=== 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: e4f847e0ea5227630c4e07962f9d27e060298d5a ################# 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/01 job id: 41858711 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/02 job id: 41858713 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/03 job id: 41858714 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/04 job id: 41858715 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/05 job id: 41858717 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/06 job id: 41858718 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/07 job id: 41858720 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/08 job id: 41858721 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/09 job id: 41858722 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/10 job id: 41858724 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/11 job id: 41858725 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/12 job id: 41858727 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/13 job id: 41858729 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/14 job id: 41858734 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/15 job id: 41858736 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/16 job id: 41858738 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/17 job id: 41858739 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/18 job id: 41858740 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/19 job id: 41858742 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/20 job id: 41858743 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/21 job id: 41858745 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/22 job id: 41858746 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/23 job id: 41858747 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/24 job id: 41858751 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/25 job id: 41858754 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/26 job id: 41858756 --- 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 0.035 134.615 134.616 farming_run 1 2.0 134.055 134.056 134.588 134.592 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.461109E+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 0.010 0.024 115.951 115.952 qs_energies 1 2.0 0.000 0.000 115.726 115.727 mp2_main 1 3.0 0.000 0.000 113.787 113.788 mp2_gpw_main 1 4.0 0.020 0.026 112.951 112.953 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.977 94.482 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.454 55.959 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.739 47.208 get_2c_integrals 1 6.0 0.000 0.001 37.252 38.350 integrate_v_rspace 273 8.0 0.438 0.451 25.131 30.355 pw_transfer 6555 10.6 0.376 0.386 27.381 28.331 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.997 26.844 grid_integrate_task_list 273 9.0 20.953 26.643 20.953 26.643 fft_wrap_pw1pw2_100 2178 12.4 1.215 1.442 23.555 24.402 compute_2c_integrals 1 7.0 0.002 0.003 19.791 19.792 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.979 19.512 mp2_eri_2c_integrate_gpw 1 9.0 2.370 2.435 18.976 19.510 rpa_ri_compute_en 1 5.0 0.000 0.000 18.864 19.036 cp_fm_cholesky_decompose 12 8.2 17.393 18.478 17.393 18.478 cholesky_decomp 1 7.0 0.000 0.000 16.309 17.391 fft3d_s 5443 13.4 16.128 16.732 16.150 16.754 ao_to_mo_and_store_B_mult_1 272 7.0 10.875 15.604 10.875 15.604 calculate_wavefunction 272 8.0 5.409 5.567 12.511 13.159 rpa_num_int 1 6.0 0.000 0.000 10.559 10.559 rpa_num_int_RPA_matrix_operati 8 7.0 0.047 0.047 10.518 10.530 calc_mat_Q 8 8.0 0.000 0.000 9.337 9.419 contract_S_to_Q 8 9.0 0.000 0.000 8.758 8.839 calc_potential_gpw 544 9.5 0.005 0.006 8.287 8.695 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.275 8.617 potential_pw2rs 545 10.0 0.108 0.111 7.719 8.491 parallel_gemm_fm 14 9.1 0.000 0.000 8.347 8.413 parallel_gemm_fm_cosma 14 10.1 8.347 8.413 8.347 8.413 collocate_single_gaussian 272 10.0 0.040 0.043 7.516 7.800 create_integ_mat 1 6.0 0.022 0.027 7.705 7.706 array2fm 1 7.0 0.000 0.000 6.695 7.103 pw_scatter_s 2720 13.7 4.404 4.614 4.404 4.614 pw_gather_s 2722 13.2 3.876 4.199 3.876 4.199 array2fm_buffer_send 1 8.0 2.932 3.098 2.932 3.098 pw_poisson_solve 545 10.5 1.136 1.184 2.226 2.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.950694, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.037 398.423 398.424 farming_run 1 2.0 397.686 397.689 398.384 398.386 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.222869E+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 25853873. 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 0.013 0.048 210.983 210.984 qs_energies 1 2.0 0.000 0.000 210.768 210.773 scf_env_do_scf 1 3.0 0.000 0.000 106.111 106.111 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.143 105.151 rebuild_ks_matrix 4 6.0 0.000 0.000 105.142 105.149 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.142 105.149 hfx_ks_matrix 4 8.0 0.001 0.001 104.773 104.777 integrate_four_center 4 9.0 0.143 0.452 104.773 104.776 mp2_main 1 3.0 0.000 0.000 104.356 104.361 mp2_gpw_main 1 4.0 0.036 0.071 103.369 103.378 integrate_four_center_main 4 10.0 0.097 0.580 96.714 99.304 integrate_four_center_bin 263 11.0 96.617 99.287 96.617 99.287 init_scf_loop 1 4.0 0.000 0.000 91.856 91.856 mp2_ri_gpw_compute_in 1 5.0 0.064 0.072 75.723 77.457 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 55.229 56.986 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.209 47.283 integrate_v_rspace 95 8.0 0.398 0.570 28.526 33.424 pw_transfer 2240 10.6 0.147 0.167 29.915 30.312 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 27.494 29.885 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.932 29.308 ao_to_mo_and_store_B_mult_1 91 7.0 11.332 29.198 11.332 29.198 grid_integrate_task_list 95 9.0 23.819 28.931 23.819 28.931 fft_wrap_pw1pw2_100 730 12.4 1.262 1.431 26.672 27.126 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.912 25.089 25.100 get_2c_integrals 1 6.0 0.001 0.008 20.366 20.435 compute_2c_integrals 1 7.0 0.003 0.004 19.344 19.353 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.841 19.215 mp2_eri_2c_integrate_gpw 1 9.0 1.754 1.881 18.840 19.214 fft3d_s 1823 13.4 18.443 18.850 18.457 18.862 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.254 14.254 calculate_wavefunction 91 8.0 2.047 2.083 9.779 10.027 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.589 8.777 9.292 potential_pw2rs 186 10.0 0.034 0.035 8.634 9.276 local_gemm 172 8.0 8.221 8.730 8.221 8.730 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.218 8.632 mp2_ri_gpw_compute_en_comm 22 7.0 0.496 0.515 8.075 8.571 collocate_single_gaussian 91 10.0 0.017 0.022 7.867 8.164 calc_potential_gpw 182 9.5 0.002 0.002 7.897 8.079 mp_sendrecv_dm3 2068 8.0 6.118 6.583 6.118 6.583 mp_sync 38 10.4 2.767 6.496 2.767 6.496 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.429 6.349 6.429 pw_gather_s 912 13.2 4.934 5.337 4.934 5.337 pw_scatter_s 910 13.7 3.920 4.264 3.920 4.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=103.376573, 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 450.453504E+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 0.036 0.108 56.424 56.425 qs_mol_dyn_low 1 2.0 0.004 0.006 55.885 56.037 qs_forces 11 3.9 0.005 0.023 55.816 55.818 qs_energies 11 4.9 0.003 0.013 54.352 54.369 scf_env_do_scf 11 5.9 0.000 0.001 46.636 46.637 scf_env_do_scf_inner_loop 108 6.5 0.002 0.011 44.489 44.490 dbcsr_multiply_generic 2286 12.5 0.091 0.096 36.770 37.209 qs_scf_new_mos 108 7.5 0.000 0.001 34.017 34.358 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.017 34.358 ot_scf_mini 108 9.5 0.002 0.003 32.280 32.494 multiply_cannon 2286 13.5 0.187 0.191 28.377 29.928 multiply_cannon_loop 2286 14.5 1.501 1.573 27.590 29.156 velocity_verlet 10 3.0 0.001 0.002 26.762 26.763 ot_mini 108 10.5 0.001 0.001 20.727 20.966 qs_ot_get_derivative 108 11.5 0.001 0.001 17.607 17.793 mp_waitall_1 267858 16.1 9.775 16.444 9.775 16.444 multiply_cannon_metrocomm3 54864 15.5 0.066 0.070 6.579 14.427 multiply_cannon_multrec 54864 15.5 4.142 6.530 8.244 12.007 rebuild_ks_matrix 119 8.3 0.000 0.000 8.027 8.148 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 8.027 8.148 mp_sum_l 7207 12.9 6.127 7.880 6.127 7.880 multiply_cannon_sync_h2d 54864 15.5 5.830 7.358 5.830 7.358 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.110 7.222 qs_ot_get_p 119 10.4 0.001 0.001 6.752 7.111 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.963 6.409 init_scf_run 11 5.9 0.000 0.001 6.334 6.334 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.334 6.334 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.879 6.001 dbcsr_mm_accdrv_process 76910 16.1 1.196 1.836 4.024 5.634 sum_up_and_integrate 119 10.3 0.013 0.017 4.634 4.644 integrate_v_rspace 119 11.3 0.002 0.002 4.621 4.631 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.363 4.471 calculate_rho_elec 119 8.7 0.011 0.016 4.363 4.470 calculate_dm_sparse 119 9.5 0.000 0.000 4.009 4.162 calculate_first_density_matrix 1 7.0 0.001 0.002 4.090 4.094 qs_ot_p2m_diag 50 11.0 0.004 0.007 3.736 3.837 jit_kernel_multiply 13 15.8 2.766 3.717 2.766 3.717 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 2.055 3.580 rs_pw_transfer 974 11.9 0.012 0.014 3.083 3.174 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.943 3.140 apply_single 119 13.6 0.000 0.000 2.943 3.140 acc_transpose_blocks 54864 15.5 0.226 0.243 2.690 3.115 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.910 2.913 ot_diis_step 108 11.5 0.006 0.009 2.763 2.765 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.756 2.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.599 2.657 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.589 2.656 density_rs2pw 119 9.7 0.004 0.004 2.423 2.510 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.456 2.459 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.428 2.454 cp_fm_redistribute_end 50 14.0 2.232 2.433 2.238 2.437 cp_fm_diag_elpa_base 50 14.0 0.196 2.367 0.197 2.377 wfi_extrapolate 11 7.9 0.001 0.003 2.186 2.186 init_scf_loop 11 6.9 0.000 0.001 2.132 2.132 grid_integrate_task_list 119 12.3 2.027 2.119 2.027 2.119 mp_sum_d 4129 12.0 1.495 2.098 1.495 2.098 make_m2s 4572 13.5 0.053 0.055 2.023 2.094 make_images 4572 14.5 0.132 0.138 1.941 2.012 potential_pw2rs 119 12.3 0.004 0.004 1.985 2.001 acc_transpose_blocks_kernels 54864 16.5 0.252 0.379 1.702 1.977 pw_transfer 1439 11.6 0.052 0.055 1.812 1.887 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.737 1.815 jit_kernel_transpose 5 15.5 1.449 1.602 1.449 1.602 fft3d_ps 1201 14.6 0.352 0.455 1.515 1.588 mp_alltoall_d11v 2130 13.8 1.419 1.563 1.419 1.563 mp_waitany 12084 13.8 1.396 1.543 1.396 1.543 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.336 1.413 grid_collocate_task_list 119 9.7 1.288 1.362 1.288 1.362 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.869 1.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.425000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 486.883328E+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 879662. 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 0.017 0.035 44.932 44.933 qs_mol_dyn_low 1 2.0 0.003 0.004 44.430 44.438 qs_forces 11 3.9 0.002 0.002 42.844 42.844 qs_energies 11 4.9 0.002 0.004 40.983 40.986 scf_env_do_scf 11 5.9 0.001 0.001 32.871 32.871 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.187 30.190 dbcsr_multiply_generic 2286 12.5 0.098 0.107 25.124 25.492 multiply_cannon 2286 13.5 0.209 0.220 19.622 21.566 qs_scf_new_mos 108 7.5 0.001 0.001 20.804 21.044 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.804 21.044 velocity_verlet 10 3.0 0.002 0.002 20.599 20.632 multiply_cannon_loop 2286 14.5 0.901 0.976 18.466 20.415 ot_scf_mini 108 9.5 0.002 0.003 19.903 20.078 mp_waitall_1 217478 16.2 7.316 14.061 7.316 14.061 ot_mini 108 10.5 0.001 0.001 12.633 12.869 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 5.177 11.839 qs_ot_get_derivative 108 11.5 0.001 0.001 10.130 10.303 multiply_cannon_multrec 27432 15.5 1.978 4.305 6.115 9.099 rebuild_ks_matrix 119 8.3 0.000 0.000 7.356 7.494 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.356 7.493 dbcsr_mm_accdrv_process 47894 16.0 3.197 5.392 4.067 6.894 init_scf_run 11 5.9 0.000 0.001 6.797 6.798 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.797 6.798 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.434 6.557 calculate_first_density_matrix 1 7.0 0.000 0.000 5.266 5.267 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.045 5.145 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.308 4.649 apply_single 119 13.6 0.000 0.000 3.308 4.649 qs_ot_get_p 119 10.4 0.001 0.001 4.349 4.589 mp_sum_l 7207 12.9 2.351 4.570 2.351 4.570 sum_up_and_integrate 119 10.3 0.025 0.028 4.359 4.367 integrate_v_rspace 119 11.3 0.002 0.003 4.334 4.344 acc_transpose_blocks 27432 15.5 0.107 0.111 3.201 4.