=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 99544d10ee311fd730c849cdf817355d5af06832 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 1.16.2, LIBGRPP 20231215 # # 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. # # Last update: 25.01.2024 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/01 job id: 51711931 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/02 job id: 51711932 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/03 job id: 51711933 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/04 job id: 51711934 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/05 job id: 51711935 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/06 job id: 51711936 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/07 job id: 51711937 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/08 job id: 51711938 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/09 job id: 51711939 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/10 job id: 51711940 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/11 job id: 51711941 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/12 job id: 51711942 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/13 job id: 51711943 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/14 job id: 51711944 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/15 job id: 51711945 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/16 job id: 51711946 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/17 job id: 51711947 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/18 job id: 51711948 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/19 job id: 51711949 --- 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/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/20 job id: 51711950 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/21 job id: 51711951 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/22 job id: 51711952 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/23 job id: 51711953 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/24 job id: 51711954 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/25 job id: 51711955 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/26 job id: 51711956 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/27 job id: 51711957 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 134.042 134.043 farming_run 1 2.0 133.460 133.491 134.013 134.018 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.478169E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.023 115.325 115.326 qs_energies 1 2.0 0.012 0.014 115.141 115.142 mp2_main 1 3.0 0.000 0.000 113.051 113.064 mp2_gpw_main 1 4.0 0.028 0.034 112.045 112.047 mp2_ri_gpw_compute_in 1 5.0 0.173 0.174 93.185 93.497 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.717 55.031 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.165 41.117 46.111 get_2c_integrals 1 6.0 0.008 0.009 37.839 38.294 integrate_v_rspace 273 8.0 0.443 0.457 24.672 29.531 pw_transfer 6555 10.6 0.380 0.407 26.996 27.488 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.611 26.134 grid_integrate_task_list 273 9.0 20.553 25.893 20.553 25.893 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 23.156 23.665 compute_2c_integrals 1 7.0 0.002 0.002 19.514 19.516 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.710 19.249 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.428 18.708 19.248 rpa_ri_compute_en 1 5.0 0.021 0.053 18.755 18.972 cp_fm_cholesky_decompose 12 8.2 18.246 18.697 18.246 18.697 cholesky_decomp 1 7.0 0.000 0.000 17.155 17.610 fft3d_s 5443 13.4 16.212 16.461 16.234 16.482 ao_to_mo_and_store_B_mult_1 272 7.0 10.774 15.310 10.774 15.310 calculate_wavefunction 272 8.0 5.376 5.516 12.330 13.047 rpa_num_int 1 6.0 0.000 0.001 10.655 10.664 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.628 10.645 calc_mat_Q 8 8.0 0.000 0.000 9.489 9.584 contract_S_to_Q 8 9.0 0.000 0.000 8.911 9.010 calc_potential_gpw 544 9.5 0.004 0.005 8.280 8.860 parallel_gemm_fm 14 9.1 0.000 0.000 8.502 8.620 parallel_gemm_fm_cosma 14 10.1 8.502 8.620 8.502 8.620 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.135 8.507 potential_pw2rs 545 10.0 0.106 0.108 7.543 8.168 create_integ_mat 1 6.0 0.015 0.028 7.666 7.675 collocate_single_gaussian 272 10.0 0.039 0.042 7.385 7.661 array2fm 1 7.0 0.000 0.000 6.561 7.101 pw_scatter_s 2720 13.7 4.356 4.448 4.356 4.448 pw_gather_s 2722 13.2 3.479 3.811 3.479 3.811 array2fm_buffer_send 1 8.0 2.875 3.386 2.875 3.386 pw_poisson_solve 545 10.5 1.122 1.203 2.212 2.533 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.046622, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2788.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.046 520.972 520.975 farming_run 1 2.0 520.147 520.152 520.919 520.921 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.234616E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 18854673. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 318.104 318.105 qs_energies 1 2.0 0.000 0.001 317.898 317.900 mp2_main 1 3.0 0.000 0.000 203.260 203.262 mp2_gpw_main 1 4.0 0.054 0.082 202.422 202.428 mp2_ri_gpw_compute_en 1 5.0 0.038 0.042 128.519 130.655 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.797 1.941 126.892 126.896 scf_env_do_scf 1 3.0 0.000 0.000 114.343 114.344 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.471 113.478 rebuild_ks_matrix 4 6.0 0.000 0.000 113.469 113.476 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.469 113.476 hfx_ks_matrix 4 8.0 0.001 0.001 113.127 113.132 integrate_four_center 4 9.0 0.152 0.462 113.126 113.131 mp2_ri_gpw_compute_en_expansio 172 7.0 0.468 0.511 111.338 111.958 local_gemm 172 8.0 110.870 111.478 110.870 111.478 integrate_four_center_main 4 10.0 0.133 0.629 101.824 104.370 integrate_four_center_bin 268 11.0 101.691 104.364 101.691 104.364 init_scf_loop 1 4.0 0.000 0.000 96.445 96.445 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 73.752 74.745 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.724 54.727 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.473 46.307 integrate_v_rspace 95 8.0 0.400 0.560 27.947 32.643 pw_transfer 2240 10.6 0.146 0.164 29.453 29.873 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.476 28.894 grid_integrate_task_list 95 9.0 23.371 28.222 23.371 28.222 ao_to_mo_and_store_B_mult_1 91 7.0 10.586 28.184 10.586 28.184 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.236 26.586 get_2c_integrals 1 6.0 0.000 0.000 19.940 19.970 compute_2c_integrals 1 7.0 0.002 0.003 18.926 18.934 fft3d_s 1823 13.4 18.486 18.859 18.500 18.873 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.570 18.812 mp2_eri_2c_integrate_gpw 1 9.0 1.752 1.848 18.569 18.812 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.897 17.897 calculate_wavefunction 91 8.0 2.009 2.039 9.613 9.845 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.516 7.664 9.109 potential_pw2rs 186 10.0 0.033 0.034 8.386 8.983 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.111 8.461 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.084 collocate_single_gaussian 91 10.0 0.016 0.020 7.726 7.945 mp_sync 37 10.5 3.275 7.898 3.275 7.898 mp_sendrecv_dm3 2068 8.0 5.698 7.183 5.698 7.183 integrate_four_center_load 4 10.0 0.000 0.000 6.764 6.770 hfx_load_balance 1 11.0 0.000 0.000 6.764 6.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=202.418738, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1494.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 444.514304E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.060 54.881 54.882 qs_mol_dyn_low 1 2.0 0.003 0.003 54.356 54.366 qs_forces 11 3.9 0.022 0.026 54.296 54.297 qs_energies 11 4.9 0.002 0.002 52.789 52.809 scf_env_do_scf 11 5.9 0.000 0.001 46.557 46.557 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.563 44.563 dbcsr_multiply_generic 2286 12.5 0.095 0.098 34.454 34.930 qs_scf_new_mos 108 7.5 0.000 0.001 34.622 34.928 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.621 34.928 ot_scf_mini 108 9.5 0.002 0.002 32.874 33.040 multiply_cannon 2286 13.5 0.186 0.193 26.715 28.282 multiply_cannon_loop 2286 14.5 1.812 1.930 26.074 27.655 velocity_verlet 10 3.0 0.001 0.002 26.472 26.473 ot_mini 108 10.5 0.001 0.001 19.988 20.251 qs_ot_get_derivative 108 11.5 0.001 0.001 17.001 17.171 mp_waitall_1 245248 16.5 8.829 15.659 8.829 15.659 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.345 13.622 multiply_cannon_multrec 54864 15.5 3.631 5.806 7.746 11.128 qs_ot_get_p 119 10.4 0.001 0.001 8.087 8.359 rebuild_ks_matrix 119 8.3 0.000 0.000 7.791 7.974 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.791 7.973 mp_sum_l 7287 12.8 5.860 7.688 5.860 7.688 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.855 7.017 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.882 6.361 multiply_cannon_sync_h2d 54864 15.5 5.139 6.331 5.139 6.331 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.755 5.865 dbcsr_mm_accdrv_process 76910 16.1 1.837 2.916 4.030 5.770 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.275 5.353 init_scf_run 11 5.9 0.000 0.001 4.937 4.937 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.937 4.937 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.439 4.439 sum_up_and_integrate 119 10.3 0.001 0.002 4.401 4.408 integrate_v_rspace 119 11.3 0.002 0.003 4.390 4.398 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.251 4.252 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.086 4.238 calculate_rho_elec 119 8.7 0.011 0.016 4.086 4.238 cp_fm_redistribute_end 50 14.0 2.169 4.228 2.175 4.231 cp_fm_diag_elpa_base 50 14.0 2.050 4.122 2.054 4.128 calculate_dm_sparse 119 9.5 0.000 0.000 3.064 3.262 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.903 3.082 apply_single 119 13.6 0.000 0.000 2.902 3.082 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.857 2.902 acc_transpose_blocks 54864 15.5 0.221 0.248 2.259 2.823 jit_kernel_multiply 13 15.8 2.129 2.764 2.129 2.764 calculate_first_density_matrix 1 7.0 0.000 0.000 2.729 2.735 ot_diis_step 108 11.5 0.006 0.006 2.684 2.685 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.567 2.