.105 20.112 20.280 qs_scf_new_mos 108 7.5 0.001 0.001 18.801 18.895 qs_scf_loop_do_ot 108 8.5 0.001 0.002 18.800 18.895 ot_scf_mini 108 9.5 0.004 0.015 17.675 17.751 multiply_cannon 2286 13.5 0.230 0.270 15.211 15.709 multiply_cannon_loop 2286 14.5 0.936 0.969 14.066 14.641 ot_mini 108 10.5 0.001 0.002 10.476 10.574 multiply_cannon_multrec 27432 15.5 2.352 3.010 8.672 8.984 qs_ot_get_derivative 108 11.5 0.001 0.007 8.583 8.661 dbcsr_mm_accdrv_process 47916 15.9 5.345 6.908 6.227 7.414 rebuild_ks_matrix 119 8.3 0.000 0.000 6.983 7.034 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.037 6.983 7.034 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.258 6.307 init_scf_loop 11 6.9 0.001 0.007 5.575 5.625 prepare_preconditioner 11 7.9 0.000 0.000 4.501 4.512 make_preconditioner 11 8.9 0.003 0.015 4.501 4.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.095 4.437 mp_waitall_1 145218 16.4 3.781 4.416 3.781 4.416 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.334 4.365 calculate_rho_elec 119 8.7 0.040 0.046 4.333 4.364 qs_ot_get_p 119 10.4 0.002 0.007 4.106 4.177 sum_up_and_integrate 119 10.3 0.035 0.046 4.155 4.162 integrate_v_rspace 119 11.3 0.002 0.003 4.119 4.129 init_scf_run 11 5.9 0.000 0.001 3.815 3.816 scf_env_initial_rho_setup 11 6.9 0.001 0.004 3.815 3.816 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.879 3.449 rs_pw_transfer 974 11.9 0.009 0.010 3.010 3.164 make_m2s 4572 13.5 0.054 0.056 2.781 2.941 density_rs2pw 119 9.7 0.003 0.004 2.706 2.851 make_images 4572 14.5 0.274 0.334 2.675 2.833 apply_preconditioner_dbcsr 119 12.6 0.000 0.003 2.309 2.807 apply_single 119 13.6 0.000 0.000 2.309 2.807 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.516 2.528 calculate_first_density_matrix 1 7.0 0.000 0.001 2.336 2.338 calculate_dm_sparse 119 9.5 0.000 0.000 2.255 2.309 mp_sum_l 7207 12.9 1.698 2.236 1.698 2.236 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.172 2.174 pw_transfer 1439 11.6 0.063 0.066 2.128 2.167 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.139 2.166 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.038 2.081 cp_fm_cholesky_decompose 22 10.9 2.012 2.017 2.012 2.017 potential_pw2rs 119 12.3 0.008 0.010 1.891 1.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.900 1.902 jit_kernel_multiply 10 16.0 0.822 1.898 0.822 1.898 grid_integrate_task_list 119 12.3 1.817 1.894 1.817 1.894 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.847 1.849 ot_diis_step 108 11.5 0.014 0.021 1.838 1.839 cp_fm_redistribute_end 50 14.0 1.213 1.804 1.215 1.806 multiply_cannon_metrocomm3 27432 15.5 0.037 0.038 1.075 1.789 cp_fm_diag_elpa_base 50 14.0 0.562 1.723 0.585 1.762 fft3d_ps 1201 14.6 0.534 0.587 1.720 1.751 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.667 1.686 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.650 1.675 fft_wrap_pw1pw2_140 487 13.2 0.084 0.091 1.548 1.594 acc_transpose_blocks 27432 15.5 0.111 0.114 1.471 1.539 wfi_extrapolate 11 7.9 0.001 0.005 1.402 1.403 make_images_data 4572 15.5 0.045 0.048 1.137 1.353 grid_collocate_task_list 119 9.7 1.223 1.326 1.223 1.326 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.252 1.262 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.959 1.231 dbcsr_complete_redistribute 329 12.2 0.126 0.151 0.929 1.215 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.153 1.155 cp_fm_upper_to_full 72 13.5 0.824 1.148 0.824 1.148 mp_alltoall_d11v 2130 13.8 0.988 1.122 0.988 1.122 cp_fm_cholesky_invert 11 10.9 1.116 1.121 1.116 1.121 mp_alltoall_z22v 1201 16.6 1.068 1.101 1.068 1.101 multiply_cannon_sync_h2d 27432 15.5 0.992 1.097 0.992 1.097 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.663 0.941 acc_transpose_blocks_kernels 27432 16.5 0.266 0.274 0.851 0.905 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.870 multiply_cannon_metrocomm1 27432 15.5 0.032 0.033 0.408 0.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.831 0.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=41.325000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 620.228608E+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 0.028 0.045 34.696 34.697 qs_mol_dyn_low 1 2.0 0.003 0.003 34.458 34.465 qs_forces 11 3.9 0.002 0.002 34.386 34.386 qs_energies 11 4.9 0.001 0.002 32.656 32.659 scf_env_do_scf 11 5.9 0.000 0.001 27.723 27.723 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.180 25.181 velocity_verlet 10 3.0 0.001 0.001 18.393 18.396 dbcsr_multiply_generic 2286 12.5 0.093 0.095 15.862 15.964 qs_scf_new_mos 108 7.5 0.001 0.001 15.550 15.587 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.549 15.586 ot_scf_mini 108 9.5 0.002 0.002 14.799 14.836 multiply_cannon 2286 13.5 0.233 0.242 11.471 12.236 multiply_cannon_loop 2286 14.5 0.331 0.344 10.339 10.579 ot_mini 108 10.5 0.001 0.001 8.818 8.857 qs_ot_get_derivative 108 11.5 0.001 0.001 7.240 7.278 rebuild_ks_matrix 119 8.3 0.000 0.000 6.947 6.984 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.946 6.984 multiply_cannon_multrec 9144 15.5 1.697 1.951 6.184 6.418 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.284 6.318 mp_waitall_1 121218 16.5 4.718 5.259 4.718 5.259 dbcsr_mm_accdrv_process 12550 15.8 2.937 3.631 4.386 4.458 sum_up_and_integrate 119 10.3 0.037 0.041 4.429 4.433 integrate_v_rspace 119 11.3 0.003 0.003 4.391 4.397 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.152 4.160 calculate_rho_elec 119 8.7 0.059 0.061 4.151 4.160 qs_ot_get_p 119 10.4 0.001 0.001 3.838 3.880 init_scf_run 11 5.9 0.000 0.001 3.448 3.448 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.447 3.448 make_m2s 4572 13.5 0.034 0.035 3.114 3.316 make_images 4572 14.5 0.272 0.309 3.024 3.226 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.010 3.026 rs_pw_transfer 974 11.9 0.008 0.008 2.865 2.933 density_rs2pw 119 9.7 0.004 0.004 2.498 2.557 init_scf_loop 11 6.9 0.000 0.000 2.522 2.523 calculate_first_density_matrix 1 7.0 0.000 0.000 2.316 2.317 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.247 2.251 pw_transfer 1439 11.6 0.063 0.065 2.225 2.244 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.135 2.153 jit_kernel_multiply 10 15.6 1.411 2.087 1.411 2.087 potential_pw2rs 119 12.3 0.010 0.011 2.061 2.082 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.004 2.007 make_images_data 4572 15.5 0.038 0.042 1.547 2.000 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.906 1.927 grid_integrate_task_list 119 12.3 1.842 1.926 1.842 1.926 calculate_dm_sparse 119 9.5 0.000 0.000 1.876 1.900 fft3d_ps 1201 14.6 0.542 0.552 1.803 1.824 hybrid_alltoall_any 4725 16.4 0.062 0.175 1.346 1.799 prepare_preconditioner 11 7.9 0.000 0.000 1.787 1.792 make_preconditioner 11 8.9 0.000 0.000 1.787 1.792 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.674 1.703 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.647 1.665 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 1.526 1.652 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.637 1.638 ot_diis_step 108 11.5 0.013 0.013 1.548 1.550 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.482 1.492 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.471 1.480 make_images_sizes 4572 15.5 0.005 0.005 0.898 1.468 mp_alltoall_i44 4572 16.5 0.893 1.463 0.893 1.463 cp_fm_redistribute_end 50 14.0 0.727 1.394 0.729 1.395 mp_alltoall_d11v 2130 13.8 1.261 1.391 1.261 1.391 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.346 1.373 apply_single 119 13.6 0.000 0.000 1.346 1.373 grid_collocate_task_list 119 9.7 1.272 1.363 1.272 1.363 cp_fm_diag_elpa_base 50 14.0 0.604 1.244 0.647 1.299 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.265 1.266 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.223 1.229 mp_alltoall_z22v 1201 16.6 1.134 1.168 1.134 1.168 mp_allgather_i34 2286 14.5 0.462 1.152 0.462 1.152 wfi_extrapolate 11 7.9 0.001 0.001 1.085 1.085 acc_transpose_blocks 9144 15.5 0.038 0.040 1.004 1.020 mp_sum_l 7207 12.9 0.840 0.988 0.840 0.988 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.923 arnoldi_extremal 119 11.4 0.001 0.001 0.864 0.896 arnoldi_normal_ev 119 12.4 0.003 0.003 0.863 0.894 cp_fm_cholesky_invert 11 10.9 0.876 0.879 0.876 0.879 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.867 0.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.786 0.789 acc_transpose_blocks_kernels 9144 16.5 0.115 0.119 0.766 0.787 multiply_cannon_sync_h2d 9144 15.5 0.704 0.783 0.704 0.783 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.713 0.770 yz_to_x 606 15.1 0.053 0.056 0.675 0.700 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=34.697000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=585.545455, yerr=6.853846 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 752.984064E+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 0.016 0.043 52.666 52.667 qs_mol_dyn_low 1 2.0 0.003 0.003 52.411 52.418 qs_forces 11 3.9 0.002 0.002 52.347 52.348 qs_energies 11 4.9 0.002 0.005 50.307 50.311 scf_env_do_scf 11 5.9 0.001 0.001 42.732 42.732 velocity_verlet 10 3.0 0.001 0.001 32.268 32.275 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 31.847 31.856 dbcsr_multiply_generic 2286 12.5 0.102 0.104 21.447 21.605 qs_scf_new_mos 108 7.5 0.001 0.001 21.377 21.479 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.376 21.478 ot_scf_mini 108 9.5 0.002 0.002 20.063 20.163 multiply_cannon 2286 13.5 0.306 0.318 15.360 16.776 multiply_cannon_loop 2286 14.5 0.345 0.352 13.643 14.611 ot_mini 108 10.5 0.001 0.001 11.609 11.743 init_scf_loop 11 6.9 0.000 0.000 10.791 10.797 qs_ot_get_derivative 108 11.5 0.001 0.001 9.233 9.335 prepare_preconditioner 11 7.9 0.000 0.000 8.941 8.955 make_preconditioner 11 8.9 0.000 0.000 8.941 8.955 multiply_cannon_multrec 9144 15.5 3.413 4.869 8.590 8.673 rebuild_ks_matrix 119 8.3 0.000 0.000 8.140 8.300 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.139 8.300 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.717 8.269 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.417 7.563 mp_waitall_1 97218 16.6 5.603 6.887 5.603 6.887 dbcsr_mm_accdrv_process 12550 15.8 3.930 5.655 5.053 6.392 init_scf_run 11 5.9 0.000 0.001 5.464 5.465 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.464 5.465 qs_ot_get_p 119 10.4 0.001 0.001 5.173 5.364 cp_fm_upper_to_full 72 14.2 