=== This is the CP2K Performance-Test === Updating a07564ec1..99df2e665 Fast-forward arch/Linux-gnu-x86_64.psmp | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 99df2e665baa944797556ab754e3afa8e99d62d4 ################# 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/01 job id: 41585587 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/02 job id: 41585589 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/03 job id: 41585675 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/04 job id: 41585745 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/05 job id: 41585820 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/06 job id: 41585897 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/07 job id: 41585946 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/08 job id: 41585955 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/09 job id: 41585958 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/10 job id: 41585960 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/11 job id: 41585963 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/12 job id: 41585964 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/13 job id: 41585966 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/14 job id: 41585968 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/15 job id: 41585970 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/16 job id: 41585972 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/17 job id: 41585974 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/18 job id: 41585976 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/19 job id: 41585979 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/20 job id: 41585981 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/21 job id: 41585983 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/22 job id: 41585984 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/23 job id: 41585985 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/24 job id: 41585986 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/25 job id: 41585987 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/26 job id: 41585988 --- 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/99df2e665baa944797556ab754e3afa8e99d62d4_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 0.022 0.052 137.231 137.232 farming_run 1 2.0 134.703 134.716 137.185 137.188 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.451966E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 116.974 116.975 qs_energies 1 2.0 0.000 0.000 116.729 116.730 mp2_main 1 3.0 0.000 0.000 114.898 114.899 mp2_gpw_main 1 4.0 0.027 0.034 114.081 114.082 mp2_ri_gpw_compute_in 1 5.0 0.172 0.182 95.489 95.777 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 56.272 56.562 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.175 42.582 47.784 get_2c_integrals 1 6.0 0.000 0.000 38.171 39.045 integrate_v_rspace 273 8.0 0.434 0.447 25.227 30.220 pw_transfer 6555 10.6 0.377 0.396 27.612 28.258 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.255 26.655 grid_integrate_task_list 273 9.0 20.992 26.510 20.992 26.510 fft_wrap_pw1pw2_100 2178 12.4 1.201 1.428 23.814 24.236 compute_2c_integrals 1 7.0 0.021 0.021 20.585 20.586 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 19.682 20.291 mp2_eri_2c_integrate_gpw 1 9.0 2.373 2.434 19.679 20.288 rpa_ri_compute_en 1 5.0 0.000 0.000 18.474 18.639 cp_fm_cholesky_decompose 12 8.2 17.544 18.433 17.544 18.433 cholesky_decomp 1 7.0 0.000 0.000 16.425 17.317 fft3d_s 5443 13.4 16.303 16.805 16.325 16.825 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.591 10.853 15.591 calculate_wavefunction 272 8.0 5.396 5.556 12.570 13.213 rpa_num_int 1 6.0 0.000 0.000 10.585 10.594 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.532 10.532 calc_potential_gpw 544 9.5 0.004 0.005 9.621 10.125 calc_mat_Q 8 8.0 0.000 0.000 9.360 9.448 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.948 9.320 contract_S_to_Q 8 9.0 0.000 0.000 8.781 8.874 potential_pw2rs 545 10.0 0.106 0.108 7.809 8.526 parallel_gemm_fm 14 9.1 0.000 0.000 8.371 8.432 parallel_gemm_fm_cosma 14 10.1 8.371 8.432 8.371 8.432 collocate_single_gaussian 272 10.0 0.039 0.041 7.540 7.873 create_integ_mat 1 6.0 0.014 0.026 7.506 7.515 array2fm 1 7.0 0.000 0.000 6.596 6.975 pw_scatter_s 2720 13.7 4.408 4.559 4.408 4.559 pw_gather_s 2722 13.2 3.975 4.287 3.975 4.287 pw_poisson_solve 545 10.5 2.486 2.528 3.546 3.771 array2fm_buffer_send 1 8.0 2.975 3.152 2.975 3.152 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.080870, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 0.027 0.037 428.202 428.203 farming_run 1 2.0 427.310 427.314 428.169 428.171 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.226428E+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 13059059. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.048 210.237 210.237 qs_energies 1 2.0 0.000 0.000 210.007 210.013 scf_env_do_scf 1 3.0 0.000 0.000 106.357 106.358 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.501 105.508 rebuild_ks_matrix 4 6.0 0.000 0.000 105.484 105.492 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 105.484 105.492 hfx_ks_matrix 4 8.0 0.001 0.001 105.101 105.104 integrate_four_center 4 9.0 0.143 0.457 105.100 105.103 mp2_main 1 3.0 0.000 0.000 103.360 103.366 mp2_gpw_main 1 4.0 0.032 0.046 102.518 102.527 integrate_four_center_main 4 10.0 0.117 0.536 96.783 99.099 integrate_four_center_bin 263 11.0 96.666 99.087 96.666 99.087 init_scf_loop 1 4.0 0.000 0.000 92.113 92.113 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 75.804 76.783 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.917 55.896 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.162 42.752 47.491 integrate_v_rspace 95 8.0 0.397 0.565 28.686 33.236 pw_transfer 2240 10.6 0.147 0.177 30.167 30.647 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 29.125 29.564 grid_integrate_task_list 95 9.0 23.964 28.748 23.964 28.748 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.566 28.315 ao_to_mo_and_store_B_mult_1 91 7.0 10.474 27.514 10.474 27.514 fft_wrap_pw1pw2_100 730 12.4 1.309 1.492 26.850 27.332 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.917 24.942 24.951 get_2c_integrals 1 6.0 0.000 0.000 20.801 20.822 compute_2c_integrals 1 7.0 0.002 0.003 19.787 19.791 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 19.377 19.662 mp2_eri_2c_integrate_gpw 1 9.0 1.727 1.880 19.375 19.661 fft3d_s 1823 13.4 18.532 18.882 18.545 18.895 scf_env_do_scf_inner_loop 4 4.0 0.000 0.016 14.243 14.243 calculate_wavefunction 91 8.0 2.024 2.055 9.806 10.126 potential_pw2rs 186 10.0 0.033 0.034 8.698 9.469 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.583 8.785 9.213 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.690 9.107 calc_potential_gpw 182 9.5 0.002 0.002 8.690 8.973 local_gemm 172 8.0 8.229 8.640 8.229 8.640 mp2_ri_gpw_compute_en_comm 22 7.0 0.496 0.517 7.924 8.345 collocate_single_gaussian 91 10.0 0.016 0.021 7.951 8.159 mp_sync 38 10.4 2.994 6.533 2.994 6.533 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.443 6.343 6.443 mp_sendrecv_dm3 2068 8.0 5.964 6.400 5.964 6.400 pw_gather_s 912 13.2 4.938 5.610 4.938 5.610 pw_scatter_s 910 13.7 3.980 4.484 3.980 4.484 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.519416, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 449.765376E+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 0.017 0.047 85.322 85.328 qs_mol_dyn_low 1 2.0 0.004 0.013 84.935 84.946 qs_forces 11 3.9 0.010 0.035 84.816 84.818 qs_energies 11 4.9 0.003 0.017 83.280 83.304 dbcsr_multiply_generic 2286 12.5 0.096 0.101 66.118 66.503 multiply_cannon 2286 13.5 0.184 0.190 58.580 60.020 multiply_cannon_loop 2286 14.5 1.507 1.580 57.904 59.353 scf_env_do_scf 11 5.9 0.000 0.001 44.414 44.415 scf_env_do_scf_inner_loop 108 6.5 0.007 0.049 42.297 42.297 init_scf_run 11 5.9 0.000 0.001 37.405 37.405 scf_env_initial_rho_setup 11 6.9 0.001 0.010 37.404 37.405 calculate_first_density_matrix 1 7.0 0.002 0.011 35.198 35.217 qs_scf_new_mos 108 7.5 0.000 0.001 32.094 32.396 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.094 32.396 ot_scf_mini 108 9.5 0.002 0.006 30.456 30.639 mp_waitall_1 267858 16.1 25.519 29.981 25.519 29.981 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 21.528 27.930 velocity_verlet 10 3.0 0.010 0.035 26.097 26.098 calculate_dm_sparse 119 9.5 0.001 0.001 20.493 20.644 ot_mini 108 10.5 0.001 0.004 19.730 19.983 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 3.219 19.917 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 19.018 19.046 acc_transpose_blocks 54864 15.5 0.225 0.244 17.063 18.737 acc_transpose_blocks_kernels 54864 16.5 0.255 0.389 16.081 17.928 jit_kernel_transpose 5 15.5 15.827 17.896 15.827 17.896 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 17.223 17.227 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 16.890 16.948 qs_ot_get_derivative 108 11.5 0.001 0.006 16.719 16.907 multiply_cannon_multrec 54864 15.5 4.182 6.473 8.023 11.678 rebuild_ks_matrix 119 8.3 0.000 0.000 8.066 8.197 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.014 8.065 8.197 multiply_cannon_sync_h2d 54864 15.5 5.870 7.680 5.870 7.680 mp_sum_l 7207 12.9 5.726 7.345 5.726 7.345 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.130 7.254 qs_ot_get_p 119 10.4 0.001 0.001 6.211 6.495 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.564 6.055 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.537 5.654 dbcsr_mm_accdrv_process 76910 16.1 1.179 1.832 3.763 5.338 sum_up_and_integrate 119 10.3 0.009 0.021 4.662 4.682 integrate_v_rspace 119 11.3 0.002 0.002 4.652 4.663 qs_rho_update_rho_low 119 7.7 0.001 0.008 4.131 4.225 calculate_rho_elec 119 8.7 0.011 0.016 4.130 4.225 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.401 3.454 jit_kernel_multiply 13 15.8 2.523 3.424 2.523 3.424 rs_pw_transfer 974 11.9 0.012 0.013 3.024 3.135 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.909 3.093 apply_single 119 13.6 0.000 0.000 2.908 3.093 ot_diis_step 108 11.5 0.006 0.009 2.713 2.713 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.582 2.582 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.347 2.417 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.388 2.389 cp_fm_redistribute_end 50 14.0 2.170 2.367 2.175 2.369 cp_fm_diag_elpa_base 50 14.0 0.192 2.327 0.193 2.336 density_rs2pw 119 9.7 0.004 0.004 2.188 2.288 wfi_extrapolate 11 7.9 0.001 0.007 2.134 2.134 grid_integrate_task_list 119 12.3 2.025 2.117 2.025 2.117 init_scf_loop 11 6.9 0.000 0.000 2.101 2.101 mp_sum_d 4129 12.0 1.461 2.063 1.461 2.063 potential_pw2rs 119 12.3 0.004 0.004 2.037 2.049 pw_transfer 1439 11.6 0.051 0.057 1.679 1.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=85.328000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.657472E+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 389620. 