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.379 2.380 density_rs2pw 119 9.7 0.004 0.004 2.065 2.172 wfi_extrapolate 11 7.9 0.001 0.001 2.137 2.138 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.076 2.127 grid_integrate_task_list 119 12.3 2.016 2.109 2.016 2.109 init_scf_loop 11 6.9 0.000 0.000 1.977 1.978 mp_sum_d 4137 12.0 1.349 1.944 1.349 1.944 potential_pw2rs 119 12.3 0.004 0.004 1.780 1.794 pw_transfer 1439 11.6 0.051 0.057 1.604 1.672 make_m2s 4572 13.5 0.053 0.055 1.565 1.604 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.529 1.598 mp_alltoall_d11v 2130 13.8 1.332 1.530 1.332 1.530 make_images 4572 14.5 0.132 0.138 1.484 1.522 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.483 1.506 acc_transpose_blocks_sync 164592 16.5 1.193 1.447 1.193 1.447 grid_collocate_task_list 119 9.7 1.348 1.415 1.348 1.415 mp_waitany 12084 13.8 1.235 1.403 1.235 1.403 transfer_rs2pw 487 10.6 0.005 0.006 1.311 1.394 transfer_pw2rs 487 13.2 0.006 0.006 1.307 1.316 fft3d_ps 1201 14.6 0.368 0.473 1.240 1.304 fft_wrap_pw1pw2_140 487 13.2 0.109 0.115 1.188 1.259 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.742 1.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.882000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=423.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 480.305152E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 254532. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 38.447 38.448 qs_mol_dyn_low 1 2.0 0.003 0.009 38.218 38.226 qs_forces 11 3.9 0.010 0.018 38.110 38.112 qs_energies 11 4.9 0.001 0.002 36.440 36.452 scf_env_do_scf 11 5.9 0.001 0.001 31.465 31.466 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.096 29.097 dbcsr_multiply_generic 2286 12.5 0.102 0.105 21.901 22.253 qs_scf_new_mos 108 7.5 0.001 0.001 20.539 20.768 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.538 20.767 ot_scf_mini 108 9.5 0.002 0.003 19.616 19.781 velocity_verlet 10 3.0 0.002 0.005 18.244 18.247 multiply_cannon 2286 13.5 0.208 0.218 16.774 18.116 multiply_cannon_loop 2286 14.5 1.187 1.244 15.628 17.134 ot_mini 108 10.5 0.001 0.001 12.107 12.339 mp_waitall_1 200699 16.5 5.441 10.554 5.441 10.554 qs_ot_get_derivative 108 11.5 0.001 0.001 9.754 9.918 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 3.967 9.299 multiply_cannon_multrec 27432 15.5 1.828 4.228 6.141 8.997 rebuild_ks_matrix 119 8.3 0.000 0.000 6.718 6.870 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.718 6.870 dbcsr_mm_accdrv_process 47894 16.0 3.550 5.638 4.233 6.296 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.906 6.045 qs_ot_get_p 119 10.4 0.001 0.002 4.602 4.825 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.654 4.568 mp_sum_l 7287 12.8 2.100 4.116 2.100 4.116 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.059 4.038 apply_single 119 13.6 0.000 0.000 3.059 4.038 init_scf_run 11 5.9 0.000 0.001 3.763 3.763 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.763 3.763 sum_up_and_integrate 119 10.3 0.001 0.001 3.662 3.668 integrate_v_rspace 119 11.3 0.002 0.003 3.648 3.653 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.578 3.609 calculate_rho_elec 119 8.7 0.021 0.024 3.577 3.609 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.971 2.989 make_m2s 4572 13.5 0.051 0.053 2.582 2.897 make_images 4572 14.5 0.205 0.244 2.493 2.811 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.576 2.576 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.303 2.389 init_scf_loop 11 6.9 0.000 0.000 2.349 2.350 ot_diis_step 108 11.5 0.011 0.011 2.304 2.305 calculate_first_density_matrix 1 7.0 0.000 0.001 2.269 2.273 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.265 2.265 cp_fm_redistribute_end 50 14.0 1.153 2.240 1.156 2.243 multiply_cannon_sync_h2d 27432 15.5 1.689 2.232 1.689 2.232 cp_fm_diag_elpa_base 50 14.0 1.055 2.144 1.084 2.184 calculate_dm_sparse 119 9.5 0.000 0.001 2.059 2.142 density_rs2pw 119 9.7 0.004 0.004 1.881 1.956 acc_transpose_blocks 27432 15.5 0.113 0.119 1.613 1.935 grid_integrate_task_list 119 12.3 1.847 1.933 1.847 1.933 pw_transfer 1439 11.6 0.064 0.068 1.828 1.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 1.854 1.856 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.738 1.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.700 make_images_data 4572 15.5 0.047 0.054 1.223 1.563 jit_kernel_multiply 8 16.4 0.624 1.537 0.624 1.537 prepare_preconditioner 11 7.9 0.000 0.000 1.477 1.503 make_preconditioner 11 8.9 0.000 0.000 1.477 1.503 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.383 1.438 wfi_extrapolate 11 7.9 0.001 0.001 1.434 1.434 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.079 1.426 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.381 1.417 potential_pw2rs 119 12.3 0.006 0.006 1.403 1.409 fft3d_ps 1201 14.6 0.521 0.579 1.369 1.401 grid_collocate_task_list 119 9.7 1.288 1.338 1.288 1.338 mp_alltoall_d11v 2130 13.8 1.170 1.308 1.170 1.308 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.256 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.204 1.248 transfer_rs2pw 487 10.6 0.005 0.005 1.071 1.136 mp_sum_d 4137 12.0 0.531 0.996 0.531 0.996 mp_allgather_i34 2286 14.5 0.569 0.970 0.569 0.970 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.949 0.958 acc_transpose_blocks_sync 82296 16.5 0.818 0.936 0.818 0.936 transfer_pw2rs 487 13.2 0.004 0.005 0.880 0.887 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.882 acc_transpose_blocks_kernels 27432 16.5 0.188 0.274 0.654 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.448000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=456.909091, yerr=1.781447 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 511.139840E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 31.862 31.864 qs_mol_dyn_low 1 2.0 0.004 0.010 31.594 31.603 qs_forces 11 3.9 0.002 0.003 31.533 31.535 qs_energies 11 4.9 0.001 0.002 29.998 30.000 scf_env_do_scf 11 5.9 0.001 0.002 25.403 25.403 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.016 23.016 dbcsr_multiply_generic 2286 12.5 0.096 0.099 16.923 16.995 qs_scf_new_mos 108 7.5 0.001 0.001 15.401 15.417 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.401 15.417 velocity_verlet 10 3.0 0.002 0.005 15.019 15.020 ot_scf_mini 108 9.5 0.002 0.002 14.644 14.658 multiply_cannon 2286 13.5 0.194 0.198 13.702 14.424 multiply_cannon_loop 2286 14.5 0.859 0.891 12.934 13.686 ot_mini 108 10.5 0.001 0.001 9.111 9.127 qs_ot_get_derivative 108 11.5 0.001 0.001 7.628 7.641 multiply_cannon_multrec 18288 15.5 1.868 2.912 6.984 7.357 rebuild_ks_matrix 119 8.3 0.000 0.000 5.933 5.949 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.933 5.949 dbcsr_mm_accdrv_process 38222 16.0 4.953 5.837 5.022 5.904 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.222 5.237 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.911 3.529 sum_up_and_integrate 119 10.3 0.001 0.001 3.513 3.517 integrate_v_rspace 119 11.3 0.003 0.003 3.500 3.506 mp_waitall_1 158411 16.6 2.611 3.487 2.611 3.487 init_scf_run 11 5.9 0.000 0.001 3.404 3.404 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.403 3.404 qs_ot_get_p 119 10.4 0.001 0.002 3.299 3.315 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.230 3.236 calculate_rho_elec 119 8.7 0.031 0.031 3.229 3.235 init_scf_loop 11 6.9 0.000 0.000 2.371 2.371 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.024 2.339 apply_single 119 13.6 0.000 0.000 2.023 2.339 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.204 2.211 calculate_first_density_matrix 1 7.0 0.000 0.001 2.187 2.188 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.453 2.164 make_m2s 4572 13.5 0.043 0.045 1.788 1.938 grid_integrate_task_list 119 12.3 1.799 1.917 1.799 1.917 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.910 1.910 density_rs2pw 119 9.7 0.004 0.004 1.798 1.893 make_images 4572 14.5 0.192 0.204 1.703 1.852 calculate_dm_sparse 119 9.5 0.000 0.001 1.845 1.850 pw_transfer 1439 11.6 0.065 0.068 1.837 1.845 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.745 1.754 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.702 1.710 acc_transpose_blocks 18288 15.5 0.079 0.081 1.623 1.699 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.671 1.682 cp_fm_diag_elpa_base 50 14.0 1.648 1.663 1.670 1.681 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.658 1.659 prepare_preconditioner 11 7.9 0.000 0.000 1.624 1.626 make_preconditioner 11 8.9 0.000 0.000 1.624 1.626 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.476 1.560 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.509 1.515 ot_diis_step 108 11.5 0.011 0.011 1.469 1.469 fft_wrap_pw1pw2_140 487 13.2 0.178 0.182 1.414 1.424 mp_sum_l 7287 12.8 1.046 1.402 1.046 1.402 potential_pw2rs 119 12.3 0.007 0.008 1.315 1.320 grid_collocate_task_list 119 9.7 1.233 1.317 1.233 1.317 fft3d_ps 1201 14.6 0.548 0.567 1.300 1.310 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.184 1.187 wfi_extrapolate 11 7.9 0.001 0.001 1.165 1.165 multiply_cannon_sync_h2d 18288 15.5 1.057 1.159 1.057 1.159 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.438 1.075 transfer_rs2pw 487 10.6 0.005 0.005 0.963 1.056 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.963 0.965 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.938 0.956 make_images_data 4572 15.5 0.047 0.052 0.767 0.941 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.663 0.862 acc_transpose_blocks_sync 54864 16.5 0.738 0.818 0.738 0.818 mp_alltoall_d11v 2130 13.8 0.652 0.813 0.652 0.813 transfer_pw2rs 487 13.2 0.004 0.004 0.794 0.798 acc_transpose_blocks_kernels 18288 16.5 0.216 0.223 0.786 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.783 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.673 0.725 cp_fm_cholesky_invert 11 10.9 0.686 0.690 0.686 0.690 mp_alltoall_z22v 1201 16.6 0.605 0.673 0.605 0.673 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.864000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=487.181818, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 549.474304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 35.784 35.785 qs_mol_dyn_low 1 2.0 0.003 0.003 35.615 35.625 qs_forces 11 3.9 0.002 0.002 35.556 35.557 qs_energies 11 4.9 0.001 0.001 33.852 33.858 scf_env_do_scf 11 5.9 0.000 0.001 28.951 28.952 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.622 25.623 dbcsr_multiply_generic 2286 12.5 0.103 0.105 19.634 19.736 velocity_verlet 10 3.0 0.001 0.002 18.212 18.214 qs_scf_new_mos 108 7.5 0.001 0.001 17.626 17.682 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.625 17.681 ot_scf_mini 108 9.5 0.002 0.003 16.599 16.656 multiply_cannon 2286 13.5 0.217 0.225 16.030 16.394 