3.311 4.925 3.311 4.925 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.771 4.852 calculate_rho_elec 119 8.7 0.118 0.121 4.770 4.851 sum_up_and_integrate 119 10.3 0.064 0.065 4.706 4.716 integrate_v_rspace 119 11.3 0.003 0.003 4.642 4.650 dbcsr_complete_redistribute 329 12.2 0.279 0.283 3.699 4.632 make_m2s 4572 13.5 0.037 0.038 3.867 4.282 make_images 4572 14.5 0.357 0.388 3.746 4.162 copy_fm_to_dbcsr 176 11.2 0.001 0.001 2.871 3.806 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.298 3.763 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 2.546 3.488 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.219 3.142 wfi_extrapolate 11 7.9 0.001 0.001 2.978 2.979 pw_transfer 1439 11.6 0.066 0.067 2.864 2.872 mp_alltoall_i22 627 13.8 1.849 2.823 1.849 2.823 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.562 2.819 apply_single 119 13.6 0.000 0.000 2.561 2.819 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.785 2.786 mp_alltoall_d11v 2130 13.8 2.671 2.785 2.671 2.785 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.768 2.775 make_images_data 4572 15.5 0.043 0.045 2.133 2.770 mp_sum_l 7207 12.9 1.671 2.702 1.671 2.702 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.670 2.674 density_rs2pw 119 9.7 0.003 0.003 2.618 2.645 hybrid_alltoall_any 4725 16.4 0.087 0.149 2.032 2.613 calculate_dm_sparse 119 9.5 0.000 0.000 2.503 2.567 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.486 2.487 fft3d_ps 1201 14.6 0.571 0.582 2.401 2.408 calculate_first_density_matrix 1 7.0 0.000 0.000 2.320 2.385 ot_diis_step 108 11.5 0.014 0.014 2.328 2.331 cp_fm_cholesky_invert 11 10.9 2.313 2.317 2.313 2.317 rs_pw_transfer 974 11.9 0.009 0.009 2.248 2.273 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 2.131 2.140 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.102 2.107 cp_fm_diag_elpa_base 50 14.0 1.940 2.003 2.098 2.103 grid_integrate_task_list 119 12.3 2.065 2.098 2.065 2.098 potential_pw2rs 119 12.3 0.014 0.015 1.947 1.953 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.903 1.947 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.850 1.911 jit_kernel_multiply 10 15.4 1.096 1.769 1.096 1.769 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.757 1.758 mp_alltoall_z22v 1201 16.6 1.693 1.721 1.693 1.721 copy_dbcsr_to_fm 153 11.3 0.002 0.002 1.619 1.657 mp_allgather_i34 2286 14.5 0.695 1.559 0.695 1.559 grid_collocate_task_list 119 9.7 1.493 1.521 1.493 1.521 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.485 1.520 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.364 1.423 make_images_sizes 4572 15.5 0.005 0.005 0.785 1.287 mp_alltoall_i44 4572 16.5 0.780 1.283 0.780 1.283 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.092 1.108 arnoldi_extremal 119 11.4 0.001 0.001 1.077 1.103 arnoldi_normal_ev 119 12.4 0.005 0.006 1.076 1.102 yz_to_x 606 15.1 0.057 0.059 1.050 1.067 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=52.667000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=710.000000, yerr=11.908744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.190080E+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 2461858. 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 0.054 0.119 116.605 116.608 qs_mol_dyn_low 1 2.0 0.004 0.006 116.154 116.163 qs_forces 11 3.9 0.002 0.004 116.048 116.050 qs_energies 11 4.9 0.003 0.011 111.119 111.142 scf_env_do_scf 11 5.9 0.001 0.006 99.332 99.336 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 89.346 89.348 dbcsr_multiply_generic 2055 12.4 0.106 0.111 61.801 62.438 qs_scf_new_mos 99 7.5 0.001 0.001 61.701 61.851 qs_scf_loop_do_ot 99 8.5 0.001 0.001 61.701 61.850 velocity_verlet 10 3.0 0.001 0.003 61.505 61.507 ot_scf_mini 99 9.5 0.002 0.003 57.999 58.156 multiply_cannon 2055 13.4 0.176 0.181 47.355 49.084 multiply_cannon_loop 2055 14.4 1.546 1.581 45.493 46.820 ot_mini 99 10.5 0.001 0.002 32.416 32.543 qs_ot_get_derivative 99 11.5 0.001 0.002 24.895 25.014 mp_waitall_1 241148 16.1 22.843 24.941 22.843 24.941 rebuild_ks_matrix 110 8.3 0.000 0.000 20.865 20.990 qs_ks_build_kohn_sham_matrix 110 9.3 0.023 0.061 20.865 20.989 multiply_cannon_multrec 49320 15.4 11.754 12.437 16.918 17.718 qs_ks_update_qs_env 110 7.6 0.001 0.001 17.141 17.241 qs_ot_get_p 110 10.4 0.001 0.002 14.941 15.059 qs_rho_update_rho_low 110 7.6 0.001 0.004 14.920 15.027 calculate_rho_elec 110 8.6 0.020 0.026 14.919 15.027 sum_up_and_integrate 110 10.3 0.037 0.047 12.088 12.101 integrate_v_rspace 110 11.3 0.003 0.009 12.050 12.073 rs_pw_transfer 902 11.9 0.012 0.013 10.375 10.866 density_rs2pw 110 9.6 0.005 0.011 10.621 10.827 multiply_cannon_sync_h2d 49320 15.4 9.589 10.299 9.589 10.299 init_scf_loop 11 6.9 0.008 0.033 9.944 9.945 pw_transfer 1331 11.6 0.054 0.063 9.305 9.576 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 9.218 9.491 init_scf_run 11 5.9 0.000 0.001 9.468 9.472 scf_env_initial_rho_setup 11 6.9 0.001 0.007 9.468 9.472 fft3d_ps 1111 14.6 0.750 0.842 8.759 9.024 multiply_cannon_metrocomm3 49320 15.4 0.075 0.080 7.734 8.876 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.339 8.864 multiply_cannon_metrocomm1 49320 15.4 0.058 0.061 6.420 8.195 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.623 8.170 apply_single 110 13.6 0.000 0.000 7.622 8.170 mp_sum_l 6514 12.8 6.957 8.085 6.957 8.085 qs_ot_p2m_diag 48 11.0 0.012 0.020 7.905 7.952 fft_wrap_pw1pw2_140 451 13.1 0.165 0.191 7.082 7.356 potential_pw2rs 110 12.3 0.006 0.007 7.077 7.322 mp_alltoall_z22v 1111 16.6 6.948 7.309 6.948 7.309 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.237 7.296 make_m2s 4110 13.4 0.061 0.066 7.070 7.295 ot_diis_step 99 11.5 0.012 0.032 7.236 7.245 make_images 4110 14.4 0.179 0.193 6.973 7.202 prepare_preconditioner 11 7.9 0.000 0.000 6.584 6.600 make_preconditioner 11 8.9 0.001 0.004 6.584 6.600 cp_dbcsr_syevd 48 12.0 0.003 0.004 6.455 6.460 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.835 5.958 mp_alltoall_d11v 2046 13.8 5.139 5.619 5.139 5.619 calculate_dm_sparse 110 9.5 0.000 0.001 5.316 5.520 dbcsr_mm_accdrv_process 87628 16.1 2.090 2.208 5.046 5.381 cp_fm_diag_elpa 48 13.0 0.001 0.008 4.992 5.000 qs_ot_get_orbitals 99 10.5 0.000 0.001 4.849 4.928 cp_fm_redistribute_end 48 14.0 4.226 4.813 4.233 4.819 wfi_extrapolate 11 7.9 0.002 0.004 4.770 4.773 cp_fm_diag_elpa_base 48 14.0 0.554 4.450 0.560 4.496 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.916 3.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.890 3.921 x_to_yz 330 15.4 0.078 0.098 3.906 3.920 cp_fm_cholesky_invert 11 10.9 3.769 3.801 3.769 3.801 yz_to_x 341 15.4 0.061 0.073 3.173 3.499 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.368 3.418 make_images_sizes 4110 15.4 0.004 0.005 2.669 3.412 mp_alltoall_i44 4110 16.4 2.665 3.408 2.665 3.408 grid_integrate_task_list 110 12.3 3.212 3.397 3.212 3.397 make_images_data 4110 15.4 0.043 0.046 2.861 3.365 mp_waitany 14300 13.8 2.972 3.245 2.972 3.245 arnoldi_extremal 110 11.4 0.001 0.002 2.830 3.015 arnoldi_normal_ev 110 12.4 0.002 0.006 2.829 3.013 hybrid_alltoall_any 4261 16.3 0.081 0.464 2.423 2.996 mp_sum_d 3883 11.9 2.381 2.927 2.381 2.927 calculate_first_density_matrix 1 7.0 0.000 0.002 2.892 2.912 jit_kernel_multiply 13 15.9 2.672 2.733 2.672 2.733 rs_scatter_matrices 121 9.7 0.103 0.124 2.503 2.718 rs_pw_transfer_RS2PW_20 110 11.6 0.075 0.084 2.345 2.652 dbcsr_complete_redistribute 325 12.2 0.136 0.181 2.232 2.488 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=116.608000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=3.146558 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 585.003008E+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 2594795. 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 0.016 0.033 82.530 82.531 qs_mol_dyn_low 1 2.0 0.003 0.003 82.157 82.189 qs_forces 11 3.9 0.002 0.002 82.089 82.090 qs_energies 11 4.9 0.002 0.005 78.805 78.809 scf_env_do_scf 11 5.9 0.000 0.001 69.394 69.397 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 61.284 61.284 dbcsr_multiply_generic 2055 12.4 0.114 0.118 43.297 43.765 velocity_verlet 10 3.0 0.001 0.001 43.353 43.354 qs_scf_new_mos 99 7.5 0.001 0.001 40.005 40.252 qs_scf_loop_do_ot 99 8.5 0.001 0.001 40.005 40.251 ot_scf_mini 99 9.5 0.003 0.003 38.123 38.345 multiply_cannon 2055 13.4 0.221 0.242 34.081 36.095 multiply_cannon_loop 2055 14.4 0.928 0.948 32.198 33.565 ot_mini 99 10.5 0.001 0.001 22.128 22.379 rebuild_ks_matrix 110 8.3 0.000 0.001 16.793 17.065 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 16.792 17.065 multiply_cannon_multrec 24660 15.4 7.490 9.442 14.051 16.233 qs_ot_get_derivative 99 11.5 0.001 0.001 15.869 16.089 qs_ks_update_qs_env 110 7.6 0.001 0.001 15.209 15.479 mp_waitall_1 186928 16.3 11.926 15.266 11.926 15.266 multiply_cannon_metrocomm3 24660 15.4 0.069 0.072 6.848 10.808 qs_ot_get_p 110 10.4 0.001 0.001 8.828 8.967 sum_up_and_integrate 110 10.3 0.052 0.060 8.555 8.566 integrate_v_rspace 110 11.3 0.002 0.003 8.503 8.516 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.415 8.451 calculate_rho_elec 110 8.6 0.039 0.047 8.414 8.450 multiply_cannon_sync_h2d 24660 15.4 6.877 8.209 6.877 8.209 init_scf_loop 11 6.9 0.000 0.000 8.072 8.073 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.988 7.713 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.613 7.236 apply_single 110 13.6 0.000 0.001 6.612 7.236 init_scf_run 11 5.9 0.000 0.001 7.006 7.007 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.006 7.007 dbcsr_mm_accdrv_process 52282 16.1 4.818 5.926 6.406 6.720 ot_diis_step 99 11.5 0.010 0.010 6.165 6.166 qs_ot_p2m_diag 48 11.0 0.028 0.043 5.939 5.962 prepare_preconditioner 11 7.9 0.000 0.000 5.888 5.905 make_preconditioner 11 8.9 0.000 0.000 5.888 5.905 make_m2s 4110 13.4 0.057 0.060 5.400 5.822 make_images 4110 14.4 0.402 0.445 5.291 5.710 pw_transfer 1331 11.6 0.065 0.068 5.368 5.604 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.449 5.602 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 5.261 5.499 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.433 5.435 density_rs2pw 110 9.6 0.004 0.004 4.977 5.219 fft3d_ps 1111 14.6 1.071 1.263 4.621 4.844 rs_pw_transfer 902 11.9 0.012 0.014 4.519 4.831 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.693 4.698 cp_fm_redistribute_end 48 14.0 3.497 4.650 3.499 4.650 cp_fm_diag_elpa_base 48 14.0 1.109 4.494 1.144 4.582 mp_sum_l 6514 12.8 3.193 4.219 3.193 4.219 fft_wrap_pw1pw2_140 451 13.1 0.198 0.218 3.966 4.188 potential_pw2rs 110 12.3 0.008 0.009 3.781 3.903 mp_alltoall_d11v 2046 13.8 3.485 3.880 3.485 3.880 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.677 3.745 calculate_dm_sparse 110 9.5 0.001 0.001 3.502 3.585 wfi_extrapolate 11 7.9 0.001 0.001 3.495 3.495 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.422 3.423 mp_alltoall_z22v 1111 16.6 3.138 3.326 3.138 3.326 make_images_data 4110 15.4 0.047 0.051 2.845 3.310 grid_integrate_task_list 110 12.3 3.126 3.286 3.126 3.286 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.995 3.062 hybrid_alltoall_any 4261 16.3 0.102 0.439 2.428 3.059 calculate_first_density_matrix 1 7.0 0.000 0.001 3.021 3.031 cp_fm_cholesky_invert 11 10.9 2.718 2.728 2.718 2.728 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.296 2.326 mp_allgather_i34 2055 14.4 1.211 2.319 1.211 2.319 grid_collocate_task_list 110 9.6 2.126 2.291 2.126 2.291 jit_kernel_multiply 9 16.4 1.233 1.973 1.233 1.973 rs_gather_matrices 110 12.3 0.129 0.140 1.548 1.914 multiply_cannon_metrocomm1 24660 15.4 0.036 0.037 0.790 1.904 qs_energies_init_hamiltonians 11 5.9 0.006 0.011 1.893 1.894 yz_to_x 451 15.2 0.088 0.102 1.739 1.885 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.853 1.873 mp_sum_d 3883 11.9 1.395 1.818 1.395 1.