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 0.050 0.116 40.132 40.134 qs_mol_dyn_low 1 2.0 0.007 0.016 39.477 39.518 qs_forces 11 3.9 0.007 0.026 39.336 39.341 qs_energies 11 4.9 0.009 0.034 37.636 37.645 scf_env_do_scf 11 5.9 0.001 0.004 31.653 31.655 scf_env_do_scf_inner_loop 108 6.5 0.012 0.076 28.833 28.834 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.504 21.904 qs_scf_new_mos 108 7.5 0.001 0.001 19.837 20.100 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.836 20.100 ot_scf_mini 108 9.5 0.003 0.007 18.940 19.128 velocity_verlet 10 3.0 0.005 0.012 18.307 18.312 multiply_cannon 2286 13.5 0.216 0.233 16.369 18.005 multiply_cannon_loop 2286 14.5 0.904 0.978 15.221 16.613 ot_mini 108 10.5 0.005 0.021 11.843 12.091 mp_waitall_1 217478 16.2 5.980 11.278 5.980 11.278 qs_ot_get_derivative 108 11.5 0.002 0.023 9.302 9.493 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.144 9.479 multiply_cannon_multrec 27432 15.5 1.967 4.540 5.903 8.675 rebuild_ks_matrix 119 8.3 0.000 0.000 7.221 7.383 qs_ks_build_kohn_sham_matrix 119 9.3 0.022 0.088 7.221 7.383 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.423 6.577 dbcsr_mm_accdrv_process 47894 16.0 3.047 5.262 3.867 5.777 init_scf_run 11 5.9 0.000 0.001 4.500 4.501 scf_env_initial_rho_setup 11 6.9 0.005 0.021 4.500 4.501 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.617 4.466 qs_ot_get_p 119 10.4 0.002 0.008 4.177 4.416 sum_up_and_integrate 119 10.3 0.021 0.063 4.218 4.225 integrate_v_rspace 119 11.3 0.002 0.003 4.198 4.211 mp_sum_l 7207 12.9 2.180 4.202 2.180 4.202 apply_preconditioner_dbcsr 119 12.6 0.001 0.008 3.065 4.126 apply_single 119 13.6 0.000 0.000 3.064 4.126 qs_rho_update_rho_low 119 7.7 0.008 0.043 3.695 3.767 calculate_rho_elec 119 8.7 0.021 0.024 3.687 3.766 calculate_first_density_matrix 1 7.0 0.008 0.032 2.943 2.945 multiply_cannon_sync_h2d 27432 15.5 2.186 2.926 2.186 2.926 init_scf_loop 11 6.9 0.003 0.019 2.797 2.800 make_m2s 4572 13.5 0.052 0.054 2.547 2.788 qs_ot_p2m_diag 50 11.0 0.010 0.025 2.734 2.753 rs_pw_transfer 974 11.9 0.010 0.011 2.656 2.738 make_images 4572 14.5 0.200 0.237 2.458 2.696 ot_diis_step 108 11.5 0.026 0.082 2.487 2.490 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.298 2.301 calculate_dm_sparse 119 9.5 0.009 0.050 2.170 2.250 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.124 2.221 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.021 2.027 density_rs2pw 119 9.7 0.004 0.004 1.946 2.007 jit_kernel_multiply 10 16.1 0.767 1.977 0.767 1.977 potential_pw2rs 119 12.3 0.006 0.006 1.954 1.969 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.957 1.957 cp_fm_redistribute_end 50 14.0 1.614 1.932 1.618 1.933 grid_integrate_task_list 119 12.3 1.827 1.932 1.827 1.932 cp_fm_diag_elpa_base 50 14.0 0.307 1.872 0.314 1.902 pw_transfer 1439 11.6 0.064 0.066 1.768 1.812 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.771 1.809 prepare_preconditioner 11 7.9 0.000 0.000 1.713 1.745 make_preconditioner 11 8.9 0.001 0.005 1.713 1.745 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.679 1.724 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.613 1.677 mp_alltoall_d11v 2130 13.8 1.324 1.598 1.324 1.598 make_images_data 4572 15.5 0.045 0.051 1.163 1.589 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.008 1.500 acc_transpose_blocks 27432 15.5 0.106 0.110 1.200 1.495 wfi_extrapolate 11 7.9 0.002 0.005 1.448 1.448 fft3d_ps 1201 14.6 0.496 0.549 1.385 1.424 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.375 1.394 mp_allgather_i34 2286 14.5 0.581 1.364 0.581 1.364 fft_wrap_pw1pw2_140 487 13.2 0.077 0.084 1.297 1.341 grid_collocate_task_list 119 9.7 1.240 1.307 1.240 1.307 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.184 1.231 mp_sum_d 4129 12.0 0.746 1.229 0.746 1.229 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.091 1.097 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.510 0.908 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.891 0.904 acc_transpose_blocks_kernels 27432 16.5 0.183 0.272 0.671 0.887 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.134000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.727273, yerr=1.863082 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 523.948032E+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 0.078 0.119 34.868 34.869 qs_mol_dyn_low 1 2.0 0.003 0.003 34.445 34.453 qs_forces 11 3.9 0.002 0.002 34.383 34.388 qs_energies 11 4.9 0.002 0.002 32.809 32.815 scf_env_do_scf 11 5.9 0.000 0.001 27.191 27.191 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.593 24.593 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.685 18.779 qs_scf_new_mos 108 7.5 0.001 0.001 16.643 16.669 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.642 16.669 ot_scf_mini 108 9.5 0.002 0.002 15.913 15.929 multiply_cannon 2286 13.5 0.194 0.202 14.884 15.669 velocity_verlet 10 3.0 0.001 0.002 15.246 15.247 multiply_cannon_loop 2286 14.5 0.639 0.670 14.053 14.798 ot_mini 108 10.5 0.001 0.001 10.558 10.569 qs_ot_get_derivative 108 11.5 0.001 0.001 9.043 9.058 multiply_cannon_multrec 18288 15.5 1.918 2.836 7.354 8.571 dbcsr_mm_accdrv_process 38222 16.0 4.498 6.755 5.351 6.816 rebuild_ks_matrix 119 8.3 0.000 0.000 6.324 6.341 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.324 6.341 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.587 5.604 mp_waitall_1 169478 16.3 4.048 5.452 4.048 5.452 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.118 5.126 init_scf_run 11 5.9 0.000 0.001 4.417 4.417 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.417 4.417 sum_up_and_integrate 119 10.3 0.018 0.018 3.958 3.962 integrate_v_rspace 119 11.3 0.002 0.003 3.940 3.947 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 2.534 3.659 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.238 3.245 calculate_rho_elec 119 8.7 0.030 0.031 3.238 3.244 qs_ot_get_p 119 10.4 0.001 0.001 3.180 3.203 calculate_first_density_matrix 1 7.0 0.000 0.000 3.186 3.187 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.428 2.975 apply_single 119 13.6 0.000 0.000 2.428 2.975 init_scf_loop 11 6.9 0.000 0.000 2.575 2.576 rs_pw_transfer 974 11.9 0.009 0.010 2.378 2.496 calculate_dm_sparse 119 9.5 0.000 0.000 2.137 2.147 mp_sum_l 7207 12.9 1.557 2.127 1.557 2.127 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.063 2.071 make_m2s 4572 13.5 0.044 0.045 1.879 2.038 density_rs2pw 119 9.7 0.004 0.004 1.842 1.960 make_images 4572 14.5 0.190 0.200 1.794 1.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.891 1.893 grid_integrate_task_list 119 12.3 1.793 1.880 1.793 1.880 jit_kernel_multiply 7 16.1 0.801 1.862 0.801 1.862 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.800 1.800 prepare_preconditioner 11 7.9 0.000 0.000 1.780 1.783 make_preconditioner 11 8.9 0.000 0.000 1.780 1.783 potential_pw2rs 119 12.3 0.007 0.008 1.771 1.778 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.741 1.751 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.637 1.717 pw_transfer 1439 11.6 0.064 0.066 1.663 1.672 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.594 1.601 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.573 1.583 multiply_cannon_sync_h2d 18288 15.5 1.387 1.573 1.387 1.573 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.562 1.563 cp_fm_redistribute_end 50 14.0 1.161 1.542 1.162 1.543 cp_fm_diag_elpa_base 50 14.0 0.364 1.479 0.379 1.517 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.503 1.507 ot_diis_step 108 11.5 0.011 0.011 1.492 1.493 acc_transpose_blocks 18288 15.5 0.074 0.076 1.352 1.390 grid_collocate_task_list 119 9.7 1.205 1.274 1.205 1.274 fft3d_ps 1201 14.6 0.501 0.521 1.259 1.269 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.209 1.218 wfi_extrapolate 11 7.9 0.001 0.001 1.187 1.188 make_images_data 4572 15.5 0.044 0.048 0.802 0.993 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.960 0.965 acc_transpose_blocks_kernels 18288 16.5 0.212 0.221 0.913 0.942 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.890 0.912 hybrid_alltoall_any 4725 16.4 0.054 0.113 0.700 0.912 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.810 mp_alltoall_d11v 2130 13.8 0.684 0.802 0.684 0.802 parallel_gemm_fm 81 9.0 0.000 0.000 0.791 0.794 parallel_gemm_fm_cosma 81 10.0 0.791 0.794 0.791 0.794 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.392 0.788 make_basis_sm 11 9.8 0.005 0.034 0.761 0.762 cp_fm_cholesky_invert 11 10.9 0.749 0.753 0.749 0.753 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.737 jit_kernel_transpose 5 15.6 0.701 0.727 0.701 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.869000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.909091, yerr=2.810841 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.212224E+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 0.088 0.181 37.629 37.630 qs_mol_dyn_low 1 2.0 0.013 0.044 36.844 36.860 qs_forces 11 3.9 0.013 0.060 36.708 36.720 qs_energies 11 4.9 0.018 0.101 34.951 34.976 scf_env_do_scf 11 5.9 0.000 0.001 28.535 28.537 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.955 24.955 dbcsr_multiply_generic 2286 12.5 0.100 0.102 19.592 19.708 velocity_verlet 10 3.0 0.030 0.073 17.567 17.569 qs_scf_new_mos 108 7.5 0.001 0.001 16.904 16.946 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.903 16.945 multiply_cannon 2286 13.5 0.235 0.282 16.064 16.509 ot_scf_mini 108 9.5 0.002 0.003 15.973 16.019 multiply_cannon_loop 2286 14.5 0.936 0.963 15.126 15.593 multiply_cannon_multrec 27432 15.5 2.382 3.108 9.605 10.451 ot_mini 108 10.5 0.001 0.001 10.182 10.243 dbcsr_mm_accdrv_process 47916 15.9 5.317 7.459 7.128 8.784 qs_ot_get_derivative 108 11.5 0.001 0.001 8.377 8.423 rebuild_ks_matrix 119 8.3 0.000 0.000 6.460 6.518 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.036 6.460 6.518 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.747 5.798 init_scf_run 11 5.9 0.000 0.001 4.707 4.708 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.707 4.708 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.740 4.068 sum_up_and_integrate 119 10.3 0.020 0.022 3.735 3.743 integrate_v_rspace 119 11.3 0.002 0.003 3.715 3.723 jit_kernel_multiply 12 16.0 1.749 3.605 1.749 3.605 init_scf_loop 11 6.9 0.003 0.026 3.560 3.561 mp_waitall_1 145218 16.4 2.467 3.519 2.467 3.519 qs_ot_get_p 119 10.4 0.001 0.001 3.281 3.356 calculate_first_density_matrix 1 7.0 0.003 0.013 3.345 3.349 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.253 3.282 calculate_rho_elec 119 8.7 0.040 0.046 3.253 3.282 prepare_preconditioner 11 7.9 0.000 0.000 2.591 2.601 make_preconditioner 11 8.9 0.003 0.021 2.591 2.601 calculate_dm_sparse 119 9.5 0.003 0.009 2.465 2.519 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.195 2.518 rs_pw_transfer 974 11.9 0.009 0.010 2.209 2.364 make_m2s 4572 13.5 0.054 0.056 2.191 2.305 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 1.209 2.209 make_images 4572 14.5 0.273 0.333 2.083 2.195 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.961 2.142 apply_single 119 13.6 0.000 0.000 1.961 2.142 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.126 2.127 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.943 1.953 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.913 1.938 grid_integrate_task_list 119 12.3 1.832 1.919 1.832 1.919 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.839 1.853 density_rs2pw 119 9.7 0.003 0.004 1.740 1.829 ot_diis_step 108 11.5 0.012 0.012 1.764 1.764 pw_transfer 1439 11.6 0.063 0.066 1.674 1.703 acc_transpose_blocks 27432 15.5 0.111 0.113 1.661 1.696 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.651 1.664 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.630 1.630 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.585 1.616 potential_pw2rs 119 12.3 0.009 0.009 1.524 1.527 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.402 1.403 cp_fm_redistribute_end 50 14.0 0.921 1.375 0.923 1.375 cp_fm_diag_elpa_base 50 14.0 0.430 1.315 0.450 1.354 qs_energies_init_hamiltonians 11 5.9 0.007 0.009 1.323 1.346 grid_collocate_task_list 119 9.7 1.221 1.319 1.221 1.319 fft_wrap_pw1pw2_140 487 13.2 0.087 0.095 1.281 1.314 mp_sum_l 7207 12.9 0.958 1.302 0.958 1.302 wfi_extrapolate 11 7.9 0.001 0.001 1.293 1.293 fft3d_ps 1201 14.6 0.528 0.585 1.263 1.289 dbcsr_complete_redistribute 329 12.2 0.172 0.234 0.899 1.173 cp_fm_upper_to_full 72 13.5 0.809 1.119 0.809 1.119 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.090 1.109 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 1.037 1.061 multiply_cannon_sync_h2d 27432 15.5 0.994 1.056 0.994 1.056 make_images_data 4572 15.5 0.044 0.049 0.847 0.969 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.694 0.963 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.715 0.911 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.876 rs_pw_transfer_RS2PW_140 130 11.5 0.154 0.164 0.720 0.876 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.790 0.867 mp_alltoall_d11v 2130 13.8 0.733 0.821 0.733 0.821 mp_waitany 7280 13.7 0.673 0.819 0.673 0.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.814 jit_kernel_transpose 5 15.6 0.767 0.794 0.767 0.794 cp_fm_cholesky_invert 11 10.9 0.756 0.759 0.756 0.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.630000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.818182, yerr=2.757409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 618.561536E+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 0.047 0.148 43.723 43.724 qs_mol_dyn_low 1 2.0 0.003 0.004 43.217 43.224 qs_forces 11 3.9 0.002 0.002 43.107 43.107 qs_energies 11 4.9 0.001 0.002 41.421 41.424 dbcsr_multiply_generic 2286 12.5 0.108 0.158 27.791 27.870 multiply_cannon 2286 13.5 0.236 0.246 25.246 25.666 multiply_cannon_loop 2286 14.5 0.330 0.341 24.354 24.557 scf_env_do_scf 11 5.9 0.000 0.001 20.587 20.587 init_scf_run 11 5.9 0.000 0.001 19.317 19.317 scf_env_initial_rho_setup 11 6.9 0.000 0.000 19.316 19.317 calculate_first_density_matrix 1 7.0 0.000 0.000 18.216 18.217 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.087 18.088 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 17.493 17.