multiply_cannon_loop 2286 14.5 1.505 1.575 15.102 15.408 ot_mini 108 10.5 0.001 0.001 10.220 10.289 multiply_cannon_multrec 27432 15.5 2.446 3.120 9.015 9.322 qs_ot_get_derivative 108 11.5 0.001 0.001 8.359 8.417 dbcsr_mm_accdrv_process 47916 15.9 6.116 7.756 6.467 7.818 rebuild_ks_matrix 119 8.3 0.000 0.000 6.285 6.330 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.284 6.329 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.565 5.606 qs_ot_get_p 119 10.4 0.001 0.002 3.553 3.631 init_scf_run 11 5.9 0.000 0.001 3.575 3.575 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.575 3.575 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.129 3.533 sum_up_and_integrate 119 10.3 0.001 0.001 3.459 3.467 integrate_v_rspace 119 11.3 0.003 0.003 3.448 3.456 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.304 3.346 calculate_rho_elec 119 8.7 0.040 0.046 3.303 3.345 init_scf_loop 11 6.9 0.000 0.000 3.310 3.310 prepare_preconditioner 11 7.9 0.000 0.000 2.487 2.495 make_preconditioner 11 8.9 0.000 0.000 2.487 2.495 acc_transpose_blocks 27432 15.5 0.118 0.121 2.365 2.472 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.095 2.426 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.294 2.324 make_m2s 4572 13.5 0.053 0.055 2.220 2.322 mp_waitall_1 137007 16.6 1.703 2.270 1.703 2.270 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.105 2.225 apply_single 119 13.6 0.000 0.000 2.104 2.224 make_images 4572 14.5 0.273 0.338 2.113 2.213 calculate_first_density_matrix 1 7.0 0.000 0.000 2.206 2.209 calculate_dm_sparse 119 9.5 0.000 0.000 2.131 2.179 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.162 2.171 pw_transfer 1439 11.6 0.064 0.069 1.965 2.001 grid_integrate_task_list 119 12.3 1.835 1.927 1.835 1.927 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.874 1.913 density_rs2pw 119 9.7 0.004 0.004 1.780 1.889 ot_diis_step 108 11.5 0.012 0.012 1.817 1.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.788 1.789 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.770 1.770 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.623 1.635 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.560 1.601 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.535 1.545 cp_fm_diag_elpa_base 50 14.0 1.501 1.519 1.533 1.543 acc_transpose_blocks_sync 82296 16.5 1.383 1.484 1.383 1.484 fft3d_ps 1201 14.6 0.592 0.656 1.363 1.386 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.778 1.369 grid_collocate_task_list 119 9.7 1.247 1.342 1.247 1.342 wfi_extrapolate 11 7.9 0.001 0.001 1.317 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.273 1.290 potential_pw2rs 119 12.3 0.009 0.009 1.267 1.269 mp_sum_l 7287 12.8 0.949 1.264 0.949 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.242 cp_fm_upper_to_full 72 14.2 0.816 1.165 0.816 1.165 jit_kernel_multiply 6 15.8 0.284 1.097 0.284 1.097 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.070 1.071 dbcsr_complete_redistribute 329 12.2 0.123 0.151 0.748 1.030 transfer_rs2pw 487 10.6 0.004 0.005 0.871 0.984 make_images_data 4572 15.5 0.048 0.053 0.819 0.935 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.878 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.691 0.874 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.833 0.851 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.557 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.811 mp_alltoall_d11v 2130 13.8 0.685 0.793 0.685 0.793 cp_fm_cholesky_invert 11 10.9 0.723 0.726 0.723 0.726 mp_alltoall_i22 627 13.8 0.428 0.725 0.428 0.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.785000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=521.909091, yerr=3.203820 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 612.651008E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 28.848 28.849 qs_mol_dyn_low 1 2.0 0.003 0.003 28.586 28.672 qs_forces 11 3.9 0.002 0.002 28.523 28.523 qs_energies 11 4.9 0.001 0.001 26.810 26.813 scf_env_do_scf 11 5.9 0.000 0.001 22.107 22.107 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.559 19.560 velocity_verlet 10 3.0 0.001 0.002 14.843 14.846 dbcsr_multiply_generic 2286 12.5 0.096 0.099 13.085 13.152 qs_scf_new_mos 108 7.5 0.001 0.001 11.849 11.871 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.849 11.870 ot_scf_mini 108 9.5 0.002 0.002 11.135 11.160 multiply_cannon 2286 13.5 0.225 0.233 10.479 10.919 multiply_cannon_loop 2286 14.5 0.642 0.664 9.579 9.773 ot_mini 108 10.5 0.001 0.001 6.389 6.423 multiply_cannon_multrec 9144 15.5 1.670 1.912 5.974 6.270 rebuild_ks_matrix 119 8.3 0.000 0.000 5.814 5.838 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.814 5.838 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.172 5.193 qs_ot_get_derivative 108 11.5 0.001 0.001 5.090 5.116 dbcsr_mm_accdrv_process 12550 15.8 3.543 4.248 4.196 4.296 sum_up_and_integrate 119 10.3 0.001 0.001 3.466 3.470 integrate_v_rspace 119 11.3 0.003 0.003 3.456 3.460 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.386 3.396 calculate_rho_elec 119 8.7 0.060 0.061 3.386 3.395 init_scf_run 11 5.9 0.000 0.001 3.258 3.258 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.258 3.258 qs_ot_get_p 119 10.4 0.031 0.031 2.883 2.917 init_scf_loop 11 6.9 0.000 0.000 2.526 2.527 pw_transfer 1439 11.6 0.065 0.067 2.131 2.141 calculate_first_density_matrix 1 7.0 0.000 0.000 2.112 2.112 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.039 2.051 mp_waitall_1 115863 16.7 1.433 1.953 1.433 1.953 grid_integrate_task_list 119 12.3 1.870 1.925 1.870 1.925 make_m2s 4572 13.5 0.033 0.035 1.784 1.916 density_rs2pw 119 9.7 0.003 0.003 1.787 1.875 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.868 1.871 make_images 4572 14.5 0.268 0.299 1.694 1.824 calculate_dm_sparse 119 9.5 0.000 0.000 1.800 1.817 prepare_preconditioner 11 7.9 0.000 0.000 1.764 1.768 make_preconditioner 11 8.9 0.000 0.000 1.764 1.768 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.728 1.750 jit_kernel_multiply 10 15.5 0.614 1.737 0.614 1.737 fft_wrap_pw1pw2_140 487 13.2 0.326 0.335 1.690 1.703 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.655 1.681 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.626 1.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.520 1.521 acc_transpose_blocks 9144 15.5 0.043 0.044 1.417 1.448 fft3d_ps 1201 14.6 0.652 0.667 1.389 1.399 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.360 1.372 grid_collocate_task_list 119 9.7 1.299 1.366 1.299 1.366 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.354 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.342 1.350 cp_fm_diag_elpa_base 50 14.0 1.314 1.332 1.340 1.349 ot_diis_step 108 11.5 0.013 0.013 1.284 1.284 potential_pw2rs 119 12.3 0.010 0.010 1.271 1.274 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.225 1.256 apply_single 119 13.6 0.000 0.000 1.225 1.256 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.228 1.228 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.184 1.189 wfi_extrapolate 11 7.9 0.001 0.001 1.093 1.093 hybrid_alltoall_any 4725 16.4 0.065 0.178 0.763 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.929 make_images_data 4572 15.5 0.042 0.047 0.763 0.922 cp_fm_cholesky_invert 11 10.9 0.900 0.903 0.900 0.903 mp_alltoall_d11v 2130 13.8 0.771 0.897 0.771 0.897 transfer_rs2pw 487 10.6 0.004 0.004 0.787 0.875 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.832 0.837 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.763 0.766 acc_transpose_blocks_sync 27432 16.5 0.717 0.746 0.717 0.746 qs_env_update_s_mstruct 11 6.9 0.003 0.005 0.677 0.722 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.376 0.710 transfer_pw2rs 487 13.2 0.003 0.004 0.662 0.664 mp_alltoall_z22v 1201 16.6 0.611 0.645 0.611 0.645 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.641 0.643 mp_allgather_i34 2286 14.5 0.214 0.617 0.214 0.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.849000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=581.090909, yerr=5.961308 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 766.894080E+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 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 42.296 42.296 qs_mol_dyn_low 1 2.0 0.003 0.003 42.098 42.109 qs_forces 11 3.9 0.002 0.002 42.033 42.033 qs_energies 11 4.9 0.002 0.002 40.024 40.028 scf_env_do_scf 11 5.9 0.001 0.001 34.433 34.433 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.661 26.662 velocity_verlet 10 3.0 0.002 0.002 23.976 23.982 dbcsr_multiply_generic 2286 12.5 0.104 0.105 18.705 18.953 qs_scf_new_mos 108 7.5 0.001 0.001 17.013 17.112 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.012 17.111 ot_scf_mini 108 9.5 0.002 0.002 15.855 15.954 multiply_cannon 2286 13.5 0.295 0.300 14.899 15.864 multiply_cannon_loop 2286 14.5 0.856 0.867 13.660 14.650 ot_mini 108 10.5 0.001 0.001 9.643 9.757 multiply_cannon_multrec 9144 15.5 3.305 4.601 8.762 8.854 init_scf_loop 11 6.9 0.000 0.000 7.746 7.749 qs_ot_get_derivative 108 11.5 0.001 0.001 7.604 7.702 rebuild_ks_matrix 119 8.3 0.000 0.000 7.242 7.380 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.241 7.380 prepare_preconditioner 11 7.9 0.000 0.000 6.789 6.802 make_preconditioner 11 8.9 0.000 0.000 6.789 6.802 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.306 6.683 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.527 6.653 dbcsr_mm_accdrv_process 12550 15.8 4.900 6.522 5.324 6.555 cp_fm_upper_to_full 72 14.2 3.200 4.632 3.200 4.632 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.183 4.190 calculate_rho_elec 119 8.7 0.117 0.120 4.183 4.190 sum_up_and_integrate 119 10.3 0.001 0.001 3.845 3.850 integrate_v_rspace 119 11.3 0.004 0.004 3.835 3.839 init_scf_run 11 5.9 0.000 0.001 3.544 3.544 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.544 3.544 qs_ot_get_p 119 10.4 0.001 0.002 3.299 3.432 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.787 3.190 mp_waitall_1 94719 16.7 2.138 3.145 2.138 3.145 pw_transfer 1439 11.6 0.068 0.068 3.000 3.006 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.902 2.910 dbcsr_complete_redistribute 329 12.2 0.282 0.285 1.997 2.827 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.704 2.535 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.195 2.506 apply_single 119 13.6 0.000 0.000 2.195 2.506 fft_wrap_pw1pw2_140 487 13.2 0.654 0.665 2.482 2.493 make_m2s 4572 13.5 0.037 0.037 2.276 2.422 mp_alltoall_i22 627 13.8 1.451 2.317 1.451 2.317 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.478 2.302 make_images 4572 14.5 0.355 0.391 2.155 2.301 calculate_dm_sparse 119 9.5 0.000 0.000 2.224 2.247 density_rs2pw 119 9.7 0.003 0.004 2.214 2.238 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.273 2.205 calculate_first_density_matrix 1 7.0 0.000 0.000 2.094 2.094 grid_integrate_task_list 119 12.3 2.072 2.094 2.072 2.094 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.020 2.069 ot_diis_step 108 11.5 0.014 0.015 2.013 2.013 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.939 1.940 acc_transpose_blocks 9144 15.5 0.044 0.046 1.801 1.825 fft3d_ps 1201 14.6 0.869 0.886 1.818 1.