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.238 4.240 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.022 4.062 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.820 3.855 calculate_rho_elec 119 8.7 0.021 0.024 3.820 3.855 acc_transpose_blocks_kernels 27432 16.5 0.181 0.267 2.671 3.633 jit_kernel_transpose 5 15.5 2.490 3.367 2.490 3.367 make_m2s 4572 13.5 0.052 0.053 2.718 2.947 multiply_cannon_sync_h2d 27432 15.5 2.194 2.890 2.194 2.890 rs_pw_transfer 974 11.9 0.010 0.012 2.774 2.888 make_images 4572 14.5 0.201 0.238 2.632 2.861 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.829 2.847 calculate_dm_sparse 119 9.5 0.000 0.000 2.670 2.748 init_scf_loop 11 6.9 0.001 0.004 2.661 2.671 jit_kernel_multiply 11 16.2 0.818 2.488 0.818 2.488 ot_diis_step 108 11.5 0.011 0.011 2.453 2.455 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.394 2.395 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.236 2.328 density_rs2pw 119 9.7 0.004 0.004 2.143 2.246 potential_pw2rs 119 12.3 0.006 0.007 2.086 2.098 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.989 1.990 pw_transfer 1439 11.6 0.063 0.066 1.944 1.979 cp_fm_redistribute_end 50 14.0 1.648 1.965 1.652 1.968 grid_integrate_task_list 119 12.3 1.830 1.916 1.830 1.916 cp_fm_diag_elpa_base 50 14.0 0.305 1.868 0.313 1.899 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.854 1.889 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.876 1.883 prepare_preconditioner 11 7.9 0.000 0.000 1.728 1.756 make_preconditioner 11 8.9 0.005 0.038 1.728 1.756 mp_sum_dm 438 4.9 1.655 1.715 1.655 1.715 make_images_data 4572 15.5 0.045 0.051 1.231 1.691 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.619 1.683 fft3d_ps 1201 14.6 0.496 0.552 1.559 1.592 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.063 1.591 md_write_output 11 3.9 0.001 0.036 0.034 1.571 md_output 10 3.0 0.000 0.000 0.035 1.569 update_particle_set 20 4.0 0.000 0.000 1.535 1.569 write_trajectory 44 4.9 0.008 0.379 0.033 1.535 mp_alltoall_d11v 2130 13.8 1.377 1.502 1.377 1.502 wfi_extrapolate 11 7.9 0.001 0.001 1.474 1.474 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.376 1.412 mp_allgather_i34 2286 14.5 0.596 1.403 0.596 1.403 grid_collocate_task_list 119 9.7 1.240 1.342 1.240 1.342 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.239 write_particle_coordinates 11 5.9 0.024 1.155 0.024 1.155 mp_sum_d 4129 12.0 0.625 1.088 0.625 1.088 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.000 1.001 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.982 0.995 make_images_sizes 4572 15.5 0.005 0.005 0.670 0.952 mp_alltoall_i44 4572 16.5 0.665 0.948 0.665 0.948 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=44.933000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.181818, yerr=1.799908 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.564160E+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 0.040 0.057 36.176 36.178 qs_mol_dyn_low 1 2.0 0.004 0.005 35.713 35.720 qs_forces 11 3.9 0.003 0.007 35.649 35.650 qs_energies 11 4.9 0.004 0.012 34.071 34.074 scf_env_do_scf 11 5.9 0.001 0.001 26.415 26.415 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.773 23.776 dbcsr_multiply_generic 2286 12.5 0.091 0.098 20.308 20.404 multiply_cannon 2286 13.5 0.196 0.205 16.306 17.469 multiply_cannon_loop 2286 14.5 0.639 0.670 15.510 16.575 qs_scf_new_mos 108 7.5 0.001 0.001 15.710 15.733 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.709 15.732 velocity_verlet 10 3.0 0.002 0.002 15.208 15.209 ot_scf_mini 108 9.5 0.003 0.004 14.980 14.993 ot_mini 108 10.5 0.001 0.001 9.742 9.754 multiply_cannon_multrec 18288 15.5 1.956 2.978 7.226 8.259 qs_ot_get_derivative 108 11.5 0.001 0.003 8.238 8.251 init_scf_run 11 5.9 0.000 0.001 6.437 6.437 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.436 6.437 rebuild_ks_matrix 119 8.3 0.000 0.000 6.353 6.367 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.353 6.367 dbcsr_mm_accdrv_process 38222 16.0 3.945 5.803 5.185 6.063 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.599 5.611 calculate_first_density_matrix 1 7.0 0.000 0.002 5.222 5.223 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.086 4.359 mp_waitall_1 169478 16.3 3.050 4.141 3.050 4.141 sum_up_and_integrate 119 10.3 0.031 0.034 3.983 3.988 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.979 3.980 acc_transpose_blocks 18288 15.5 0.075 0.077 3.945 3.971 integrate_v_rspace 119 11.3 0.003 0.004 3.952 3.959 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.822 3.828 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 3.504 3.518 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.338 3.346 calculate_rho_elec 119 8.7 0.030 0.031 3.338 3.346 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.690 3.318 apply_single 119 13.6 0.000 0.000 2.689 3.318 jit_kernel_transpose 5 15.6 3.296 3.300 3.296 3.300 qs_ot_get_p 119 10.4 0.001 0.001 3.101 3.125 jit_kernel_multiply 10 15.9 1.188 3.110 1.188 3.110 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.573 2.731 init_scf_loop 11 6.9 0.002 0.005 2.623 2.629 calculate_dm_sparse 119 9.5 0.000 0.000 2.501 2.516 rs_pw_transfer 974 11.9 0.010 0.011 2.307 2.460 mp_sum_l 7207 12.9 1.765 2.402 1.765 2.402 qs_ot_p2m_diag 50 11.0 0.012 0.014 2.012 2.018 density_rs2pw 119 9.7 0.004 0.004 1.881 2.012 make_m2s 4572 13.5 0.044 0.045 1.859 2.012 make_images 4572 14.5 0.192 0.204 1.776 1.928 grid_integrate_task_list 119 12.3 1.797 1.894 1.797 1.894 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.864 1.869 prepare_preconditioner 11 7.9 0.000 0.000 1.813 1.815 make_preconditioner 11 8.9 0.014 0.039 1.813 1.815 pw_transfer 1439 11.6 0.063 0.067 1.778 1.789 potential_pw2rs 119 12.3 0.007 0.008 1.745 1.755 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.752 1.752 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.652 1.750 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.688 1.699 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.581 1.592 multiply_cannon_sync_h2d 18288 15.5 1.385 1.550 1.385 1.550 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.520 1.520 cp_fm_redistribute_end 50 14.0 1.130 1.501 1.131 1.502 ot_diis_step 108 11.5 0.012 0.014 1.483 1.485 cp_fm_diag_elpa_base 50 14.0 0.354 1.437 0.369 1.478 fft3d_ps 1201 14.6 0.505 0.521 1.374 1.388 grid_collocate_task_list 119 9.7 1.206 1.314 1.206 1.314 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.291 1.302 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.349 1.258 wfi_extrapolate 11 7.9 0.001 0.002 1.157 1.157 make_images_data 4572 15.5 0.045 0.049 0.819 1.004 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.971 0.972 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.726 0.932 mp_alltoall_d11v 2130 13.8 0.792 0.899 0.792 0.899 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.865 0.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.829 mp_alltoall_z22v 1201 16.6 0.707 0.770 0.707 0.770 cp_fm_cholesky_invert 11 10.9 0.724 0.728 0.724 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.178000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.454545, yerr=2.387986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 554.147840E+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 0.078 0.180 36.903 36.904 qs_mol_dyn_low 1 2.0 0.004 0.006 36.202 36.503 qs_forces 11 3.9 0.002 0.003 36.125 36.126 qs_energies 11 4.9 0.002 0.005 34.417 34.423 scf_env_do_scf 11 5.9 0.001 0.001 28.725 28.726 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.302 25.307 dbcsr_multiply_generic 2286 12.5 0.099 0.118 19.921 20.052 velocity_verlet 10 3.0 0.001 0.002 17.553 17.555 qs_scf_new_mos 108 7.5 0.001 0.001 17.256 17.310 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.255 17.309 multiply_cannon 2286 13.5 0.227 0.262 16.365 17.081 ot_scf_mini 108 9.5 0.003 0.003 16.322 16.378 multiply_cannon_loop 2286 14.5 0.936 0.972 15.422 16.004 multiply_cannon_multrec 27432 15.5 2.370 3.099 9.760 10.790 ot_mini 108 10.5 0.001 0.001 10.456 10.528 dbcsr_mm_accdrv_process 47916 15.9 5.981 7.607 7.293 9.112 qs_ot_get_derivative 108 11.5 0.001 0.002 8.623 8.680 rebuild_ks_matrix 119 8.3 0.000 0.000 6.453 6.507 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.453 6.507 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.726 5.774 init_scf_run 11 5.9 0.000 0.001 4.326 4.326 scf_env_initial_rho_setup 11 6.9 0.000 0.002 4.326 4.326 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.002 4.269 mp_waitall_1 145218 16.4 2.854 4.028 2.854 4.028 sum_up_and_integrate 119 10.3 0.035 0.040 3.745 3.753 integrate_v_rspace 119 11.3 0.002 0.003 3.710 3.718 init_scf_loop 11 6.9 0.001 0.003 3.403 3.408 qs_ot_get_p 119 10.4 0.001 0.001 3.284 3.358 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.205 3.234 calculate_rho_elec 119 8.7 0.040 0.046 3.204 3.233 calculate_first_density_matrix 1 7.0 0.000 0.002 2.960 2.962 multiply_cannon_metrocomm3 27432 15.5 0.037 0.038 1.525 2.834 jit_kernel_multiply 9 16.0 1.251 2.777 1.251 2.777 prepare_preconditioner 11 7.9 0.000 0.000 2.547 2.556 make_preconditioner 11 8.9 0.000 0.001 2.547 2.556 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.503 2.505 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.139 2.479 make_m2s 4572 13.5 0.053 0.057 2.257 2.376 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.292 2.310 make_images 4572 14.5 0.274 0.335 2.150 2.269 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.910 2.155 apply_single 119 13.6 0.000 0.000 1.910 2.155 calculate_dm_sparse 119 9.5 0.000 0.000 2.031 2.092 rs_pw_transfer 974 11.9 0.009 0.010 1.956 2.082 qs_ot_p2m_diag 50 11.0 0.016 0.023 1.979 1.988 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.947 1.975 grid_integrate_task_list 119 12.3 1.829 1.905 1.829 1.905 density_rs2pw 119 9.7 0.003 0.004 1.696 1.811 ot_diis_step 108 11.5 0.015 0.023 1.788 1.788 pw_transfer 1439 11.6 0.063 0.067 1.730 1.759 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.640 1.672 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.642 1.643 potential_pw2rs 119 12.3 0.008 0.010 1.526 1.530 acc_transpose_blocks 27432 15.5 0.112 0.114 1.460 1.490 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.412 1.413 cp_fm_redistribute_end 50 14.0 0.933 1.391 0.935 1.392 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.346 1.382 cp_fm_diag_elpa_base 50 14.0 0.434 1.326 0.454 1.366 fft3d_ps 1201 14.6 0.532 0.583 1.321 1.343 mp_sum_l 7207 12.9 0.904 1.329 0.904 1.329 wfi_extrapolate 11 7.9 0.001 0.002 1.319 1.319 grid_collocate_task_list 119 9.7 1.221 1.305 1.221 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.217 1.228 cp_fm_upper_to_full 72 13.5 0.824 1.148 0.824 1.148 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.097 1.112 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.107 1.108 dbcsr_complete_redistribute 329 12.2 0.123 0.152 0.807 1.090 multiply_cannon_sync_h2d 27432 15.5 0.997 1.070 0.997 1.070 make_images_data 4572 15.5 0.045 0.048 0.880 1.017 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.763 0.926 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.614 0.891 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.870 acc_transpose_blocks_kernels 27432 16.5 0.266 0.274 0.835 0.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.817 0.822 mp_alltoall_d11v 2130 13.8 0.747 0.818 0.747 0.818 mp_alltoall_i22 627 13.8 0.443 0.749 0.443 0.749 cp_fm_cholesky_invert 11 10.9 0.736 0.739 0.736 0.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.904000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.090909, yerr=3.987584 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 600.788992E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 29.208 29.209 qs_mol_dyn_low 1 2.0 0.003 0.004 29.035 29.042 qs_forces 11 3.9 0.002 0.002 28.796 28.797 qs_energies 11 4.9 0.002 0.003 27.071 27.074 scf_env_do_scf 11 5.9 0.000 0.001 21.785 21.786 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.158 19.159 velocity_verlet 10 3.0 0.002 0.002 14.568 14.582 dbcsr_multiply_generic 2286 12.5 0.089 0.092 13.347 13.428 qs_scf_new_mos 108 7.5 0.001 0.001 11.512 11.541 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.511 11.540 multiply_cannon 2286 13.5 0.231 0.237 10.687 11.117 ot_scf_mini 108 9.5 0.002 0.002 10.836 10.866 multiply_cannon_loop 2286 14.5 0.330 0.343 9.741 9.959 multiply_cannon_multrec 9144 15.5 1.736 2.285 5.907 6.406 ot_mini 108 10.5 0.001 0.001 6.279 6.314 rebuild_ks_matrix 119 8.3 0.000 0.000 5.893 5.921 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.892 5.921 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.244 5.270 qs_ot_get_derivative 108 11.5 0.001 0.001 4.975 5.005 dbcsr_mm_accdrv_process 12550 15.8 3.344 4.087 4.071 4.130 init_scf_run 11 5.9 0.000 0.001 3.811 3.811 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.810 3.810 sum_up_and_integrate 119 10.3 0.038 0.042 3.603 3.608 integrate_v_rspace 119 11.3 0.003 0.003 3.565 3.571 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.279 3.287 calculate_rho_elec 119 8.7 0.060 0.061 3.279 3.286 qs_ot_get_p 119 10.4 0.001 0.001 2.754 2.801 mp_waitall_1 121218 16.5 2.168 2.738 2.168 2.738 calculate_first_density_matrix 1 7.0 0.000 0.000 2.690 2.691 init_scf_loop 11 6.9 0.000 0.000 2.608 2.609 make_m2s 4572 13.5 0.033 0.034 1.840 2.030 calculate_dm_sparse 119 9.5 0.000 0.000 2.010 2.029 make_images 4572 14.5 0.268 0.303 1.753 1.941 grid_integrate_task_list 119 12.3 1.843 1.912 1.843 1.