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 17.220 17.226 acc_transpose_blocks 9144 15.5 0.038 0.039 16.651 16.668 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 16.410 16.424 jit_kernel_transpose 5 15.6 16.292 16.307 16.292 16.307 velocity_verlet 10 3.0 0.021 0.058 13.927 13.930 qs_scf_new_mos 108 7.5 0.001 0.001 10.677 10.700 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.676 10.699 ot_scf_mini 108 9.5 0.002 0.002 10.015 10.039 multiply_cannon_multrec 9144 15.5 1.595 1.837 5.804 6.004 rebuild_ks_matrix 119 8.3 0.000 0.000 5.773 5.798 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.772 5.798 ot_mini 108 10.5 0.001 0.001 5.610 5.641 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.146 5.169 qs_ot_get_derivative 108 11.5 0.001 0.001 4.336 4.361 dbcsr_mm_accdrv_process 12550 15.8 3.136 4.143 4.110 4.191 sum_up_and_integrate 119 10.3 0.022 0.024 3.499 3.504 integrate_v_rspace 119 11.3 0.003 0.003 3.477 3.482 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.119 3.125 calculate_rho_elec 119 8.7 0.059 0.061 3.119 3.124 qs_ot_get_p 119 10.4 0.001 0.001 2.677 2.717 init_scf_loop 11 6.9 0.000 0.000 2.480 2.481 mp_waitall_1 121218 16.5 1.506 1.925 1.506 1.925 make_m2s 4572 13.5 0.034 0.035 1.755 1.920 grid_integrate_task_list 119 12.3 1.851 1.918 1.851 1.918 make_images 4572 14.5 0.266 0.300 1.665 1.829 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.761 1.763 rs_pw_transfer 974 11.9 0.008 0.008 1.656 1.730 calculate_dm_sparse 119 9.5 0.000 0.000 1.710 1.727 prepare_preconditioner 11 7.9 0.000 0.000 1.711 1.715 make_preconditioner 11 8.9 0.000 0.000 1.711 1.715 pw_transfer 1439 11.6 0.064 0.066 1.670 1.679 density_rs2pw 119 9.7 0.003 0.004 1.580 1.666 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.604 1.629 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.580 1.589 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.569 1.569 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.520 1.534 jit_kernel_multiply 8 15.6 0.936 1.443 0.936 1.443 grid_collocate_task_list 119 9.7 1.271 1.353 1.271 1.353 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.325 1.325 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.311 1.312 potential_pw2rs 119 12.3 0.010 0.011 1.306 1.310 cp_fm_redistribute_end 50 14.0 0.658 1.304 0.659 1.305 fft_wrap_pw1pw2_140 487 13.2 0.085 0.088 1.294 1.303 cp_fm_diag_elpa_base 50 14.0 0.601 1.233 0.644 1.290 ot_diis_step 108 11.5 0.012 0.013 1.261 1.261 fft3d_ps 1201 14.6 0.538 0.549 1.245 1.251 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.137 1.158 apply_single 119 13.6 0.000 0.000 1.136 1.158 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.140 1.145 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.105 1.118 wfi_extrapolate 11 7.9 0.001 0.001 1.056 1.056 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.730 0.945 make_images_data 4572 15.5 0.038 0.042 0.780 0.939 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=43.724000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=582.181818, yerr=9.003213 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 738.430976E+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 0.085 0.110 43.469 43.470 qs_mol_dyn_low 1 2.0 0.004 0.009 42.938 42.946 qs_forces 11 3.9 0.006 0.007 42.805 42.806 qs_energies 11 4.9 0.001 0.002 40.870 40.873 scf_env_do_scf 11 5.9 0.001 0.001 33.807 33.808 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.999 26.000 velocity_verlet 10 3.0 0.034 0.073 22.727 22.733 dbcsr_multiply_generic 2286 12.5 0.107 0.109 19.677 19.917 qs_scf_new_mos 108 7.5 0.001 0.001 17.193 17.284 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.192 17.283 multiply_cannon 2286 13.5 0.307 0.318 15.444 16.721 ot_scf_mini 108 9.5 0.002 0.002 16.112 16.210 multiply_cannon_loop 2286 14.5 0.344 0.350 14.176 15.487 ot_mini 108 10.5 0.001 0.001 10.178 10.290 multiply_cannon_multrec 9144 15.5 3.376 4.808 9.503 9.959 qs_ot_get_derivative 108 11.5 0.001 0.001 8.180 8.279 init_scf_loop 11 6.9 0.000 0.000 7.778 7.780 dbcsr_mm_accdrv_process 12550 15.8 4.592 5.971 6.004 7.676 rebuild_ks_matrix 119 8.3 0.000 0.000 6.867 6.996 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.866 6.995 prepare_preconditioner 11 7.9 0.000 0.000 6.813 6.826 make_preconditioner 11 8.9 0.000 0.000 6.813 6.826 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.387 6.705 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.210 6.328 init_scf_run 11 5.9 0.000 0.001 4.885 4.885 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.885 4.885 mp_waitall_1 97218 16.6 3.122 4.674 3.122 4.674 cp_fm_upper_to_full 72 14.2 3.117 4.487 3.117 4.487 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.568 4.346 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.626 3.671 calculate_rho_elec 119 8.7 0.118 0.121 3.626 3.671 sum_up_and_integrate 119 10.3 0.038 0.038 3.636 3.644 integrate_v_rspace 119 11.3 0.003 0.003 3.598 3.605 calculate_first_density_matrix 1 7.0 0.000 0.000 3.487 3.496 qs_ot_get_p 119 10.4 0.001 0.001 3.194 3.325 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.810 3.203 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.518 3.112 apply_single 119 13.6 0.000 0.000 2.518 3.112 jit_kernel_multiply 10 15.5 1.386 3.031 1.386 3.031 dbcsr_complete_redistribute 329 12.2 0.325 0.331 2.022 2.827 calculate_dm_sparse 119 9.5 0.000 0.000 2.563 2.592 make_m2s 4572 13.5 0.038 0.038 2.320 2.499 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.698 2.496 make_images 4572 14.5 0.350 0.380 2.199 2.378 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.343 2.344 mp_sum_l 7207 12.9 1.377 2.325 1.377 2.325 mp_alltoall_i22 627 13.8 1.404 2.234 1.404 2.234 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.421 2.212 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.072 2.115 pw_transfer 1439 11.6 0.067 0.068 2.019 2.023 grid_integrate_task_list 119 12.3 2.001 2.017 2.001 2.017 ot_diis_step 108 11.5 0.014 0.014 1.971 1.972 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.946 1.948 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.923 1.928 qs_energies_init_hamiltonians 11 5.9 0.006 0.007 1.906 1.906 density_rs2pw 119 9.7 0.003 0.003 1.796 1.812 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.686 1.736 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.669 1.670 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.631 1.647 fft_wrap_pw1pw2_140 487 13.2 0.087 0.093 1.602 1.608 fft3d_ps 1201 14.6 0.567 0.579 1.556 1.561 rs_pw_transfer 974 11.9 0.009 0.009 1.465 1.520 cp_fm_cholesky_invert 11 10.9 1.510 1.513 1.510 1.513 grid_collocate_task_list 119 9.7 1.443 1.454 1.443 1.454 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.393 1.394 cp_fm_diag_elpa_base 50 14.0 1.249 1.301 1.391 1.392 wfi_extrapolate 11 7.9 0.001 0.001 1.336 1.336 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.060 1.329 make_images_data 4572 15.5 0.042 0.045 1.085 1.327 acc_transpose_blocks 9144 15.5 0.038 0.039 1.254 1.260 potential_pw2rs 119 12.3 0.014 0.014 1.229 1.232 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.157 1.214 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.144 1.162 multiply_cannon_sync_h2d 9144 15.5 1.039 1.042 1.039 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.982 1.019 mp_alltoall_d11v 2130 13.8 0.980 1.014 0.980 1.014 acc_transpose_blocks_kernels 9144 16.5 0.119 0.120 1.004 1.010 qs_create_task_list 11 7.9 0.031 0.031 0.979 0.995 generate_qs_task_list 11 8.9 0.374 0.394 0.948 0.964 jit_kernel_transpose 5 15.6 0.886 0.891 0.886 0.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.867 0.879 mp_alltoall_z22v 1201 16.6 0.854 0.873 0.854 0.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.470000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=697.000000, yerr=10.054398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.178368E+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 1274957. 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 0.017 0.034 83.977 83.978 qs_mol_dyn_low 1 2.0 0.003 0.004 83.526 83.536 qs_forces 11 3.9 0.003 0.003 83.454 83.454 qs_energies 11 4.9 0.002 0.003 80.554 80.567 scf_env_do_scf 11 5.9 0.000 0.001 71.325 71.328 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.710 65.710 dbcsr_multiply_generic 2055 12.4 0.105 0.107 52.875 53.122 qs_scf_new_mos 99 7.5 0.000 0.001 48.676 48.770 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.676 48.770 ot_scf_mini 99 9.5 0.002 0.002 46.283 46.378 multiply_cannon 2055 13.4 0.179 0.186 43.262 44.246 multiply_cannon_loop 2055 14.4 1.569 1.611 42.287 43.257 velocity_verlet 10 3.0 0.004 0.010 42.054 42.055 ot_mini 99 10.5 0.001 0.001 28.409 28.493 qs_ot_get_derivative 99 11.5 0.001 0.001 21.622 21.700 multiply_cannon_multrec 49320 15.4 12.088 12.831 17.574 18.775 rebuild_ks_matrix 110 8.3 0.000 0.001 14.550 14.683 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.549 14.683 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.777 12.899 mp_waitall_1 241148 16.1 11.901 12.731 11.901 12.731 multiply_cannon_sync_h2d 49320 15.4 9.974 10.628 9.974 10.628 qs_ot_get_p 110 10.4 0.001 0.001 9.402 9.510 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.049 8.090 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.238 7.839 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.214 7.697 apply_single 110 13.6 0.000 0.000 7.213 7.697 init_scf_run 11 5.9 0.000 0.001 7.158 7.158 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.158 7.158 sum_up_and_integrate 110 10.3 0.023 0.028 7.123 7.138 integrate_v_rspace 110 11.3 0.002 0.003 7.100 7.120 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.937 6.989 ot_diis_step 99 11.5 0.006 0.006 6.522 6.522 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.201 6.345 calculate_rho_elec 110 8.6 0.020 0.024 6.201 6.344 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.140 6.164 dbcsr_mm_accdrv_process 87628 16.1 2.084 2.210 5.365 5.868 mp_sum_l 6514 12.8 5.248 5.866 5.248 5.866 init_scf_loop 11 6.9 0.000 0.000 5.587 5.587 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.087 5.088 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.581 4.583 cp_fm_redistribute_end 48 14.0 3.989 4.554 3.994 4.556 cp_fm_diag_elpa_base 48 14.0 0.554 4.447 0.558 4.473 rs_pw_transfer 902 11.9 0.012 0.014 4.085 4.262 wfi_extrapolate 11 7.9 0.001 0.001 4.027 4.027 make_m2s 4110 13.4 0.061 0.066 3.919 4.008 calculate_dm_sparse 110 9.5 0.000 0.001 3.880 3.964 make_images 4110 14.4 0.179 0.192 3.823 3.914 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.821 3.792 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.658 3.662 density_rs2pw 110 9.6 0.004 0.004 3.344 3.509 prepare_preconditioner 11 7.9 0.000 0.000 3.427 3.445 make_preconditioner 11 8.9 0.000 0.000 3.426 3.445 grid_integrate_task_list 110 12.3 3.258 3.423 3.258 3.423 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.257 3.311 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.254 3.306 jit_kernel_multiply 13 15.9 2.996 3.256 2.996 3.256 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.216 3.254 calculate_first_density_matrix 1 7.0 0.000 0.000 3.049 3.053 pw_transfer 1331 11.6 0.054 0.065 2.814 2.867 potential_pw2rs 110 12.3 0.006 0.007 2.780 2.794 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.722 2.779 fft_wrap_pw1pw2_140 451 13.1 0.169 0.187 2.261 2.317 acc_transpose_blocks 49320 15.4 0.212 0.225 2.221 2.317 mp_alltoall_d11v 2046 13.8 1.979 2.313 1.979 2.313 fft3d_ps 1111 14.6 0.752 0.836 2.260 2.306 grid_collocate_task_list 110 9.6 2.083 2.194 2.083 2.194 mp_sum_d 3883 11.9 1.494 1.979 1.494 1.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.913 1.930 cp_fm_cholesky_invert 11 10.9 1.897 1.901 1.897 1.901 mp_waitany 14300 13.8 1.667 1.882 1.667 1.882 make_images_data 4110 15.4 0.043 0.046 1.700 1.829 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.760 1.777 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.483 1.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.978000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=2.499587 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.592832E+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 2358156. 