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.787 1.787 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.768 mp_sum_l 7287 12.8 0.978 1.702 0.978 1.702 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.589 1.590 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.548 1.588 grid_collocate_task_list 119 9.7 1.516 1.545 1.516 1.545 wfi_extrapolate 11 7.9 0.001 0.001 1.387 1.387 cp_fm_cholesky_invert 11 10.9 1.373 1.376 1.373 1.376 potential_pw2rs 119 12.3 0.013 0.014 1.358 1.359 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.321 1.321 cp_fm_diag_elpa_base 50 14.0 1.176 1.229 1.319 1.320 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.300 1.318 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.229 1.236 hybrid_alltoall_any 4725 16.4 0.091 0.149 1.005 1.207 mp_alltoall_d11v 2130 13.8 1.133 1.153 1.133 1.153 acc_transpose_blocks_sync 27432 16.5 1.095 1.116 1.095 1.116 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.098 1.115 make_images_data 4572 15.5 0.047 0.050 0.940 1.112 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.992 1.055 qs_create_task_list 11 7.9 0.001 0.001 0.927 0.938 generate_qs_task_list 11 8.9 0.367 0.386 0.926 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.923 0.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.296000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.272727, yerr=15.926484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 493.158400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 81.733 81.734 qs_mol_dyn_low 1 2.0 0.003 0.003 81.489 81.500 qs_forces 11 3.9 0.003 0.003 81.385 81.386 qs_energies 11 4.9 0.002 0.002 78.547 78.559 scf_env_do_scf 11 5.9 0.000 0.001 69.785 69.788 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.264 64.265 dbcsr_multiply_generic 2055 12.4 0.105 0.109 50.720 50.963 qs_scf_new_mos 99 7.5 0.000 0.001 47.023 47.128 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.022 47.128 ot_scf_mini 99 9.5 0.002 0.002 44.657 44.727 multiply_cannon 2055 13.4 0.181 0.187 42.380 43.262 velocity_verlet 10 3.0 0.001 0.001 43.183 43.184 multiply_cannon_loop 2055 14.4 1.787 1.840 41.427 42.346 ot_mini 99 10.5 0.001 0.001 26.348 26.407 qs_ot_get_derivative 99 11.5 0.001 0.001 19.598 19.684 multiply_cannon_multrec 49320 15.4 11.411 12.277 17.426 18.186 rebuild_ks_matrix 110 8.3 0.000 0.000 14.600 14.741 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.599 14.740 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.814 12.944 mp_waitall_1 220248 16.4 10.642 11.444 10.642 11.444 multiply_cannon_sync_h2d 49320 15.4 9.647 10.195 9.647 10.195 qs_ot_get_p 110 10.4 0.001 0.001 9.822 9.980 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.652 8.214 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.220 7.718 apply_single 110 13.6 0.000 0.000 7.219 7.718 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.389 7.435 sum_up_and_integrate 110 10.3 0.001 0.003 7.095 7.108 integrate_v_rspace 110 11.3 0.003 0.003 7.070 7.089 init_scf_run 11 5.9 0.000 0.001 6.742 6.743 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.742 6.742 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.599 6.620 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.438 6.576 calculate_rho_elec 110 8.6 0.020 0.024 6.437 6.576 ot_diis_step 99 11.5 0.005 0.006 6.562 6.562 dbcsr_mm_accdrv_process 87628 16.1 3.067 3.155 5.884 6.159 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.724 5.724 init_scf_loop 11 6.9 0.000 0.000 5.492 5.493 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.266 5.301 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.144 5.171 cp_fm_diag_elpa_base 48 14.0 5.130 5.155 5.142 5.169 mp_sum_l 6594 12.7 3.880 4.595 3.880 4.595 make_m2s 4110 13.4 0.060 0.065 4.000 4.100 wfi_extrapolate 11 7.9 0.001 0.001 4.056 4.056 make_images 4110 14.4 0.177 0.189 3.905 4.007 calculate_dm_sparse 110 9.5 0.001 0.001 3.730 3.825 density_rs2pw 110 9.6 0.004 0.004 3.466 3.638 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.511 3.514 grid_integrate_task_list 110 12.3 3.265 3.407 3.265 3.407 prepare_preconditioner 11 7.9 0.000 0.000 3.322 3.342 make_preconditioner 11 8.9 0.000 0.000 3.322 3.342 pw_transfer 1331 11.6 0.054 0.062 3.216 3.301 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.323 3.286 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.180 3.230 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.183 3.224 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.128 3.216 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.093 3.140 fft_wrap_pw1pw2_140 451 13.1 0.309 0.330 2.629 2.721 potential_pw2rs 110 12.3 0.005 0.006 2.685 2.703 acc_transpose_blocks 49320 15.4 0.208 0.215 2.541 2.611 mp_alltoall_d11v 2046 13.8 2.103 2.603 2.103 2.603 calculate_first_density_matrix 1 7.0 0.000 0.000 2.585 2.588 jit_kernel_multiply 13 15.9 2.537 2.579 2.537 2.579 fft3d_ps 1111 14.6 0.797 0.889 2.424 2.487 grid_collocate_task_list 110 9.6 2.156 2.291 2.156 2.291 transfer_rs2pw 451 10.6 0.005 0.006 1.935 2.145 mp_waitany 14300 13.8 1.815 2.081 1.815 2.081 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.915 1.929 make_images_data 4110 15.4 0.044 0.047 1.795 1.924 transfer_pw2rs 451 13.1 0.006 0.007 1.809 1.821 cp_fm_cholesky_invert 11 10.9 1.810 1.815 1.810 1.815 mp_sum_d 3891 11.9 1.290 1.813 1.290 1.813 hybrid_alltoall_any 4261 16.3 0.083 0.482 1.551 1.776 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.638 1.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.734000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=467.727273, yerr=2.525899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 583.421952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2017478. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 68.788 68.788 qs_mol_dyn_low 1 2.0 0.004 0.007 68.579 68.589 qs_forces 11 3.9 0.003 0.003 68.475 68.476 qs_energies 11 4.9 0.001 0.002 65.143 65.147 scf_env_do_scf 11 5.9 0.001 0.001 56.668 56.671 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.111 49.111 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.017 38.180 velocity_verlet 10 3.0 0.003 0.011 36.242 36.244 qs_scf_new_mos 99 7.5 0.001 0.001 33.223 33.331 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.222 33.331 multiply_cannon 2055 13.4 0.223 0.242 31.460 32.555 ot_scf_mini 99 9.5 0.003 0.003 31.550 31.662 multiply_cannon_loop 2055 14.4 1.160 1.186 30.148 30.943 ot_mini 99 10.5 0.001 0.001 18.409 18.520 multiply_cannon_multrec 24660 15.4 7.011 8.989 14.070 15.911 rebuild_ks_matrix 110 8.3 0.000 0.000 13.593 13.699 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.592 13.699 qs_ot_get_derivative 99 11.5 0.001 0.001 12.570 12.677 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.957 12.049 mp_waitall_1 176588 16.5 7.517 9.973 7.517 9.973 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.133 7.788 multiply_cannon_sync_h2d 24660 15.4 6.486 7.634 6.486 7.634 init_scf_loop 11 6.9 0.000 0.000 7.523 7.523 dbcsr_mm_accdrv_process 52282 16.1 5.441 6.303 6.887 7.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.508 7.110 apply_single 110 13.6 0.000 0.001 6.507 7.110 qs_ot_get_p 110 10.4 0.001 0.002 6.410 6.534 sum_up_and_integrate 110 10.3 0.002 0.003 6.377 6.387 integrate_v_rspace 110 11.3 0.003 0.003 6.350 6.361 init_scf_run 11 5.9 0.000 0.001 6.100 6.101 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.100 6.101 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.886 5.896 calculate_rho_elec 110 8.6 0.039 0.047 5.885 5.895 ot_diis_step 99 11.5 0.010 0.010 5.789 5.789 prepare_preconditioner 11 7.9 0.000 0.000 5.499 5.521 make_preconditioner 11 8.9 0.000 0.000 5.499 5.521 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.731 5.445 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.084 5.238 make_m2s 4110 13.4 0.056 0.059 4.224 4.724 make_images 4110 14.4 0.407 0.462 4.112 4.607 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.535 4.554 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.073 4.073 pw_transfer 1331 11.6 0.066 0.074 3.562 3.701 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.455 3.597 wfi_extrapolate 11 7.9 0.001 0.001 3.544 3.545 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.486 3.501 cp_fm_diag_elpa_base 48 14.0 3.440 3.458 3.483 3.498 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.263 3.325 grid_integrate_task_list 110 12.3 3.168 3.320 3.168 3.320 density_rs2pw 110 9.6 0.004 0.004 3.128 3.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.241 3.243 fft_wrap_pw1pw2_140 451 13.1 0.359 0.380 2.953 3.100 calculate_dm_sparse 110 9.5 0.001 0.001 2.975 3.001 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.904 2.947 make_images_data 4110 15.4 0.049 0.053 2.302 2.821 hybrid_alltoall_any 4261 16.3 0.105 0.447 2.014 2.779 fft3d_ps 1111 14.6 1.116 1.348 2.545 2.705 cp_fm_cholesky_invert 11 10.9 2.578 2.585 2.578 2.585 calculate_first_density_matrix 1 7.0 0.000 0.001 2.461 2.463 grid_collocate_task_list 110 9.6 2.170 2.326 2.170 2.326 potential_pw2rs 110 12.3 0.008 0.009 2.244 2.253 mp_sum_l 6594 12.7 1.699 2.252 1.699 2.252 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.978 1.996 acc_transpose_blocks 24660 15.4 0.116 0.119 1.972 1.996 mp_alltoall_d11v 2046 13.8 1.758 1.909 1.758 1.909 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.860 1.