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.763 1.776 dbcsr_complete_redistribute 325 12.2 0.283 0.371 1.416 1.683 acc_transpose_blocks 24660 15.4 0.105 0.108 1.627 1.658 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=82.531000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.090909, yerr=5.743843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 659.275776E+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 0.030 0.119 76.792 76.793 qs_mol_dyn_low 1 2.0 0.003 0.003 76.411 76.421 qs_forces 11 3.9 0.007 0.012 76.321 76.321 qs_energies 11 4.9 0.002 0.004 68.811 68.844 scf_env_do_scf 11 5.9 0.001 0.001 58.498 58.499 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.080 49.082 velocity_verlet 10 3.0 0.001 0.001 41.501 41.504 dbcsr_multiply_generic 2055 12.4 0.109 0.112 32.749 33.273 qs_scf_new_mos 99 7.5 0.001 0.001 32.552 32.718 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.551 32.718 ot_scf_mini 99 9.5 0.002 0.003 31.248 31.427 multiply_cannon 2055 13.4 0.212 0.223 24.116 25.675 multiply_cannon_loop 2055 14.4 0.619 0.633 22.467 23.966 rebuild_ks_matrix 110 8.3 0.000 0.000 17.561 17.798 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 17.561 17.797 ot_mini 99 10.5 0.001 0.001 16.144 16.337 mp_waitall_1 146766 16.3 11.795 15.903 11.795 15.903 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.421 12.601 qs_ot_get_derivative 99 11.5 0.001 0.001 11.177 11.332 multiply_cannon_multrec 16440 15.4 3.837 5.069 9.623 10.730 sum_up_and_integrate 110 10.3 0.059 0.060 10.691 10.718 integrate_v_rspace 110 11.3 0.003 0.003 10.632 10.659 qs_ot_get_p 110 10.4 0.001 0.001 9.511 9.703 init_scf_loop 11 6.9 0.000 0.002 9.376 9.382 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 5.506 9.097 init_scf_run 11 5.9 0.000 0.001 7.882 7.882 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.881 7.882 prepare_preconditioner 11 7.9 0.000 0.000 7.528 7.556 make_preconditioner 11 8.9 0.000 0.002 7.528 7.556 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.888 7.227 qs_ot_p2m_diag 48 11.0 0.042 0.044 6.842 6.849 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.424 6.440 calculate_rho_elec 110 8.6 0.058 0.058 6.424 6.440 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.413 6.418 make_m2s 4110 13.4 0.050 0.052 5.749 6.294 make_images 4110 14.4 0.395 0.512 5.634 6.180 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.321 5.852 apply_single 110 13.6 0.000 0.000 5.321 5.852 dbcsr_mm_accdrv_process 34862 16.1 4.600 5.215 5.645 5.799 rs_pw_transfer 902 11.9 0.011 0.011 5.155 5.495 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.475 5.486 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 5.373 5.443 cp_fm_redistribute_end 48 14.0 3.422 5.381 3.426 5.385 wfi_extrapolate 11 7.9 0.001 0.001 5.334 5.334 cp_fm_diag_elpa_base 48 14.0 1.867 5.071 1.937 5.190 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.448 5.122 potential_pw2rs 110 12.3 0.010 0.010 4.948 5.077 ot_diis_step 99 11.5 0.011 0.011 4.874 4.879 pw_transfer 1331 11.6 0.065 0.071 4.614 4.636 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.508 4.531 multiply_cannon_sync_h2d 16440 15.4 3.606 4.336 3.606 4.336 mp_alltoall_d11v 2046 13.8 3.974 4.289 3.974 4.289 make_images_data 4110 15.4 0.043 0.048 3.397 4.124 fft3d_ps 1111 14.6 1.064 1.076 3.825 3.847 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.854 3.824 density_rs2pw 110 9.6 0.004 0.004 3.522 3.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.645 3.647 fft_wrap_pw1pw2_140 451 13.1 0.208 0.210 3.411 3.433 grid_integrate_task_list 110 12.3 3.157 3.374 3.157 3.374 mp_sum_l 6514 12.8 2.349 3.219 2.349 3.219 cp_fm_cholesky_invert 11 10.9 3.187 3.197 3.187 3.197 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.827 3.012 calculate_dm_sparse 110 9.5 0.001 0.001 2.722 2.762 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.588 2.664 mp_allgather_i34 2055 14.4 0.970 2.633 0.970 2.633 mp_alltoall_z22v 1111 16.6 2.548 2.578 2.548 2.578 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.894 2.474 make_basis_sm 11 9.8 0.000 0.001 2.445 2.462 calculate_first_density_matrix 1 7.0 0.000 0.000 2.457 2.459 dbcsr_complete_redistribute 325 12.2 0.317 0.354 1.936 2.397 grid_collocate_task_list 110 9.6 2.180 2.381 2.180 2.381 mp_irecv_dv 48980 15.7 0.825 2.347 0.825 2.347 reorthogonalize_vectors 10 9.0 0.000 0.000 2.255 2.272 cp_fm_cholesky_decompose 22 10.9 2.238 2.260 2.238 2.260 rs_gather_matrices 110 12.3 0.137 0.150 1.844 2.088 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.964 1.970 cp_fm_upper_to_full 70 13.6 1.411 1.930 1.411 1.930 jit_kernel_multiply 10 16.3 0.652 1.901 0.652 1.901 mp_waitany 17072 13.8 1.658 1.896 1.658 1.896 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.547 1.703 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.624 1.649 make_images_sizes 4110 15.4 0.005 0.005 1.001 1.622 x_to_yz 550 16.1 0.101 0.104 1.606 1.618 mp_alltoall_i44 4110 16.4 0.996 1.617 0.996 1.617 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.115 1.572 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=76.793000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.272727, yerr=8.528998 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 723.976192E+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 0.102 0.179 76.444 76.446 qs_mol_dyn_low 1 2.0 0.003 0.003 75.825 75.844 qs_forces 11 3.9 0.003 0.007 75.711 75.713 qs_energies 11 4.9 0.002 0.004 71.890 71.895 scf_env_do_scf 11 5.9 0.001 0.001 62.719 62.722 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 46.669 46.671 velocity_verlet 10 3.0 0.001 0.001 44.188 44.195 dbcsr_multiply_generic 2055 12.4 0.116 0.121 32.142 32.358 qs_scf_new_mos 99 7.5 0.001 0.001 31.430 31.529 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.429 31.529 ot_scf_mini 99 9.5 0.003 0.004 29.745 29.854 multiply_cannon 2055 13.4 0.246 0.263 23.434 24.388 multiply_cannon_loop 2055 14.4 0.882 0.908 21.705 22.351 ot_mini 99 10.5 0.001 0.001 16.926 17.043 init_scf_loop 11 6.9 0.001 0.003 16.001 16.008 prepare_preconditioner 11 7.9 0.000 0.000 14.130 14.149 make_preconditioner 11 8.9 0.000 0.001 14.130 14.149 make_full_inverse_cholesky 11 9.9 0.000 0.000 12.243 13.741 multiply_cannon_multrec 24660 15.4 4.141 6.696 12.572 13.598 qs_ot_get_derivative 99 11.5 0.001 0.001 12.564 12.676 rebuild_ks_matrix 110 8.3 0.000 0.000 12.380 12.511 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.380 12.511 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.970 11.055 dbcsr_mm_accdrv_process 52304 16.0 6.802 8.650 8.286 9.200 mp_waitall_1 126806 16.4 6.185 8.294 6.185 8.294 qs_ot_get_p 110 10.4 0.001 0.003 7.658 7.791 sum_up_and_integrate 110 10.3 0.068 0.071 6.703 6.716 integrate_v_rspace 110 11.3 0.003 0.003 6.634 6.650 make_m2s 4110 13.4 0.059 0.061 6.466 6.641 make_images 4110 14.4 0.582 0.706 6.326 6.499 init_scf_run 11 5.9 0.000 0.001 6.239 6.239 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.239 6.239 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.124 6.134 calculate_rho_elec 110 8.6 0.077 0.081 6.123 6.134 qs_ot_p2m_diag 48 11.0 0.054 0.062 5.162 5.178 cp_fm_cholesky_invert 11 10.9 5.129 5.145 5.129 5.145 cp_fm_upper_to_full 70 13.8 3.618 5.063 3.618 5.063 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.658 4.660 dbcsr_complete_redistribute 325 12.2 0.411 0.461 3.209 4.505 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.207 4.307 apply_single 110 13.6 0.000 0.000 4.206 4.307 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.222 4.283 ot_diis_step 99 11.5 0.011 0.011 4.277 4.278 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.946 4.152 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.991 3.995 cp_fm_redistribute_end 48 14.0 1.983 3.939 1.985 3.941 cp_fm_diag_elpa_base 48 14.0 1.858 3.771 1.950 3.904 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.429 3.720 pw_transfer 1331 11.6 0.064 0.071 3.623 3.660 make_images_data 4110 15.4 0.046 0.049 3.295 3.655 wfi_extrapolate 11 7.9 0.001 0.001 3.636 3.636 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.518 3.557 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.484 3.542 hybrid_alltoall_any 4261 16.3 0.120 0.455 2.817 3.441 grid_integrate_task_list 110 12.3 3.273 3.411 3.273 3.411 density_rs2pw 110 9.6 0.004 0.004 3.215 3.379 multiply_cannon_sync_h2d 24660 15.4 3.161 3.338 3.161 3.338 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.877 3.155 mp_alltoall_i22 605 13.7 1.787 3.053 1.787 3.053 calculate_dm_sparse 110 9.5 0.001 0.001 3.013 3.036 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.030 3.034 rs_pw_transfer 902 11.9 0.010 0.011 2.901 3.030 cp_fm_cholesky_decompose 22 10.9 2.935 2.976 2.935 2.976 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.908 2.951 fft3d_ps 1111 14.6 1.065 1.097 2.822 2.850 mp_alltoall_d11v 2046 13.8 2.430 2.636 2.430 2.636 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.490 2.532 calculate_first_density_matrix 1 7.0 0.000 0.000 2.506 2.508 potential_pw2rs 110 12.3 0.013 0.013 2.392 2.413 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.355 2.358 grid_collocate_task_list 110 9.6 2.220 2.340 2.220 2.340 mp_sum_l 6514 12.8 1.581 2.046 1.581 2.046 jit_kernel_multiply 11 15.8 1.152 1.982 1.152 1.982 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.887 1.