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.833 1.834 prepare_preconditioner 11 7.9 0.000 0.000 1.822 1.828 make_preconditioner 11 8.9 0.000 0.000 1.822 1.828 pw_transfer 1439 11.6 0.063 0.066 1.783 1.792 rs_pw_transfer 974 11.9 0.008 0.008 1.736 1.780 qs_ot_p2m_diag 50 11.0 0.023 0.023 1.765 1.768 jit_kernel_multiply 10 16.0 0.689 1.759 0.689 1.759 density_rs2pw 119 9.7 0.004 0.004 1.687 1.747 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.701 1.730 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.692 1.703 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.661 1.670 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.550 1.551 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.531 1.549 acc_transpose_blocks 9144 15.5 0.038 0.040 1.494 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.430 1.434 fft_wrap_pw1pw2_140 487 13.2 0.083 0.086 1.380 1.391 potential_pw2rs 119 12.3 0.010 0.011 1.372 1.377 fft3d_ps 1201 14.6 0.559 0.574 1.358 1.368 grid_collocate_task_list 119 9.7 1.272 1.328 1.272 1.328 ot_diis_step 108 11.5 0.013 0.013 1.289 1.290 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.264 1.265 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.262 1.263 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 1.253 1.258 cp_fm_redistribute_end 50 14.0 0.628 1.241 0.629 1.242 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.224 1.239 cp_fm_diag_elpa_base 50 14.0 0.568 1.167 0.611 1.224 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.164 1.187 apply_single 119 13.6 0.000 0.000 1.164 1.187 jit_kernel_transpose 5 15.6 1.137 1.139 1.137 1.139 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.831 1.119 make_images_data 4572 15.5 0.039 0.042 0.872 1.108 wfi_extrapolate 11 7.9 0.001 0.001 1.076 1.076 cp_fm_cholesky_invert 11 10.9 0.938 0.941 0.938 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.920 mp_alltoall_d11v 2130 13.8 0.825 0.906 0.825 0.906 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.681 0.866 multiply_cannon_sync_h2d 9144 15.5 0.704 0.788 0.704 0.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.775 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.720 0.768 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.719 0.727 mp_allgather_i34 2286 14.5 0.275 0.708 0.275 0.708 mp_alltoall_z22v 1201 16.6 0.671 0.700 0.671 0.700 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.337 0.673 qs_create_task_list 11 7.9 0.001 0.003 0.563 0.584 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.209000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.545455, yerr=5.105499 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 747.372544E+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 0.039 43.511 43.511 qs_mol_dyn_low 1 2.0 0.003 0.004 42.495 42.502 qs_forces 11 3.9 0.002 0.002 41.582 41.583 qs_energies 11 4.9 0.002 0.002 39.639 39.643 scf_env_do_scf 11 5.9 0.001 0.001 32.932 32.932 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.076 25.077 velocity_verlet 10 3.0 0.002 0.002 23.812 23.923 dbcsr_multiply_generic 2286 12.5 0.097 0.098 18.566 18.745 qs_scf_new_mos 108 7.5 0.001 0.001 16.084 16.188 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.083 16.187 multiply_cannon 2286 13.5 0.304 0.311 14.580 15.493 ot_scf_mini 108 9.5 0.002 0.002 14.987 15.088 multiply_cannon_loop 2286 14.5 0.343 0.352 13.308 14.210 ot_mini 108 10.5 0.001 0.001 8.811 8.932 multiply_cannon_multrec 9144 15.5 3.427 4.772 8.541 8.635 init_scf_loop 11 6.9 0.000 0.000 7.827 7.831 rebuild_ks_matrix 119 8.3 0.000 0.000 6.888 7.033 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.888 7.033 prepare_preconditioner 11 7.9 0.000 0.000 6.909 6.923 make_preconditioner 11 8.9 0.000 0.000 6.909 6.923 qs_ot_get_derivative 108 11.5 0.001 0.001 6.765 6.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.468 6.800 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.233 6.364 dbcsr_mm_accdrv_process 12550 15.8 4.021 5.388 4.990 6.355 init_scf_run 11 5.9 0.000 0.001 4.690 4.690 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.689 4.689 cp_fm_upper_to_full 72 14.2 3.197 4.582 3.197 4.582 sum_up_and_integrate 119 10.3 0.065 0.066 3.724 3.732 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.721 3.727 calculate_rho_elec 119 8.7 0.118 0.121 3.721 3.727 integrate_v_rspace 119 11.3 0.004 0.004 3.660 3.666 mp_waitall_1 97218 16.6 2.622 3.582 2.622 3.582 qs_ot_get_p 119 10.4 0.001 0.001 3.401 3.541 calculate_first_density_matrix 1 7.0 0.000 0.000 3.246 3.248 calculate_dm_sparse 119 9.5 0.000 0.000 3.180 3.195 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.424 2.863 dbcsr_complete_redistribute 329 12.2 0.290 0.295 1.995 2.829 make_m2s 4572 13.5 0.037 0.037 2.363 2.541 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.513 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.224 2.458 apply_single 119 13.6 0.000 0.000 2.224 2.457 make_images 4572 14.5 0.352 0.382 2.244 2.421 mp_alltoall_i22 627 13.8 1.479 2.286 1.479 2.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.267 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.232 2.239 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.285 2.188 pw_transfer 1439 11.6 0.067 0.067 2.090 2.094 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.061 2.062 grid_integrate_task_list 119 12.3 2.006 2.019 2.006 2.019 ot_diis_step 108 11.5 0.014 0.014 2.015 2.015 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.994 1.998 acc_transpose_blocks 9144 15.5 0.038 0.039 1.944 1.967 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.963 1.964 density_rs2pw 119 9.7 0.003 0.003 1.817 1.837 mp_sum_l 7207 12.9 1.085 1.811 1.085 1.811 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.787 1.787 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.753 1.753 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.690 1.742 acc_transpose_blocks_kernels 9144 16.5 0.117 0.118 1.694 1.718 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.633 1.666 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.659 1.663 fft3d_ps 1201 14.6 0.567 0.577 1.628 1.630 jit_kernel_transpose 5 15.6 1.577 1.603 1.577 1.603 cp_fm_cholesky_invert 11 10.9 1.495 1.498 1.495 1.498 grid_collocate_task_list 119 9.7 1.449 1.475 1.449 1.475 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.211 1.471 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.464 1.467 cp_fm_diag_elpa_base 50 14.0 1.317 1.367 1.463 1.466 rs_pw_transfer 974 11.9 0.009 0.009 1.422 1.463 jit_kernel_multiply 8 15.5 0.941 1.441 0.941 1.441 make_images_data 4572 15.5 0.042 0.046 1.159 1.396 wfi_extrapolate 11 7.9 0.001 0.001 1.382 1.382 potential_pw2rs 119 12.3 0.014 0.015 1.254 1.257 mp_alltoall_d11v 2130 13.8 1.168 1.186 1.168 1.186 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.162 1.183 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.094 1.112 multiply_cannon_sync_h2d 9144 15.5 1.042 1.049 1.042 1.049 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.986 1.035 qs_create_task_list 11 7.9 0.000 0.001 0.943 0.956 generate_qs_task_list 11 8.9 0.372 0.391 0.943 0.955 mp_alltoall_z22v 1201 16.6 0.923 0.938 0.923 0.938 mp_sum_dm 438 4.9 0.803 0.927 0.803 0.927 md_write_output 11 3.9 0.111 0.881 0.113 0.894 md_output 10 3.0 0.000 0.000 0.114 0.891 update_particle_set 20 4.0 0.000 0.000 0.779 0.890 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.867 0.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.511000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=701.090909, yerr=15.299760 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 500.269056E+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 2050640. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.042 84.919 84.920 qs_mol_dyn_low 1 2.0 0.003 0.004 84.669 84.679 qs_forces 11 3.9 0.002 0.003 84.598 84.599 qs_energies 11 4.9 0.002 0.003 81.735 81.744 scf_env_do_scf 11 5.9 0.000 0.001 72.900 72.902 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 67.143 67.143 dbcsr_multiply_generic 2055 12.4 0.101 0.104 53.706 54.067 qs_scf_new_mos 99 7.5 0.000 0.001 49.869 49.968 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.868 49.967 ot_scf_mini 99 9.5 0.002 0.002 47.446 47.537 multiply_cannon 2055 13.4 0.180 0.186 43.620 44.940 multiply_cannon_loop 2055 14.4 1.557 1.591 42.574 43.943 velocity_verlet 10 3.0 0.001 0.002 42.550 42.551 ot_mini 99 10.5 0.001 0.001 28.969 29.090 qs_ot_get_derivative 99 11.5 0.001 0.001 22.098 22.183 multiply_cannon_multrec 49320 15.4 12.099 12.981 17.461 18.795 rebuild_ks_matrix 110 8.3 0.000 0.001 14.678 14.847 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.678 14.846 mp_waitall_1 241148 16.1 12.698 14.220 12.698 14.220 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.884 13.021 multiply_cannon_sync_h2d 49320 15.4 9.940 10.531 9.940 10.531 qs_ot_get_p 110 10.4 0.001 0.001 9.848 9.950 multiply_cannon_metrocomm3 49320 15.4 0.076 0.082 7.520 9.560 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.063 8.610 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.170 7.630 apply_single 110 13.6 0.000 0.001 7.169 7.630 sum_up_and_integrate 110 10.3 0.038 0.045 7.195 7.211 integrate_v_rspace 110 11.3 0.003 0.003 7.157 7.184 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.993 7.059 init_scf_run 11 5.9 0.000 0.001 6.818 6.818 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.817 6.818 ot_diis_step 99 11.5 0.005 0.005 6.547 6.548 mp_sum_l 6514 12.8 5.511 6.545 5.511 6.545 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.333 6.478 calculate_rho_elec 110 8.6 0.020 0.024 6.332 6.477 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.334 6.367 dbcsr_mm_accdrv_process 87628 16.1 2.062 2.178 5.241 6.058 init_scf_loop 11 6.9 0.000 0.000 5.727 5.727 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.195 5.195 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.651 4.652 cp_fm_redistribute_end 48 14.0 4.048 4.621 4.053 4.622 cp_fm_diag_elpa_base 48 14.0 0.562 4.505 0.566 4.537 make_m2s 4110 13.4 0.059 0.064 4.132 4.244 make_images 4110 14.4 0.178 0.192 4.038 4.154 rs_pw_transfer 902 11.9 0.012 0.014 3.924 4.117 wfi_extrapolate 11 7.9 0.001 0.001 4.097 4.097 multiply_cannon_metrocomm1 49320 15.4 0.059 0.064 2.853 4.091 calculate_dm_sparse 110 9.5 0.000 0.001 3.788 3.890 density_rs2pw 110 9.6 0.004 0.004 3.409 3.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.603 3.608 prepare_preconditioner 11 7.9 0.000 0.000 3.554 3.573 make_preconditioner 11 8.9 0.000 0.000 3.554 3.573 jit_kernel_multiply 13 15.9 2.897 3.452 2.897 3.452 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.388 3.442 grid_integrate_task_list 110 12.3 3.257 3.428 3.257 3.428 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.306 3.359 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.260 3.312 pw_transfer 1331 11.6 0.054 0.062 3.013 3.079 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.926 2.994 potential_pw2rs 110 12.3 0.006 0.007 2.739 2.768 mp_alltoall_d11v 2046 13.8 2.225 2.744 2.225 2.744 calculate_first_density_matrix 1 7.0 0.000 0.000 2.636 2.640 fft3d_ps 1111 14.6 0.740 0.824 2.462 2.523 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 2.448 2.517 acc_transpose_blocks 49320 15.4 0.211 0.226 2.165 2.245 grid_collocate_task_list 110 9.6 2.088 2.243 2.088 2.243 mp_waitany 14300 13.8 1.866 2.164 1.866 2.164 mp_sum_d 3883 11.9 1.549 2.090 1.549 2.090 make_images_data 4110 15.4 0.044 0.047 1.864 2.007 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.935 1.968 cp_fm_cholesky_invert 11 10.9 1.917 1.922 1.917 1.922 hybrid_alltoall_any 4261 16.3 0.082 0.487 1.612 1.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.920000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.272727, yerr=2.799055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.014720E+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 3215589. 