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.793 1.802 jit_kernel_multiply 9 16.3 1.076 1.755 1.076 1.755 mp_allgather_i34 2055 14.4 0.619 1.661 0.619 1.661 multiply_cannon_metrocomm4 22605 15.4 0.074 0.080 0.745 1.594 transfer_rs2pw 451 10.6 0.006 0.007 1.419 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.541 1.551 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.492 dbcsr_complete_redistribute 325 12.2 0.241 0.311 1.148 1.415 mp_irecv_dv 57340 16.2 0.617 1.377 0.617 1.377 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.788000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.000000, yerr=7.122308 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.713472E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.038 59.322 59.323 qs_mol_dyn_low 1 2.0 0.004 0.011 58.855 58.865 qs_forces 11 3.9 0.003 0.003 58.783 58.786 qs_energies 11 4.9 0.002 0.008 55.608 55.614 scf_env_do_scf 11 5.9 0.001 0.002 47.718 47.718 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.347 39.347 velocity_verlet 10 3.0 0.002 0.005 32.117 32.119 dbcsr_multiply_generic 2055 12.4 0.109 0.114 28.595 28.823 qs_scf_new_mos 99 7.5 0.001 0.001 24.835 24.935 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.835 24.935 ot_scf_mini 99 9.5 0.002 0.003 23.586 23.708 multiply_cannon 2055 13.4 0.212 0.232 22.492 23.678 multiply_cannon_loop 2055 14.4 0.814 0.840 21.325 22.324 ot_mini 99 10.5 0.001 0.001 13.590 13.712 rebuild_ks_matrix 110 8.3 0.000 0.000 12.069 12.189 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 12.068 12.189 multiply_cannon_multrec 16440 15.4 3.691 4.772 9.901 10.958 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.619 10.730 mp_waitall_1 139946 16.5 6.742 10.086 6.742 10.086 qs_ot_get_derivative 99 11.5 0.001 0.001 9.173 9.293 init_scf_loop 11 6.9 0.001 0.003 8.338 8.339 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.260 7.440 prepare_preconditioner 11 7.9 0.000 0.000 6.595 6.613 make_preconditioner 11 8.9 0.000 0.001 6.595 6.613 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.903 6.246 sum_up_and_integrate 110 10.3 0.001 0.002 6.213 6.226 dbcsr_mm_accdrv_process 34862 16.1 5.371 5.699 6.055 6.211 integrate_v_rspace 110 11.3 0.003 0.003 6.187 6.201 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.671 5.678 calculate_rho_elec 110 8.6 0.059 0.060 5.671 5.677 init_scf_run 11 5.9 0.000 0.001 5.464 5.464 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.464 5.464 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.870 5.364 apply_single 110 13.6 0.000 0.000 4.870 5.364 qs_ot_get_p 110 10.4 0.001 0.002 5.188 5.311 make_m2s 4110 13.4 0.049 0.051 4.096 4.479 ot_diis_step 99 11.5 0.011 0.011 4.387 4.387 make_images 4110 14.4 0.401 0.524 3.980 4.364 multiply_cannon_sync_h2d 16440 15.4 3.265 3.877 3.265 3.877 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.189 3.777 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.585 3.589 pw_transfer 1331 11.6 0.065 0.070 3.562 3.571 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.455 3.466 grid_integrate_task_list 110 12.3 3.221 3.353 3.221 3.353 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.220 3.220 density_rs2pw 110 9.6 0.004 0.004 2.915 3.093 fft_wrap_pw1pw2_140 451 13.1 0.461 0.475 3.008 3.026 wfi_extrapolate 11 7.9 0.001 0.001 2.923 2.923 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.799 2.800 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.072 2.798 make_images_data 4110 15.4 0.047 0.051 2.307 2.760 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.702 2.711 cp_fm_diag_elpa_base 48 14.0 2.637 2.669 2.700 2.710 calculate_dm_sparse 110 9.5 0.001 0.001 2.523 2.558 cp_fm_cholesky_invert 11 10.9 2.545 2.551 2.545 2.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.468 2.514 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.406 2.480 calculate_first_density_matrix 1 7.0 0.000 0.002 2.451 2.452 fft3d_ps 1111 14.6 1.121 1.131 2.370 2.387 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.869 2.370 grid_collocate_task_list 110 9.6 2.218 2.358 2.218 2.358 mp_irecv_dv 48980 15.7 0.792 2.231 0.792 2.231 potential_pw2rs 110 12.3 0.011 0.011 2.054 2.057 mp_sum_l 6594 12.7 1.388 2.032 1.388 2.032 mp_alltoall_d11v 2046 13.8 1.785 2.006 1.785 2.006 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.993 1.996 dbcsr_complete_redistribute 325 12.2 0.323 0.366 1.418 1.860 cp_fm_upper_to_full 70 14.2 1.384 1.747 1.384 1.747 acc_transpose_blocks 16440 15.4 0.078 0.081 1.616 1.682 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.642 1.658 cp_fm_cholesky_decompose 22 10.9 1.514 1.536 1.514 1.536 mp_allgather_i34 2055 14.4 0.478 1.497 0.478 1.497 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.371 1.490 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.472 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.990 1.444 transfer_rs2pw 451 10.6 0.005 0.006 1.246 1.410 mp_waitany 17072 13.8 1.132 1.317 1.132 1.317 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.259 1.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.323000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.909091, yerr=9.248632 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 736.915456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.038 65.263 65.264 qs_mol_dyn_low 1 2.0 0.004 0.011 64.938 64.948 qs_forces 11 3.9 0.003 0.003 64.772 64.773 qs_energies 11 4.9 0.002 0.006 61.385 61.389 scf_env_do_scf 11 5.9 0.001 0.002 53.032 53.035 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.576 41.577 velocity_verlet 10 3.0 0.002 0.008 36.954 36.960 dbcsr_multiply_generic 2055 12.4 0.117 0.124 30.448 30.619 qs_scf_new_mos 99 7.5 0.001 0.001 27.060 27.192 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.059 27.191 ot_scf_mini 99 9.5 0.003 0.003 25.374 25.478 multiply_cannon 2055 13.4 0.238 0.256 23.385 24.713 multiply_cannon_loop 2055 14.4 1.396 1.506 22.041 22.648 ot_mini 99 10.5 0.001 0.001 14.587 14.717 multiply_cannon_multrec 24660 15.4 4.108 6.730 13.059 14.322 rebuild_ks_matrix 110 8.3 0.000 0.000 11.923 12.020 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 11.923 12.019 init_scf_loop 11 6.9 0.001 0.004 11.417 11.418 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.528 10.612 qs_ot_get_derivative 99 11.5 0.001 0.001 10.450 10.557 dbcsr_mm_accdrv_process 52304 16.0 7.704 8.985 8.793 9.734 prepare_preconditioner 11 7.9 0.000 0.000 9.687 9.703 make_preconditioner 11 8.9 0.000 0.001 9.686 9.703 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.943 9.402 sum_up_and_integrate 110 10.3 0.001 0.002 6.173 6.185 integrate_v_rspace 110 11.3 0.003 0.003 6.147 6.159 mp_waitall_1 121746 16.5 4.141 5.905 4.141 5.905 qs_ot_get_p 110 10.4 0.001 0.002 5.659 5.803 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.786 5.798 calculate_rho_elec 110 8.6 0.078 0.081 5.785 5.797 make_m2s 4110 13.4 0.059 0.061 5.453 5.748 make_images 4110 14.4 0.577 0.699 5.314 5.604 init_scf_run 11 5.9 0.000 0.001 5.538 5.538 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.537 5.538 cp_fm_upper_to_full 70 14.2 3.371 4.873 3.371 4.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.030 4.115 apply_single 110 13.6 0.000 0.000 4.030 4.115 ot_diis_step 99 11.5 0.011 0.011 4.095 4.096 qs_ot_p2m_diag 48 11.0 0.056 0.064 3.833 3.848 pw_transfer 1331 11.6 0.065 0.073 3.713 3.745 dbcsr_complete_redistribute 325 12.2 0.417 0.466 2.590 3.729 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.607 3.641 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.477 3.526 grid_integrate_task_list 110 12.3 3.254 3.412 3.254 3.412 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.299 3.359 cp_dbcsr_syevd 48 12.0 0.005 0.016 3.277 3.277 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.079 3.203 fft_wrap_pw1pw2_140 451 13.1 0.532 0.544 3.123 3.157 density_rs2pw 110 9.6 0.004 0.004 2.907 3.083 make_images_data 4110 15.4 0.049 0.053 2.739 3.047 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.013 hybrid_alltoall_any 4261 16.3 0.122 0.455 2.269 2.977 wfi_extrapolate 11 7.9 0.001 0.001 2.961 2.961 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.307 2.923 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.735 2.847 mp_alltoall_i22 605 13.7 1.650 2.810 1.650 2.810 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.805 2.808 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.746 2.755 cp_fm_diag_elpa_base 48 14.0 2.598 2.659 2.742 2.752 cp_fm_cholesky_invert 11 10.9 2.554 2.562 2.554 2.562 acc_transpose_blocks 24660 15.4 0.111 0.114 2.376 2.558 multiply_cannon_sync_h2d 24660 15.4 2.377 2.553 2.377 2.553 calculate_first_density_matrix 1 7.0 0.000 0.002 2.482 2.485 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.483 grid_collocate_task_list 110 9.6 2.261 2.471 2.261 2.471 fft3d_ps 1111 14.6 1.150 1.188 2.432 2.447 qs_energies_init_hamiltonians 11 5.9 0.002 0.014 2.318 2.320 potential_pw2rs 110 12.3 0.012 0.013 1.990 1.992 mp_alltoall_d11v 2046 13.8 1.731 1.925 1.731 1.925 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.784 1.818 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.709 cp_fm_cholesky_decompose 22 10.9 1.608 1.670 1.608 1.670 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.636 1.646 mp_sum_l 6594 12.7 0.925 1.638 0.925 1.638 mp_allgather_i34 2055 14.4 0.423 1.617 0.423 1.617 multiply_cannon_metrocomm4 20550 15.4 0.060 0.062 0.846 1.596 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.530 1.550 acc_transpose_blocks_sync 73980 16.4 1.359 1.520 1.359 1.520 mp_irecv_dv 62702 16.1 0.739 1.512 0.739 1.512 transfer_rs2pw 451 10.6 0.005 0.006 1.195 1.383 qs_env_update_s_mstruct 11 6.9 0.001 0.007 1.250 1.353 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.264000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.000000, yerr=12.158050 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 844.525568E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 54.486 54.487 qs_mol_dyn_low 1 2.0 0.003 0.003 54.253 54.264 qs_forces 11 3.9 0.003 0.003 54.187 54.187 qs_energies 11 4.9 0.001 0.002 50.528 50.534 scf_env_do_scf 11 5.9 0.000 0.001 42.328 42.329 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.751 34.752 velocity_verlet 10 3.0 0.002 0.002 30.881 30.884 dbcsr_multiply_generic 2055 12.4 0.107 0.109 22.924 23.036 qs_scf_new_mos 99 7.5 0.001 0.001 20.221 20.276 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.221 20.275 ot_scf_mini 99 9.5 0.002 0.002 18.974 18.995 multiply_cannon 2055 13.4 0.236 0.245 17.463 18.813 multiply_cannon_loop 2055 14.4 0.598 0.624 16.189 16.463 rebuild_ks_matrix 110 8.3 0.000 0.000 11.611 11.621 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.611 11.621 ot_mini 99 10.5 0.001 0.001 10.442 10.453 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.327 10.339 multiply_cannon_multrec 8220 15.4 3.164 4.300 7.584 8.404 init_scf_loop 11 6.9 0.000 