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.659 1.706 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.704 mp_alltoall_z22v 1111 16.6 1.541 1.607 1.541 1.607 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.573 1.588 mp_allgather_i34 2055 14.4 0.826 1.542 0.826 1.542 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=76.446000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=685.272727, yerr=6.876718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 831.275008E+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 0.015 0.050 67.077 67.077 qs_mol_dyn_low 1 2.0 0.003 0.003 66.761 66.771 qs_forces 11 3.9 0.002 0.002 66.695 66.696 qs_energies 11 4.9 0.002 0.002 62.966 62.972 scf_env_do_scf 11 5.9 0.000 0.001 52.657 52.657 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.128 44.129 velocity_verlet 10 3.0 0.001 0.001 37.815 37.818 dbcsr_multiply_generic 2055 12.4 0.108 0.113 27.637 28.086 qs_scf_new_mos 99 7.5 0.001 0.001 27.380 27.436 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.379 27.436 ot_scf_mini 99 9.5 0.002 0.002 25.918 25.965 multiply_cannon 2055 13.4 0.249 0.261 19.695 21.099 multiply_cannon_loop 2055 14.4 0.323 0.336 17.706 18.382 ot_mini 99 10.5 0.001 0.001 13.156 13.189 rebuild_ks_matrix 110 8.3 0.000 0.000 12.820 12.882 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.820 12.882 mp_waitall_1 106626 16.5 9.705 12.221 9.705 12.221 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.514 11.566 qs_ot_get_derivative 99 11.5 0.001 0.001 8.696 8.770 init_scf_loop 11 6.9 0.000 0.000 8.464 8.466 multiply_cannon_multrec 8220 15.4 3.221 4.532 7.471 8.387 qs_ot_get_p 110 10.4 0.001 0.001 7.953 8.081 init_scf_run 11 5.9 0.000 0.001 7.175 7.175 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.175 7.175 sum_up_and_integrate 110 10.3 0.079 0.081 7.128 7.143 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.054 7.074 calculate_rho_elec 110 8.6 0.115 0.115 7.054 7.073 integrate_v_rspace 110 11.3 0.003 0.003 7.049 7.062 prepare_preconditioner 11 7.9 0.000 0.000 6.739 6.742 make_preconditioner 11 8.9 0.000 0.000 6.739 6.742 make_m2s 4110 13.4 0.038 0.040 5.881 6.422 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.262 6.350 make_images 4110 14.4 0.641 0.696 5.751 6.292 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.640 5.543 qs_ot_p2m_diag 48 11.0 0.081 0.084 5.316 5.325 dbcsr_mm_accdrv_process 17442 15.9 2.864 3.958 4.118 5.023 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.888 4.889 wfi_extrapolate 11 7.9 0.001 0.001 4.622 4.622 ot_diis_step 99 11.5 0.012 0.012 4.373 4.375 pw_transfer 1331 11.6 0.065 0.069 4.324 4.346 make_images_data 4110 15.4 0.037 0.043 3.443 4.344 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.218 4.243 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.118 4.187 apply_single 110 13.6 0.000 0.000 4.118 4.187 hybrid_alltoall_any 4261 16.3 0.199 0.858 3.082 4.055 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.882 3.887 density_rs2pw 110 9.6 0.004 0.004 3.721 3.848 cp_fm_redistribute_end 48 14.0 1.026 3.800 1.043 3.805 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.751 3.754 cp_fm_diag_elpa_base 48 14.0 2.559 3.506 2.741 3.661 grid_integrate_task_list 110 12.3 3.346 3.517 3.346 3.517 fft3d_ps 1111 14.6 1.116 1.138 3.477 3.502 fft_wrap_pw1pw2_140 451 13.1 0.211 0.213 3.468 3.495 rs_pw_transfer 902 11.9 0.010 0.010 3.165 3.350 cp_fm_cholesky_invert 11 10.9 3.295 3.300 3.295 3.300 multiply_cannon_sync_h2d 8220 15.4 2.897 2.996 2.897 2.996 mp_alltoall_d11v 2046 13.8 2.736 2.944 2.736 2.944 mp_allgather_i34 2055 14.4 1.168 2.934 1.168 2.934 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.889 2.902 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.702 2.873 calculate_dm_sparse 110 9.5 0.001 0.001 2.757 2.805 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 1.921 2.699 potential_pw2rs 110 12.3 0.015 0.016 2.665 2.684 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.666 2.667 calculate_first_density_matrix 1 7.0 0.000 0.000 2.463 2.464 grid_collocate_task_list 110 9.6 2.332 2.437 2.332 2.437 dbcsr_complete_redistribute 325 12.2 0.609 0.732 2.197 2.310 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.266 2.286 mp_alltoall_z22v 1111 16.6 2.126 2.160 2.126 2.160 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.739 2.043 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 2.003 cp_fm_cholesky_decompose 22 10.9 1.967 1.983 1.967 1.983 make_images_sizes 4110 15.4 0.005 0.005 1.023 1.738 mp_alltoall_i44 4110 16.4 1.018 1.734 1.018 1.734 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.653 1.661 mp_sum_l 6514 12.8 1.373 1.635 1.373 1.635 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.510 1.627 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.505 1.521 mp_waitany 9240 13.8 1.305 1.464 1.305 1.464 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.424 1.445 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=67.077000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=785.181818, yerr=11.590107 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 1.307967E+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 0.017 0.038 99.399 99.400 qs_mol_dyn_low 1 2.0 0.003 0.003 99.039 99.047 qs_forces 11 3.9 0.002 0.002 97.733 97.734 qs_energies 11 4.9 0.005 0.005 92.922 92.940 scf_env_do_scf 11 5.9 0.000 0.001 80.997 80.998 velocity_verlet 10 3.0 0.001 0.001 62.721 62.881 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.098 49.101 init_scf_loop 11 6.9 0.000 0.000 31.804 31.808 qs_scf_new_mos 99 7.5 0.001 0.001 31.309 31.366 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.309 31.365 dbcsr_multiply_generic 2055 12.4 0.120 0.123 30.681 30.785 prepare_preconditioner 11 7.9 0.000 0.000 29.730 29.738 make_preconditioner 11 8.9 0.000 0.000 29.730 29.738 ot_scf_mini 99 9.5 0.002 0.002 29.437 29.470 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.783 29.193 multiply_cannon 2055 13.4 0.348 0.371 22.627 23.455 multiply_cannon_loop 2055 14.4 0.343 0.349 20.727 21.264 cp_fm_upper_to_full 70 14.2 12.993 18.656 12.993 18.656 rebuild_ks_matrix 110 8.3 0.001 0.001 14.233 14.257 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.232 14.257 ot_mini 99 10.5 0.001 0.001 14.109 14.143 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.647 12.672 dbcsr_complete_redistribute 325 12.2 1.022 1.051 8.057 11.379 mp_waitall_1 87304 16.6 9.613 10.577 9.613 10.577 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.746 10.073 multiply_cannon_multrec 8220 15.4 4.378 4.576 9.840 9.988 qs_ot_get_p 110 10.4 0.001 0.001 9.721 9.769 qs_ot_get_derivative 99 11.5 0.001 0.001 9.417 9.442 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.931 9.231 mp_alltoall_i22 605 13.7 5.658 8.988 5.658 8.988 cp_fm_cholesky_invert 11 10.9 7.787 7.797 7.787 7.797 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.382 7.424 calculate_rho_elec 110 8.6 0.227 0.227 7.382 7.424 sum_up_and_integrate 110 10.3 0.150 0.151 7.350 7.371 init_scf_run 11 5.9 0.000 0.001 7.314 7.314 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.313 7.314 integrate_v_rspace 110 11.3 0.003 0.004 7.199 7.221 qs_ot_p2m_diag 48 11.0 0.150 0.155 7.142 7.149 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.646 6.649 make_m2s 4110 13.4 0.043 0.043 6.030 6.593 make_images 4110 14.4 0.882 0.923 5.842 6.406 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.505 6.161 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.837 5.843 cp_fm_diag_elpa_base 48 14.0 5.277 5.481 5.834 5.839 dbcsr_mm_accdrv_process 11614 15.7 3.266 3.614 5.319 5.606 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.749 5.241 apply_single 110 13.6 0.000 0.000 4.749 5.240 ot_diis_step 99 11.5 0.015 0.016 4.655 4.656 wfi_extrapolate 11 7.9 0.001 0.001 4.559 4.560 pw_transfer 1331 11.6 0.074 0.074 4.448 4.466 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.332 4.351 make_images_data 4110 15.4 0.041 0.044 3.385 4.185 hybrid_alltoall_any 4261 16.3 0.257 0.554 3.332 4.155 multiply_cannon_sync_h2d 8220 15.4 3.956 3.962 3.956 3.962 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.772 3.777 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.762 3.764 grid_integrate_task_list 110 12.3 3.684 3.743 3.684 3.743 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.660 3.676 density_rs2pw 110 9.6 0.004 0.004 3.644 3.667 fft3d_ps 1111 14.6 1.272 1.285 3.541 3.561 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.997 3.450 calculate_dm_sparse 110 9.5 0.001 0.001 3.393 3.424 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.911 2.926 mp_alltoall_d11v 2046 13.8 2.862 2.911 2.862 2.911 rs_pw_transfer 902 11.9 0.010 0.011 2.784 2.831 grid_collocate_task_list 110 9.6 2.642 2.668 2.642 2.668 calculate_first_density_matrix 1 7.0 0.000 0.000 2.630 2.630 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.559 2.580 potential_pw2rs 110 12.3 0.022 0.022 2.555 2.569 cp_fm_cholesky_decompose 22 10.9 2.545 2.568 2.545 2.568 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.307 2.378 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.246 2.282 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.195 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.081 2.152 jit_kernel_multiply 10 15.3 1.850 2.114 1.850 2.114 mp_alltoall_z22v 1111 16.6 2.019 2.035 2.019 2.035 qs_create_task_list 11 7.9 0.001 0.001 1.964 2.008 generate_qs_task_list 11 8.9 0.737 0.789 1.963 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=99.400000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1182.000000, yerr=50.831272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 626.778112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 56729. MP_Allreduce 11258 785. MP_Sync 168 MP_Alltoall 2210 3227166. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.125 233.477 233.479 qs_mol_dyn_low 1 2.0 0.004 0.006 232.901 232.917 qs_forces 11 3.9 0.004 0.009 232.369 232.373 qs_energies 11 4.9 0.002 0.006 226.761 226.780 scf_env_do_scf 11 5.9 0.001 0.001 208.725 208.729 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 182.272 182.278 qs_scf_new_mos 116 7.6 0.001 0.001 139.054 139.357 qs_scf_loop_do_ot 116 8.6 0.001 0.001 139.053 139.357 velocity_verlet 10 3.0 0.001 0.003 138.955 138.961 dbcsr_multiply_generic 2485 12.5 0.174 0.177 133.436 134.717 ot_scf_mini 116 9.6 0.003 0.004 131.758 132.016 multiply_cannon 2485 13.5 0.237 0.245 104.916 107.146 multiply_cannon_loop 2485 14.5 2.078 2.151 101.629 103.840 ot_mini 116 10.6 0.001 0.002 70.448 70.766 qs_ot_get_derivative 116 11.6 0.002 0.012 45.200 45.568 multiply_cannon_multrec 59640 15.5 32.527 34.767 40.900 42.544 mp_waitall_1 288940 16.2 37.862 40.342 37.862 40.342 qs_ot_get_p 127 10.4 0.001 0.001 37.461 37.704 rebuild_ks_matrix 127 8.3 0.001 0.001 34.192 34.458 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.037 34.192 34.457 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.826 31.055 multiply_cannon_sync_h2d 59640 15.5 26.686 29.182 26.686 29.182 qs_ot_p2m_diag 82 11.4 0.076 0.090 26.428 26.509 init_scf_loop 11 6.9 0.000 0.000 26.375 26.390 apply_preconditioner_dbcsr 127 12.6 0.001 0.007 24.666 25.767 apply_single 127 13.6 0.001 0.001 24.666 25.767 ot_diis_step 116 11.6 0.009 0.014 24.827 24.832 cp_dbcsr_syevd 82 12.4 0.004 0.005 23.294 23.318 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 21.284 21.549 prepare_preconditioner 11 7.9 0.000 0.000 21.486 21.533 make_preconditioner 11 8.9 0.000 0.000 21.486 21.533 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.687 20.883 multiply_cannon_metrocomm3 59640 15.5 0.109 0.113 15.978 18.474 make_m2s 4970 13.5 0.102 0.109 17.606 18.260 make_images 4970 14.5 0.401 0.421 17.427 18.076 cp_fm_diag_elpa 82 13.4 0.000 0.001 16.627 16.781 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.540 16.715 calculate_rho_elec 127 8.7 0.044 0.063 16.540 16.714 cp_fm_redistribute_end 82 14.4 13.054 16.517 13.106 16.560 cp_fm_diag_elpa_base 82 14.4 3.270 15.819 3.307 15.954 sum_up_and_integrate 127 10.3 0.089 0.108 14.956 14.971 integrate_v_rspace 127 11.3 0.003 0.005 14.867 14.884 init_scf_run 11 5.9 0.000 0.001 13.785 13.786 scf_env_initial_rho_setup 11 6.9 0.001 0.003 13.785 13.785 multiply_cannon_metrocomm1 59640 15.5 0.084 0.087 10.122 12.177 mp_sum_l 7804 13.0 9.758 11.264 9.758 11.264 density_rs2pw 127 9.7 0.006 0.007 9.905 10.391 wfi_extrapolate 11 7.9 0.002 0.005 10.255 10.256 make_images_data 4970 15.5 0.068 0.076 8.359 9.889 cp_fm_cholesky_invert 11 10.9 9.811 9.820 9.811 9.820 rs_pw_transfer 1038 11.9 0.017 0.018 9.180 9.803 calculate_dm_sparse 127 9.5 0.001 0.001 9.443 9.521 hybrid_alltoall_any 5155 16.4 0.287 2.233 7.499 9.022 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.887 8.988 cp_fm_cholesky_decompose 22 10.9 8.794 8.823 8.794 8.823 dbcsr_mm_accdrv_process 123452 16.2 3.249 3.427 7.946 8.673 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.303 8.418 pw_transfer 1535 11.6 0.074 0.106 7.720 8.206 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 7.517 7.985 mp_alltoall_d11v 2401 14.1 6.334 7.772 6.334 7.772 grid_integrate_task_list 127 12.3 6.953 7.429 6.953 7.429 fft3d_ps 1281 14.7 2.067 2.541 6.326 6.725 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.659 6.668 fft_wrap_pw1pw2_140 519 13.2 0.434 0.503 5.769 6.139 check_diag 82 13.4 2.216 2.279 5.875 6.025 make_images_sizes 4970 15.5 0.006 0.007 3.743 5.890 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.802 5.884 mp_alltoall_i44 4970 16.5 3.737 5.883 3.737 5.883 mp_sum_d 4447 12.1 4.377 5.558 4.377 5.558 potential_pw2rs 127 12.3 0.009 0.010 5.428 5.544 grid_collocate_task_list 127 9.7 4.689 5.070 4.689 5.070 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=233.479000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=591.818182, yerr=6.249628 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 828.301312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 6256179. 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 0.028 0.055 218.018 218.022 qs_mol_dyn_low 1 2.0 0.004 0.006 217.541 217.557 qs_forces 11 3.9 0.006 0.025 217.395 217.397 qs_energies 11 4.9 0.002 0.005 209.401 209.431 scf_env_do_scf 11 5.9 0.001 0.001 189.799 189.809 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 153.054 153.057 velocity_verlet 10 3.0 0.001 0.002 136.955 136.959 dbcsr_multiply_generic 2527 12.6 0.186 0.197 110.080 113.035 qs_scf_new_mos 118 7.6 0.001 0.001 109.522 110.186 qs_scf_loop_do_ot 118 8.6 0.001 0.001 109.521 110.185 ot_scf_mini 118 9.6 0.004 0.005 104.104 104.809 multiply_cannon 2527 13.6 0.479 0.537 85.134 92.095 multiply_cannon_loop 2527 14.6 1.263 1.303 81.216 87.140 ot_mini 118 10.6 0.001 0.002 57.369 58.010 mp_waitall_1 228564 16.4 34.693 48.351 34.693 48.351 rebuild_ks_matrix 129 8.3 0.001 0.001 37.582 38.539 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.024 37.582 38.538 multiply_cannon_multrec 30324 15.6 21.782 26.611 31.521 36.829 init_scf_loop 11 6.9 0.001 0.005 36.653 36.657 multiply_cannon_metrocomm3 30324 15.6 0.095 0.101 21.241 36.521 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.133 34.046 prepare_preconditioner 11 7.9 0.000 0.000 32.218 32.308 make_preconditioner 11 8.9 0.000 0.001 32.218 32.308 qs_ot_get_derivative 118 11.6 0.002 0.004 31.488 32.079 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.053 30.730 apply_preconditioner_dbcsr 129 12.6 0.001 0.011 24.474 27.017 apply_single 129 13.6 0.001 0.001 24.473 27.017 qs_ot_get_p 129 10.4 0.001 0.001 25.793 26.476 ot_diis_step 118 11.6 0.016 0.019 25.690 25.696 multiply_cannon_sync_h2d 30324 15.6 19.157 21.866 19.157 21.866 cp_fm_cholesky_invert 11 10.9 19.138 19.151 19.138 19.151 qs_ot_p2m_diag 83 11.4 0.187 0.215 18.662 18.716 make_m2s 5054 13.6 0.090 0.096 16.606 18.285 make_images 5054 14.6 1.179 1.367 16.397 18.074 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.391 17.393 sum_up_and_integrate 129 10.3 0.117 0.133 17.019 17.065 integrate_v_rspace 129 11.3 0.003 0.004 16.901 16.953 qs_rho_update_rho_low 129 7.7 0.001 0.002 14.151 14.187 calculate_rho_elec 129 8.7 0.088 0.106 14.150 14.186 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.111 14.120 cp_fm_redistribute_end 83 14.4 8.168 13.918 8.185 13.925 cp_fm_diag_elpa_base 83 14.4 5.470 13.386 5.710 13.781 init_scf_run 11 5.9 0.000 0.001 13.553 13.555 scf_env_initial_rho_setup 11 6.9 0.001 0.003 13.553 13.554 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.825 13.453 make_images_data 5054 15.6 0.067 0.074 10.222 12.155 multiply_cannon_metrocomm4 27797 15.6 0.098 0.117 3.818 10.735 hybrid_alltoall_any 5240 16.5 0.345 1.537 8.373 10.716 mp_irecv_dv 70031 16.3 3.624 10.341 3.624 10.341 wfi_extrapolate 11 7.9 0.019 0.050 10.052 10.053 dbcsr_mm_accdrv_process 62734 16.2 4.545 5.515 9.206 9.993 mp_sum_l 7930 13.1 6.571 9.434 6.571 9.434 pw_transfer 1559 11.6 0.085 0.101 8.808 8.914 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.582 8.689 mp_alltoall_d11v 2423 14.1 7.239 8.367 7.239 8.367 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 7.346 8.263 density_rs2pw 129 9.7 0.006 0.006 7.621 8.167 grid_integrate_task_list 129 12.3 7.189 7.710 7.189 7.710 rs_pw_transfer 1054 12.0 0.014 0.017 6.984 7.561 fft_wrap_pw1pw2_140 527 13.2 0.470 0.518 7.432 7.552 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.504 7.518 calculate_dm_sparse 129 9.5 0.001 0.001 7.284 7.486 cp_fm_cholesky_decompose 22 10.9 7.136 7.213 7.136 7.213 fft3d_ps 1301 14.7 2.789 2.937 6.949 7.021 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 6.070 6.664 mp_allgather_i34 2527 14.6 2.482 6.592 2.482 6.592 qs_ot_get_orbitals 118 10.6 0.001 0.001 6.141 6.284 mp_sum_d 4493 12.2 4.343 6.013 4.343 6.013 potential_pw2rs 129 12.3 0.015 0.016 5.611 5.651 dbcsr_dot_sd 1330 12.0 0.780 0.856 3.996 5.604 grid_collocate_task_list 129 9.7 4.949 5.391 4.949 5.391 dbcsr_complete_redistribute 395 12.7 0.763 0.858 4.212 5.311 rs_gather_matrices 129 12.3 0.361 0.394 3.984 5.022 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 4.877 4.918 multiply_cannon_metrocomm1 30324 15.6 0.048 0.050 1.962 4.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=218.022000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.363636, yerr=4.637255 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 929.808384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4129 56602. MP_Allreduce 11361 981. MP_Sync 172 MP_Alltoall 1724 9394234. MP_ISendRecv 15996 75008. MP_Wait 29984 MP_comm_split 84 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.033 204.171 204.173 qs_mol_dyn_low 1 2.0 0.004 0.007 203.702 203.716 qs_forces 11 3.9 0.004 0.006 203.374 203.378 qs_energies 11 4.9 0.001 0.002 196.835 196.849 scf_env_do_scf 11 5.9 0.001 0.001 179.817 179.820 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 137.052 137.053 velocity_verlet 10 3.0 0.001 0.002 131.433 131.443 qs_scf_new_mos 118 7.6 0.001 0.001 97.898 98.701 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.897 98.700 dbcsr_multiply_generic 2529 12.6 0.183 0.187 93.161 94.659 ot_scf_mini 118 9.6 0.004 0.004 93.669 94.520 multiply_cannon 2529 13.6 0.502 0.526 67.375 74.446 multiply_cannon_loop 2529 14.6 0.867 0.901 63.302 68.269 ot_mini 118 10.6 0.001 0.004 53.595 54.433 init_scf_loop 11 6.9 0.000 0.000 42.663 42.665 mp_waitall_1 180004 16.5 31.957 42.127 31.957 42.127 prepare_preconditioner 11 7.9 0.000 0.000 38.351 38.398 make_preconditioner 11 8.9 0.000 0.000 38.351 38.398 make_full_inverse_cholesky 11 9.9 0.000 0.000 34.974 36.571 rebuild_ks_matrix 129 8.3 0.001 0.001 32.474 33.509 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 32.474 33.508 qs_ot_get_derivative 118 11.6 0.002 0.012 32.487 33.319 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.538 30.544 multiply_cannon_metrocomm3 20232 15.6 0.058 0.062 18.325 30.355 multiply_cannon_multrec 20232 15.6 13.433 16.983 22.256 25.785 qs_ot_get_p 129 10.4 0.001 0.001 24.894 25.364 apply_preconditioner_dbcsr 129 12.6 0.001 0.007 21.323 22.886 apply_single 129 13.6 0.001 0.001 21.323 22.885 ot_diis_step 118 11.6 0.019 0.023 20.935 20.936 