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 0.037 0.063 69.577 69.578 qs_mol_dyn_low 1 2.0 0.003 0.020 69.075 69.085 qs_forces 11 3.9 0.012 0.038 68.979 68.981 qs_energies 11 4.9 0.016 0.056 65.599 65.613 scf_env_do_scf 11 5.9 0.001 0.001 56.566 56.569 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.331 48.335 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.632 37.826 velocity_verlet 10 3.0 0.006 0.011 35.953 35.955 qs_scf_new_mos 99 7.5 0.001 0.001 32.431 32.563 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.430 32.562 multiply_cannon 2055 13.4 0.221 0.249 31.002 32.368 ot_scf_mini 99 9.5 0.003 0.003 30.778 30.896 multiply_cannon_loop 2055 14.4 0.932 0.953 29.654 30.639 ot_mini 99 10.5 0.001 0.001 18.115 18.246 multiply_cannon_multrec 24660 15.4 7.609 9.672 13.799 15.695 rebuild_ks_matrix 110 8.3 0.000 0.001 14.042 14.188 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.042 14.188 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.370 12.497 qs_ot_get_derivative 99 11.5 0.001 0.001 12.345 12.466 mp_waitall_1 186928 16.3 7.793 10.248 7.793 10.248 init_scf_loop 11 6.9 0.002 0.011 8.198 8.223 multiply_cannon_sync_h2d 24660 15.4 7.001 8.071 7.001 8.071 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.156 7.666 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.454 7.243 apply_single 110 13.6 0.000 0.001 6.454 7.243 sum_up_and_integrate 110 10.3 0.028 0.033 6.591 6.613 integrate_v_rspace 110 11.3 0.002 0.003 6.562 6.586 init_scf_run 11 5.9 0.000 0.001 6.571 6.572 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.571 6.572 qs_ot_get_p 110 10.4 0.001 0.004 6.287 6.443 dbcsr_mm_accdrv_process 52282 16.1 4.693 5.630 6.029 6.319 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.904 5.913 calculate_rho_elec 110 8.6 0.039 0.047 5.903 5.913 ot_diis_step 99 11.5 0.010 0.010 5.722 5.722 prepare_preconditioner 11 7.9 0.000 0.000 5.604 5.623 make_preconditioner 11 8.9 0.004 0.032 5.604 5.623 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.658 5.363 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.155 5.315 make_m2s 4110 13.4 0.057 0.060 4.188 4.682 make_images 4110 14.4 0.396 0.439 4.079 4.569 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.119 4.139 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.686 3.687 wfi_extrapolate 11 7.9 0.001 0.001 3.512 3.512 pw_transfer 1331 11.6 0.066 0.072 3.315 3.457 rs_pw_transfer 902 11.9 0.012 0.013 3.198 3.433 density_rs2pw 110 9.6 0.004 0.004 3.193 3.381 grid_integrate_task_list 110 12.3 3.168 3.368 3.168 3.368 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.347 3.349 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.204 3.347 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.180 3.251 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.140 3.142 cp_fm_redistribute_end 48 14.0 2.326 3.094 2.328 3.094 cp_fm_diag_elpa_base 48 14.0 0.732 2.969 0.763 3.056 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.980 3.033 calculate_dm_sparse 110 9.5 0.001 0.001 2.958 2.992 make_images_data 4110 15.4 0.047 0.051 2.354 2.858 calculate_first_density_matrix 1 7.0 0.000 0.000 2.845 2.848 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.687 2.828 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.016 2.806 fft3d_ps 1111 14.6 1.069 1.267 2.560 2.690 cp_fm_cholesky_invert 11 10.9 2.579 2.586 2.579 2.586 potential_pw2rs 110 12.3 0.009 0.009 2.495 2.511 mp_sum_l 6514 12.8 1.861 2.448 1.861 2.448 grid_collocate_task_list 110 9.6 2.134 2.280 2.134 2.280 mp_alltoall_d11v 2046 13.8 1.720 2.076 1.720 2.076 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.944 1.968 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.874 1.885 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.843 1.865 jit_kernel_multiply 9 16.3 0.981 1.747 0.981 1.747 mp_allgather_i34 2055 14.4 0.658 1.683 0.658 1.683 mp_sum_d 3883 11.9 1.245 1.609 1.245 1.609 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.795 1.573 acc_transpose_blocks 24660 15.4 0.106 0.109 1.526 1.562 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.550 dbcsr_complete_redistribute 325 12.2 0.249 0.320 1.213 1.482 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.372 1.479 mp_irecv_dv 57340 16.2 0.671 1.454 0.671 1.454 mp_waitany 10164 13.8 1.178 1.403 1.178 1.403 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.578000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.090909, yerr=6.947263 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.734528E+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 0.019 0.047 59.828 59.829 qs_mol_dyn_low 1 2.0 0.005 0.017 59.401 59.411 qs_forces 11 3.9 0.006 0.031 59.328 59.333 qs_energies 11 4.9 0.001 0.002 56.137 56.148 scf_env_do_scf 11 5.9 0.000 0.001 48.161 48.161 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.589 39.589 velocity_verlet 10 3.0 0.001 0.002 32.411 32.413 dbcsr_multiply_generic 2055 12.4 0.107 0.109 28.578 28.841 qs_scf_new_mos 99 7.5 0.001 0.001 25.070 25.170 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.070 25.169 ot_scf_mini 99 9.5 0.002 0.003 23.850 23.970 multiply_cannon 2055 13.4 0.211 0.220 22.265 23.399 multiply_cannon_loop 2055 14.4 0.620 0.635 21.022 22.211 ot_mini 99 10.5 0.001 0.001 13.589 13.715 rebuild_ks_matrix 110 8.3 0.000 0.000 12.217 12.347 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.217 12.347 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.773 10.890 multiply_cannon_multrec 16440 15.4 3.954 5.144 9.724 10.794 mp_waitall_1 146766 16.3 7.169 10.602 7.169 10.602 qs_ot_get_derivative 99 11.5 0.001 0.001 9.086 9.206 init_scf_loop 11 6.9 0.000 0.000 8.536 8.538 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.320 7.549 prepare_preconditioner 11 7.9 0.000 0.000 6.780 6.796 make_preconditioner 11 8.9 0.000 0.000 6.780 6.796 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.136 6.481 sum_up_and_integrate 110 10.3 0.033 0.034 6.440 6.454 integrate_v_rspace 110 11.3 0.002 0.003 6.406 6.422 dbcsr_mm_accdrv_process 34862 16.1 4.804 5.341 5.624 5.768 qs_ot_get_p 110 10.4 0.001 0.001 5.417 5.566 init_scf_run 11 5.9 0.000 0.001 5.554 5.554 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.554 5.554 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.525 5.545 calculate_rho_elec 110 8.6 0.058 0.058 5.524 5.545 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.025 5.486 apply_single 110 13.6 0.000 0.000 5.024 5.485 make_m2s 4110 13.4 0.050 0.052 4.216 4.639 make_images 4110 14.4 0.390 0.507 4.101 4.522 ot_diis_step 99 11.5 0.010 0.011 4.474 4.474 multiply_cannon_sync_h2d 16440 15.4 3.682 4.370 3.682 4.370 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.795 3.798 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.125 3.749 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.458 3.458 grid_integrate_task_list 110 12.3 3.172 3.361 3.172 3.361 pw_transfer 1331 11.6 0.066 0.073 3.105 3.113 rs_pw_transfer 902 11.9 0.011 0.011 2.868 3.010 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.994 3.004 make_images_data 4110 15.4 0.044 0.048 2.457 2.984 density_rs2pw 110 9.6 0.004 0.004 2.832 2.954 wfi_extrapolate 11 7.9 0.001 0.001 2.951 2.951 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.185 2.936 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.935 2.936 cp_fm_redistribute_end 48 14.0 1.825 2.907 1.827 2.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.881 2.883 cp_fm_diag_elpa_base 48 14.0 1.018 2.768 1.077 2.879 cp_fm_cholesky_invert 11 10.9 2.654 2.661 2.654 2.661 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.554 2.608 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.547 2.558 calculate_dm_sparse 110 9.5 0.001 0.001 2.511 2.545 calculate_first_density_matrix 1 7.0 0.000 0.000 2.518 2.521 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.879 2.409 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.311 2.381 grid_collocate_task_list 110 9.6 2.176 2.336 2.176 2.336 fft3d_ps 1111 14.6 1.059 1.067 2.306 2.319 potential_pw2rs 110 12.3 0.011 0.011 2.294 2.305 mp_irecv_dv 48980 15.7 0.809 2.287 0.809 2.287 mp_sum_l 6514 12.8 1.485 2.100 1.485 2.100 mp_alltoall_d11v 2046 13.8 1.782 2.098 1.782 2.098 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.993 2.002 dbcsr_complete_redistribute 325 12.2 0.323 0.348 1.420 1.877 cp_fm_upper_to_full 70 13.6 1.386 1.863 1.386 1.863 mp_allgather_i34 2055 14.4 0.566 1.691 0.566 1.691 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.643 1.662 cp_fm_cholesky_decompose 22 10.9 1.615 1.638 1.615 1.638 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.472 1.484 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.482 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.946 1.399 mp_waitany 17072 13.8 1.142 1.332 1.142 1.332 acc_transpose_blocks 16440 15.4 0.071 0.073 1.291 1.326 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.251 1.259 rs_gather_matrices 110 12.3 0.138 0.151 0.891 1.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.829000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.636364, yerr=8.488203 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 730.099712E+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 0.072 0.131 65.444 65.445 qs_mol_dyn_low 1 2.0 0.003 0.005 64.969 64.979 qs_forces 11 3.9 0.020 0.043 64.881 64.882 qs_energies 11 4.9 0.003 0.015 61.454 61.475 scf_env_do_scf 11 5.9 0.001 0.001 52.937 52.940 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.941 40.942 velocity_verlet 10 3.0 0.002 0.002 37.032 37.037 dbcsr_multiply_generic 2055 12.4 0.120 0.136 29.451 29.665 qs_scf_new_mos 99 7.5 0.001 0.001 26.461 26.560 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.461 26.559 ot_scf_mini 99 9.5 0.003 0.004 24.825 24.938 multiply_cannon 2055 13.4 0.245 0.261 22.407 23.363 multiply_cannon_loop 2055 14.4 0.884 0.904 20.996 21.622 ot_mini 99 10.5 0.001 0.001 14.164 14.274 multiply_cannon_multrec 24660 15.4 4.204 6.753 12.637 13.863 rebuild_ks_matrix 110 8.3 0.000 0.000 11.977 12.095 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 11.976 12.094 init_scf_loop 11 6.9 0.001 0.002 11.953 11.956 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.570 10.673 prepare_preconditioner 11 7.9 0.000 0.000 10.199 10.213 make_preconditioner 11 8.9 0.000 0.001 10.199 10.213 qs_ot_get_derivative 99 11.5 0.001 0.001 9.994 10.101 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.386 9.875 dbcsr_mm_accdrv_process 52304 16.0 7.042 8.332 8.286 9.170 mp_waitall_1 126806 16.4 4.374 6.589 4.374 6.589 sum_up_and_integrate 110 10.3 0.038 0.040 6.388 6.406 integrate_v_rspace 110 11.3 0.003 0.003 6.351 6.368 qs_ot_get_p 110 10.4 0.003 0.019 5.708 5.846 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.681 5.688 calculate_rho_elec 110 8.6 0.077 0.081 5.680 5.688 make_m2s 4110 13.4 0.059 0.061 5.356 5.685 init_scf_run 11 5.9 0.000 0.001 5.632 5.632 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.631 5.632 make_images 4110 14.4 0.575 0.697 5.215 5.540 cp_fm_upper_to_full 70 13.8 3.355 4.767 3.355 4.767 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.034 4.146 apply_single 110 13.6 0.000 0.001 4.034 4.146 ot_diis_step 99 11.5 0.011 0.013 4.131 4.131 dbcsr_complete_redistribute 325 12.2 0.441 0.515 2.776 3.933 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.837 3.853 grid_integrate_task_list 110 12.3 3.295 3.452 3.295 3.452 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.328 3.376 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.472 3.373 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.220 3.370 multiply_cannon_sync_h2d 24660 15.4 3.199 3.366 3.199 3.366 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.353 3.354 pw_transfer 1331 11.6 0.066 0.074 3.284 3.312 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.173 3.204 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.047 3.106 hybrid_alltoall_any 4261 16.3 0.119 0.455 2.387 2.992 make_images_data 4110 15.4 0.046 0.051 2.656 2.991 density_rs2pw 110 9.6 0.004 0.004 2.827 2.969 calculate_dm_sparse 110 9.5 0.001 0.001 2.924 2.964 wfi_extrapolate 11 7.9 0.002 0.013 2.946 2.947 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.804 2.934 mp_alltoall_i22 605 13.7 1.679 2.893 1.679 2.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.872 2.874 cp_fm_cholesky_invert 11 10.9 2.813 2.823 2.813 2.823 rs_pw_transfer 902 11.9 0.010 0.011 2.636 2.814 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.783 2.785 cp_fm_redistribute_end 48 14.0 1.386 2.752 1.388 2.753 cp_fm_diag_elpa_base 48 14.0 1.283 2.620 1.363 2.730 fft_wrap_pw1pw2_140 451 13.1 0.205 0.216 2.689 2.721 calculate_first_density_matrix 1 7.0 0.000 0.001 2.596 2.599 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.476 2.520 fft3d_ps 1111 14.6 1.061 1.089 2.471 2.492 qs_energies_init_hamiltonians 11 5.9 0.009 0.016 2.359 2.378 grid_collocate_task_list 110 9.6 2.226 2.362 2.226 2.362 potential_pw2rs 110 12.3 0.013 0.013 2.210 2.224 mp_alltoall_d11v 2046 13.8 1.814 1.971 1.814 1.971 jit_kernel_multiply 10 16.0 0.911 1.826 0.911 1.826 cp_fm_cholesky_decompose 22 10.9 1.748 1.808 1.748 1.808 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.725 1.761 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.605 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.651 1.665 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.840 1.542 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.525 1.541 acc_transpose_blocks 24660 15.4 0.103 0.105 1.506 1.528 mp_sum_l 6514 12.8 1.012 1.500 1.012 1.500 mp_irecv_dv 62702 16.1 0.742 1.465 0.742 1.465 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.271 1.374 mp_waitany 13376 13.8 1.099 1.329 1.099 1.