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 0.027 0.064 71.035 71.036 qs_mol_dyn_low 1 2.0 0.003 0.006 70.714 70.723 qs_forces 11 3.9 0.002 0.002 70.645 70.647 qs_energies 11 4.9 0.004 0.004 67.320 67.324 scf_env_do_scf 11 5.9 0.000 0.001 57.667 57.671 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.679 49.680 dbcsr_multiply_generic 2055 12.4 0.109 0.113 39.099 39.275 velocity_verlet 10 3.0 0.001 0.002 36.468 36.469 qs_scf_new_mos 99 7.5 0.001 0.001 33.550 33.704 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.550 33.703 multiply_cannon 2055 13.4 0.222 0.248 32.263 33.508 ot_scf_mini 99 9.5 0.003 0.003 31.868 32.006 multiply_cannon_loop 2055 14.4 0.931 0.952 30.963 31.923 ot_mini 99 10.5 0.001 0.001 18.565 18.716 multiply_cannon_multrec 24660 15.4 7.606 8.922 14.324 15.634 rebuild_ks_matrix 110 8.3 0.000 0.001 13.829 13.940 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.829 13.940 qs_ot_get_derivative 99 11.5 0.001 0.001 12.669 12.810 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.197 12.291 mp_waitall_1 186928 16.3 8.288 10.924 8.288 10.924 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.384 8.458 multiply_cannon_sync_h2d 24660 15.4 7.035 8.090 7.035 8.090 init_scf_loop 11 6.9 0.000 0.000 7.949 7.949 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.573 7.331 apply_single 110 13.6 0.000 0.001 6.573 7.330 init_scf_run 11 5.9 0.000 0.001 7.270 7.270 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.269 7.270 dbcsr_mm_accdrv_process 52282 16.1 4.890 5.987 6.559 6.844 sum_up_and_integrate 110 10.3 0.054 0.060 6.746 6.758 qs_ot_get_p 110 10.4 0.001 0.001 6.575 6.746 integrate_v_rspace 110 11.3 0.002 0.003 6.692 6.706 prepare_preconditioner 11 7.9 0.000 0.000 5.939 5.955 make_preconditioner 11 8.9 0.000 0.000 5.939 5.955 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.857 5.867 calculate_rho_elec 110 8.6 0.039 0.047 5.856 5.866 ot_diis_step 99 11.5 0.010 0.010 5.835 5.835 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.503 5.664 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.764 5.501 make_m2s 4110 13.4 0.056 0.059 4.324 4.762 make_images 4110 14.4 0.403 0.447 4.221 4.656 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.580 4.605 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.116 4.116 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 4.098 4.101 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.692 3.732 calculate_first_density_matrix 1 7.0 0.000 0.003 3.647 3.648 pw_transfer 1331 11.6 0.066 0.071 3.418 3.561 wfi_extrapolate 11 7.9 0.001 0.001 3.532 3.533 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.527 3.529 cp_fm_redistribute_end 48 14.0 2.626 3.492 2.629 3.492 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.311 3.453 cp_fm_diag_elpa_base 48 14.0 0.826 3.358 0.859 3.441 grid_integrate_task_list 110 12.3 3.163 3.391 3.163 3.391 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.302 3.382 density_rs2pw 110 9.6 0.004 0.004 3.134 3.301 calculate_dm_sparse 110 9.5 0.001 0.001 3.227 3.259 rs_pw_transfer 902 11.9 0.012 0.014 3.107 3.249 jit_kernel_multiply 13 16.4 1.317 3.012 1.317 3.012 make_images_data 4110 15.4 0.048 0.051 2.427 2.876 fft_wrap_pw1pw2_140 451 13.1 0.199 0.217 2.733 2.873 fft3d_ps 1111 14.6 1.069 1.263 2.668 2.798 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.152 2.793 cp_fm_cholesky_invert 11 10.9 2.751 2.763 2.751 2.763 mp_sum_l 6514 12.8 1.921 2.620 1.921 2.620 potential_pw2rs 110 12.3 0.009 0.009 2.586 2.611 grid_collocate_task_list 110 9.6 2.124 2.296 2.124 2.296 acc_transpose_blocks 24660 15.4 0.107 0.110 2.056 2.138 mp_alltoall_d11v 2046 13.8 1.891 2.065 1.891 2.065 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.034 2.054 qs_energies_init_hamiltonians 11 5.9 0.013 0.032 1.871 1.873 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.797 1.810 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.778 1.790 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.782 1.731 mp_allgather_i34 2055 14.4 0.629 1.674 0.629 1.674 dbcsr_complete_redistribute 325 12.2 0.273 0.341 1.274 1.540 mp_irecv_dv 57340 16.2 0.661 1.522 0.661 1.522 acc_transpose_blocks_kernels 24660 16.4 0.307 0.319 1.445 1.521 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.478 cp_fm_cholesky_decompose 22 10.9 1.452 1.458 1.452 1.458 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.036000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.181818, yerr=7.208810 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.398080E+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 0.033 0.097 60.648 60.649 qs_mol_dyn_low 1 2.0 0.003 0.005 60.222 60.232 qs_forces 11 3.9 0.003 0.007 60.132 60.133 qs_energies 11 4.9 0.005 0.027 56.944 56.947 scf_env_do_scf 11 5.9 0.001 0.001 48.834 48.834 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.406 40.409 velocity_verlet 10 3.0 0.002 0.006 32.042 32.044 dbcsr_multiply_generic 2055 12.4 0.104 0.106 29.663 29.892 qs_scf_new_mos 99 7.5 0.001 0.001 25.949 26.039 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.948 26.038 ot_scf_mini 99 9.5 0.003 0.004 24.729 24.840 multiply_cannon 2055 13.4 0.213 0.224 23.421 24.546 multiply_cannon_loop 2055 14.4 0.618 0.633 22.213 23.256 ot_mini 99 10.5 0.001 0.001 14.698 14.808 rebuild_ks_matrix 110 8.3 0.000 0.000 12.112 12.230 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 12.111 12.229 multiply_cannon_multrec 16440 15.4 3.870 4.979 10.010 11.735 mp_waitall_1 146766 16.3 8.061 11.488 8.061 11.488 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.690 10.795 qs_ot_get_derivative 99 11.5 0.001 0.001 10.279 10.388 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 5.309 8.553 init_scf_loop 11 6.9 0.000 0.002 8.393 8.397 dbcsr_mm_accdrv_process 34862 16.1 4.628 6.156 5.992 7.091 prepare_preconditioner 11 7.9 0.000 0.000 6.626 6.641 make_preconditioner 11 8.9 0.000 0.001 6.626 6.641 sum_up_and_integrate 110 10.3 0.060 0.061 6.362 6.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.978 6.321 integrate_v_rspace 110 11.3 0.003 0.003 6.301 6.316 init_scf_run 11 5.9 0.000 0.001 5.562 5.562 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.561 5.562 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.498 5.505 calculate_rho_elec 110 8.6 0.058 0.058 5.497 5.505 qs_ot_get_p 110 10.4 0.001 0.001 5.251 5.393 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.934 5.389 apply_single 110 13.6 0.000 0.000 4.934 5.389 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.383 5.025 make_m2s 4110 13.4 0.049 0.051 4.159 4.535 make_images 4110 14.4 0.396 0.514 4.047 4.424 ot_diis_step 99 11.5 0.011 0.014 4.391 4.391 multiply_cannon_sync_h2d 16440 15.4 3.683 4.132 3.683 4.132 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.650 3.655 grid_integrate_task_list 110 12.3 3.196 3.358 3.196 3.358 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.304 3.305 pw_transfer 1331 11.6 0.065 0.072 3.177 3.185 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.071 3.080 density_rs2pw 110 9.6 0.004 0.004 2.770 3.013 wfi_extrapolate 11 7.9 0.001 0.001 2.925 2.925 make_images_data 4110 15.4 0.044 0.048 2.399 2.906 rs_pw_transfer 902 11.9 0.011 0.012 2.660 2.893 jit_kernel_multiply 8 16.4 0.974 2.868 0.974 2.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.827 2.828 hybrid_alltoall_any 4261 16.3 0.106 0.376 2.084 2.796 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.776 2.778 cp_fm_redistribute_end 48 14.0 1.728 2.753 1.731 2.754 cp_fm_diag_elpa_base 48 14.0 0.960 2.609 1.018 2.719 fft_wrap_pw1pw2_140 451 13.1 0.211 0.216 2.626 2.638 calculate_dm_sparse 110 9.5 0.001 0.001 2.584 2.616 cp_fm_cholesky_invert 11 10.9 2.608 2.614 2.608 2.614 calculate_first_density_matrix 1 7.0 0.001 0.006 2.555 2.557 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.464 2.508 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.883 2.427 fft3d_ps 1111 14.6 1.062 1.073 2.383 2.397 grid_collocate_task_list 110 9.6 2.179 2.354 2.179 2.354 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.278 2.338 mp_irecv_dv 48980 15.7 0.814 2.299 0.814 2.299 potential_pw2rs 110 12.3 0.010 0.011 2.182 2.188 mp_sum_l 6514 12.8 1.516 2.115 1.516 2.115 qs_energies_init_hamiltonians 11 5.9 0.012 0.016 2.057 2.071 dbcsr_complete_redistribute 325 12.2 0.350 0.404 1.483 1.940 mp_alltoall_d11v 2046 13.8 1.737 1.904 1.737 1.904 cp_fm_upper_to_full 70 13.6 1.377 1.841 1.377 1.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.621 1.635 mp_allgather_i34 2055 14.4 0.527 1.567 0.527 1.567 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.005 1.544 1.557 cp_fm_cholesky_decompose 22 10.9 1.526 1.542 1.526 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.499 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.000 1.442 mp_waitany 17072 13.8 1.183 1.376 1.183 1.376 acc_transpose_blocks 16440 15.4 0.072 0.074 1.217 1.334 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.251 1.258 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.142 1.225 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.649000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.636364, yerr=9.256671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 733.487104E+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 0.013 0.033 65.780 65.780 qs_mol_dyn_low 1 2.0 0.003 0.003 65.235 65.463 qs_forces 11 3.9 0.002 0.002 65.157 65.157 qs_energies 11 4.9 0.001 0.002 61.778 61.782 scf_env_do_scf 11 5.9 0.000 0.001 52.864 52.867 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.221 41.222 velocity_verlet 10 3.0 0.004 0.005 36.543 36.545 dbcsr_multiply_generic 2055 12.4 0.111 0.115 31.029 31.275 qs_scf_new_mos 99 7.5 0.001 0.001 26.879 27.001 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.879 27.000 ot_scf_mini 99 9.5 0.003 0.003 25.254 25.366 multiply_cannon 2055 13.4 0.243 0.260 23.913 24.958 multiply_cannon_loop 2055 14.4 0.883 0.906 22.566 23.217 multiply_cannon_multrec 24660 15.4 4.330 7.235 13.930 15.181 ot_mini 99 10.5 0.001 0.001 14.889 15.026 rebuild_ks_matrix 110 8.3 0.000 0.000 11.888 12.000 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.888 12.000 init_scf_loop 11 6.9 0.000 0.000 11.601 11.602 qs_ot_get_derivative 99 11.5 0.001 0.001 10.740 10.857 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.502 10.601 dbcsr_mm_accdrv_process 52304 16.0 7.510 9.712 9.452 10.547 prepare_preconditioner 11 7.9 0.000 0.000 9.883 9.904 make_preconditioner 11 8.9 0.000 0.000 9.883 9.904 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.103 9.570 mp_waitall_1 126806 16.4 4.734 6.554 4.734 6.554 sum_up_and_integrate 110 10.3 0.069 0.072 6.274 6.288 integrate_v_rspace 110 11.3 0.003 0.003 6.205 6.218 init_scf_run 11 5.9 0.000 0.001 6.179 6.180 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.179 6.179 make_m2s 4110 13.4 0.058 0.059 5.531 5.819 make_images 4110 14.4 0.580 0.700 5.392 5.676 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.596 5.609 calculate_rho_elec 110 8.6 0.077 0.081 5.596 5.608 qs_ot_get_p 110 10.4 0.001 0.001 5.333 5.496 cp_fm_upper_to_full 70 13.8 3.317 4.714 3.317 4.714 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.040 4.127 apply_single 110 13.6 0.000 0.000 4.040 4.127 ot_diis_step 99 11.5 0.011 0.012 4.112 4.112 dbcsr_complete_redistribute 325 12.2 0.419 0.469 2.657 3.797 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.663 3.721 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.575 3.589 calculate_dm_sparse 110 9.5 0.001 0.001 3.382 3.410 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.525 3.396 grid_integrate_task_list 110 12.3 3.242 3.391 3.242 3.391 multiply_cannon_sync_h2d 24660 15.4 3.164 3.342 3.164 3.342 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.134 3.264 pw_transfer 1331 11.6 0.065 0.072 3.212 3.232 make_images_data 4110 15.4 0.046 0.050 2.860 3.166 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.429 3.134 calculate_first_density_matrix 1 7.0 0.000 0.000 3.129 3.130 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.107 3.130 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.101 3.102 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.095 3.096 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.986 3.051 wfi_extrapolate 11 7.9 0.001 0.001 2.967 2.967 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.881 density_rs2pw 110 9.6 0.004 0.004 2.743 2.879 mp_alltoall_i22 605 13.7 1.655 2.846 1.655 2.846 jit_kernel_multiply 10 15.8 1.609 2.823 1.609 2.