qs_ot_p2m_diag 84 11.4 0.268 0.275 20.289 20.300 cp_dbcsr_syevd 84 12.4 0.005 0.005 19.211 19.212 make_m2s 5058 13.6 0.081 0.086 17.688 18.659 make_images 5058 14.6 1.217 1.302 17.454 18.417 cp_fm_cholesky_invert 11 10.9 17.893 17.913 17.893 17.913 multiply_cannon_sync_h2d 20232 15.6 14.329 15.888 14.329 15.888 sum_up_and_integrate 129 10.3 0.135 0.146 15.854 15.883 cp_fm_diag_elpa 84 13.4 0.000 0.001 15.817 15.820 integrate_v_rspace 129 11.3 0.003 0.004 15.719 15.744 cp_fm_redistribute_end 84 14.4 6.008 15.735 6.029 15.739 cp_fm_diag_elpa_base 84 14.4 9.238 15.069 9.676 15.578 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 13.707 14.412 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.304 13.343 calculate_rho_elec 129 8.7 0.133 0.148 13.303 13.342 make_images_data 5058 15.6 0.061 0.068 10.875 12.275 init_scf_run 11 5.9 0.000 0.001 11.983 11.983 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.983 11.983 hybrid_alltoall_any 5245 16.5 0.434 1.997 9.222 11.072 cp_fm_cholesky_decompose 22 10.9 9.740 9.795 9.740 9.795 multiply_cannon_metrocomm4 17703 15.6 0.064 0.073 3.510 9.081 pw_transfer 1559 11.6 0.086 0.107 8.878 8.989 mp_sum_l 7936 13.1 6.263 8.980 6.263 8.980 wfi_extrapolate 11 7.9 0.001 0.004 8.858 8.858 mp_irecv_dv 50659 16.2 3.388 8.835 3.388 8.835 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.652 8.770 dbcsr_complete_redistribute 397 12.7 1.167 1.217 6.325 8.621 dbcsr_mm_accdrv_process 41846 16.2 4.633 5.402 8.276 8.434 cp_fm_upper_to_full 106 14.5 6.374 8.389 6.374 8.389 grid_integrate_task_list 129 12.3 7.331 7.855 7.331 7.855 fft_wrap_pw1pw2_140 527 13.2 0.482 0.534 7.309 7.428 density_rs2pw 129 9.7 0.006 0.006 6.707 7.114 copy_fm_to_dbcsr 210 11.7 0.002 0.002 4.699 6.987 fft3d_ps 1301 14.7 2.715 2.921 6.907 6.970 mp_alltoall_d11v 2429 14.1 5.799 6.672 5.799 6.672 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.909 6.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.351 6.360 potential_pw2rs 129 12.3 0.021 0.023 5.970 6.017 mp_allgather_i34 2529 14.6 2.584 6.015 2.584 6.015 rs_pw_transfer 1054 12.0 0.014 0.014 5.661 5.988 calculate_dm_sparse 129 9.5 0.001 0.001 5.860 5.965 grid_collocate_task_list 129 9.7 5.100 5.737 5.100 5.737 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.358 5.643 mp_alltoall_i22 720 14.1 2.639 5.115 2.639 5.115 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.748 4.975 multiply_cannon_metrocomm1 20232 15.6 0.038 0.040 2.306 4.649 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.204 4.256 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=204.173000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=880.000000, yerr=7.147791 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/18/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 4.320339E+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 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 1.135378E+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 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57164. MP_Allreduce 11256 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 0.031 0.081 222.992 222.993 qs_mol_dyn_low 1 2.0 0.003 0.005 222.419 222.432 qs_forces 11 3.9 0.006 0.026 220.194 220.198 qs_energies 11 4.9 0.002 0.004 213.176 213.190 scf_env_do_scf 11 5.9 0.001 0.001 191.172 191.186 velocity_verlet 10 3.0 0.001 0.002 146.811 146.902 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.170 133.172 qs_scf_new_mos 116 7.6 0.001 0.001 94.582 94.890 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.582 94.889 ot_scf_mini 116 9.6 0.003 0.004 90.016 90.321 dbcsr_multiply_generic 2485 12.5 0.188 0.196 82.326 83.091 multiply_cannon 2485 13.5 0.549 0.582 55.756 58.823 init_scf_loop 11 6.9 0.007 0.032 57.879 57.883 multiply_cannon_loop 2485 14.5 1.171 1.202 51.703 53.608 prepare_preconditioner 11 7.9 0.000 0.000 51.897 51.930 make_preconditioner 11 8.9 0.000 0.001 51.897 51.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 45.299 50.415 ot_mini 116 10.6 0.001 0.002 43.957 44.264 rebuild_ks_matrix 127 8.3 0.001 0.001 32.550 32.866 qs_ks_build_kohn_sham_matrix 127 9.3 0.033 0.085 32.549 32.865 qs_ot_get_p 127 10.4 0.001 0.001 31.853 32.182 multiply_cannon_multrec 29820 15.5 13.980 18.879 26.067 30.595 mp_waitall_1 152434 16.5 20.982 30.233 20.982 30.233 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.789 30.074 qs_ot_p2m_diag 82 11.4 0.338 0.384 26.307 26.358 cp_dbcsr_syevd 82 12.4 0.005 0.005 24.999 25.001 qs_ot_get_derivative 116 11.6 0.001 0.003 24.319 24.616 make_m2s 4970 13.5 0.094 0.098 21.768 22.924 make_images 4970 14.5 1.963 2.256 21.467 22.624 cp_fm_diag_elpa 82 13.4 0.000 0.001 21.702 21.707 cp_fm_redistribute_end 82 14.4 3.687 21.502 3.722 21.509 cp_fm_cholesky_invert 11 10.9 21.411 21.423 21.411 21.423 cp_fm_diag_elpa_base 82 14.4 17.005 20.767 17.691 21.268 ot_diis_step 116 11.6 0.019 0.024 19.497 19.505 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.984 19.501 apply_single 127 13.6 0.001 0.001 18.984 19.501 qs_rho_update_rho_low 127 7.7 0.038 0.190 17.185 17.267 calculate_rho_elec 127 8.7 0.174 0.189 17.147 17.266 cp_fm_upper_to_full 104 14.7 11.451 16.988 11.451 16.988 sum_up_and_integrate 127 10.3 0.140 0.154 16.817 16.840 integrate_v_rspace 127 11.3 0.003 0.004 16.676 16.701 init_scf_run 11 5.9 0.000 0.001 15.831 15.835 scf_env_initial_rho_setup 11 6.9 0.043 0.229 15.831 15.835 multiply_cannon_metrocomm3 29820 15.5 0.045 0.047 6.483 15.538 dbcsr_complete_redistribute 393 12.7 1.510 1.637 10.433 14.296 make_images_data 4970 15.5 0.062 0.067 11.873 13.938 hybrid_alltoall_any 5155 16.4 0.521 2.183 10.610 12.677 copy_fm_to_dbcsr 208 11.6 0.001 0.002 8.538 12.411 multiply_cannon_sync_h2d 29820 15.5 11.594 12.350 11.594 12.350 dbcsr_mm_accdrv_process 61748 16.2 7.418 8.203 11.673 12.190 pw_transfer 1535 11.6 0.084 0.099 11.579 11.649 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 11.356 11.430 cp_fm_cholesky_decompose 22 10.9 11.305 11.406 11.305 11.406 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.628 10.848 wfi_extrapolate 11 7.9 0.001 0.004 10.593 10.595 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.575 10.318 fft3d_ps 1281 14.7 2.769 2.824 9.603 9.668 density_rs2pw 127 9.7 0.005 0.006 9.031 9.655 fft_wrap_pw1pw2_140 519 13.2 0.475 0.484 9.558 9.652 mp_alltoall_i22 712 14.1 5.782 9.550 5.782 9.550 mp_alltoall_d11v 2401 14.1 8.152 9.411 8.152 9.411 grid_integrate_task_list 127 12.3 7.420 7.793 7.420 7.793 rs_pw_transfer 1038 11.9 0.013 0.014 6.533 7.162 calculate_dm_sparse 127 9.5 0.001 0.001 6.873 6.982 potential_pw2rs 127 12.3 0.022 0.023 6.419 6.480 multiply_cannon_metrocomm4 24850 15.5 0.075 0.085 2.754 6.416 mp_alltoall_z22v 1281 16.7 6.256 6.319 6.256 6.319 mp_irecv_dv 75445 16.2 2.617 6.179 2.617 6.179 grid_collocate_task_list 127 9.7 5.181 5.667 5.181 5.667 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.610 5.639 qs_energies_init_hamiltonians 11 5.9 0.004 0.022 4.612 4.613 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.446 4.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.449 4.536 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=222.993000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.363636, yerr=18.762984 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.496912E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 58365. MP_Allreduce 10975 1176. MP_Sync 86 MP_Alltoall 1700 18828162. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.040 206.000 206.000 qs_mol_dyn_low 1 2.0 0.003 0.003 205.443 205.460 qs_forces 11 3.9 0.003 0.003 205.127 205.135 qs_energies 11 4.9 0.001 0.002 194.993 195.006 scf_env_do_scf 11 5.9 0.001 0.001 172.144 172.156 velocity_verlet 10 3.0 0.001 0.001 132.997 133.015 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 129.338 129.339 qs_scf_new_mos 116 7.6 0.001 0.001 91.333 91.446 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.332 91.445 ot_scf_mini 116 9.6 0.003 0.004 86.793 86.847 dbcsr_multiply_generic 2485 12.5 0.180 0.187 80.530 81.529 multiply_cannon 2485 13.5 0.582 0.609 57.065 62.099 multiply_cannon_loop 2485 14.5 0.446 0.454 51.430 53.078 ot_mini 116 10.6 0.001 0.001 43.176 43.234 init_scf_loop 11 6.9 0.000 0.000 42.620 42.623 prepare_preconditioner 11 7.9 0.000 0.000 38.566 38.596 make_preconditioner 11 8.9 0.000 0.000 38.566 38.596 mp_waitall_1 128490 16.6 31.050 38.224 31.050 38.224 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.265 36.550 rebuild_ks_matrix 127 8.3 0.001 0.001 32.064 32.195 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 32.063 32.194 qs_ot_get_p 127 10.4 0.001 0.001 29.488 29.581 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.730 27.823 qs_ot_p2m_diag 82 11.4 0.489 0.494 22.758 22.777 qs_ot_get_derivative 116 11.6 0.002 0.002 22.666 22.713 cp_fm_cholesky_invert 11 10.9 22.612 22.619 22.612 22.619 cp_dbcsr_syevd 82 12.4 0.005 0.005 21.601 21.613 make_m2s 4970 13.5 0.065 0.070 18.967 21.460 make_images 4970 14.5 2.304 2.865 18.665 21.162 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.620 21.087 ot_diis_step 116 11.6 0.020 0.020 20.424 20.425 multiply_cannon_multrec 9940 15.5 10.269 13.901 17.548 20.407 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.570 20.195 apply_single 127 13.6 0.001 0.001 19.570 20.195 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.210 18.214 cp_fm_diag_elpa_base 82 14.4 17.841 18.021 18.195 18.199 sum_up_and_integrate 127 10.3 0.178 0.187 16.768 16.816 integrate_v_rspace 127 11.3 0.004 0.004 16.590 16.645 init_scf_run 11 5.9 0.000 0.001 15.809 15.810 scf_env_initial_rho_setup 11 6.9 0.001 0.001 15.809 15.810 make_images_data 4970 15.5 0.051 0.060 11.777 14.710 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.232 14.260 calculate_rho_elec 127 8.7 0.256 0.267 14.232 14.259 hybrid_alltoall_any 5155 16.4 0.806 3.531 11.378 13.991 wfi_extrapolate 11 7.9 0.001 0.001 12.840 12.841 multiply_cannon_sync_h2d 9940 15.5 11.449 11.866 11.449 11.866 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.394 10.442 cp_fm_cholesky_decompose 22 10.9 10.164 10.263 10.164 10.263 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 6.178 9.785 pw_transfer 1535 11.6 0.083 0.091 9.685 9.719 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.466 9.508 mp_allgather_i34 2485 14.5 3.845 