0.000 7.533 7.536 mp_waitall_1 103326 16.6 5.628 7.153 5.628 7.153 qs_ot_get_derivative 99 11.5 0.001 0.001 6.768 6.787 sum_up_and_integrate 110 10.3 0.001 0.002 6.171 6.185 integrate_v_rspace 110 11.3 0.003 0.003 6.144 6.157 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.976 5.989 calculate_rho_elec 110 8.6 0.113 0.114 5.975 5.989 prepare_preconditioner 11 7.9 0.000 0.000 5.867 5.871 make_preconditioner 11 8.9 0.000 0.000 5.867 5.871 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.472 5.543 dbcsr_mm_accdrv_process 17442 15.9 3.103 4.475 4.281 5.205 init_scf_run 11 5.9 0.000 0.001 5.078 5.078 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.078 5.078 qs_ot_get_p 110 10.4 0.016 0.019 4.609 4.623 make_m2s 4110 13.4 0.037 0.039 4.135 4.413 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.930 4.320 make_images 4110 14.4 0.647 0.709 4.006 4.285 pw_transfer 1331 11.6 0.065 0.070 4.045 4.052 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.937 3.948 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.714 3.759 apply_single 110 13.6 0.000 0.000 3.714 3.759 ot_diis_step 99 11.5 0.012 0.012 3.652 3.652 grid_integrate_task_list 110 12.3 3.387 3.519 3.387 3.519 fft_wrap_pw1pw2_140 451 13.1 0.719 0.732 3.431 3.445 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.250 3.253 density_rs2pw 110 9.6 0.004 0.004 2.978 3.089 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.916 2.917 cp_fm_cholesky_invert 11 10.9 2.821 2.825 2.821 2.825 wfi_extrapolate 11 7.9 0.001 0.001 2.682 2.682 hybrid_alltoall_any 4261 16.3 0.199 0.850 2.226 2.675 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.671 2.672 make_images_data 4110 15.4 0.041 0.047 2.226 2.578 calculate_dm_sparse 110 9.5 0.001 0.001 2.494 2.531 fft3d_ps 1111 14.6 1.290 1.298 2.481 2.497 grid_collocate_task_list 110 9.6 2.367 2.482 2.367 2.482 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.477 2.478 multiply_cannon_sync_h2d 8220 15.4 2.384 2.443 2.384 2.443 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.378 2.385 cp_fm_diag_elpa_base 48 14.0 2.322 2.343 2.376 2.384 calculate_first_density_matrix 1 7.0 0.000 0.000 2.302 2.303 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.123 2.140 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 2.017 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.982 2.008 potential_pw2rs 110 12.3 0.015 0.015 1.972 1.976 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.877 1.889 mp_alltoall_d11v 2046 13.8 1.620 1.763 1.620 1.763 cp_fm_cholesky_decompose 22 10.9 1.655 1.674 1.655 1.674 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.610 1.614 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.500 1.614 mp_allgather_i34 2055 14.4 0.444 1.533 0.444 1.533 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.449 1.458 dbcsr_complete_redistribute 325 12.2 0.555 0.584 1.359 1.441 acc_transpose_blocks 8220 15.4 0.041 0.042 1.348 1.392 qs_create_task_list 11 7.9 0.001 0.001 1.210 1.306 generate_qs_task_list 11 8.9 0.372 0.440 1.209 1.305 transfer_rs2pw 451 10.6 0.005 0.005 1.084 1.183 jit_kernel_multiply 7 15.6 0.856 1.173 0.856 1.173 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.134 1.164 mp_waitany 9240 13.8 1.034 1.151 1.034 1.151 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.487000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=800.636364, yerr=9.177767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 1.355117E+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 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.041 87.338 87.338 qs_mol_dyn_low 1 2.0 0.003 0.003 86.997 87.006 qs_forces 11 3.9 0.003 0.003 86.928 86.929 qs_energies 11 4.9 0.001 0.001 82.796 82.798 scf_env_do_scf 11 5.9 0.000 0.001 72.782 72.783 velocity_verlet 10 3.0 0.002 0.002 55.825 55.832 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.506 44.507 dbcsr_multiply_generic 2055 12.4 0.128 0.134 29.480 29.590 init_scf_loop 11 6.9 0.000 0.000 28.207 28.210 qs_scf_new_mos 99 7.5 0.001 0.001 26.666 26.717 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.665 26.717 prepare_preconditioner 11 7.9 0.000 0.000 26.182 26.189 make_preconditioner 11 8.9 0.000 0.000 26.182 26.189 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.376 25.673 ot_scf_mini 99 9.5 0.002 0.002 24.862 24.907 multiply_cannon 2055 13.4 0.329 0.358 22.500 23.166 multiply_cannon_loop 2055 14.4 0.825 0.836 20.754 21.062 cp_fm_upper_to_full 70 14.2 12.828 18.360 12.828 18.360 ot_mini 99 10.5 0.001 0.001 14.128 14.172 rebuild_ks_matrix 110 8.3 0.000 0.000 13.900 13.947 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.899 13.947 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.613 12.655 dbcsr_complete_redistribute 325 12.2 0.999 1.020 7.324 10.544 multiply_cannon_multrec 8220 15.4 4.074 4.255 9.761 9.871 qs_ot_get_derivative 99 11.5 0.001 0.001 9.511 9.557 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.341 9.556 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.790 8.983 mp_alltoall_i22 605 13.7 5.426 8.619 5.426 8.619 mp_waitall_1 84994 16.7 7.440 8.254 7.440 8.254 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.435 7.468 calculate_rho_elec 110 8.6 0.223 0.224 7.435 7.468 sum_up_and_integrate 110 10.3 0.002 0.002 6.866 6.881 integrate_v_rspace 110 11.3 0.004 0.004 6.838 6.853 pw_transfer 1331 11.6 0.074 0.075 5.773 5.781 init_scf_run 11 5.9 0.000 0.001 5.753 5.754 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.753 5.753 dbcsr_mm_accdrv_process 11614 15.7 3.877 4.243 5.538 5.736 make_m2s 4110 13.4 0.042 0.043 5.206 5.670 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.656 5.664 qs_ot_get_p 110 10.4 0.006 0.012 5.461 5.513 make_images 4110 14.4 0.887 0.943 5.018 5.480 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.749 5.217 apply_single 110 13.6 0.000 0.000 4.749 5.217 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.956 5.181 cp_fm_cholesky_invert 11 10.9 5.117 5.121 5.117 5.121 fft_wrap_pw1pw2_140 451 13.1 1.365 1.371 4.982 4.988 ot_diis_step 99 11.5 0.015 0.016 4.597 4.597 density_rs2pw 110 9.6 0.004 0.004 3.906 3.934 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.908 3.914 grid_integrate_task_list 110 12.3 3.683 3.748 3.683 3.748 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.661 3.661 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.993 3.449 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.393 3.394 fft3d_ps 1111 14.6 1.883 1.891 3.364 3.369 hybrid_alltoall_any 4261 16.3 0.265 0.567 2.759 3.355 wfi_extrapolate 11 7.9 0.001 0.001 3.315 3.315 make_images_data 4110 15.4 0.045 0.049 2.673 3.250 calculate_dm_sparse 110 9.5 0.001 0.001 3.175 3.205 multiply_cannon_sync_h2d 8220 15.4 3.122 3.138 3.122 3.138 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.889 2.892 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.817 2.817 cp_fm_diag_elpa_base 48 14.0 2.287 2.478 2.816 2.816 grid_collocate_task_list 110 9.6 2.671 2.688 2.671 2.688 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.619 2.644 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.400 2.421 potential_pw2rs 110 12.3 0.021 0.021 2.400 2.403 calculate_first_density_matrix 1 7.0 0.000 0.000 2.320 2.321 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.184 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.171 mp_alltoall_d11v 2046 13.8 1.962 2.000 1.962 2.000 cp_fm_cholesky_decompose 22 10.9 1.941 1.956 1.941 1.956 qs_create_task_list 11 7.9 0.000 0.001 1.880 1.925 generate_qs_task_list 11 8.9 0.730 0.786 1.879 1.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.885 1.890 acc_transpose_blocks 8220 15.4 0.042 0.043 1.727 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.338000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1226.454545, yerr=42.294129 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 622.874624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2592764. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 207.352 207.352 qs_mol_dyn_low 1 2.0 0.003 0.003 206.917 206.931 qs_forces 11 3.9 0.005 0.005 206.831 206.832 qs_energies 11 4.9 0.001 0.002 201.218 201.235 scf_env_do_scf 11 5.9 0.001 0.001 184.667 184.671 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.119 164.121 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.623 126.402 qs_scf_new_mos 117 7.6 0.001 0.001 124.607 124.911 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.606 124.910 velocity_verlet 10 3.0 0.001 0.002 123.997 123.998 ot_scf_mini 117 9.6 0.003 0.003 117.917 118.191 multiply_cannon 2507 13.6 0.234 0.240 101.837 103.333 multiply_cannon_loop 2507 14.6 2.392 2.444 99.642 101.242 ot_mini 117 10.6 0.001 0.001 66.184 66.464 multiply_cannon_multrec 60168 15.6 31.984 33.537 41.784 43.555 qs_ot_get_derivative 117 11.6 0.001 0.001 41.216 41.473 rebuild_ks_matrix 128 8.3 0.001 0.001 33.615 33.827 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.614 33.827 mp_waitall_1 267128 16.5 28.749 31.616 28.749 31.616 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.183 30.380 qs_ot_get_p 128 10.4 0.001 0.002 29.515 29.782 multiply_cannon_sync_h2d 60168 15.6 26.492 28.469 26.492 28.469 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.454 25.588 apply_single 128 13.6 0.001 0.001 24.454 25.587 ot_diis_step 117 11.6 0.007 0.008 24.714 24.715 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.927 22.989 init_scf_loop 11 6.9 0.000 0.001 20.473 20.474 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.121 20.122 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.841 20.080 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 16.198 17.959 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.128 17.162 cp_fm_diag_elpa_base 83 14.4 17.059 17.094 17.124 17.159 prepare_preconditioner 11 7.9 0.000 0.000 15.883 15.917 make_preconditioner 11 8.9 0.000 0.000 15.883 15.917 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.096 15.258 make_m2s 5014 13.6 0.104 0.113 13.729 14.072 sum_up_and_integrate 128 10.3 0.002 0.004 13.980 13.996 integrate_v_rspace 128 11.3 0.003 0.004 13.921 13.940 make_images 5014 14.6 0.396 0.416 13.545 13.899 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.173 13.306 calculate_rho_elec 128 8.7 0.045 0.064 13.172 13.305 init_scf_run 11 5.9 0.000 0.001 12.404 12.405 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.404 12.405 mp_sum_l 7950 12.9 8.799 10.147 8.799 10.147 dbcsr_mm_accdrv_process 124484 16.2 4.744 4.870 9.368 9.889 wfi_extrapolate 11 7.9 0.001 0.001 9.145 9.145 cp_fm_cholesky_invert 11 10.9 9.035 9.042 9.035 9.042 calculate_dm_sparse 128 9.5 0.001 0.001 8.518 8.609 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.104 8.251 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.097 8.190 multiply_cannon_metrocomm1 60168 15.6 0.095 0.099 6.113 8.167 pw_transfer 1547 11.6 0.073 0.094 7.622 7.814 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.420 7.607 make_images_data 5014 15.6 0.067 0.073 6.701 7.579 grid_integrate_task_list 128 12.3 7.042 7.392 7.042 7.392 density_rs2pw 128 9.7 0.006 0.007 6.809 7.380 hybrid_alltoall_any 5200 16.5 0.294 2.265 5.868 7.198 fft_wrap_pw1pw2_140 523 13.2 0.856 0.895 6.513 6.708 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.642 6.651 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.748 5.830 fft3d_ps 1291 14.7 2.197 2.841 5.430 5.756 mp_alltoall_d11v 2415 14.1 4.327 5.447 4.327 5.447 grid_collocate_task_list 128 9.7 4.849 5.271 4.849 5.271 cp_fm_cholesky_decompose 22 10.9 4.648 4.662 4.648 4.662 potential_pw2rs 128 12.3 0.009 0.010 4.502 4.526 mp_sum_d 4467 12.1 3.437 4.329 3.437 4.