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.445000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.545455, yerr=9.355248 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 838.328320E+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 0.018 0.041 54.873 54.874 qs_mol_dyn_low 1 2.0 0.003 0.004 54.245 54.520 qs_forces 11 3.9 0.007 0.008 54.155 54.155 qs_energies 11 4.9 0.005 0.015 50.516 50.523 scf_env_do_scf 11 5.9 0.000 0.001 41.912 41.912 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.215 34.215 velocity_verlet 10 3.0 0.010 0.012 30.476 30.479 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.989 23.146 qs_scf_new_mos 99 7.5 0.001 0.001 20.137 20.194 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.136 20.193 multiply_cannon 2055 13.4 0.248 0.262 17.611 19.054 ot_scf_mini 99 9.5 0.002 0.002 18.908 18.948 multiply_cannon_loop 2055 14.4 0.322 0.336 16.266 16.612 rebuild_ks_matrix 110 8.3 0.000 0.000 11.493 11.537 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.493 11.536 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.231 10.268 ot_mini 99 10.5 0.001 0.001 10.189 10.224 multiply_cannon_multrec 8220 15.4 3.222 4.533 7.461 8.489 mp_waitall_1 106626 16.5 6.035 7.756 6.035 7.756 init_scf_loop 11 6.9 0.000 0.000 7.649 7.651 qs_ot_get_derivative 99 11.5 0.001 0.001 6.425 6.465 sum_up_and_integrate 110 10.3 0.048 0.049 6.186 6.199 integrate_v_rspace 110 11.3 0.003 0.003 6.138 6.151 prepare_preconditioner 11 7.9 0.000 0.000 6.027 6.031 make_preconditioner 11 8.9 0.000 0.000 6.027 6.031 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.609 5.682 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.558 5.571 calculate_rho_elec 110 8.6 0.115 0.116 5.557 5.570 init_scf_run 11 5.9 0.000 0.001 5.465 5.465 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.465 5.465 dbcsr_mm_accdrv_process 17442 15.9 2.874 4.006 4.109 5.031 qs_ot_get_p 110 10.4 0.001 0.001 4.810 4.846 make_m2s 4110 13.4 0.039 0.040 4.151 4.435 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.951 4.369 make_images 4110 14.4 0.631 0.689 4.021 4.309 ot_diis_step 99 11.5 0.012 0.012 3.743 3.743 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.661 3.712 apply_single 110 13.6 0.000 0.000 3.661 3.712 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.477 3.481 grid_integrate_task_list 110 12.3 3.375 3.477 3.375 3.477 pw_transfer 1331 11.6 0.066 0.070 3.175 3.182 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.176 3.177 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.064 3.075 multiply_cannon_sync_h2d 8220 15.4 2.921 3.073 2.921 3.073 cp_fm_cholesky_invert 11 10.9 2.905 2.909 2.905 2.909 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.898 2.899 density_rs2pw 110 9.6 0.004 0.004 2.647 2.825 make_images_data 4110 15.4 0.038 0.043 2.375 2.805 hybrid_alltoall_any 4261 16.3 0.199 0.866 2.300 2.744 calculate_first_density_matrix 1 7.0 0.000 0.000 2.695 2.696 wfi_extrapolate 11 7.9 0.001 0.001 2.685 2.685 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.678 2.683 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.644 2.645 fft_wrap_pw1pw2_140 451 13.1 0.214 0.221 2.619 2.632 cp_fm_redistribute_end 48 14.0 0.672 2.616 0.676 2.617 cp_fm_diag_elpa_base 48 14.0 1.756 2.418 1.933 2.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.539 2.565 calculate_dm_sparse 110 9.5 0.001 0.001 2.502 2.546 grid_collocate_task_list 110 9.6 2.321 2.493 2.321 2.493 rs_pw_transfer 902 11.9 0.010 0.011 2.299 2.488 fft3d_ps 1111 14.6 1.111 1.141 2.319 2.332 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 2.002 potential_pw2rs 110 12.3 0.015 0.016 1.989 1.997 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.870 1.892 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.747 1.767 mp_alltoall_d11v 2046 13.8 1.570 1.767 1.570 1.767 cp_fm_cholesky_decompose 22 10.9 1.685 1.705 1.685 1.705 mp_allgather_i34 2055 14.4 0.530 1.637 0.530 1.637 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.521 1.637 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.592 1.598 jit_kernel_multiply 9 15.7 0.926 1.534 0.926 1.534 dbcsr_complete_redistribute 325 12.2 0.562 0.586 1.444 1.531 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.473 1.484 qs_create_task_list 11 7.9 0.006 0.014 1.243 1.341 generate_qs_task_list 11 8.9 0.380 0.449 1.237 1.339 acc_transpose_blocks 8220 15.4 0.036 0.037 1.259 1.271 mp_waitany 9240 13.8 1.049 1.247 1.049 1.247 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.135 1.161 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.476 1.149 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.776 1.140 mp_irecv_dv 24056 15.7 0.452 1.107 0.452 1.107 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.874000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=788.181818, yerr=16.044360 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 1.191473E+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 0.024 0.041 84.516 84.517 qs_mol_dyn_low 1 2.0 0.003 0.004 84.196 84.207 qs_forces 11 3.9 0.011 0.021 84.121 84.123 qs_energies 11 4.9 0.002 0.002 80.057 80.069 scf_env_do_scf 11 5.9 0.001 0.001 70.108 70.108 velocity_verlet 10 3.0 0.002 0.002 54.191 54.198 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.855 41.855 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.316 28.448 init_scf_loop 11 6.9 0.000 0.000 28.180 28.181 prepare_preconditioner 11 7.9 0.000 0.000 26.248 26.259 make_preconditioner 11 8.9 0.000 0.000 26.248 26.259 qs_scf_new_mos 99 7.5 0.001 0.001 25.753 25.808 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.753 25.807 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.430 25.724 ot_scf_mini 99 9.5 0.002 0.002 24.040 24.089 multiply_cannon 2055 13.4 0.344 0.369 21.408 21.943 multiply_cannon_loop 2055 14.4 0.340 0.344 19.716 20.040 cp_fm_upper_to_full 70 14.2 12.571 18.115 12.571 18.115 ot_mini 99 10.5 0.001 0.001 13.390 13.439 rebuild_ks_matrix 110 8.3 0.001 0.001 13.042 13.106 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.042 13.106 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.860 11.918 dbcsr_complete_redistribute 325 12.2 1.027 1.048 7.396 10.533 multiply_cannon_multrec 8220 15.4 4.365 4.546 9.585 9.682 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.366 9.504 qs_ot_get_derivative 99 11.5 0.001 0.001 8.866 8.917 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.803 8.911 mp_alltoall_i22 605 13.7 5.435 8.599 5.435 8.599 mp_waitall_1 87304 16.6 7.537 8.314 7.537 8.314 sum_up_and_integrate 110 10.3 0.090 0.091 6.515 6.529 integrate_v_rspace 110 11.3 0.003 0.003 6.425 6.439 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.255 6.295 calculate_rho_elec 110 8.6 0.227 0.227 6.254 6.294 init_scf_run 11 5.9 0.000 0.001 5.713 5.714 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.713 5.713 qs_ot_get_p 110 10.4 0.001 0.001 5.558 5.621 make_m2s 4110 13.4 0.043 0.043 5.134 5.566 make_images 4110 14.4 0.873 0.909 4.949 5.380 dbcsr_mm_accdrv_process 11614 15.7 3.208 3.608 5.077 5.344 cp_fm_cholesky_invert 11 10.9 5.317 5.321 5.317 5.321 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.606 5.086 apply_single 110 13.6 0.000 0.000 4.606 5.086 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.633 4.899 ot_diis_step 99 11.5 0.015 0.015 4.501 4.501 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.024 4.029 multiply_cannon_sync_h2d 8220 15.4 3.948 3.953 3.948 3.953 grid_integrate_task_list 110 12.3 3.660 3.716 3.660 3.716 pw_transfer 1331 11.6 0.073 0.074 3.665 3.671 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.644 3.655 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.583 3.584 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.540 3.546 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.755 3.419 make_images_data 4110 15.4 0.041 0.045 2.803 3.358 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.825 3.293 wfi_extrapolate 11 7.9 0.001 0.001 3.154 3.154 calculate_dm_sparse 110 9.5 0.001 0.001 3.113 3.137 fft_wrap_pw1pw2_140 451 13.1 0.217 0.219 3.068 3.077 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.039 3.040 cp_fm_diag_elpa_base 48 14.0 2.514 2.710 3.038 3.038 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.993 2.996 density_rs2pw 110 9.6 0.004 0.004 2.900 2.917 fft3d_ps 1111 14.6 1.260 1.274 2.748 2.752 grid_collocate_task_list 110 9.6 2.629 2.655 2.629 2.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.508 2.532 calculate_first_density_matrix 1 7.0 0.000 0.000 2.457 2.458 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.371 2.397 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.184 2.249 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.201 rs_pw_transfer 902 11.9 0.011 0.011 2.123 2.165 potential_pw2rs 110 12.3 0.021 0.021 2.034 2.037 cp_fm_cholesky_decompose 22 10.9 2.000 2.021 2.000 2.021 qs_create_task_list 11 7.9 0.000 0.001 1.907 1.953 generate_qs_task_list 11 8.9 0.744 0.799 1.907 1.952 jit_kernel_multiply 10 15.4 1.668 1.808 1.668 1.808 mp_alltoall_d11v 2046 13.8 1.749 1.792 1.749 1.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.791 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.684 1.729 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.695 1.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=84.517000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1118.545455, yerr=22.022153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 626.278400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 56784. MP_Allreduce 11248 786. MP_Sync 168 MP_Alltoall 2210 2588608. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 201.067 201.068 qs_mol_dyn_low 1 2.0 0.003 0.003 200.629 200.642 qs_forces 11 3.9 0.023 0.026 200.536 200.537 qs_energies 11 4.9 0.002 0.003 194.980 195.013 scf_env_do_scf 11 5.9 0.001 0.001 178.521 178.525 scf_env_do_scf_inner_loop 116 6.6 0.002 0.008 157.924 157.925 dbcsr_multiply_generic 2485 12.5 0.174 0.177 122.513 123.119 velocity_verlet 10 3.0 0.001 0.002 120.640 120.644 qs_scf_new_mos 116 7.6 0.001 0.001 119.299 119.586 qs_scf_loop_do_ot 116 8.6 0.001 0.001 119.298 119.585 ot_scf_mini 116 9.6 0.003 0.003 112.787 113.105 multiply_cannon 2485 13.5 0.236 0.245 99.953 102.524 multiply_cannon_loop 2485 14.5 2.080 2.164 97.771 99.602 ot_mini 116 10.6 0.001 0.001 64.379 64.670 multiply_cannon_multrec 59640 15.5 32.950 35.653 41.155 42.844 qs_ot_get_derivative 116 11.6 0.001 0.001 39.889 40.165 rebuild_ks_matrix 127 8.3 0.001 0.001 33.319 33.709 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.318 33.708 mp_waitall_1 288940 16.2 27.802 30.991 27.802 30.991 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.940 30.297 multiply_cannon_sync_h2d 59640 15.5 27.367 30.168 27.367 30.168 qs_ot_get_p 127 10.4 0.001 0.001 26.741 27.160 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.084 24.545 apply_single 127 13.6 0.001 0.001 24.084 24.544 ot_diis_step 116 11.6 0.007 0.008 24.161 24.162 init_scf_loop 11 6.9 0.000 0.000 20.523 20.524 qs_ot_p2m_diag 82 11.4 0.077 0.097 20.123 20.214 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.160 18.387 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.542 17.543 multiply_cannon_metrocomm3 59640 15.5 0.110 0.116 15.495 17.418 prepare_preconditioner 11 7.9 0.000 0.000 15.909 15.945 make_preconditioner 11 8.9 0.000 0.000 15.909 15.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.152 15.310 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.786 14.791 cp_fm_redistribute_end 82 14.4 11.675 14.714 11.687 14.718 cp_fm_diag_elpa_base 82 14.4 2.987 14.436 3.019 14.550 sum_up_and_integrate 127 10.3 0.052 0.063 14.102 14.118 integrate_v_rspace 127 11.3 0.003 0.004 14.049 14.067 make_m2s 4970 13.5 0.103 0.111 13.357 13.715 make_images 4970 14.5 0.401 0.418 13.177 13.544 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.547 12.604 calculate_rho_elec 127 8.7 0.044 0.063 12.546 12.603 init_scf_run 11 5.9 0.000 0.001 12.329 12.330 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.329 12.329 mp_sum_l 7804 13.0 7.951 9.377 7.951 9.377 cp_fm_cholesky_invert 11 10.9 9.141 9.148 9.141 9.148 wfi_extrapolate 11 7.9 0.001 0.001 9.045 9.045 calculate_dm_sparse 127 9.5 0.001 0.002 8.312 8.417 dbcsr_mm_accdrv_process 123452 16.2 3.183 3.397 7.772 8.301 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.956 8.082 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.746 7.834 multiply_cannon_metrocomm1 59640 15.5 0.086 0.091 5.865 7.647 grid_integrate_task_list 127 12.3 7.071 7.495 7.071 7.495 make_images_data 4970 15.5 0.068 0.074 6.470 7.469 density_rs2pw 127 9.7 0.006 0.007 6.479 7.053 pw_transfer 1535 11.6 0.074 0.102 6.590 6.879 hybrid_alltoall_any 5155 16.4 0.288 2.236 5.667 6.834 rs_pw_transfer 1038 11.9 0.017 0.019 6.182 6.744 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.654 6.662 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 6.378 6.642 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.811 5.928 fft_wrap_pw1pw2_140 519 13.2 0.434 0.483 5.505 5.688 fft3d_ps 1281 14.7 2.082 2.575 5.185 5.415 grid_collocate_task_list 127 9.7 4.687 5.131 4.687 5.131 mp_alltoall_d11v 2401 14.1 3.999 4.999 3.999 4.999 potential_pw2rs 127 12.3 0.009 0.011 4.722 4.755 cp_fm_cholesky_decompose 22 10.9 4.596 4.609 4.596 4.609 mp_sum_d 4445 12.1 3.410 4.105 3.410 4.105 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=201.068000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.545455, yerr=6.272068 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.166473E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 829.308928E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069299E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842592400 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57311. MP_Allreduce 11232 947. MP_Sync 168 MP_Alltoall 1955 4612762. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.184 188.893 188.895 qs_mol_dyn_low 1 2.0 0.005 0.010 188.274 188.286 qs_forces 11 3.9 0.013 0.044 188.122 188.125 qs_energies 11 4.9 0.003 0.012 181.225 181.238 scf_env_do_scf 11 5.9 0.001 0.002 164.597 164.607 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.711 131.715 velocity_verlet 10 3.0 0.006 0.013 117.941 117.944 dbcsr_multiply_generic 2485 12.5 0.183 0.190 97.006 98.256 qs_scf_new_mos 116 7.6 0.001 0.001 93.414 94.018 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.413 94.017 ot_scf_mini 116 9.6 0.004 0.005 88.652 89.286 multiply_cannon 2485 13.5 0.474 0.522 77.164 81.445 multiply_cannon_loop 2485 14.5 1.250 1.286 73.871 76.258 ot_mini 116 10.6 0.004 0.011 49.316 49.885 mp_waitall_1 224796 16.4 24.542 38.396 24.542 38.396 multiply_cannon_multrec 29820 15.5 21.939 26.245 31.493 36.182 rebuild_ks_matrix 127 8.3 0.001 0.001 32.784 33.321 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.023 32.783 33.320 init_scf_loop 11 6.9 0.001 0.003 32.798 32.803 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.411 29.907 prepare_preconditioner 11 7.9 0.000 0.000 28.350 28.430 make_preconditioner 11 8.9 0.000 0.001 28.350 28.430 qs_ot_get_derivative 116 11.6 0.001 0.002 27.545 28.168 multiply_cannon_metrocomm3 29820 15.5 0.094 0.099 15.541 27.793 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.052 27.558 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 21.840 22.995 apply_single 127 13.6 0.001 0.001 21.839 22.995 multiply_cannon_sync_h2d 29820 15.5 19.298 22.005 19.298 22.005 qs_ot_get_p 127 10.4 0.001 0.002 21.207 21.928 ot_diis_step 116 11.6 0.023 0.039 21.594 21.598 cp_fm_cholesky_invert 11 10.9 16.545 16.557 16.545 16.557 qs_ot_p2m_diag 82 11.4 0.187 0.215 16.399 16.433 make_m2s 4970 13.5 0.088 0.094 14.115 15.706 make_images 4970 14.5 1.136 1.326 13.908 15.501 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.271 15.274 sum_up_and_integrate 127 10.3 0.061 0.069 14.301 14.346 integrate_v_rspace 127 11.3 0.003 0.004 14.240 14.288 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.761 12.789 calculate_rho_elec 127 8.7 0.087 0.105 12.761 12.788 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.227 12.234 cp_fm_redistribute_end 82 14.4 7.139 12.162 7.151 12.163 cp_fm_diag_elpa_base 82 14.4 4.783 11.648 4.996 12.066 init_scf_run 11 5.9 0.000 0.001 11.743 11.745 scf_env_initial_rho_setup 11 6.9 0.000 0.002 11.743 11.744 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.798 11.254 multiply_cannon_metrocomm4 27335 15.5 0.095 0.114 3.634 10.529 make_images_data 4970 15.5 0.065 0.073 8.613 10.435 mp_irecv_dv 68888 16.3 3.441 10.132 3.441 10.132 hybrid_alltoall_any 5155 16.4 0.338 1.480 7.321 9.828 dbcsr_mm_accdrv_process 61726 16.2 4.475 5.259 9.020 9.610 wfi_extrapolate 11 7.9 0.003 0.008 8.349 8.350 pw_transfer 1535 11.6 0.082 0.097 7.655 7.763 grid_integrate_task_list 127 12.3 7.136 7.549 7.136 7.549 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.428 7.533 density_rs2pw 127 9.7 0.006 0.006 6.672 7.091 cp_fm_cholesky_decompose 22 10.9 6.953 7.042 6.953 7.042 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.163 6.895 fft_wrap_pw1pw2_140 519 13.2 0.456 0.494 6.484 6.599 calculate_dm_sparse 127 9.5 0.001 0.002 6.419 6.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.388 6.395 rs_pw_transfer 1038 11.9 0.014 0.015 5.760 6.170 mp_sum_l 7804 13.0 4.007 5.938 4.007 5.938 fft3d_ps 1281 14.7 2.751 2.919 5.822 5.889 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.550 5.716 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.288 5.359 grid_collocate_task_list 127 9.7 4.871 5.266 4.871 5.266 potential_pw2rs 127 12.3 0.016 0.018 4.903 4.940 mp_allgather_i34 2485 14.5 1.877 4.869 1.877 4.869 mp_alltoall_d11v 2401 14.1 3.955 4.433 3.955 4.433 mp_sum_d 4442 12.1 2.670 4.044 2.670 4.044 dbcsr_complete_redistribute 393 12.7 0.767 0.877 3.091 3.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.790 3.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.895000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.181818, yerr=3.562894 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/17/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 2.928533E+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 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 942.620672E+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 931531083728 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 0.040 0.085 174.938 174.942 qs_mol_dyn_low 1 2.0 0.003 0.004 174.485 174.498 qs_forces 11 3.9 0.003 0.003 174.378 174.384 qs_energies 11 4.9 0.020 0.043 167.869 167.880 scf_env_do_scf 11 5.9 0.001 0.002 152.206 152.207 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.326 117.327 velocity_verlet 10 3.0 0.002 0.002 111.990 111.992 dbcsr_multiply_generic 2507 12.6 0.186 0.210 81.391 82.428 qs_scf_new_mos 117 7.6 0.001 0.001 81.035 81.376 qs_scf_loop_do_ot 117 8.6 0.001 0.002 81.034 81.375 ot_scf_mini 117 9.6 0.004 0.004 76.915 77.275 multiply_cannon 2507 13.6 0.498 0.525 61.721 66.119 multiply_cannon_loop 2507 14.6 0.861 0.888 58.533 60.846 ot_mini 117 10.6 0.001 0.001 42.301 42.676 init_scf_loop 11 6.9 0.000 0.001 34.782 34.782 mp_waitall_1 178456 16.5 24.882 34.289 24.882 34.289 prepare_preconditioner 11 7.9 0.000 0.000 30.794 30.843 make_preconditioner 11 8.9 0.000 0.002 30.794 30.843 rebuild_ks_matrix 128 8.3 0.001 0.001 29.946 30.474 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 29.945 30.474 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.467 29.843 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.036 27.513 multiply_cannon_multrec 20056 15.6 13.647 16.536 22.298 25.289 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 14.965 24.027 qs_ot_get_derivative 117 11.6 0.001 0.002 22.566 22.929 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.807 20.770 apply_single 128 13.6 0.001 0.001 19.807 20.770 qs_ot_get_p 128 10.4 0.001 0.001 20.020 20.444 ot_diis_step 117 11.6 0.018 0.018 19.630 19.630 multiply_cannon_sync_h2d 20056 15.6 14.238 15.906 14.238 15.906 qs_ot_p2m_diag 83 11.4 0.266 0.272 15.804 15.813 make_m2s 5014 13.6 0.080 0.087 14.620 15.621 make_images 5014 14.6 1.164 1.254 14.387 15.385 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.795 14.797 cp_fm_cholesky_invert 11 10.9 14.357 14.365 14.357 14.365 sum_up_and_integrate 128 10.3 0.070 0.076 13.960 13.984 integrate_v_rspace 128 11.3 0.003 0.004 13.890 13.914 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.863 12.891 calculate_rho_elec 128 8.7 0.132 0.146 12.862 12.890 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.767 11.768 cp_fm_redistribute_end 83 14.4 4.432 11.699 4.447 11.700 cp_fm_diag_elpa_base 83 14.4 6.828 11.102 7.233 11.602 init_scf_run 11 5.9 0.000 0.001 10.577 10.577 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.577 10.577 make_images_data 5014 15.6 0.060 0.068 9.043 10.391 hybrid_alltoall_any 5200 16.5 0.433 1.991 7.859 9.474 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.402 9.356 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.860 9.119 mp_irecv_dv 50230 16.2 3.278 9.106 3.278 9.106 dbcsr_mm_accdrv_process 41502 16.2 4.398 5.191 8.102 8.267 grid_integrate_task_list 128 12.3 7.307 7.726 7.307 7.726 cp_fm_upper_to_full 105 14.5 5.754 7.517 5.754 7.517 pw_transfer 1547 11.6 0.083 0.102 7.352 7.466 cp_fm_cholesky_decompose 22 10.9 7.307 7.328 7.307 7.328 wfi_extrapolate 11 7.9 0.001 0.001 7.318 7.318 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.126 7.246 density_rs2pw 128 9.7 0.005 0.006 6.372 6.648 fft_wrap_pw1pw2_140 523 13.2 0.470 0.518 6.266 6.401 dbcsr_complete_redistribute 395 12.7 1.167 1.193 4.412 6.179 calculate_dm_sparse 128 9.5 0.001 0.001 5.711 5.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.608 5.615 grid_collocate_task_list 128 9.7 5.075 5.522 5.075 5.522 rs_pw_transfer 1046 11.9 0.013 0.014 5.239 5.518 fft3d_ps 1291 14.7 2.694 2.914 5.402 5.483 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.513 5.169 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.243 5.008 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.733 4.863 mp_alltoall_d11v 2415 14.1 4.016 4.744 4.016 4.744 mp_sum_l 7870 13.0 3.258 4.669 3.258 4.669 mp_allgather_i34 2507 14.6 1.713 4.592 1.713 4.592 potential_pw2rs 128 12.3 0.021 0.023 4.465 4.478 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.029 4.062 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.308 4.047 mp_alltoall_i22 716 14.1 1.904 3.871 1.904 3.871 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.728 3.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.942000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.090909, yerr=13.740812 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.171145E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 1067. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.065 185.562 185.563 qs_mol_dyn_low 1 2.0 0.003 0.004 185.085 185.098 qs_forces 11 3.9 0.014 0.026 184.963 184.969 qs_energies 11 4.9 0.002 0.004 177.917 177.941 scf_env_do_scf 11 5.9 0.001 0.001 161.417 161.426 velocity_verlet 10 3.0 0.002 0.003 123.250 123.253 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.750 114.752 dbcsr_multiply_generic 2507 12.6 0.190 0.193 78.811 79.556 qs_scf_new_mos 117 7.6 0.001 0.001 79.231 79.497 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.230 79.496 ot_scf_mini 117 9.6 0.003 0.004 74.732 75.005 multiply_cannon 2507 13.6 0.558 0.595 54.548 57.448 multiply_cannon_loop 2507 14.6 1.178 1.217 50.850 52.592 init_scf_loop 11 6.9 0.004 0.027 46.544 46.545 prepare_preconditioner 11 7.9 0.000 0.000 42.409 42.432 make_preconditioner 11 8.9 0.000 0.001 42.409 42.432 ot_mini 117 10.6 0.001 0.001 41.752 42.005 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.820 40.926 multiply_cannon_multrec 30084 15.6 14.132 19.533 26.028 30.857 rebuild_ks_matrix 128 8.3 0.001 0.001 29.041 29.246 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 29.041 29.245 mp_waitall_1 153770 16.5 17.276 27.209 17.276 27.209 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.247 26.426 qs_ot_get_derivative 117 11.6 0.001 0.002 22.337 22.610 make_m2s 5014 13.6 0.095 0.098 20.013 21.144 make_images 5014 14.6 1.933 2.137 19.709 20.845 qs_ot_get_p 128 10.4 0.001 0.001 19.343 19.635 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.852 19.338 apply_single 128 13.6 0.001 0.001 18.851 19.338 ot_diis_step 117 11.6 0.017 0.018 19.289 19.291 cp_fm_upper_to_full 105 14.7 11.293 16.747 11.293 16.747 cp_fm_cholesky_invert 11 10.9 16.015 16.024 16.015 16.024 multiply_cannon_metrocomm3 30084 15.6 0.046 0.049 6.319 15.332 qs_ot_p2m_diag 83 11.4 0.343 0.391 15.204 15.256 sum_up_and_integrate 128 10.3 0.079 0.086 13.943 13.971 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.943 13.946 integrate_v_rspace 128 11.3 0.003 0.004 13.863 13.892 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.061 13.114 calculate_rho_elec 128 8.7 0.174 0.190 13.061 13.114 dbcsr_complete_redistribute 395 12.7 1.504 1.621 9.202 12.998 multiply_cannon_sync_h2d 30084 15.6 11.759 12.890 11.759 12.890 make_images_data 5014 15.6 0.063 0.068 10.833 12.606 dbcsr_mm_accdrv_process 62264 16.2 7.424 8.366 11.475 11.980 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.810 11.616 hybrid_alltoall_any 5200 16.5 0.515 2.167 9.596 11.498 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.885 10.888 cp_fm_redistribute_end 83 14.4 1.863 10.807 1.875 10.813 cp_fm_diag_elpa_base 83 14.4 8.304 10.200 8.903 10.701 init_scf_run 11 5.9 0.000 0.001 10.580 10.581 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.580 10.581 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.567 10.279 mp_alltoall_i22 716 14.1 5.708 9.544 5.708 9.544 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.242 9.443 grid_integrate_task_list 128 12.3 7.525 7.904 7.525 7.904 pw_transfer 1547 11.6 0.083 0.097 7.730 7.799 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.502 7.574 cp_fm_cholesky_decompose 22 10.9 7.462 7.548 7.462 7.548 wfi_extrapolate 11 7.9 0.001 0.001 7.459 7.459 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.790 6.996 fft_wrap_pw1pw2_140 523 13.2 0.474 0.483 6.654 6.736 mp_irecv_dv 76098 16.2 2.645 6.734 2.645 6.734 density_rs2pw 128 9.7 0.005 0.006 6.243 6.626 calculate_dm_sparse 128 9.5 0.001 0.001 6.157 6.263 fft3d_ps 1291 14.7 2.796 2.851 5.738 5.788 grid_collocate_task_list 128 9.7 5.208 5.560 5.208 5.560 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.394 5.446 mp_alltoall_d11v 2415 14.1 4.774 5.207 4.774 5.207 rs_pw_transfer 1046 11.9 0.013 0.014 4.738 5.084 qs_energies_init_hamiltonians 11 5.9 0.009 0.016 4.466 4.478 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.325 4.411 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.313 4.400 potential_pw2rs 128 12.3 0.024 0.024 4.261 4.271 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.147 4.218 mp_allgather_i34 2507 14.6 1.796 3.907 1.796 3.907 dbcsr_special_finalize 12535 15.6 0.067 0.069 3.502 3.