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.713 2.750 cp_fm_cholesky_invert 11 10.9 2.658 2.665 2.658 2.665 fft_wrap_pw1pw2_140 451 13.1 0.202 0.215 2.641 2.665 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.588 2.590 cp_fm_redistribute_end 48 14.0 1.289 2.562 1.291 2.564 cp_fm_diag_elpa_base 48 14.0 1.190 2.429 1.269 2.543 rs_pw_transfer 902 11.9 0.011 0.011 2.388 2.538 fft3d_ps 1111 14.6 1.061 1.094 2.409 2.425 grid_collocate_task_list 110 9.6 2.219 2.393 2.219 2.393 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.249 2.249 mp_alltoall_d11v 2046 13.8 1.851 2.081 1.851 2.081 potential_pw2rs 110 12.3 0.013 0.013 2.063 2.072 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.992 2.008 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.686 1.718 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.703 cp_fm_cholesky_decompose 22 10.9 1.639 1.697 1.639 1.697 acc_transpose_blocks 24660 15.4 0.105 0.108 1.647 1.681 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.643 mp_sum_l 6514 12.8 0.905 1.594 0.905 1.594 multiply_cannon_metrocomm4 20550 15.4 0.056 0.058 0.848 1.529 mp_irecv_dv 62702 16.1 0.751 1.450 0.751 1.450 mp_allgather_i34 2055 14.4 0.444 1.434 0.444 1.434 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.780000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.454545, yerr=11.515710 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 829.849600E+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 0.027 0.038 55.860 55.861 qs_mol_dyn_low 1 2.0 0.003 0.003 55.024 55.043 qs_forces 11 3.9 0.002 0.002 54.750 54.750 qs_energies 11 4.9 0.005 0.006 51.143 51.147 scf_env_do_scf 11 5.9 0.000 0.001 42.592 42.592 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.913 34.913 velocity_verlet 10 3.0 0.003 0.003 30.746 30.762 dbcsr_multiply_generic 2055 12.4 0.102 0.103 23.516 23.648 qs_scf_new_mos 99 7.5 0.001 0.001 20.711 20.743 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.711 20.743 ot_scf_mini 99 9.5 0.002 0.002 19.459 19.482 multiply_cannon 2055 13.4 0.249 0.268 18.010 19.439 multiply_cannon_loop 2055 14.4 0.322 0.336 16.658 16.980 rebuild_ks_matrix 110 8.3 0.000 0.000 11.492 11.532 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.492 11.532 ot_mini 99 10.5 0.001 0.001 10.683 10.704 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.247 10.281 multiply_cannon_multrec 8220 15.4 3.474 4.829 7.807 8.686 mp_waitall_1 106626 16.5 6.215 7.844 6.215 7.844 init_scf_loop 11 6.9 0.000 0.000 7.632 7.634 qs_ot_get_derivative 99 11.5 0.001 0.001 6.907 6.927 sum_up_and_integrate 110 10.3 0.080 0.082 6.205 6.218 integrate_v_rspace 110 11.3 0.003 0.003 6.124 6.137 prepare_preconditioner 11 7.9 0.000 0.000 6.012 6.018 make_preconditioner 11 8.9 0.000 0.000 6.012 6.018 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.664 5.676 calculate_rho_elec 110 8.6 0.115 0.115 5.663 5.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.584 5.671 init_scf_run 11 5.9 0.000 0.001 5.421 5.421 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.421 5.421 dbcsr_mm_accdrv_process 17442 15.9 2.800 4.107 4.204 5.324 qs_ot_get_p 110 10.4 0.001 0.001 4.838 4.864 make_m2s 4110 13.4 0.038 0.040 4.270 4.532 make_images 4110 14.4 0.640 0.687 4.142 4.405 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.894 4.248 ot_diis_step 99 11.5 0.012 0.012 3.756 3.756 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.677 3.693 apply_single 110 13.6 0.000 0.000 3.677 3.693 grid_integrate_task_list 110 12.3 3.365 3.512 3.365 3.512 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.505 3.508 pw_transfer 1331 11.6 0.065 0.071 3.300 3.315 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.193 3.211 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.206 3.207 multiply_cannon_sync_h2d 8220 15.4 2.928 3.038 2.928 3.038 calculate_dm_sparse 110 9.5 0.001 0.001 2.862 2.896 cp_fm_cholesky_invert 11 10.9 2.892 2.896 2.892 2.896 make_images_data 4110 15.4 0.038 0.044 2.444 2.854 density_rs2pw 110 9.6 0.004 0.004 2.699 2.835 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.362 2.830 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 2.724 2.743 wfi_extrapolate 11 7.9 0.001 0.001 2.697 2.697 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.695 2.696 cp_fm_redistribute_end 48 14.0 0.684 2.669 0.689 2.670 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.663 2.663 calculate_first_density_matrix 1 7.0 0.000 0.000 2.637 2.638 cp_fm_diag_elpa_base 48 14.0 1.799 2.478 1.973 2.637 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.536 2.537 grid_collocate_task_list 110 9.6 2.330 2.494 2.330 2.494 fft3d_ps 1111 14.6 1.122 1.151 2.450 2.465 rs_pw_transfer 902 11.9 0.010 0.011 2.213 2.357 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.161 2.179 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.988 potential_pw2rs 110 12.3 0.015 0.016 1.954 1.957 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.876 1.887 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.811 1.821 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.806 1.818 mp_alltoall_d11v 2046 13.8 1.637 1.818 1.637 1.818 jit_kernel_multiply 10 15.8 1.094 1.801 1.094 1.801 cp_fm_cholesky_decompose 22 10.9 1.654 1.664 1.654 1.664 mp_allgather_i34 2055 14.4 0.534 1.645 0.534 1.645 qs_env_update_s_mstruct 11 6.9 0.003 0.005 1.508 1.623 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.577 1.582 dbcsr_complete_redistribute 325 12.2 0.575 0.591 1.463 1.568 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.891 1.351 qs_create_task_list 11 7.9 0.001 0.001 1.228 1.326 generate_qs_task_list 11 8.9 0.379 0.447 1.227 1.325 acc_transpose_blocks 8220 15.4 0.036 0.037 1.234 1.243 mp_waitany 9240 13.8 1.061 1.206 1.061 1.206 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.140 1.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.861000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.000000, yerr=12.240024 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 1.371247E+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 0.038 0.054 87.464 87.465 qs_mol_dyn_low 1 2.0 0.003 0.003 86.843 86.856 qs_forces 11 3.9 0.002 0.002 86.778 86.779 qs_energies 11 4.9 0.029 0.057 82.735 82.740 scf_env_do_scf 11 5.9 0.001 0.001 72.146 72.147 velocity_verlet 10 3.0 0.002 0.002 55.223 55.229 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.544 43.545 dbcsr_multiply_generic 2055 12.4 0.117 0.119 29.555 29.681 init_scf_loop 11 6.9 0.000 0.000 28.529 28.531 qs_scf_new_mos 99 7.5 0.001 0.001 26.790 26.837 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.789 26.836 prepare_preconditioner 11 7.9 0.000 0.000 26.519 26.534 make_preconditioner 11 8.9 0.000 0.000 26.519 26.534 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.657 25.985 ot_scf_mini 99 9.5 0.002 0.002 25.016 25.041 multiply_cannon 2055 13.4 0.352 0.378 22.457 23.127 multiply_cannon_loop 2055 14.4 0.342 0.344 20.660 20.984 cp_fm_upper_to_full 70 14.2 12.804 18.392 12.804 18.392 ot_mini 99 10.5 0.001 0.001 14.036 14.072 rebuild_ks_matrix 110 8.3 0.000 0.001 13.282 13.340 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.281 13.340 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.097 12.150 dbcsr_complete_redistribute 325 12.2 1.026 1.040 7.456 10.678 multiply_cannon_multrec 8220 15.4 4.363 4.568 9.625 9.824 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.415 9.632 qs_ot_get_derivative 99 11.5 0.001 0.001 9.469 9.494 mp_waitall_1 87304 16.6 8.002 9.230 8.002 9.230 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.846 9.031 mp_alltoall_i22 605 13.7 5.478 8.721 5.478 8.721 sum_up_and_integrate 110 10.3 0.151 0.152 6.766 6.794 integrate_v_rspace 110 11.3 0.004 0.004 6.615 6.643 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.512 6.550 calculate_rho_elec 110 8.6 0.227 0.228 6.512 6.550 init_scf_run 11 5.9 0.000 0.001 6.299 6.299 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.299 6.299 qs_ot_get_p 110 10.4 0.001 0.001 5.794 5.840 make_m2s 4110 13.4 0.042 0.043 5.289 5.819 make_images 4110 14.4 0.884 0.931 5.104 5.633 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.940 5.436 dbcsr_mm_accdrv_process 11614 15.7 3.257 3.788 5.120 5.422 cp_fm_cholesky_invert 11 10.9 5.259 5.263 5.259 5.263 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.654 5.142 apply_single 110 13.6 0.000 0.000 4.654 5.142 ot_diis_step 99 11.5 0.015 0.015 4.536 4.538 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.189 4.197 multiply_cannon_sync_h2d 8220 15.4 3.950 3.958 3.950 3.958 pw_transfer 1331 11.6 0.074 0.075 3.809 3.827 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.315 3.771 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.737 3.738 grid_integrate_task_list 110 12.3 3.661 3.719 3.661 3.719 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.693 3.711 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 3.629 3.630 hybrid_alltoall_any 4261 16.3 0.256 0.553 2.936 3.607 make_images_data 4110 15.4 0.042 0.045 2.873 3.554 calculate_dm_sparse 110 9.5 0.001 0.001 3.392 3.415 wfi_extrapolate 11 7.9 0.001 0.001 3.253 3.254 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.205 3.223 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.195 3.198 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.135 3.135 cp_fm_diag_elpa_base 48 14.0 2.593 2.799 3.133 3.134 density_rs2pw 110 9.6 0.004 0.004 2.981 3.004 calculate_first_density_matrix 1 7.0 0.000 0.000 2.944 2.944 fft3d_ps 1111 14.6 1.266 1.275 2.901 2.922 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.687 2.710 grid_collocate_task_list 110 9.6 2.629 2.658 2.629 2.658 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.404 2.419 rs_pw_transfer 902 11.9 0.011 0.011 2.198 2.243 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.175 2.232 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.077 2.188 potential_pw2rs 110 12.3 0.022 0.022 2.154 2.173 mp_alltoall_d11v 2046 13.8 2.101 2.138 2.101 2.138 cp_fm_cholesky_decompose 22 10.9 2.065 2.092 2.065 2.092 qs_create_task_list 11 7.9 0.001 0.001 1.899 1.943 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.925 1.943 generate_qs_task_list 11 8.9 0.739 0.793 1.898 1.943 jit_kernel_multiply 10 15.4 1.658 1.903 1.658 1.903 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.773 1.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.796 acc_transpose_blocks 8220 15.4 0.035 0.035 1.620 1.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.465000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1228.727273, yerr=69.495704 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.706176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 3266308. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 204.554 204.555 qs_mol_dyn_low 1 2.0 0.003 0.004 204.212 204.226 qs_forces 11 3.9 0.003 0.003 204.129 204.129 qs_energies 11 4.9 0.003 0.008 198.633 198.645 scf_env_do_scf 11 5.9 0.001 0.001 181.952 181.955 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 160.779 160.781 dbcsr_multiply_generic 2507 12.6 0.169 0.174 124.890 125.613 velocity_verlet 10 3.0 0.001 0.001 123.009 123.010 qs_scf_new_mos 117 7.6 0.001 0.001 121.981 122.235 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.981 122.234 ot_scf_mini 117 9.6 0.003 0.003 115.337 115.527 multiply_cannon 2507 13.6 0.239 0.247 101.114 103.321 multiply_cannon_loop 2507 14.6 2.117 2.181 98.815 100.442 ot_mini 117 10.6 0.001 0.001 66.017 66.204 multiply_cannon_multrec 60168 15.6 32.966 34.728 41.709 43.264 qs_ot_get_derivative 117 11.6 0.001 0.001 41.100 41.319 rebuild_ks_matrix 128 8.3 0.001 0.001 33.214 33.470 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.213 33.470 mp_waitall_1 291448 16.2 28.814 31.963 28.814 31.963 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.895 30.110 multiply_cannon_sync_h2d 60168 15.6 27.369 29.774 27.369 29.774 qs_ot_get_p 128 10.4 0.001 0.001 27.313 27.566 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.298 24.973 apply_single 128 13.6 0.001 0.001 24.298 24.973 ot_diis_step 117 11.6 0.007 0.008 24.518 24.519 init_scf_loop 11 6.9 0.000 0.000 21.097 21.099 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.401 20.471 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.312 19.465 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.755 17.757 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 15.529 17.548 prepare_preconditioner 11 7.9 0.000 0.000 16.519 16.587 make_preconditioner 11 8.9 0.000 0.000 16.519 16.586 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.725 15.914 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.824 14.845 cp_fm_redistribute_end 83 14.4 11.721 14.758 11.741 14.763 cp_fm_diag_elpa_base 83 14.4 2.969 14.346 3.000 14.463 make_m2s 5014 13.6 0.101 0.109 14.096 14.409 make_images 5014 14.6 0.408 0.424 13.924 14.243 sum_up_and_integrate 128 10.3 0.092 0.110 13.993 14.007 integrate_v_rspace 128 11.3 0.004 0.004 13.901 13.916 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.722 12.808 calculate_rho_elec 128 8.7 0.045 0.064 12.721 12.807 init_scf_run 11 5.9 0.000 0.001 12.575 12.576 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.575 12.575 cp_fm_cholesky_invert 11 10.9 9.507 9.515 9.507 9.515 mp_sum_l 7870 13.0 8.414 9.514 8.414 9.514 wfi_extrapolate 11 7.9 0.001 0.001 9.007 9.007 dbcsr_mm_accdrv_process 124484 16.2 3.334 3.490 8.297 8.871 calculate_dm_sparse 128 9.5 0.001 0.001 8.448 8.547 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.947 8.074 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.952 8.016 make_images_data 5014 15.6 0.073 0.079 6.971 7.957 multiply_cannon_metrocomm1 60168 15.6 0.085 0.090 6.126 7.956 hybrid_alltoall_any 5200 16.5 0.291 2.252 6.113 7.579 grid_integrate_task_list 128 12.3 7.129 7.481 7.129 7.481 pw_transfer 1547 11.6 0.076 0.105 6.820 7.