9.345 3.845 9.345 mp_alltoall_d11v 2401 14.1 8.820 9.325 8.820 9.325 dbcsr_complete_redistribute 393 12.7 2.120 2.171 7.896 8.175 fft_wrap_pw1pw2_140 519 13.2 0.496 0.516 8.098 8.137 grid_integrate_task_list 127 12.3 7.694 8.041 7.694 8.041 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.963 7.978 fft3d_ps 1281 14.7 2.701 2.785 7.643 7.695 dbcsr_mm_accdrv_process 20590 16.1 2.668 3.428 6.917 7.546 density_rs2pw 127 9.7 0.005 0.006 6.787 6.988 calculate_dm_sparse 127 9.5 0.001 0.001 6.432 6.530 potential_pw2rs 127 12.3 0.026 0.027 5.900 5.924 grid_collocate_task_list 127 9.7 5.450 5.759 5.450 5.759 rs_pw_transfer 1038 11.9 0.012 0.013 5.342 5.564 copy_fm_to_dbcsr 208 11.6 0.001 0.002 5.220 5.557 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.367 5.368 copy_dbcsr_to_fm 185 11.7 0.004 0.004 5.216 5.337 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 5.239 5.280 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.702 4.842 mp_alltoall_z22v 1281 16.7 4.328 4.477 4.328 4.477 make_basis_sm 11 9.8 0.001 0.001 4.275 4.308 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=206.000000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1392.454545, yerr=38.146520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_performance_tests/20/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 11.696233E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.023245E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59285. MP_Allreduce 10977 1514. MP_Sync 87 MP_Alltoall 1712 36974159. MP_ISendRecv 3584 218624. MP_Wait 11594 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.051 318.388 318.389 qs_mol_dyn_low 1 2.0 0.003 0.003 317.687 317.699 qs_forces 11 3.9 0.021 0.021 317.589 317.593 qs_energies 11 4.9 0.002 0.002 308.635 308.644 scf_env_do_scf 11 5.9 0.001 0.001 285.135 285.152 velocity_verlet 10 3.0 0.001 0.001 227.405 227.413 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 153.986 153.988 init_scf_loop 11 6.9 0.000 0.000 130.871 130.875 prepare_preconditioner 11 7.9 0.000 0.000 126.000 126.041 make_preconditioner 11 8.9 0.000 0.000 126.000 126.041 make_full_inverse_cholesky 11 9.9 0.000 0.000 101.776 123.147 qs_scf_new_mos 117 7.6 0.001 0.001 107.048 107.168 qs_scf_loop_do_ot 117 8.6 0.001 0.001 107.047 107.167 ot_scf_mini 117 9.6 0.004 0.004 101.886 102.013 dbcsr_multiply_generic 2507 12.6 0.214 0.221 90.772 91.406 cp_fm_upper_to_full 105 14.8 53.884 77.018 53.884 77.018 multiply_cannon 2507 13.6 0.709 0.781 62.802 63.864 multiply_cannon_loop 2507 14.6 0.472 0.478 58.493 60.068 ot_mini 117 10.6 0.001 0.001 50.709 50.874 dbcsr_complete_redistribute 395 12.7 4.003 4.049 30.667 43.607 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.693 39.605 mp_waitall_1 105466 16.7 33.933 38.517 33.933 38.517 rebuild_ks_matrix 128 8.3 0.001 0.001 37.104 37.261 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.018 37.103 37.261 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.179 37.007 cp_fm_cholesky_invert 11 10.9 36.747 36.764 36.747 36.764 mp_alltoall_i22 716 14.1 22.057 35.232 22.057 35.232 qs_ot_get_p 128 10.4 0.001 0.001 34.752 34.876 qs_ks_update_qs_env 128 7.6 0.001 0.001 34.540 34.705 qs_ot_p2m_diag 83 11.4 0.877 0.883 29.796 29.835 qs_ot_get_derivative 117 11.6 0.002 0.002 28.995 29.129 cp_dbcsr_syevd 83 12.4 0.006 0.006 27.910 27.917 cp_fm_diag_elpa 83 13.4 0.000 0.000 24.414 24.424 cp_fm_diag_elpa_base 83 14.4 19.969 21.638 24.408 24.418 make_m2s 5014 13.6 0.076 0.078 22.176 23.444 multiply_cannon_metrocomm3 10028 15.6 0.023 0.023 21.780 23.295 make_images 5014 14.6 3.766 3.962 21.697 22.968 ot_diis_step 117 11.6 0.022 0.023 21.652 21.654 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.613 20.850 apply_single 128 13.6 0.001 0.001 20.612 20.850 sum_up_and_integrate 128 10.3 0.321 0.322 18.247 18.351 multiply_cannon_multrec 10028 15.6 10.527 12.287 18.227 18.326 integrate_v_rspace 128 11.3 0.004 0.004 17.926 18.029 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.242 17.257 calculate_rho_elec 128 8.7 0.483 0.483 17.241 17.256 multiply_cannon_sync_h2d 10028 15.6 15.671 15.680 15.671 15.680 make_images_data 5014 15.6 0.059 0.064 12.145 14.672 hybrid_alltoall_any 5200 16.5 1.304 3.047 12.018 14.413 init_scf_run 11 5.9 0.000 0.001 13.129 13.129 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.129 13.129 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.775 11.889 pw_transfer 1547 11.6 0.090 0.091 10.968 10.993 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 10.735 10.761 cp_fm_cholesky_decompose 22 10.9 10.459 10.524 10.459 10.524 wfi_extrapolate 11 7.9 0.001 0.001 9.740 9.741 fft_wrap_pw1pw2_140 523 13.2 0.540 0.542 9.259 9.288 dbcsr_mm_accdrv_process 20762 16.1 3.754 5.459 7.459 9.275 mp_alltoall_d11v 2415 14.1 9.000 9.157 9.000 9.157 fft3d_ps 1291 14.7 2.761 2.767 8.771 8.797 grid_integrate_task_list 128 12.3 8.553 8.727 8.553 8.727 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 8.232 8.234 density_rs2pw 128 9.7 0.005 0.005 7.790 7.827 calculate_dm_sparse 128 9.5 0.001 0.001 7.230 7.331 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.909 7.044 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.671 6.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=318.389000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2660.727273, yerr=179.932862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 1.260253E+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 18005791. 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 0.021 0.058 97.316 97.317 qs_energies 1 2.0 0.000 0.000 96.822 96.834 ls_scf 1 3.0 0.001 0.011 95.926 95.940 dbcsr_multiply_generic 111 6.7 0.014 0.016 84.613 84.885 multiply_cannon 111 7.7 0.018 0.021 61.611 63.710 multiply_cannon_loop 111 8.7 0.209 0.223 57.948 60.217 ls_scf_main 1 4.0 0.001 0.012 59.358 59.359 density_matrix_trs4 2 5.0 0.002 0.018 53.917 54.006 mp_waitall_1 11316 10.9 32.733 36.150 32.733 36.150 ls_scf_init_scf 1 4.0 0.000 0.001 31.640 31.641 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.553 30.613 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.488 27.517 make_m2s 222 7.7 0.008 0.010 17.946 18.553 make_images 222 8.7 0.099 0.108 17.924 18.532 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 14.922 17.621 multiply_cannon_multrec 2664 9.7 8.115 8.841 15.448 17.188 multiply_cannon_sync_h2d 2664 9.7 13.245 15.166 13.245 15.166 make_images_data 222 9.7 0.004 0.005 11.946 12.999 hybrid_alltoall_any 227 10.6 0.215 1.843 9.588 11.749 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 6.185 9.241 dbcsr_mm_accdrv_process 4760 10.4 0.503 0.615 6.952 7.908 dbcsr_mm_accdrv_process_sort 4760 11.4 6.243 7.140 6.243 7.140 mp_sum_l 807 5.4 4.558 6.256 4.558 6.256 make_images_sizes 222 9.7 0.000 0.000 1.433 6.179 mp_alltoall_i44 222 10.7 1.432 6.178 1.432 6.178 calculate_norms 4752 9.8 5.471 6.140 5.471 6.140 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.755 5.917 ls_scf_post 1 4.0 0.000 0.001 4.926 4.941 ls_scf_store_result 1 5.0 0.000 0.000 4.750 4.783 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.107 4.420 arnoldi_extremal 4 6.8 0.000 0.000 3.396 3.422 arnoldi_normal_ev 4 7.8 0.001 0.003 3.396 3.422 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.045 3.355 mp_irecv_dv 6231 10.9 2.028 3.326 2.028 3.326 build_subspace 16 8.4 0.009 0.012 3.247 3.262 make_images_pack 222 9.7 2.194 2.622 2.196 2.624 dbcsr_special_finalize 555 9.7 0.005 0.005 2.228 2.614 dbcsr_merge_single_wm 555 10.7 0.444 0.570 2.221 2.606 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.335 2.591 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.469 2.065 2.471 dbcsr_sort_data 658 11.4 2.036 2.381 2.036 2.381 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.220 2.321 buffer_matrices_ensure_size 222 8.7 1.752 2.057 1.752 2.057 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=97.317000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 2.103325E+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 0.031 0.053 102.453 102.460 qs_energies 1 2.0 0.000 0.000 101.931 101.937 ls_scf 1 3.0 0.000 0.002 100.587 100.594 dbcsr_multiply_generic 111 6.7 0.015 0.016 86.029 86.543 multiply_cannon 111 7.7 0.028 0.075 58.329 62.705 ls_scf_main 1 4.0 0.000 0.003 61.406 61.416 multiply_cannon_loop 111 8.7 0.115 0.123 54.249 57.408 density_matrix_trs4 2 5.0 0.002 0.004 55.710 55.927 mp_waitall_1 9246 10.9 29.648 39.084 29.648 39.084 ls_scf_init_scf 1 4.0 0.000 0.001 33.614 33.616 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.066 32.189 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.952 28.972 multiply_cannon_multrec 1332 9.7 12.997 16.604 22.264 27.117 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 14.234 25.083 make_m2s 222 7.7 0.006 0.007 19.740 20.763 make_images 222 8.7 1.586 1.974 19.710 20.735 make_images_data 222 9.7 0.004 0.004 12.720 14.240 hybrid_alltoall_any 227 10.6 0.522 2.477 10.869 12.781 dbcsr_mm_accdrv_process 4041 10.4 0.281 0.558 8.876 10.361 dbcsr_mm_accdrv_process_sort 4041 11.4 8.445 9.935 8.445 9.935 mp_sum_l 807 5.4 6.681 9.619 6.681 9.619 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 5.295 7.830 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.216 7.682 mp_irecv_dv 3311 11.0 3.198 7.619 3.198 7.619 calculate_norms 2376 9.8 5.952 6.628 5.952 6.628 apply_matrix_preconditioner 6 5.3 0.000 0.000 6.175 6.406 multiply_cannon_sync_h2d 1332 9.7 4.779 5.990 4.779 5.990 ls_scf_post 1 4.0 0.000 0.002 5.567 5.572 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 3.223 5.410 ls_scf_store_result 1 5.0 0.000 0.000 5.247 5.377 arnoldi_extremal 4 6.8 0.000 0.002 4.780 4.807 arnoldi_normal_ev 4 7.8 0.002 0.007 4.780 4.807 build_subspace 16 8.4 0.015 0.021 4.510 4.514 mp_allgather_i34 111 8.7 1.775 4.242 1.775 4.242 make_images_sizes 222 9.7 0.000 0.000 1.117 3.679 mp_alltoall_i44 222 10.7 1.117 3.678 1.117 3.678 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.229 3.454 dbcsr_matrix_vector_mult_local 304 10.0 2.776 3.267 2.778 3.268 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.526 2.611 make_images_pack 222 9.7 2.036 2.409 2.039 2.411 dbcsr_sort_data 436 11.2 1.766 2.069 1.766 2.069 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=102.460000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1711.