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.563000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1099.090909, yerr=23.317783 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.467064E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58279. MP_Allreduce 10990 1175. MP_Sync 86 MP_Alltoall 1700 18828160. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.044 168.602 168.602 qs_mol_dyn_low 1 2.0 0.003 0.004 168.189 168.203 qs_forces 11 3.9 0.003 0.003 168.088 168.091 qs_energies 11 4.9 0.001 0.002 160.824 160.829 scf_env_do_scf 11 5.9 0.001 0.001 143.749 143.758 velocity_verlet 10 3.0 0.006 0.007 109.981 109.985 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.051 108.052 dbcsr_multiply_generic 2485 12.5 0.189 0.195 72.993 73.319 qs_scf_new_mos 116 7.6 0.001 0.001 73.078 73.181 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.077 73.180 ot_scf_mini 116 9.6 0.003 0.004 68.679 68.759 multiply_cannon 2485 13.5 0.586 0.629 53.900 57.461 multiply_cannon_loop 2485 14.5 0.446 0.461 49.311 50.204 ot_mini 116 10.6 0.001 0.001 38.815 38.881 init_scf_loop 11 6.9 0.000 0.000 35.547 35.548 mp_waitall_1 128490 16.6 24.920 32.301 24.920 32.301 prepare_preconditioner 11 7.9 0.000 0.000 31.555 31.592 make_preconditioner 11 8.9 0.000 0.000 31.555 31.592 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.478 29.730 rebuild_ks_matrix 127 8.3 0.001 0.001 28.277 28.464 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 28.277 28.463 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.805 25.974 multiply_cannon_multrec 9940 15.5 10.382 14.198 17.899 21.130 ot_diis_step 116 11.6 0.019 0.020 19.451 19.451 qs_ot_get_derivative 116 11.6 0.002 0.002 19.297 19.377 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.152 19.345 apply_single 127 13.6 0.001 0.001 19.152 19.345 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 11.888 18.873 cp_fm_cholesky_invert 11 10.9 18.105 18.111 18.105 18.111 make_m2s 4970 13.5 0.065 0.070 15.517 17.905 make_images 4970 14.5 2.277 2.810 15.214 17.603 qs_ot_get_p 127 10.4 0.001 0.001 17.291 17.375 sum_up_and_integrate 127 10.3 0.102 0.110 13.805 13.927 integrate_v_rspace 127 11.3 0.003 0.004 13.702 13.820 qs_ot_p2m_diag 82 11.4 0.490 0.496 13.648 13.665 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.938 12.967 calculate_rho_elec 127 8.7 0.255 0.266 12.937 12.966 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.532 12.533 multiply_cannon_sync_h2d 9940 15.5 11.542 12.267 11.542 12.267 make_images_data 4970 15.5 0.051 0.059 9.321 11.902 hybrid_alltoall_any 5155 16.4 0.830 3.771 9.178 11.733 init_scf_run 11 5.9 0.000 0.001 10.323 10.323 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.323 10.323 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.578 9.582 cp_fm_diag_elpa_base 82 14.4 9.344 9.415 9.575 9.579 grid_integrate_task_list 127 12.3 7.690 8.053 7.690 8.053 cp_fm_cholesky_decompose 22 10.9 7.853 7.952 7.853 7.952 dbcsr_mm_accdrv_process 20590 16.1 2.750 3.552 7.152 7.793 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.680 7.732 pw_transfer 1535 11.6 0.082 0.090 7.542 7.680 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.317 7.461 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.308 7.311 wfi_extrapolate 11 7.9 0.001 0.001 7.146 7.146 mp_allgather_i34 2485 14.5 2.803 6.627 2.803 6.627 fft_wrap_pw1pw2_140 519 13.2 0.491 0.509 6.458 6.606 calculate_dm_sparse 127 9.5 0.001 0.001 6.141 6.199 density_rs2pw 127 9.7 0.005 0.006 5.827 6.092 grid_collocate_task_list 127 9.7 5.441 5.743 5.441 5.743 fft3d_ps 1281 14.7 2.708 2.772 5.494 5.625 dbcsr_complete_redistribute 393 12.7 2.132 2.208 5.091 5.457 mp_alltoall_d11v 2401 14.1 4.518 5.286 4.518 5.286 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.250 5.257 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.216 5.217 rs_pw_transfer 1038 11.9 0.012 0.013 4.117 4.408 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.862 4.360 mp_irecv_dv 28618 15.9 1.824 4.289 1.824 4.289 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.191 4.241 potential_pw2rs 127 12.3 0.026 0.027 4.055 4.181 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.563 3.845 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.633 3.664 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.305 3.614 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.506 3.576 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.419 3.442 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.365 3.383 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.602000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1375.636364, yerr=35.886875 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 2.263450E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 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 0.045 0.054 285.691 285.692 qs_mol_dyn_low 1 2.0 0.003 0.006 285.016 285.030 qs_forces 11 3.9 0.005 0.009 284.909 284.911 qs_energies 11 4.9 0.002 0.002 276.382 276.398 scf_env_do_scf 11 5.9 0.001 0.001 254.706 254.719 velocity_verlet 10 3.0 0.029 0.031 207.183 207.191 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 131.260 131.261 init_scf_loop 11 6.9 0.000 0.000 123.192 123.194 prepare_preconditioner 11 7.9 0.000 0.000 118.582 118.611 make_preconditioner 11 8.9 0.000 0.000 118.582 118.611 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.720 115.720 qs_scf_new_mos 118 7.6 0.001 0.001 89.526 89.615 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.525 89.614 ot_scf_mini 118 9.6 0.004 0.004 84.792 84.894 dbcsr_multiply_generic 2534 12.6 0.215 0.222 81.967 82.118 cp_fm_upper_to_full 106 14.8 52.549 75.357 52.549 75.357 multiply_cannon 2534 13.6 0.687 0.726 58.085 59.266 multiply_cannon_loop 2534 14.6 0.479 0.485 54.517 54.900 ot_mini 118 10.6 0.001 0.001 43.856 43.990 dbcsr_complete_redistribute 397 12.7 3.999 4.043 29.479 42.464 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.117 39.084 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.817 36.638 mp_alltoall_i22 720 14.1 21.672 34.789 21.672 34.789 rebuild_ks_matrix 129 8.3 0.001 0.001 33.296 33.471 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 33.296 33.471 cp_fm_cholesky_invert 11 10.9 33.196 33.202 33.196 33.202 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.144 31.308 mp_waitall_1 106586 16.7 26.952 29.800 26.952 29.800 qs_ot_get_p 129 10.4 0.001 0.001 25.669 25.855 qs_ot_get_derivative 118 11.6 0.002 0.002 23.700 23.799 qs_ot_p2m_diag 84 11.4 0.889 0.894 21.684 21.712 ot_diis_step 118 11.6 0.022 0.023 20.102 20.102 make_m2s 5068 13.6 0.077 0.079 19.175 20.037 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.951 19.952 make_images 5068 14.6 3.766 3.899 18.697 19.562 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.136 19.456 apply_single 129 13.6 0.001 0.001 19.136 19.456 multiply_cannon_metrocomm3 10136 15.6 0.023 0.023 18.116 18.518 multiply_cannon_multrec 10136 15.6 10.521 12.330 17.969 18.102 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.775 16.776 cp_fm_diag_elpa_base 84 14.4 12.348 13.961 16.772 16.773 multiply_cannon_sync_h2d 10136 15.6 15.774 15.803 15.774 15.803 sum_up_and_integrate 129 10.3 0.195 0.197 15.432 15.531 integrate_v_rspace 129 11.3 0.004 0.004 15.236 15.336 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.753 14.780 calculate_rho_elec 129 8.7 0.486 0.487 14.752 14.779 hybrid_alltoall_any 5255 16.5 1.316 3.076 10.175 11.802 init_scf_run 11 5.9 0.000 0.001 11.748 11.748 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.748 11.748 make_images_data 5068 15.6 0.060 0.065 10.022 11.621 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.326 9.397 dbcsr_mm_accdrv_process 20954 16.1 3.892 5.841 7.208 9.108 cp_fm_cholesky_decompose 22 10.9 8.772 8.813 8.772 8.813 grid_integrate_task_list 129 12.3 8.604 8.767 8.604 8.767 wfi_extrapolate 11 7.9 0.001 0.001 8.508 8.508 pw_transfer 1559 11.6 0.094 0.095 8.060 8.067 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.959 7.962 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 7.818 7.824 fft_wrap_pw1pw2_140 527 13.2 0.532 0.538 6.943 6.963 calculate_dm_sparse 129 9.5 0.001 0.001 6.606 6.698 mp_alltoall_d11v 2429 14.1 6.562 6.681 6.562 6.681 grid_collocate_task_list 129 9.7 6.388 6.419 6.388 6.419 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.341 6.399 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.017 6.146 density_rs2pw 129 9.7 0.005 0.005 5.864 5.892 fft3d_ps 1301 14.7 2.787 2.809 5.858 5.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=285.692000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2130.000000, yerr=50.722059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 1.259987E+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 0.038 0.125 84.911 84.912 qs_energies 1 2.0 0.000 0.000 84.171 84.222 ls_scf 1 3.0 0.000 0.002 83.240 83.259 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.198 72.381 multiply_cannon 111 7.7 0.018 0.021 55.765 56.948 multiply_cannon_loop 111 8.7 0.211 0.228 52.376 53.652 ls_scf_main 1 4.0 0.000 0.002 52.012 52.013 density_matrix_trs4 2 5.0 0.002 0.003 46.556 46.648 ls_scf_init_scf 1 4.0 0.000 0.001 28.184 28.185 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.124 27.168 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.031 25.054 mp_waitall_1 11316 10.9 21.846 25.017 21.846 25.017 multiply_cannon_multrec 2664 9.7 8.190 8.965 15.484 17.234 multiply_cannon_sync_h2d 2664 9.7 13.955 15.470 13.955 15.470 make_m2s 222 7.7 0.008 0.010 12.865 13.408 make_images 222 8.7 0.099 0.108 12.843 13.388 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.350 12.509 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.397 8.379 make_images_data 222 9.7 0.004 0.005 7.424 7.931 dbcsr_mm_accdrv_process 4760 10.4 0.504 0.607 6.912 7.897 hybrid_alltoall_any 227 10.6 0.215 1.832 6.364 7.637 dbcsr_mm_accdrv_process_sort 4760 11.4 6.210 7.110 6.210 7.110 calculate_norms 4752 9.8 5.561 6.234 5.561 6.234 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.014 5.289 mp_sum_l 807 5.4 3.091 4.590 3.091 4.590 make_images_sizes 222 9.7 0.000 0.000 0.695 3.445 mp_alltoall_i44 222 10.7 0.695 3.444 0.695 3.444 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.203 3.444 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.029 3.272 mp_irecv_dv 6231 10.9 2.012 3.243 2.012 3.243 arnoldi_extremal 4 6.8 0.000 0.000 3.202 3.232 arnoldi_normal_ev 4 7.8 0.001 0.003 3.202 3.232 build_subspace 16 8.4 0.009 0.012 3.102 3.104 ls_scf_post 1 4.0 0.003 0.034 3.044 3.064 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.889 dbcsr_special_finalize 555 9.7 0.005 0.006 2.392 2.766 dbcsr_merge_single_wm 555 10.7 0.461 0.586 2.384 2.759 make_images_pack 222 9.7 2.207 2.622 2.209 2.624 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.306 2.552 dbcsr_sort_data 658 11.4 2.180 2.511 2.180 2.511 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.461 2.067 2.