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.015 7.023 density_rs2pw 128 9.7 0.006 0.007 6.487 7.020 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.615 6.877 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.151 6.262 rs_pw_transfer 1046 11.9 0.018 0.020 5.679 6.180 fft_wrap_pw1pw2_140 523 13.2 0.446 0.515 5.683 5.864 fft3d_ps 1291 14.7 2.091 2.560 5.404 5.604 mp_alltoall_d11v 2415 14.1 4.356 5.457 4.356 5.457 grid_collocate_task_list 128 9.7 4.720 5.183 4.720 5.183 cp_fm_cholesky_decompose 22 10.9 4.760 4.772 4.760 4.772 potential_pw2rs 128 12.3 0.010 0.011 4.454 4.481 mp_sum_d 4459 12.1 3.641 4.480 3.641 4.480 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.555000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.181818, yerr=7.732084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 825.626624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5165060. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.043 194.439 194.440 qs_mol_dyn_low 1 2.0 0.003 0.006 194.005 194.025 qs_forces 11 3.9 0.003 0.006 193.500 193.501 qs_energies 11 4.9 0.007 0.045 186.782 186.794 scf_env_do_scf 11 5.9 0.001 0.001 167.795 167.805 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.741 134.744 velocity_verlet 10 3.0 0.001 0.002 120.206 120.216 dbcsr_multiply_generic 2507 12.6 0.180 0.186 100.718 101.958 qs_scf_new_mos 117 7.6 0.001 0.001 95.840 96.279 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.839 96.278 ot_scf_mini 117 9.6 0.004 0.005 90.918 91.375 multiply_cannon 2507 13.6 0.478 0.526 80.279 85.428 multiply_cannon_loop 2507 14.6 1.258 1.300 76.955 79.251 ot_mini 117 10.6 0.001 0.002 50.386 50.954 mp_waitall_1 226760 16.4 25.435 38.818 25.435 38.818 multiply_cannon_multrec 30084 15.6 22.156 26.336 32.867 37.474 rebuild_ks_matrix 128 8.3 0.001 0.001 33.030 33.438 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 33.029 33.438 init_scf_loop 11 6.9 0.000 0.002 32.965 32.968 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.788 30.166 multiply_cannon_metrocomm3 30084 15.6 0.096 0.100 15.809 29.199 qs_ot_get_derivative 117 11.6 0.001 0.002 28.386 28.880 prepare_preconditioner 11 7.9 0.000 0.000 28.542 28.600 make_preconditioner 11 8.9 0.000 0.001 28.542 28.600 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.263 27.790 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 21.999 23.198 apply_single 128 13.6 0.001 0.001 21.998 23.198 qs_ot_get_p 128 10.4 0.001 0.001 22.062 22.724 multiply_cannon_sync_h2d 30084 15.6 19.440 22.061 19.440 22.061 ot_diis_step 117 11.6 0.019 0.052 21.822 21.852 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.233 17.277 cp_fm_cholesky_invert 11 10.9 16.646 16.658 16.646 16.658 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.044 16.046 make_m2s 5014 13.6 0.089 0.095 14.491 15.990 make_images 5014 14.6 1.175 1.366 14.292 15.793 sum_up_and_integrate 128 10.3 0.117 0.134 14.576 14.613 integrate_v_rspace 128 11.3 0.003 0.004 14.459 14.503 init_scf_run 11 5.9 0.000 0.001 14.160 14.161 scf_env_initial_rho_setup 11 6.9 0.000 0.002 14.159 14.161 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.971 12.998 calculate_rho_elec 128 8.7 0.088 0.105 12.971 12.997 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.864 12.874 cp_fm_redistribute_end 83 14.4 7.528 12.794 7.542 12.800 cp_fm_diag_elpa_base 83 14.4 5.017 12.320 5.238 12.679 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.331 11.688 multiply_cannon_metrocomm4 27577 15.6 0.095 0.111 3.699 10.801 make_images_data 5014 15.6 0.068 0.074 8.786 10.741 dbcsr_mm_accdrv_process 62242 16.2 5.038 6.087 10.171 10.738 mp_irecv_dv 69486 16.3 3.510 10.421 3.510 10.421 hybrid_alltoall_any 5200 16.5 0.344 1.519 7.366 9.885 calculate_dm_sparse 128 9.5 0.001 0.001 8.400 8.544 wfi_extrapolate 11 7.9 0.002 0.006 8.479 8.479 pw_transfer 1547 11.6 0.085 0.100 7.851 7.924 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.627 7.697 grid_integrate_task_list 128 12.3 7.148 7.567 7.148 7.567 density_rs2pw 128 9.7 0.006 0.006 6.771 7.210 cp_fm_cholesky_decompose 22 10.9 6.996 7.086 6.996 7.086 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.310 7.028 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.780 6.791 fft_wrap_pw1pw2_140 523 13.2 0.471 0.518 6.662 6.746 mp_sum_l 7870 13.0 4.176 6.403 4.176 6.403 rs_pw_transfer 1046 11.9 0.014 0.017 5.818 6.274 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.913 6.065 fft3d_ps 1291 14.7 2.775 2.918 5.997 6.046 calculate_first_density_matrix 1 7.0 0.002 0.013 5.512 5.513 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.395 5.462 mp_allgather_i34 2507 14.6 1.891 5.293 1.891 5.293 grid_collocate_task_list 128 9.7 4.911 5.283 4.911 5.283 potential_pw2rs 128 12.3 0.015 0.016 5.051 5.083 mp_alltoall_d11v 2415 14.1 4.097 4.545 4.097 4.545 mp_sum_d 4459 12.1 2.714 4.217 2.714 4.217 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 4.063 4.094 dbcsr_complete_redistribute 395 12.7 0.772 0.895 3.172 4.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.440000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.181818, yerr=4.281422 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 933.900288E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.044 176.519 176.521 qs_mol_dyn_low 1 2.0 0.003 0.006 176.025 176.039 qs_forces 11 3.9 0.003 0.005 175.918 175.922 qs_energies 11 4.9 0.003 0.008 169.440 169.451 scf_env_do_scf 11 5.9 0.001 0.001 154.149 154.150 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.090 119.093 velocity_verlet 10 3.0 0.002 0.006 112.018 112.021 dbcsr_multiply_generic 2507 12.6 0.177 0.182 82.885 84.034 qs_scf_new_mos 117 7.6 0.001 0.001 82.606 83.053 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.605 83.052 ot_scf_mini 117 9.6 0.004 0.007 78.470 78.989 multiply_cannon 2507 13.6 0.500 0.521 62.757 67.097 multiply_cannon_loop 2507 14.6 0.860 0.888 59.923 62.612 ot_mini 117 10.6 0.001 0.002 44.014 44.528 mp_waitall_1 178456 16.5 26.770 36.489 26.770 36.489 init_scf_loop 11 6.9 0.000 0.002 34.958 34.965 prepare_preconditioner 11 7.9 0.000 0.000 30.947 31.005 make_preconditioner 11 8.9 0.000 0.000 30.947 31.005 rebuild_ks_matrix 128 8.3 0.001 0.001 30.057 30.648 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.057 30.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.542 29.968 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.175 27.717 multiply_cannon_multrec 20056 15.6 13.400 16.723 22.492 26.344 multiply_cannon_metrocomm3 20056 15.6 0.058 0.063 16.353 25.766 qs_ot_get_derivative 117 11.6 0.001 0.002 24.150 24.666 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.945 20.883 apply_single 128 13.6 0.001 0.001 19.945 20.883 qs_ot_get_p 128 10.4 0.001 0.002 19.830 20.433 ot_diis_step 117 11.6 0.019 0.023 19.761 19.763 multiply_cannon_sync_h2d 20056 15.6 14.158 15.986 14.158 15.986 make_m2s 5014 13.6 0.081 0.085 15.090 15.807 qs_ot_p2m_diag 83 11.4 0.266 0.273 15.619 15.631 make_images 5014 14.6 1.202 1.296 14.863 15.572 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.604 14.605 cp_fm_cholesky_invert 11 10.9 14.460 14.468 14.460 14.468 sum_up_and_integrate 128 10.3 0.135 0.147 14.191 14.218 integrate_v_rspace 128 11.3 0.004 0.004 14.056 14.086 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.950 12.974 calculate_rho_elec 128 8.7 0.131 0.147 12.950 12.973 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.474 11.475 cp_fm_redistribute_end 83 14.4 4.344 11.418 4.358 11.421 cp_fm_diag_elpa_base 83 14.4 6.631 10.795 7.044 11.331 make_images_data 5014 15.6 0.062 0.069 9.476 10.589 init_scf_run 11 5.9 0.000 0.001 10.311 10.311 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.311 10.311 hybrid_alltoall_any 5200 16.5 0.432 1.975 8.204 9.913 dbcsr_mm_accdrv_process 41502 16.2 4.834 6.470 8.546 9.640 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.903 9.287 multiply_cannon_metrocomm4 17549 15.6 0.062 0.073 3.474 9.280 mp_irecv_dv 50230 16.2 3.353 9.035 3.353 9.035 pw_transfer 1547 11.6 0.085 0.105 7.777 7.882 grid_integrate_task_list 128 12.3 7.327 7.787 7.327 7.787 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.552 7.661 cp_fm_upper_to_full 105 14.5 5.769 7.462 5.769 7.462 wfi_extrapolate 11 7.9 0.001 0.001 7.360 7.360 cp_fm_cholesky_decompose 22 10.9 7.246 7.269 7.246 7.269 density_rs2pw 128 9.7 0.006 0.006 6.445 6.889 fft_wrap_pw1pw2_140 523 13.2 0.480 0.528 6.632 6.753 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.869 6.516 dbcsr_complete_redistribute 395 12.7 1.175 1.210 4.645 6.384 fft3d_ps 1291 14.7 2.710 2.914 5.811 5.879 calculate_dm_sparse 128 9.5 0.001 0.001 5.737 5.824 rs_pw_transfer 1046 11.9 0.014 0.015 5.092 5.547 grid_collocate_task_list 128 9.7 5.078 5.482 5.078 5.482 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.409 5.414 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.454 5.188 mp_sum_l 7870 13.0 3.249 4.681 3.249 4.681 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.486 4.660 mp_alltoall_d11v 2415 14.1 4.202 4.647 4.202 4.647 mp_allgather_i34 2507 14.6 1.348 4.620 1.348 4.620 potential_pw2rs 128 12.3 0.021 0.023 4.590 4.612 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.385 4.102 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.044 4.071 mp_alltoall_i22 716 14.1 2.000 3.903 2.000 3.903 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 3.740 3.741 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.521000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.181818, yerr=9.749762 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.153692E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57136. MP_Allreduce 11261 1067. MP_Sync 168 MP_Alltoall 1700 12496371. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.136 187.139 187.140 qs_mol_dyn_low 1 2.0 0.005 0.014 186.638 186.653 qs_forces 11 3.9 0.009 0.051 186.479 186.497 qs_energies 11 4.9 0.004 0.017 179.403 179.413 scf_env_do_scf 11 5.9 0.001 0.002 162.250 162.261 velocity_verlet 10 3.0 0.002 0.004 123.288 123.293 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 114.965 114.967 qs_scf_new_mos 116 7.6 0.001 0.001 80.106 80.349 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.105 80.348 dbcsr_multiply_generic 2485 12.5 0.185 0.191 79.739 80.326 ot_scf_mini 116 9.6 0.004 0.005 75.631 75.932 multiply_cannon 2485 13.5 0.558 0.590 55.384 58.201 multiply_cannon_loop 2485 14.5 1.168 1.205 51.435 52.815 init_scf_loop 11 6.9 0.001 0.002 47.165 47.167 prepare_preconditioner 11 7.9 0.000 0.000 43.086 43.104 make_preconditioner 11 8.9 0.000 0.001 43.086 43.104 ot_mini 116 10.6 0.001 0.001 42.442 42.727 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.580 41.683 multiply_cannon_multrec 29820 15.5 14.138 19.393 26.486 31.253 rebuild_ks_matrix 127 8.3 0.001 0.001 28.718 28.929 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 28.718 28.928 mp_waitall_1 152434 16.5 17.253 26.691 17.253 26.691 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.901 26.094 qs_ot_get_derivative 116 11.6 0.001 0.002 22.938 23.242 make_m2s 4970 13.5 0.094 0.098 19.985 21.311 make_images 4970 14.5 1.976 2.303 19.685 21.012 qs_ot_get_p 127 10.4 0.001 0.002 19.533 19.842 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.959 19.423 apply_single 127 13.6 0.001 0.001 18.959 19.422 ot_diis_step 116 11.6 0.018 0.022 19.369 19.371 cp_fm_upper_to_full 104 14.7 11.243 16.682 11.243 16.682 cp_fm_cholesky_invert 11 10.9 16.479 16.489 16.479 16.489 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.387 15.665 qs_ot_p2m_diag 82 11.4 0.339 0.385 15.434 15.489 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.176 14.178 sum_up_and_integrate 127 10.3 0.141 0.154 13.940 13.977 integrate_v_rspace 127 11.3 0.004 0.004 13.799 13.841 dbcsr_complete_redistribute 393 12.7 1.517 1.644 9.051 12.920 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.781 12.814 calculate_rho_elec 127 8.7 0.174 0.189 12.780 12.813 multiply_cannon_sync_h2d 29820 15.5 11.685 12.584 11.685 12.584 dbcsr_mm_accdrv_process 61748 16.2 7.511 8.551 11.928 12.513 make_images_data 4970 15.5 0.064 0.068 10.670 12.257 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.713 11.541 init_scf_run 11 5.9 0.000 0.001 11.169 11.171 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.169 11.170 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.162 11.165 cp_fm_redistribute_end 82 14.4 1.912 11.094 1.926 11.100 hybrid_alltoall_any 5155 16.4 0.525 2.203 9.594 10.994 cp_fm_diag_elpa_base 82 14.4 8.550 10.490 9.139 10.978 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.484 10.211 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.240 9.448 mp_alltoall_i22 712 14.1 5.649 9.391 5.649 9.391 cp_fm_cholesky_decompose 22 10.9 7.827 7.919 7.827 7.919 grid_integrate_task_list 127 12.3 7.498 7.836 7.498 7.836 wfi_extrapolate 11 7.9 0.001 0.001 7.620 7.620 pw_transfer 1535 11.6 0.085 0.100 7.522 7.619 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 7.298 7.399 multiply_cannon_metrocomm4 24850 15.5 0.073 0.083 2.710 7.068 mp_irecv_dv 75445 16.2 2.570 6.808 2.570 6.808 fft_wrap_pw1pw2_140 519 13.2 0.479 0.487 6.444 6.575 calculate_dm_sparse 127 9.5 0.001 0.001 6.473 6.563 density_rs2pw 127 9.7 0.005 0.006 6.007 6.239 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.637 5.681 fft3d_ps 1281 14.7 2.775 2.864 5.536 5.625 grid_collocate_task_list 127 9.7 5.167 5.435 5.167 5.435 mp_alltoall_d11v 2401 14.1 4.584 5.280 4.584 5.280 rs_pw_transfer 1038 11.9 0.013 0.014 4.616 4.962 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.788 4.879 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.577 4.634 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.441 4.444 potential_pw2rs 127 12.3 0.023 0.023 4.281 4.315 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.152 4.