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 2.706108E+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 0.080 0.203 108.710 108.715 qs_energies 1 2.0 0.000 0.000 107.857 107.872 ls_scf 1 3.0 0.000 0.001 106.464 106.481 dbcsr_multiply_generic 111 6.7 0.015 0.016 89.447 89.799 ls_scf_main 1 4.0 0.001 0.015 66.494 66.499 multiply_cannon 111 7.7 0.043 0.101 57.690 63.246 density_matrix_trs4 2 5.0 0.002 0.004 60.667 60.814 multiply_cannon_loop 111 8.7 0.100 0.110 53.834 57.952 mp_waitall_1 7374 11.0 34.163 44.616 34.163 44.616 ls_scf_init_scf 1 4.0 0.000 0.002 36.401 36.402 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.234 35.299 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.628 32.644 make_m2s 222 7.7 0.006 0.007 23.674 25.047 make_images 222 8.7 1.979 2.319 23.636 25.008 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 13.342 24.767 multiply_cannon_multrec 888 9.7 12.644 15.205 21.224 24.366 make_images_data 222 9.7 0.003 0.004 15.498 17.330 hybrid_alltoall_any 227 10.6 0.618 2.866 13.490 16.897 mp_sum_l 807 5.4 6.575 10.482 6.575 10.482 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 5.941 10.290 dbcsr_mm_accdrv_process 3754 10.4 0.257 0.440 8.106 9.384 dbcsr_mm_accdrv_process_sort 3754 11.4 7.731 8.944 7.731 8.944 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 5.129 8.408 multiply_cannon_sync_h2d 888 9.7 5.991 7.291 5.991 7.291 arnoldi_extremal 4 6.8 0.000 0.002 6.861 6.880 arnoldi_normal_ev 4 7.8 0.001 0.008 6.861 6.880 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.439 6.594 build_subspace 16 8.4 0.014 0.021 6.542 6.551 mp_irecv_dv 2335 11.1 2.426 6.546 2.426 6.546 make_images_sizes 222 9.7 0.000 0.000 1.875 5.370 mp_alltoall_i44 222 10.7 1.875 5.369 1.875 5.369 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.932 5.110 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 4.690 5.108 calculate_norms 1584 9.8 4.261 4.617 4.261 4.617 mp_allgather_i34 111 8.7 1.703 4.427 1.703 4.427 dbcsr_matrix_vector_mult_local 304 10.0 3.044 3.608 3.046 3.610 ls_scf_post 1 4.0 0.002 0.017 3.569 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.302 3.390 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.706 2.808 mp_sum_dv 2129 10.8 2.356 2.676 2.356 2.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=108.715000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2195.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 3.314237E+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 0.067 0.075 116.585 116.586 qs_energies 1 2.0 0.000 0.000 115.957 115.965 ls_scf 1 3.0 0.000 0.000 114.323 114.330 dbcsr_multiply_generic 111 6.7 0.016 0.017 94.852 95.456 ls_scf_main 1 4.0 0.000 0.000 70.572 70.572 multiply_cannon 111 7.7 0.062 0.149 61.321 65.187 density_matrix_trs4 2 5.0 0.002 0.003 62.372 62.967 multiply_cannon_loop 111 8.7 0.114 0.125 56.270 60.780 mp_waitall_1 6438 11.0 36.832 44.851 36.832 44.851 ls_scf_init_scf 1 4.0 0.000 0.000 39.777 39.779 ls_scf_init_matrix_S 1 5.0 0.000 0.000 38.610 38.680 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 36.101 36.111 make_m2s 222 7.7 0.006 0.008 26.157 27.591 make_images 222 8.7 3.142 3.598 26.107 27.543 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 13.437 25.308 multiply_cannon_multrec 1332 9.7 14.158 17.455 22.044 24.840 make_images_data 222 9.7 0.004 0.004 16.133 18.218 hybrid_alltoall_any 227 10.6 0.796 3.811 14.696 17.657 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 8.087 14.126 mp_sum_l 807 5.4 5.800 10.034 5.800 10.034 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.408 7.524 9.043 dbcsr_mm_accdrv_process_sort 3641 11.4 7.151 8.626 7.151 8.626 arnoldi_extremal 4 6.8 0.000 0.000 7.939 7.953 arnoldi_normal_ev 4 7.8 0.001 0.004 7.939 7.953 build_subspace 16 8.4 0.014 0.021 7.591 7.601 apply_matrix_preconditioner 6 5.3 0.000 0.000 6.920 7.332 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.969 7.027 multiply_cannon_sync_h2d 1332 9.7 5.518 6.238 5.518 6.238 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.074 5.897 mp_irecv_dv 3229 10.9 2.051 5.820 2.051 5.820 make_images_sizes 222 9.7 0.000 0.000 1.415 4.595 mp_alltoall_i44 222 10.7 1.414 4.595 1.414 4.595 mp_allgather_i34 111 8.7 2.165 4.510 2.165 4.510 calculate_norms 2376 9.8 4.166 4.493 4.166 4.493 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.398 4.443 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.920 4.313 ls_scf_post 1 4.0 0.000 0.000 3.975 3.983 ls_scf_store_result 1 5.0 0.000 0.000 3.724 3.780 dbcsr_matrix_vector_mult_local 304 10.0 3.211 3.722 3.213 3.724 mp_sum_dv 2129 10.8 3.120 3.548 3.120 3.548 dbcsr_sort_data 658 11.4 3.046 3.467 3.046 3.467 dbcsr_special_finalize 555 9.7 0.006 0.007 2.769 3.207 dbcsr_merge_single_wm 555 10.7 0.514 0.644 2.761 3.199 dbcsr_sym_m_v_mult 184 10.2 0.004 0.010 2.586 2.816 dbcsr_complete_redistribute 5 7.6 0.474 0.532 1.494 2.527 dbcsr_data_release 10477 10.7 1.589 2.394 1.589 2.394 matrix_ls_to_qs 2 6.0 0.000 0.000 1.310 2.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=116.586000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2721.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 4.622574E+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 0.043 0.055 103.360 103.361 qs_energies 1 2.0 0.000 0.000 102.569 102.576 ls_scf 1 3.0 0.000 0.000 100.650 100.655 dbcsr_multiply_generic 111 6.7 0.017 0.018 81.552 81.731 ls_scf_main 1 4.0 0.000 0.000 65.135 65.136 multiply_cannon 111 7.7 0.085 0.171 57.530 62.916 density_matrix_trs4 2 5.0 0.002 0.004 57.528 57.606 multiply_cannon_loop 111 8.7 0.069 0.079 52.987 54.858 mp_waitall_1 5481 11.0 29.316 34.778 29.316 34.778 ls_scf_init_scf 1 4.0 0.000 0.000 31.909 31.913 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.688 30.759 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.391 28.403 multiply_cannon_multrec 444 9.7 14.088 16.440 21.198 23.095 make_m2s 222 7.7 0.004 0.005 18.593 21.114 make_images 222 8.7 3.722 4.426 18.531 21.054 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 13.443 18.210 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.034 15.980 make_images_data 222 9.7 0.003 0.004 10.616 13.104 hybrid_alltoall_any 227 10.6 0.790 3.760 10.178 13.024 multiply_cannon_sync_h2d 444 9.7 6.465 8.131 6.465 8.131 dbcsr_mm_accdrv_process 3003 10.4 0.183 0.450 6.811 7.914 dbcsr_mm_accdrv_process_sort 3003 11.4 6.470 7.570 6.470 7.570 mp_allgather_i34 111 8.7 2.770 7.467 2.770 7.467 arnoldi_extremal 4 6.8 0.000 0.000 5.934 5.948 arnoldi_normal_ev 4 7.8 0.002 0.005 5.934 5.948 mp_sum_l 807 5.4 3.480 5.733 3.480 5.733 build_subspace 16 8.4 0.015 0.019 5.538 5.549 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.554 4.688 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.596 4.647 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.234 4.450 dbcsr_matrix_vector_mult_local 304 10.0 3.710 4.188 3.712 4.190 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.664 4.152 mp_irecv_dv 1241 11.2 1.645 4.130 1.645 4.130 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.024 4.108 make_images_sizes 222 9.7 0.000 0.000 1.336 3.979 mp_alltoall_i44 222 10.7 1.335 3.978 1.335 3.978 calculate_norms 792 9.8 3.540 3.695 3.540 3.695 ls_scf_post 1 4.0 0.000 0.000 3.606 3.612 ls_scf_store_result 1 5.0 0.000 0.000 3.397 3.428 dbcsr_finalize 304 7.8 0.062 0.078 2.197 2.323 dbcsr_merge_all 275 8.9 0.474 0.532 2.046 2.156 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=103.361000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3600.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f51da1b24f48180e503e660b45b9ce8da4b1a63d_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 8.718447E+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 0.074 0.095 113.090 113.090 qs_energies 1 2.0 0.000 0.000 111.671 111.685 ls_scf 1 3.0 0.000 0.000 108.780 108.794 dbcsr_multiply_generic 111 6.7 0.023 0.025 81.327 81.438 ls_scf_main 1 4.0 0.000 0.000 68.166 68.167 density_matrix_trs4 2 5.0 0.002 0.004 59.026 59.077 multiply_cannon 111 7.7 0.160 0.218 51.853 54.308 multiply_cannon_loop 111 8.7 0.068 0.071 48.225 48.730 ls_scf_init_scf 1 4.0 0.000 0.000 36.526 36.527 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.186 35.208 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.802 31.811 mp_waitall_1 4569 11.1 24.432 28.033 24.432 28.033 make_m2s 222 7.7 0.005 0.005 25.138 26.157 make_images 222 8.7 4.589 4.967 25.032 26.049 multiply_cannon_multrec 444 9.7 17.919 18.718 22.665 23.216 hybrid_alltoall_any 227 10.6 1.660 3.622 13.424 17.050 make_images_data 222 9.7 0.003 0.003 13.921 16.803 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.883 12.438 multiply_cannon_sync_h2d 444 9.7 8.824 8.868 8.824 8.868 arnoldi_extremal 4 6.8 0.000 0.000 8.480 8.505 arnoldi_normal_ev 4 7.8 0.006 0.012 8.480 8.505 build_subspace 16 8.4 0.025 0.036 7.725 7.753 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 6.057 6.099 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.573 5.830 dbcsr_matrix_vector_mult_local 304 10.0 5.087 5.377 5.090 5.379 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.040 5.129 dbcsr_mm_accdrv_process 1814 10.4 0.312 0.533 4.578 4.818 make_images_sizes 222 9.7 0.000 0.000 1.885 4.470 mp_alltoall_i44 222 10.7 1.885 4.470 1.885 4.470 dbcsr_mm_accdrv_process_sort 1814 11.4 4.161 4.290 4.161 4.290 ls_scf_post 1 4.0 0.000 0.000 4.087 4.102 ls_scf_store_result 1 5.0 0.000 0.000 3.819 3.835 mp_allgather_i34 111 8.7 1.253 3.756 1.253 3.756 dbcsr_complete_redistribute 5 7.6 1.428 1.466 3.258 3.356 calculate_norms 792 9.8 3.262 3.297 3.262 3.297 dbcsr_finalize 304 7.8 0.082 0.090 3.084 3.125 dbcsr_merge_all 275 8.9 0.892 0.919 2.868 2.900 dbcsr_data_release 12724 10.6 2.335 2.863 2.335 2.863 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.861 2.861 matrix_ls_to_qs 2 6.0 0.000 0.000 2.658 2.766 mp_alltoall_d11v 48 9.2 2.513 2.571 2.513 2.571 dbcsr_sort_data 325 11.1 2.438 2.488 2.438 2.488 dbcsr_new_transposed 4 7.5 0.244 0.255 2.334 2.341 mp_sum_l 807 5.4 2.011 2.299 2.011 2.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=113.090000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6841.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: f51da1b24f48180e503e660b45b9ce8da4b1a63d Summary: empty Status: OK