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.239 2.385 buffer_matrices_ensure_size 222 8.7 1.765 2.060 1.765 2.060 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.753 1.756 rebuild_ks_matrix 3 7.3 0.000 0.000 1.744 1.746 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.744 1.746 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.912000, 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/99df2e665baa944797556ab754e3afa8e99d62d4_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 2.102964E+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 0.033 0.117 90.957 90.957 qs_energies 1 2.0 0.000 0.000 90.398 90.429 ls_scf 1 3.0 0.002 0.013 89.085 89.091 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.241 75.551 multiply_cannon 111 7.7 0.028 0.040 53.450 57.260 ls_scf_main 1 4.0 0.001 0.016 54.728 54.728 multiply_cannon_loop 111 8.7 0.116 0.122 50.154 53.576 density_matrix_trs4 2 5.0 0.002 0.003 49.071 49.283 ls_scf_init_scf 1 4.0 0.000 0.000 30.778 30.779 mp_waitall_1 9246 10.9 21.092 29.965 21.092 29.965 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.653 29.734 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.249 27.265 multiply_cannon_multrec 1332 9.7 13.096 16.776 22.281 27.091 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.039 20.838 make_m2s 222 7.7 0.006 0.007 14.983 15.623 make_images 222 8.7 1.574 1.955 14.953 15.594 dbcsr_mm_accdrv_process 4041 10.4 0.278 0.444 8.790 10.319 dbcsr_mm_accdrv_process_sort 4041 11.4 8.384 9.893 8.384 9.893 make_images_data 222 9.7 0.004 0.004 8.450 9.289 hybrid_alltoall_any 227 10.6 0.523 2.495 7.969 8.832 mp_sum_l 807 5.4 5.556 8.457 5.556 8.457 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.220 7.765 mp_irecv_dv 3311 11.0 3.200 7.713 3.200 7.713 calculate_norms 2376 9.8 5.993 6.720 5.993 6.720 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.263 6.599 multiply_cannon_sync_h2d 1332 9.7 4.812 5.901 4.812 5.901 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.059 5.259 arnoldi_extremal 4 6.8 0.000 0.000 4.646 4.671 arnoldi_normal_ev 4 7.8 0.002 0.005 4.646 4.671 build_subspace 16 8.4 0.014 0.021 4.388 4.390 ls_scf_post 1 4.0 0.001 0.005 3.578 3.581 ls_scf_store_result 1 5.0 0.000 0.000 3.261 3.397 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.114 3.345 dbcsr_matrix_vector_mult_local 304 10.0 2.730 3.212 2.732 3.214 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.231 2.892 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.505 2.587 make_images_pack 222 9.7 2.023 2.405 2.026 2.407 mp_allgather_i34 111 8.7 0.994 2.261 0.994 2.261 dbcsr_sort_data 436 11.2 1.846 2.065 1.846 2.065 dbcsr_data_new 4174 10.1 1.604 1.836 1.604 1.836 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.817 1.819 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.957000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 2.698248E+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 0.047 0.073 93.798 93.799 qs_energies 1 2.0 0.000 0.000 92.998 93.004 ls_scf 1 3.0 0.005 0.038 91.567 91.570 dbcsr_multiply_generic 111 6.7 0.016 0.018 76.270 76.599 ls_scf_main 1 4.0 0.000 0.003 57.419 57.426 multiply_cannon 111 7.7 0.038 0.082 52.991 57.236 multiply_cannon_loop 111 8.7 0.100 0.111 49.298 52.304 density_matrix_trs4 2 5.0 0.002 0.003 51.484 51.621 mp_waitall_1 7374 11.0 23.737 32.982 23.737 32.982 ls_scf_init_scf 1 4.0 0.000 0.002 30.540 30.541 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.400 29.462 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.010 27.024 multiply_cannon_multrec 888 9.7 12.578 15.395 21.099 24.473 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.240 22.054 make_m2s 222 7.7 0.006 0.007 16.773 18.080 make_images 222 8.7 1.971 2.288 16.735 18.043 make_images_data 222 9.7 0.003 0.004 9.585 10.771 hybrid_alltoall_any 227 10.6 0.619 2.865 9.230 10.533 dbcsr_mm_accdrv_process 3754 10.4 0.258 0.441 8.038 9.283 mp_sum_l 807 5.4 4.925 8.894 4.925 8.894 dbcsr_mm_accdrv_process_sort 3754 11.4 7.662 8.843 7.662 8.843 multiply_cannon_sync_h2d 888 9.7 6.086 7.373 6.086 7.373 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.439 7.145 mp_irecv_dv 2335 11.1 2.423 7.090 2.423 7.090 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.718 6.959 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.529 6.416 arnoldi_extremal 4 6.8 0.000 0.001 5.145 5.163 arnoldi_normal_ev 4 7.8 0.005 0.031 5.145 5.162 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.983 5.101 build_subspace 16 8.4 0.014 0.020 4.800 4.810 calculate_norms 1584 9.8 4.275 4.627 4.275 4.627 mp_allgather_i34 111 8.7 1.534 3.886 1.534 3.886 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.433 3.781 ls_scf_post 1 4.0 0.000 0.001 3.604 3.607 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.599 3.017 3.601 ls_scf_store_result 1 5.0 0.000 0.000 3.335 3.427 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.754 2.830 make_images_pack 222 9.7 1.813 2.126 1.816 2.128 dbcsr_sort_data 325 11.1 1.890 2.113 1.890 2.113 make_images_sizes 222 9.7 0.000 0.000 0.861 2.032 mp_alltoall_i44 222 10.7 0.860 2.032 0.860 2.032 dbcsr_data_release 9322 10.9 1.326 1.998 1.326 1.998 dbcsr_finalize 304 7.8 0.026 0.032 1.623 1.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.799000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2162.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 3.335582E+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 0.033 0.049 96.608 96.608 qs_energies 1 2.0 0.000 0.000 95.955 95.962 ls_scf 1 3.0 0.000 0.000 94.316 94.323 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.285 78.568 ls_scf_main 1 4.0 0.000 0.000 58.560 58.561 multiply_cannon 111 7.7 0.039 0.084 51.529 56.187 density_matrix_trs4 2 5.0 0.002 0.003 52.546 52.653 multiply_cannon_loop 111 8.7 0.114 0.125 46.493 49.335 ls_scf_init_scf 1 4.0 0.000 0.000 32.502 32.504 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.346 31.434 mp_waitall_1 6438 11.0 22.543 29.722 22.543 29.722 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.894 28.906 multiply_cannon_multrec 1332 9.7 14.151 17.405 21.975 24.554 make_m2s 222 7.7 0.006 0.008 20.937 22.378 make_images 222 8.7 3.131 3.588 20.887 22.330 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.135 17.273 make_images_data 222 9.7 0.004 0.004 11.606 13.282 hybrid_alltoall_any 227 10.6 0.794 3.746 10.965 12.862 dbcsr_mm_accdrv_process 3641 10.4 0.210 0.401 7.463 8.991 dbcsr_mm_accdrv_process_sort 3641 11.4 7.091 8.570 7.091 8.570 mp_sum_l 807 5.4 4.238 7.716 4.238 7.716 multiply_cannon_sync_h2d 1332 9.7 5.552 6.328 5.552 6.328 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.073 6.081 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.264 6.078 mp_irecv_dv 3229 10.9 2.050 6.011 2.050 6.011 arnoldi_extremal 4 6.8 0.000 0.000 5.209 5.224 arnoldi_normal_ev 4 7.8 0.001 0.004 5.209 5.224 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.627 4.983 build_subspace 16 8.4 0.014 0.021 4.869 4.878 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.561 4.768 calculate_norms 2376 9.8 4.187 4.544 4.187 4.544 mp_allgather_i34 111 8.7 2.166 4.458 2.166 4.458 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.558 3.880 dbcsr_matrix_vector_mult_local 304 10.0 3.173 3.695 3.175 3.697 dbcsr_sort_data 658 11.4 3.098 3.478 3.098 3.478 ls_scf_post 1 4.0 0.000 0.000 3.254 3.262 dbcsr_special_finalize 555 9.7 0.006 0.008 2.846 3.210 dbcsr_merge_single_wm 555 10.7 0.537 0.662 2.837 3.201 ls_scf_store_result 1 5.0 0.000 0.000 2.988 3.053 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.897 2.979 dbcsr_data_release 10477 10.7 1.576 2.429 1.576 2.429 dbcsr_finalize 304 7.8 0.049 0.061 1.797 1.998 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.608000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2696.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 4.624871E+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 0.051 0.084 98.091 98.092 qs_energies 1 2.0 0.000 0.000 97.301 97.307 ls_scf 1 3.0 0.000 0.000 95.381 95.386 dbcsr_multiply_generic 111 6.7 0.017 0.018 76.946 77.177 ls_scf_main 1 4.0 0.000 0.000 61.271 61.271 multiply_cannon 111 7.7 0.095 0.195 54.873 59.498 density_matrix_trs4 2 5.0 0.002 0.003 54.280 54.355 multiply_cannon_loop 111 8.7 0.070 0.076 50.274 52.047 mp_waitall_1 5481 11.0 25.661 30.599 25.661 30.599 ls_scf_init_scf 1 4.0 0.000 0.000 30.555 30.561 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.339 29.373 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.022 27.036 multiply_cannon_multrec 444 9.7 14.044 16.475 21.044 23.654 make_m2s 222 7.7 0.004 0.005 17.353 19.917 make_images 222 8.7 3.711 4.419 17.291 19.857 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.638 15.989 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.173 13.781 make_images_data 222 9.7 0.003 0.004 9.648 12.127 hybrid_alltoall_any 227 10.6 0.787 3.763 9.463 11.903 multiply_cannon_sync_h2d 444 9.7 6.596 7.969 6.596 7.969 dbcsr_mm_accdrv_process 3003 10.4 0.140 0.314 6.707 7.835 dbcsr_mm_accdrv_process_sort 3003 11.4 6.395 7.491 6.395 7.491 mp_allgather_i34 111 8.7 2.828 7.071 2.828 7.071 arnoldi_extremal 4 6.8 0.000 0.000 5.801 5.817 arnoldi_normal_ev 4 7.8 0.002 0.005 5.801 5.817 build_subspace 16 8.4 0.015 0.019 5.414 5.422 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.565 4.715 mp_sum_l 807 5.4 2.779 4.468 2.779 4.468 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.134 4.341 dbcsr_matrix_vector_mult_local 304 10.0 3.666 4.146 3.668 4.148 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.613 3.717 calculate_norms 792 9.8 3.571 3.703 3.571 3.703 mp_irecv_dv 1241 11.2 1.593 3.645 1.593 3.645 ls_scf_post 1 4.0 0.000 0.000 3.555 3.560 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.935 3.528 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.408 3.524 ls_scf_store_result 1 5.0 0.000 0.000 3.327 3.379 make_images_sizes 222 9.7 0.000 0.000 1.070 3.141 mp_alltoall_i44 222 10.7 1.070 3.140 1.070 3.140 dbcsr_finalize 304 7.8 0.062 0.077 2.196 2.284 dbcsr_merge_all 275 8.9 0.474 0.527 2.046 2.117 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.006 2.007 rebuild_ks_matrix 3 7.3 0.000 0.000 1.974 1.975 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 1.974 1.975 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.092000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3628.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99df2e665baa944797556ab754e3afa8e99d62d4_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 8.712106E+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 0.099 0.140 106.577 106.578 qs_energies 1 2.0 0.000 0.000 105.124 105.139 ls_scf 1 3.0 0.000 0.000 102.233 102.248 dbcsr_multiply_generic 111 6.7 0.023 0.027 76.556 76.670 ls_scf_main 1 4.0 0.000 0.000 64.888 64.889 density_matrix_trs4 2 5.0 0.002 0.003 56.084 56.138 multiply_cannon 111 7.7 0.119 0.173 49.422 51.086 multiply_cannon_loop 111 8.7 0.068 0.071 46.081 47.006 ls_scf_init_scf 1 4.0 0.000 0.000 33.696 33.697 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.303 32.315 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.483 29.494 mp_waitall_1 4569 11.1 21.681 25.265 21.681 25.265 make_m2s 222 7.7 0.005 0.005 23.588 24.454 make_images 222 8.7 4.563 4.950 23.482 24.346 multiply_cannon_multrec 444 9.7 17.837 18.560 22.417 23.061 hybrid_alltoall_any 227 10.6 1.657 3.617 12.620 15.445 make_images_data 222 9.7 0.003 0.003 12.868 15.163 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.220 10.837 multiply_cannon_sync_h2d 444 9.7 8.853 8.901 8.853 8.901 arnoldi_extremal 4 6.8 0.000 0.000 7.228 7.241 arnoldi_normal_ev 4 7.8 0.003 0.009 7.228 7.241 build_subspace 16 8.4 0.026 0.036 6.681 6.693 dbcsr_matrix_vector_mult 304 9.0 0.009 0.024 5.360 5.505 dbcsr_matrix_vector_mult_local 304 10.0 4.956 5.242 4.958 5.245 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.993 5.209 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.795 4.888 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.322 4.413 4.529 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.239 4.114 4.239 ls_scf_post 1 4.0 0.000 0.000 3.650 3.664 make_images_sizes 222 9.7 0.000 0.000 1.422 3.526 mp_alltoall_i44 222 10.7 1.422 3.526 1.422 3.526 ls_scf_store_result 1 5.0 0.000 0.000 3.398 3.406 calculate_norms 792 9.8 3.233 3.273 3.233 3.273 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.126 mp_allgather_i34 111 8.7 0.980 3.116 0.980 3.116 dbcsr_merge_all 275 8.9 0.892 0.917 2.866 2.907 dbcsr_data_release 12724 10.6 2.327 2.871 2.327 2.871 dbcsr_complete_redistribute 5 7.6 1.435 1.482 2.771 2.865 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.860 2.860 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.529 dbcsr_sort_data 325 11.1 2.437 2.500 2.437 2.500 dbcsr_new_transposed 4 7.5 0.244 0.252 2.278 2.291 dbcsr_frobenius_norm 74 6.6 2.055 2.135 2.192 2.234 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.220 2.223 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.206 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.206 rebuild_ks_matrix 3 7.3 0.000 0.000 2.154 2.157 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.154 2.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.578000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 99df2e665baa944797556ab754e3afa8e99d62d4 Summary: empty Status: OK