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.140000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.545455, yerr=20.123996 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.529872E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838186. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.038 171.596 171.596 qs_mol_dyn_low 1 2.0 0.003 0.004 170.749 170.760 qs_forces 11 3.9 0.003 0.003 170.640 170.645 qs_energies 11 4.9 0.028 0.054 163.352 163.359 scf_env_do_scf 11 5.9 0.001 0.001 145.866 145.885 velocity_verlet 10 3.0 0.026 0.027 111.644 111.648 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.994 109.995 dbcsr_multiply_generic 2507 12.6 0.177 0.182 74.599 75.150 qs_scf_new_mos 117 7.6 0.001 0.001 74.264 74.352 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.263 74.352 ot_scf_mini 117 9.6 0.004 0.004 69.859 69.943 multiply_cannon 2507 13.6 0.592 0.634 54.874 58.943 multiply_cannon_loop 2507 14.6 0.447 0.459 50.381 51.263 ot_mini 117 10.6 0.001 0.001 39.633 39.702 init_scf_loop 11 6.9 0.000 0.000 35.723 35.724 mp_waitall_1 129618 16.6 26.177 33.543 26.177 33.543 prepare_preconditioner 11 7.9 0.000 0.000 31.874 31.903 make_preconditioner 11 8.9 0.000 0.000 31.874 31.903 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.741 30.034 rebuild_ks_matrix 128 8.3 0.001 0.001 28.595 28.675 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.595 28.675 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.143 26.219 multiply_cannon_multrec 10028 15.6 10.450 14.672 18.064 21.032 qs_ot_get_derivative 117 11.6 0.002 0.002 19.954 20.044 ot_diis_step 117 11.6 0.020 0.021 19.597 19.598 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.077 19.439 apply_single 128 13.6 0.001 0.001 19.077 19.439 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.197 19.122 cp_fm_cholesky_invert 11 10.9 18.205 18.210 18.205 18.210 make_m2s 5014 13.6 0.066 0.070 16.113 18.027 make_images 5014 14.6 2.298 2.884 15.811 17.726 qs_ot_get_p 128 10.4 0.001 0.001 17.374 17.476 sum_up_and_integrate 128 10.3 0.182 0.192 14.080 14.129 integrate_v_rspace 128 11.3 0.004 0.004 13.898 13.955 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.577 13.596 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.217 13.252 calculate_rho_elec 128 8.7 0.259 0.269 13.216 13.251 make_images_data 5014 15.6 0.053 0.062 9.793 12.455 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.440 12.442 multiply_cannon_sync_h2d 10028 15.6 11.617 12.013 11.617 12.013 hybrid_alltoall_any 5200 16.5 0.816 3.581 9.590 11.857 init_scf_run 11 5.9 0.000 0.001 10.660 10.660 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.659 10.660 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.455 9.458 cp_fm_diag_elpa_base 83 14.4 9.211 9.287 9.452 9.455 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.149 8.209 cp_fm_cholesky_decompose 22 10.9 8.022 8.109 8.022 8.109 grid_integrate_task_list 128 12.3 7.782 8.046 7.782 8.046 dbcsr_mm_accdrv_process 20762 16.1 2.708 3.431 7.242 7.868 pw_transfer 1547 11.6 0.084 0.091 7.799 7.808 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.578 7.592 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.687 7.507 wfi_extrapolate 11 7.9 0.001 0.001 7.238 7.238 mp_allgather_i34 2507 14.6 2.686 6.796 2.686 6.796 fft_wrap_pw1pw2_140 523 13.2 0.505 0.524 6.662 6.678 density_rs2pw 128 9.7 0.005 0.006 5.995 6.281 calculate_dm_sparse 128 9.5 0.001 0.001 6.073 6.166 fft3d_ps 1291 14.7 2.732 2.814 5.732 5.767 grid_collocate_task_list 128 9.7 5.491 5.723 5.491 5.723 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.631 5.639 dbcsr_complete_redistribute 395 12.7 2.119 2.202 5.237 5.608 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.223 5.223 mp_alltoall_d11v 2415 14.1 4.775 5.163 4.775 5.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.565 4.590 rs_pw_transfer 1046 11.9 0.013 0.013 4.107 4.406 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.860 4.278 mp_irecv_dv 28860 15.9 1.824 4.199 1.824 4.199 potential_pw2rs 128 12.3 0.027 0.028 4.080 4.100 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.554 3.865 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.412 3.719 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.654 3.693 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.546 3.595 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.423 3.459 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.596000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1431.090909, yerr=57.387763 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.083932E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.045 293.083 293.084 qs_mol_dyn_low 1 2.0 0.003 0.004 292.547 292.558 qs_forces 11 3.9 0.003 0.003 292.451 292.454 qs_energies 11 4.9 0.002 0.002 283.837 283.847 scf_env_do_scf 11 5.9 0.001 0.001 261.013 261.027 velocity_verlet 10 3.0 0.003 0.003 210.830 210.839 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.688 135.688 init_scf_loop 11 6.9 0.000 0.000 125.073 125.074 prepare_preconditioner 11 7.9 0.000 0.000 120.416 120.447 make_preconditioner 11 8.9 0.000 0.000 120.416 120.447 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.198 117.524 qs_scf_new_mos 118 7.6 0.001 0.001 93.058 93.149 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.057 93.147 ot_scf_mini 118 9.6 0.004 0.004 88.190 88.210 dbcsr_multiply_generic 2535 12.6 0.211 0.218 85.985 86.567 cp_fm_upper_to_full 106 14.8 53.176 76.303 53.176 76.303 multiply_cannon 2535 13.6 0.718 0.762 60.676 61.359 multiply_cannon_loop 2535 14.6 0.477 0.484 56.851 58.417 ot_mini 118 10.6 0.001 0.001 45.302 45.331 dbcsr_complete_redistribute 397 12.7 4.013 4.056 29.924 43.262 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.525 39.812 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.172 37.271 mp_alltoall_i22 720 14.1 21.963 35.065 21.963 35.065 rebuild_ks_matrix 129 8.3 0.001 0.001 34.001 34.057 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.001 34.056 cp_fm_cholesky_invert 11 10.9 33.778 33.785 33.778 33.785 mp_waitall_1 106626 16.7 28.984 32.622 28.984 32.622 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.767 31.820 qs_ot_get_p 129 10.4 0.001 0.001 26.987 27.073 qs_ot_get_derivative 118 11.6 0.002 0.002 24.526 24.552 qs_ot_p2m_diag 84 11.4 0.890 0.897 22.813 22.843 make_m2s 5070 13.6 0.078 0.079 20.565 21.630 make_images 5070 14.6 3.826 3.947 20.088 21.153 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.983 20.985 ot_diis_step 118 11.6 0.022 0.023 20.741 20.741 multiply_cannon_metrocomm3 10140 15.6 0.023 0.024 18.987 20.332 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.943 20.160 apply_single 129 13.6 0.001 0.001 19.943 20.160 multiply_cannon_multrec 10140 15.6 10.715 12.531 18.442 18.552 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.653 17.654 cp_fm_diag_elpa_base 84 14.4 13.261 14.847 17.648 17.650 sum_up_and_integrate 129 10.3 0.324 0.326 15.932 16.019 multiply_cannon_sync_h2d 10140 15.6 15.787 15.823 15.787 15.823 integrate_v_rspace 129 11.3 0.004 0.004 15.608 15.695 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.079 15.103 calculate_rho_elec 129 8.7 0.488 0.488 15.079 15.102 hybrid_alltoall_any 5257 16.5 1.309 3.052 11.164 13.054 make_images_data 5070 15.6 0.061 0.065 11.121 13.010 init_scf_run 11 5.9 0.000 0.001 12.922 12.922 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.921 12.922 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.597 9.619 dbcsr_mm_accdrv_process 20958 16.1 3.898 5.887 7.488 9.411 cp_fm_cholesky_decompose 22 10.9 8.969 9.001 8.969 9.001 grid_integrate_task_list 129 12.3 8.603 8.779 8.603 8.779 wfi_extrapolate 11 7.9 0.001 0.001 8.617 8.617 pw_transfer 1559 11.6 0.092 0.093 8.474 8.481 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.237 8.247 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.940 7.941 calculate_dm_sparse 129 9.5 0.001 0.001 7.800 7.892 fft_wrap_pw1pw2_140 527 13.2 0.544 0.546 7.322 7.338 mp_alltoall_d11v 2429 14.1 7.131 7.271 7.131 7.271 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.480 6.542 grid_collocate_task_list 129 9.7 6.388 6.426 6.388 6.426 fft3d_ps 1301 14.7 2.782 2.798 6.263 6.268 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.058 6.122 density_rs2pw 129 9.7 0.005 0.005 6.044 6.078 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.084000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2722.727273, yerr=188.628587 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 1.260155E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.063 85.501 85.502 qs_energies 1 2.0 0.000 0.000 84.989 84.996 ls_scf 1 3.0 0.006 0.048 84.084 84.091 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.670 72.848 multiply_cannon 111 7.7 0.017 0.020 55.880 57.113 multiply_cannon_loop 111 8.7 0.213 0.233 52.512 53.926 ls_scf_main 1 4.0 0.000 0.002 52.418 52.434 density_matrix_trs4 2 5.0 0.002 0.003 46.882 46.942 ls_scf_init_scf 1 4.0 0.000 0.000 28.332 28.334 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.154 27.212 mp_waitall_1 11316 10.9 22.354 25.040 22.354 25.040 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.011 25.029 multiply_cannon_multrec 2664 9.7 8.193 8.989 15.615 17.406 multiply_cannon_sync_h2d 2664 9.7 13.651 15.543 13.651 15.543 make_m2s 222 7.7 0.009 0.012 13.084 13.554 make_images 222 8.7 0.102 0.112 13.062 13.533 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.635 12.029 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.417 8.626 make_images_data 222 9.7 0.005 0.006 7.630 8.197 dbcsr_mm_accdrv_process 4760 10.4 0.515 0.612 7.040 7.993 hybrid_alltoall_any 227 10.6 0.219 1.855 6.594 7.807 dbcsr_mm_accdrv_process_sort 4760 11.4 6.325 7.212 6.325 7.212 calculate_norms 4752 9.8 5.528 6.342 5.528 6.342 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.261 mp_sum_l 807 5.4 3.190 4.321 3.190 4.321 make_images_sizes 222 9.7 0.000 0.000 0.715 3.505 mp_alltoall_i44 222 10.7 0.715 3.505 0.715 3.505 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.064 3.429 mp_irecv_dv 6231 10.9 2.047 3.406 2.047 3.406 ls_scf_post 1 4.0 0.003 0.049 3.327 3.336 arnoldi_extremal 4 6.8 0.000 0.000 3.285 3.309 arnoldi_normal_ev 4 7.8 0.001 0.002 3.284 3.309 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.367 3.184 build_subspace 16 8.4 0.009 0.012 3.152 3.162 ls_scf_store_result 1 5.0 0.000 0.000 2.877 2.914 dbcsr_special_finalize 555 9.7 0.005 0.006 2.374 2.758 dbcsr_merge_single_wm 555 10.7 0.452 0.591 2.365 2.749 make_images_pack 222 9.7 2.215 2.637 2.217 2.639 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.328 2.571 dbcsr_sort_data 658 11.4 2.170 2.487 2.170 2.487 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.451 2.063 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.249 2.331 buffer_matrices_ensure_size 222 8.7 1.763 2.046 1.763 2.046 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.749 1.752 rebuild_ks_matrix 3 7.3 0.000 0.000 1.739 1.742 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.739 1.742 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.502000, 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/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 2.126410E+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 0.028 0.043 91.807 91.808 qs_energies 1 2.0 0.000 0.000 91.310 91.314 ls_scf 1 3.0 0.000 0.000 89.985 89.988 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.003 76.394 multiply_cannon 111 7.7 0.030 0.055 53.353 57.553 ls_scf_main 1 4.0 0.000 0.000 55.403 55.408 multiply_cannon_loop 111 8.7 0.117 0.123 50.076 53.441 density_matrix_trs4 2 5.0 0.002 0.003 49.730 49.926 ls_scf_init_scf 1 4.0 0.000 0.001 30.996 30.997 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.852 29.951 mp_waitall_1 9246 10.9 21.356 29.842 21.356 29.842 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.418 27.433 multiply_cannon_multrec 1332 9.7 13.189 15.905 22.550 26.333 multiply_cannon_metrocomm3 1332 9.7 0.007 0.009 11.680 20.886 make_m2s 222 7.7 0.007 0.008 15.693 16.351 make_images 222 8.7 1.589 1.978 15.663 16.322 dbcsr_mm_accdrv_process 4041 10.4 0.297 0.450 8.958 10.560 dbcsr_mm_accdrv_process_sort 4041 11.4 8.549 10.110 8.549 10.110 make_images_data 222 9.7 0.004 0.005 9.073 9.968 hybrid_alltoall_any 227 10.6 0.524 2.491 8.435 9.442 mp_sum_l 807 5.4 5.640 8.605 5.640 8.605 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.225 7.797 mp_irecv_dv 3311 11.0 3.206 7.738 3.206 7.738 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.382 7.004 calculate_norms 2376 9.8 6.012 6.694 6.012 6.694 multiply_cannon_sync_h2d 1332 9.7 4.809 6.067 4.809 6.067 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.035 5.228 arnoldi_extremal 4 6.8 0.000 0.001 4.671 4.697 arnoldi_normal_ev 4 7.8 0.001 0.005 4.671 4.696 build_subspace 16 8.4 0.014 0.021 4.417 4.419 ls_scf_post 1 4.0 0.000 0.000 3.585 3.588 ls_scf_store_result 1 5.0 0.000 0.000 3.262 3.401 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.194 3.391 dbcsr_matrix_vector_mult_local 304 10.0 2.778 3.238 2.780 3.240 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.218 2.822 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.551 2.652 mp_allgather_i34 111 8.7 0.968 2.449 0.968 2.449 make_images_pack 222 9.7 2.032 2.424 2.034 2.427 dbcsr_sort_data 436 11.2 1.847 2.075 1.847 2.075 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.834 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.808000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1708.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 2.672316E+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 0.074 0.176 95.159 95.160 qs_energies 1 2.0 0.000 0.000 94.087 94.101 ls_scf 1 3.0 0.001 0.006 92.645 92.659 dbcsr_multiply_generic 111 6.7 0.017 0.026 77.410 77.706 ls_scf_main 1 4.0 0.000 0.001 57.713 57.719 multiply_cannon 111 7.7 0.038 0.067 53.382 56.941 multiply_cannon_loop 111 8.7 0.100 0.106 49.798 53.849 density_matrix_trs4 2 5.0 0.002 0.003 51.712 51.902 mp_waitall_1 7374 11.0 24.065 34.330 24.065 34.330 ls_scf_init_scf 1 4.0 0.000 0.001 31.275 31.277 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.144 30.204 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.378 27.391 multiply_cannon_multrec 888 9.7 12.754 15.424 21.506 24.694 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.080 23.548 make_m2s 222 7.7 0.007 0.008 17.071 18.333 make_images 222 8.7 1.987 2.332 17.032 18.294 make_images_data 222 9.7 0.004 0.005 9.799 10.874 hybrid_alltoall_any 227 10.6 0.623 2.872 9.486 10.568 dbcsr_mm_accdrv_process 3754 10.4 0.261 0.722 8.284 9.410 dbcsr_mm_accdrv_process_sort 3754 11.4 7.811 8.995 7.811 8.995 mp_sum_l 807 5.4 5.385 8.829 5.385 8.829 multiply_cannon_sync_h2d 888 9.7 6.048 7.486 6.048 7.486 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.055 mp_irecv_dv 2335 11.1 2.455 6.995 2.455 6.995 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.808 6.806 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.076 6.749 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.359 5.572 arnoldi_extremal 4 6.8 0.000 0.002 5.135 5.156 arnoldi_normal_ev 4 7.8 0.002 0.007 5.135 5.154 build_subspace 16 8.4 0.015 0.020 4.806 4.813 calculate_norms 1584 9.8 4.253 4.612 4.253 4.612 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.499 3.797 mp_allgather_i34 111 8.7 1.415 3.780 1.415 3.780 ls_scf_post 1 4.0 0.005 0.019 3.656 3.670 dbcsr_matrix_vector_mult_local 304 10.0 3.067 3.618 3.069 3.620 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.474 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.806 2.941 make_images_pack 222 9.7 1.812 2.130 1.815 2.133 dbcsr_sort_data 325 11.1 1.881 2.085 1.881 2.085 make_images_sizes 222 9.7 0.000 0.000 0.933 2.034 mp_alltoall_i44 222 10.7 0.932 2.033 0.932 2.033 dbcsr_data_release 9322 10.9 1.319 1.997 1.319 1.997 dbcsr_finalize 304 7.8 0.027 0.033 1.623 1.908 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.160000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2210.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 3.339751E+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 0.046 0.072 97.022 97.023 qs_energies 1 2.0 0.000 0.000 96.327 96.333 ls_scf 1 3.0 0.000 0.002 94.683 94.688 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.400 78.614 ls_scf_main 1 4.0 0.000 0.001 58.857 58.860 multiply_cannon 111 7.7 0.048 0.123 51.629 56.029 density_matrix_trs4 2 5.0 0.002 0.003 52.784 52.908 multiply_cannon_loop 111 8.7 0.115 0.125 46.490 49.556 ls_scf_init_scf 1 4.0 0.000 0.002 32.562 32.566 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.397 31.485 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.930 28.942 mp_waitall_1 6438 11.0 22.430 28.898 22.430 28.898 multiply_cannon_multrec 1332 9.7 14.303 17.261 22.308 24.983 make_m2s 222 7.7 0.008 0.009 21.081 22.570 make_images 222 8.7 3.153 3.620 21.029 22.521 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.002 16.245 make_images_data 222 9.7 0.004 0.005 11.778 13.495 hybrid_alltoall_any 227 10.6 0.798 3.766 11.137 12.995 dbcsr_mm_accdrv_process 3641 10.4 0.222 0.411 7.640 9.183 dbcsr_mm_accdrv_process_sort 3641 11.4 7.264 8.766 7.264 8.766 mp_sum_l 807 5.4 4.114 7.508 4.114 7.508 multiply_cannon_sync_h2d 1332 9.7 5.507 6.323 5.507 6.323 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.117 5.882 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.069 5.863 mp_irecv_dv 3229 10.9 2.046 5.820 2.046 5.820 arnoldi_extremal 4 6.8 0.000 0.002 5.383 5.396 arnoldi_normal_ev 4 7.8 0.002 0.006 5.383 5.396 build_subspace 16 8.4 0.015 0.022 5.030 5.043 mp_allgather_i34 111 8.7 2.245 4.913 2.245 4.913 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.479 4.903 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.577 4.773 calculate_norms 2376 9.8 4.187 4.539 4.187 4.539 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.655 3.964 dbcsr_matrix_vector_mult_local 304 10.0 3.245 3.717 3.247 3.719 dbcsr_sort_data 658 11.4 3.090 3.388 3.090 3.388 ls_scf_post 1 4.0 0.000 0.001 3.263 3.269 dbcsr_special_finalize 555 9.7 0.006 0.006 2.845 3.170 dbcsr_merge_single_wm 555 10.7 0.538 0.652 2.837 3.162 ls_scf_store_result 1 5.0 0.000 0.000 3.010 3.064 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.932 2.980 dbcsr_data_release 10477 10.7 1.600 2.470 1.600 2.470 dbcsr_finalize 304 7.8 0.050 0.062 1.807 1.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.023000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 4.643361E+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 0.041 0.057 100.185 100.187 qs_energies 1 2.0 0.000 0.000 99.408 99.414 ls_scf 1 3.0 0.000 0.000 97.490 97.495 dbcsr_multiply_generic 111 6.7 0.018 0.019 79.063 79.281 ls_scf_main 1 4.0 0.000 0.000 62.746 62.747 multiply_cannon 111 7.7 0.095 0.195 56.096 60.890 density_matrix_trs4 2 5.0 0.002 0.003 55.730 55.811 multiply_cannon_loop 111 8.7 0.070 0.082 51.597 53.818 mp_waitall_1 5481 11.0 27.374 32.530 27.374 32.530 ls_scf_init_scf 1 4.0 0.000 0.000 31.091 31.097 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.881 29.908 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.693 27.707 multiply_cannon_multrec 444 9.7 14.136 16.508 21.278 24.675 make_m2s 222 7.7 0.006 0.006 17.868 20.289 make_images 222 8.7 3.733 4.454 17.805 20.227 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 12.444 17.149 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.558 15.281 make_images_data 222 9.7 0.003 0.004 10.173 12.428 hybrid_alltoall_any 227 10.6 0.789 3.761 9.882 12.265 multiply_cannon_sync_h2d 444 9.7 6.514 8.291 6.514 8.291 dbcsr_mm_accdrv_process 3003 10.4 0.184 0.394 6.851 7.997 dbcsr_mm_accdrv_process_sort 3003 11.4 6.523 7.647 6.523 7.647 mp_allgather_i34 111 8.7 2.679 7.079 2.679 7.079 arnoldi_extremal 4 6.8 0.000 0.000 5.963 5.976 arnoldi_normal_ev 4 7.8 0.001 0.004 5.963 5.976 build_subspace 16 8.4 0.015 0.020 5.535 5.544 mp_sum_l 807 5.4 3.129 5.189 3.129 5.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.609 4.750 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.213 4.397 dbcsr_matrix_vector_mult_local 304 10.0 3.730 4.168 3.732 4.170 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.227 4.049 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.581 3.846 mp_irecv_dv 1241 11.2 1.562 3.800 1.562 3.800 calculate_norms 792 9.8 3.532 3.701 3.532 3.701 ls_scf_post 1 4.0 0.000 0.000 3.652 3.658 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.480 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.387 3.470 make_images_sizes 222 9.7 0.000 0.000 1.016 3.397 mp_alltoall_i44 222 10.7 1.016 3.396 1.016 3.396 dbcsr_finalize 304 7.8 0.062 0.078 2.214 2.336 dbcsr_merge_all 275 8.9 0.476 0.533 2.061 2.167 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.187000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3614.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4f847e0ea5227630c4e07962f9d27e060298d5a_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 8.717562E+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 0.095 0.147 107.564 107.571 qs_energies 1 2.0 0.000 0.000 106.044 106.059 ls_scf 1 3.0 0.000 0.000 103.159 103.172 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.258 77.389 ls_scf_main 1 4.0 0.000 0.000 65.657 65.658 density_matrix_trs4 2 5.0 0.002 0.003 56.828 56.892 multiply_cannon 111 7.7 0.133 0.262 49.304 50.804 multiply_cannon_loop 111 8.7 0.068 0.070 45.868 47.060 ls_scf_init_scf 1 4.0 0.000 0.000 33.826 33.827 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.500 32.516 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.675 29.682 mp_waitall_1 4569 11.1 21.575 25.658 21.575 25.658 make_m2s 222 7.7 0.007 0.007 24.029 24.933 make_images 222 8.7 4.594 5.017 23.919 24.821 multiply_cannon_multrec 444 9.7 17.939 18.669 22.618 23.212 hybrid_alltoall_any 227 10.6 1.660 3.619 12.931 15.677 make_images_data 222 9.7 0.003 0.004 13.194 15.531 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.802 10.331 multiply_cannon_sync_h2d 444 9.7 8.846 8.886 8.846 8.886 arnoldi_extremal 4 6.8 0.000 0.000 7.494 7.504 arnoldi_normal_ev 4 7.8 0.002 0.009 7.494 7.504 build_subspace 16 8.4 0.026 0.036 6.949 6.961 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.589 5.696 dbcsr_matrix_vector_mult_local 304 10.0 5.101 5.371 5.104 5.373 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.022 5.290 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.778 4.871 dbcsr_mm_accdrv_process 1814 10.4 0.233 0.324 4.509 4.632 dbcsr_mm_accdrv_process_sort 1814 11.4 4.208 4.338 4.208 4.338 ls_scf_post 1 4.0 0.000 0.000 3.675 3.690 make_images_sizes 222 9.7 0.000 0.000 1.461 3.648 mp_alltoall_i44 222 10.7 1.461 3.648 1.461 3.648 ls_scf_store_result 1 5.0 0.000 0.000 3.426 3.438 calculate_norms 792 9.8 3.241 3.280 3.241 3.280 mp_allgather_i34 111 8.7 1.043 3.196 1.043 3.196 dbcsr_finalize 304 7.8 0.083 0.090 3.102 3.173 dbcsr_merge_all 275 8.9 0.892 0.913 2.874 2.922 dbcsr_complete_redistribute 5 7.6 1.449 1.478 2.784 2.874 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.855 2.855 dbcsr_data_release 12724 10.6 2.347 2.850 2.347 2.850 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.529 dbcsr_sort_data 325 11.1 2.441 2.513 2.441 2.513 mp_sum_l 807 5.4 1.601 2.459 1.601 2.459 dbcsr_new_transposed 4 7.5 0.244 0.252 2.272 2.290 dbcsr_add_d 103 6.2 0.000 0.000 2.165 2.237 dbcsr_add_anytype 103 7.2 0.860 0.892 2.165 2.237 dbcsr_frobenius_norm 74 6.6 2.055 2.130 2.190 2.234 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.185 2.186 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.571000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6814.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e4f847e0ea5227630c4e07962f9d27e060298d5a Summary: empty Status: OK