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.352000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=589.000000, yerr=6.901910 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 837.091328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 4647382. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.034 189.518 189.519 qs_mol_dyn_low 1 2.0 0.003 0.003 189.137 189.152 qs_forces 11 3.9 0.004 0.004 189.051 189.052 qs_energies 11 4.9 0.001 0.002 182.311 182.323 scf_env_do_scf 11 5.9 0.001 0.001 166.045 166.055 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.124 133.126 velocity_verlet 10 3.0 0.002 0.002 119.325 119.327 dbcsr_multiply_generic 2507 12.6 0.188 0.193 97.324 98.643 qs_scf_new_mos 117 7.6 0.001 0.001 94.579 95.186 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.578 95.185 ot_scf_mini 117 9.6 0.004 0.004 89.762 90.383 multiply_cannon 2507 13.6 0.502 0.559 77.149 81.556 multiply_cannon_loop 2507 14.6 1.566 1.643 73.728 76.577 ot_mini 117 10.6 0.001 0.001 50.494 51.120 mp_waitall_1 214728 16.6 24.243 39.245 24.243 39.245 multiply_cannon_multrec 30084 15.6 20.724 25.270 31.346 36.375 rebuild_ks_matrix 128 8.3 0.001 0.001 32.175 32.922 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.174 32.922 init_scf_loop 11 6.9 0.000 0.000 32.829 32.830 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.926 29.615 multiply_cannon_metrocomm3 30084 15.6 0.098 0.105 15.904 29.462 qs_ot_get_derivative 117 11.6 0.001 0.002 28.274 28.886 prepare_preconditioner 11 7.9 0.000 0.000 28.525 28.600 make_preconditioner 11 8.9 0.000 0.000 28.525 28.600 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.225 27.753 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.315 23.442 apply_single 128 13.6 0.001 0.001 22.315 23.441 ot_diis_step 117 11.6 0.014 0.015 22.042 22.044 qs_ot_get_p 128 10.4 0.001 0.002 21.321 22.038 multiply_cannon_sync_h2d 30084 15.6 17.768 20.043 17.768 20.043 cp_fm_cholesky_invert 11 10.9 16.682 16.697 16.682 16.697 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.440 16.474 make_m2s 5014 13.6 0.086 0.092 14.199 15.874 make_images 5014 14.6 1.174 1.368 13.983 15.652 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.234 15.235 sum_up_and_integrate 128 10.3 0.002 0.004 13.954 13.981 integrate_v_rspace 128 11.3 0.003 0.004 13.894 13.926 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.480 13.507 calculate_rho_elec 128 8.7 0.087 0.104 13.480 13.506 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.981 12.009 cp_fm_diag_elpa_base 83 14.4 11.727 11.826 11.976 12.000 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.211 11.659 init_scf_run 11 5.9 0.000 0.001 11.440 11.442 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.440 11.442 multiply_cannon_metrocomm4 27577 15.6 0.100 0.116 3.929 11.217 mp_irecv_dv 69486 16.3 3.723 10.816 3.723 10.816 dbcsr_mm_accdrv_process 62242 16.2 5.399 6.180 10.073 10.655 make_images_data 5014 15.6 0.067 0.076 8.317 10.240 hybrid_alltoall_any 5200 16.5 0.347 1.492 7.269 9.959 pw_transfer 1547 11.6 0.084 0.097 8.579 8.651 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 8.356 8.427 wfi_extrapolate 11 7.9 0.001 0.001 8.387 8.387 density_rs2pw 128 9.7 0.006 0.006 7.055 7.821 grid_integrate_task_list 128 12.3 7.170 7.501 7.170 7.501 fft_wrap_pw1pw2_140 523 13.2 0.931 0.951 7.393 7.470 cp_fm_cholesky_decompose 22 10.9 6.986 7.083 6.986 7.083 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.275 7.042 calculate_dm_sparse 128 9.5 0.001 0.001 6.477 6.609 mp_sum_l 7950 12.9 4.184 6.371 4.184 6.371 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.061 6.069 fft3d_ps 1291 14.7 2.830 3.033 5.914 5.987 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.416 5.479 grid_collocate_task_list 128 9.7 5.001 5.362 5.001 5.362 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.151 5.355 mp_allgather_i34 2507 14.6 1.940 5.101 1.940 5.101 mp_alltoall_d11v 2415 14.1 4.445 4.840 4.445 4.840 potential_pw2rs 128 12.3 0.016 0.019 4.469 4.484 mp_sum_d 4467 12.1 2.636 4.022 2.636 4.022 dbcsr_complete_redistribute 395 12.7 0.770 0.846 3.105 3.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.519000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=797.363636, yerr=2.772355 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/17/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 2.928533E+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 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 935.792640E+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 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.011 0.029 174.070 174.071 qs_mol_dyn_low 1 2.0 0.003 0.003 173.687 173.701 qs_forces 11 3.9 0.004 0.005 173.591 173.592 qs_energies 11 4.9 0.001 0.002 167.063 167.073 scf_env_do_scf 11 5.9 0.001 0.001 151.613 151.617 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.001 117.002 velocity_verlet 10 3.0 0.002 0.002 111.459 111.462 dbcsr_multiply_generic 2507 12.6 0.185 0.189 81.897 83.053 qs_scf_new_mos 117 7.6 0.001 0.001 80.092 80.476 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.091 80.475 ot_scf_mini 117 9.6 0.003 0.004 75.846 76.254 multiply_cannon 2507 13.6 0.506 0.525 62.106 66.022 multiply_cannon_loop 2507 14.6 1.128 1.196 59.125 61.478 ot_mini 117 10.6 0.001 0.001 42.652 43.059 mp_waitall_1 170520 16.6 24.696 34.717 24.696 34.717 init_scf_loop 11 6.9 0.000 0.000 34.516 34.516 rebuild_ks_matrix 128 8.3 0.001 0.001 30.122 30.657 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 30.122 30.657 prepare_preconditioner 11 7.9 0.000 0.000 30.530 30.580 make_preconditioner 11 8.9 0.000 0.000 30.530 30.580 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.173 29.564 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.150 27.634 multiply_cannon_multrec 20056 15.6 13.033 15.774 22.437 25.150 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.400 24.632 qs_ot_get_derivative 117 11.6 0.001 0.002 22.809 23.213 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.971 21.165 apply_single 128 13.6 0.001 0.001 19.970 21.165 ot_diis_step 117 11.6 0.018 0.018 19.734 19.735 qs_ot_get_p 128 10.4 0.001 0.002 18.567 19.040 make_m2s 5014 13.6 0.078 0.084 14.749 15.980 make_images 5014 14.6 1.155 1.231 14.514 15.744 multiply_cannon_sync_h2d 20056 15.6 13.516 15.164 13.516 15.164 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.340 14.349 cp_fm_cholesky_invert 11 10.9 14.152 14.161 14.152 14.161 sum_up_and_integrate 128 10.3 0.002 0.003 13.899 13.924 integrate_v_rspace 128 11.3 0.003 0.004 13.839 13.864 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.440 13.464 calculate_rho_elec 128 8.7 0.130 0.144 13.439 13.463 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.284 13.285 make_images_data 5014 15.6 0.065 0.075 8.844 10.457 init_scf_run 11 5.9 0.000 0.001 10.429 10.429 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.429 10.429 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.111 10.129 cp_fm_diag_elpa_base 83 14.4 9.703 9.839 10.108 10.126 hybrid_alltoall_any 5200 16.5 0.448 2.045 7.657 9.892 multiply_cannon_metrocomm4 17549 15.6 0.066 0.075 3.514 9.673 mp_irecv_dv 50230 16.2 3.382 9.404 3.382 9.404 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.998 9.287 dbcsr_mm_accdrv_process 41502 16.2 5.644 6.008 8.865 9.025 pw_transfer 1547 11.6 0.084 0.102 8.724 8.824 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.502 8.607 grid_integrate_task_list 128 12.3 7.312 7.905 7.312 7.905 fft_wrap_pw1pw2_140 523 13.2 1.059 1.093 7.565 7.678 wfi_extrapolate 11 7.9 0.001 0.001 7.387 7.387 density_rs2pw 128 9.7 0.006 0.006 6.852 7.295 cp_fm_upper_to_full 105 14.8 5.751 7.266 5.751 7.266 cp_fm_cholesky_decompose 22 10.9 7.197 7.226 7.197 7.226 dbcsr_complete_redistribute 395 12.7 1.163 1.196 4.564 6.331 calculate_dm_sparse 128 9.5 0.001 0.001 5.837 5.937 fft3d_ps 1291 14.7 2.820 3.053 5.818 5.901 grid_collocate_task_list 128 9.7 5.167 5.684 5.167 5.684 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.430 5.437 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.510 5.234 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.407 5.172 mp_alltoall_d11v 2415 14.1 4.230 4.926 4.230 4.926 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.541 4.684 mp_allgather_i34 2507 14.6 1.479 4.672 1.479 4.672 mp_sum_l 7950 12.9 3.207 4.508 3.207 4.508 potential_pw2rs 128 12.3 0.020 0.022 4.343 4.352 transfer_fm_to_dbcsr 11 9.9 0.020 0.025 2.338 4.076 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.958 3.985 mp_alltoall_i22 716 14.1 1.932 3.821 1.932 3.821 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.752 3.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.468 3.520 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.071000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=881.727273, yerr=11.152586 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.184035E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496376. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.061 188.615 188.616 qs_mol_dyn_low 1 2.0 0.004 0.008 188.110 188.123 qs_forces 11 3.9 0.004 0.004 187.808 187.814 qs_energies 11 4.9 0.002 0.007 180.636 180.649 scf_env_do_scf 11 5.9 0.001 0.002 163.938 163.949 velocity_verlet 10 3.0 0.002 0.006 124.518 124.529 scf_env_do_scf_inner_loop 116 6.6 0.013 0.088 117.211 117.213 qs_scf_new_mos 116 7.6 0.001 0.001 80.906 81.241 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.905 81.240 dbcsr_multiply_generic 2485 12.5 0.191 0.196 80.163 80.769 ot_scf_mini 116 9.6 0.003 0.004 76.373 76.703 multiply_cannon 2485 13.5 0.545 0.574 55.392 58.302 multiply_cannon_loop 2485 14.5 1.798 1.851 51.684 53.347 init_scf_loop 11 6.9 0.001 0.003 46.599 46.601 ot_mini 116 10.6 0.001 0.001 42.734 43.045 prepare_preconditioner 11 7.9 0.000 0.000 42.497 42.524 make_preconditioner 11 8.9 0.000 0.001 42.497 42.524 make_full_inverse_cholesky 11 9.9 0.012 0.023 36.080 41.113 multiply_cannon_multrec 29820 15.5 13.415 18.769 26.267 31.230 rebuild_ks_matrix 127 8.3 0.001 0.001 29.360 29.635 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.360 29.635 mp_waitall_1 146592 16.7 17.421 26.906 17.421 26.906 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.426 26.679 qs_ot_get_derivative 116 11.6 0.001 0.002 23.094 23.417 make_m2s 4970 13.5 0.091 0.094 20.463 21.513 make_images 4970 14.5 1.948 2.282 20.166 21.222 qs_ot_get_p 127 10.4 0.001 0.002 19.817 20.194 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.108 19.545 apply_single 127 13.6 0.001 0.001 19.107 19.545 ot_diis_step 116 11.6 0.017 0.018 19.523 19.530 cp_fm_upper_to_full 104 14.8 11.535 16.954 11.535 16.954 cp_fm_cholesky_invert 11 10.9 16.194 16.203 16.194 16.203 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.629 15.678 multiply_cannon_metrocomm3 29820 15.5 0.048 0.052 6.481 15.337 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.234 14.236 sum_up_and_integrate 127 10.3 0.002 0.003 13.989 14.017 integrate_v_rspace 127 11.3 0.003 0.004 13.930 13.959 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.801 13.859 calculate_rho_elec 127 8.7 0.171 0.186 13.800 13.858 make_images_data 4970 15.5 0.065 0.071 11.041 13.043 dbcsr_mm_accdrv_process 61748 16.2 8.303 9.297 12.420 12.929 dbcsr_complete_redistribute 393 12.7 1.490 1.600 9.110 12.844 hybrid_alltoall_any 5155 16.4 0.527 2.206 9.862 12.130 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.739 11.478 multiply_cannon_sync_h2d 29820 15.5 10.292 11.063 10.292 11.063 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.024 11.041 cp_fm_diag_elpa_base 82 14.4 10.054 10.371 11.018 11.033 init_scf_run 11 5.9 0.000 0.001 10.737 10.738 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.736 10.738 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.394 10.050 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.662 9.890 mp_alltoall_i22 712 14.1 5.612 9.297 5.612 9.297 pw_transfer 1535 11.6 0.085 0.101 9.183 9.265 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.959 9.046 fft_wrap_pw1pw2_140 519 13.2 1.243 1.286 7.973 8.076 grid_integrate_task_list 127 12.3 7.437 7.895 7.437 7.895 cp_fm_cholesky_decompose 22 10.9 7.607 7.714 7.607 7.714 wfi_extrapolate 11 7.9 0.001 0.001 7.688 7.688 density_rs2pw 127 9.7 0.005 0.006 6.822 7.237 multiply_cannon_metrocomm4 24850 15.5 0.080 0.089 2.835 7.165 mp_irecv_dv 75445 16.2 2.679 6.889 2.679 6.889 calculate_dm_sparse 127 9.5 0.001 0.001 6.216 6.303 fft3d_ps 1281 14.7 2.978 3.068 6.026 6.095 grid_collocate_task_list 127 9.7 5.247 5.558 5.247 5.558 mp_alltoall_d11v 2401 14.1 5.001 5.546 5.001 5.546 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.428 5.481 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.446 4.542 qs_energies_init_hamiltonians 11 5.9 0.007 0.014 4.481 4.482 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.368 4.480 potential_pw2rs 127 12.3 0.022 0.023 4.328 4.338 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.210 4.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.616000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1105.727273, yerr=26.505886 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 1.515966E+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 58273. MP_Allreduce 11070 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.049 169.683 169.684 qs_mol_dyn_low 1 2.0 0.003 0.003 169.214 169.228 qs_forces 11 3.9 0.004 0.005 169.096 169.099 qs_energies 11 4.9 0.002 0.002 161.689 161.694 scf_env_do_scf 11 5.9 0.001 0.001 144.501 144.512 velocity_verlet 10 3.0 0.002 0.002 110.896 110.902 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.796 108.797 qs_scf_new_mos 116 7.6 0.001 0.001 71.768 71.859 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.767 71.859 dbcsr_multiply_generic 2485 12.5 0.183 0.191 71.421 71.802 ot_scf_mini 116 9.6 0.003 0.003 67.395 67.462 multiply_cannon 2485 13.5 0.555 0.590 52.756 56.176 multiply_cannon_loop 2485 14.5 0.800 0.836 49.841 50.644 ot_mini 116 10.6 0.001 0.001 37.155 37.226 init_scf_loop 11 6.9 0.000 0.000 35.557 35.558 prepare_preconditioner 11 7.9 0.000 0.000 31.700 31.733 make_preconditioner 11 8.9 0.000 0.000 31.700 31.732 mp_waitall_1 124680 16.7 24.420 30.640 24.420 30.640 make_full_inverse_cholesky 11 9.9 0.013 0.026 29.525 29.818 rebuild_ks_matrix 127 8.3 0.001 0.001 29.152 29.244 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.152 29.243 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.563 26.646 multiply_cannon_multrec 9940 15.5 10.327 14.343 17.890 21.102 qs_ot_get_derivative 116 11.6 0.001 0.002 20.340 20.411 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.370 19.477 cp_fm_cholesky_invert 11 10.9 18.025 18.031 18.025 18.031 qs_ot_get_p 127 10.4 0.001 0.002 17.624 17.718 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.865 17.019 apply_single 127 13.6 0.001 0.001 16.865 17.019 ot_diis_step 116 11.6 0.020 0.020 16.750 16.750 make_m2s 4970 13.5 0.063 0.068 14.760 15.577 make_images 4970 14.5 2.147 2.569 14.455 15.267 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.254 14.293 calculate_rho_elec 127 8.7 0.253 0.264 14.254 14.293 sum_up_and_integrate 127 10.3 0.002 0.002 14.159 14.203 integrate_v_rspace 127 11.3 0.004 0.004 14.099 14.145 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.831 13.850 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.656 12.656 multiply_cannon_sync_h2d 9940 15.5 10.818 11.320 10.818 11.320 init_scf_run 11 5.9 0.000 0.001 10.364 10.365 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.364 10.365 pw_transfer 1535 11.6 0.084 0.092 9.932 9.966 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.778 9.790 cp_fm_diag_elpa_base 82 14.4 9.541 9.614 9.773 9.785 make_images_data 4970 15.5 0.055 0.064 8.485 9.763 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.711 9.749 hybrid_alltoall_any 5155 16.4 0.836 3.737 8.323 9.744 fft_wrap_pw1pw2_140 519 13.2 1.661 1.691 8.575 8.618 cp_fm_cholesky_decompose 22 10.9 7.938 8.118 7.938 8.118 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.014 8.076 grid_integrate_task_list 127 12.3 7.704 7.947 7.704 7.947 dbcsr_mm_accdrv_process 20590 16.1 3.378 4.469 7.214 7.876 wfi_extrapolate 11 7.9 0.001 0.001 7.398 7.398 density_rs2pw 127 9.7 0.005 0.005 6.990 7.103 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.312 7.064 fft3d_ps 1281 14.7 3.127 3.229 6.179 6.203 calculate_dm_sparse 127 9.5 0.001 0.001 6.029 6.100 grid_collocate_task_list 127 9.7 5.529 5.770 5.529 5.770 dbcsr_complete_redistribute 393 12.7 2.086 2.135 5.296 5.737 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.261 5.262 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.133 5.150 mp_alltoall_d11v 2401 14.1 4.740 5.052 4.740 5.052 mp_allgather_i34 2485 14.5 1.114 4.611 1.114 4.611 potential_pw2rs 127 12.3 0.026 0.026 4.369 4.374 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.044 4.096 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.577 3.870 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.436 3.775 multiply_cannon_metrocomm4 7455 15.5 0.026 0.028 1.704 3.773 mp_irecv_dv 28618 15.9 1.664 3.696 1.664 3.696 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.609 3.678 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.450 3.606 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.542 3.572 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.488 3.497 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.684000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1423.454545, yerr=25.367709 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.132498E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4057 58363. MP_Allreduce 11219 1499. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 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 0.021 0.047 290.881 290.881 qs_mol_dyn_low 1 2.0 0.003 0.003 290.323 290.335 qs_forces 11 3.9 0.005 0.005 290.232 290.234 qs_energies 11 4.9 0.002 0.002 281.363 281.369 scf_env_do_scf 11 5.9 0.001 0.001 258.989 258.995 velocity_verlet 10 3.0 0.002 0.002 211.358 211.366 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 134.879 134.880 init_scf_loop 11 6.9 0.000 0.000 123.839 123.840 prepare_preconditioner 11 7.9 0.000 0.000 119.002 119.022 make_preconditioner 11 8.9 0.000 0.000 119.002 119.022 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.004 116.039 qs_scf_new_mos 118 7.6 0.001 0.001 89.698 89.739 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.697 89.738 ot_scf_mini 118 9.6 0.004 0.004 84.835 84.848 dbcsr_multiply_generic 2535 12.6 0.231 0.243 82.141 82.560 cp_fm_upper_to_full 106 14.8 53.144 76.026 53.144 76.026 multiply_cannon 2535 13.6 0.684 0.730 59.122 60.731 multiply_cannon_loop 2535 14.6 1.054 1.097 55.070 56.233 ot_mini 118 10.6 0.001 0.001 44.194 44.206 dbcsr_complete_redistribute 397 12.7 3.978 4.028 29.860 43.000 copy_fm_to_dbcsr 210 11.7 0.001 0.001 26.380 39.509 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.956 36.895 mp_alltoall_i22 720 14.1 21.700 34.955 21.700 34.955 rebuild_ks_matrix 129 8.3 0.001 0.001 34.919 34.936 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 34.919 34.935 cp_fm_cholesky_invert 11 10.9 32.963 32.969 32.963 32.969 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.457 32.477 mp_waitall_1 104820 16.8 27.721 32.276 27.721 32.276 qs_ot_get_p 129 10.4 0.002 0.002 25.417 25.459 qs_ot_get_derivative 118 11.6 0.002 0.002 24.701 24.709 qs_ot_p2m_diag 84 11.4 0.891 0.896 21.355 21.382 make_m2s 5070 13.6 0.074 0.077 18.504 19.994 multiply_cannon_metrocomm3 10140 15.6 0.025 0.025 18.709 19.886 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.552 19.553 make_images 5070 14.6 3.103 3.304 18.019 19.508 ot_diis_step 118 11.6 0.022 0.022 19.469 19.469 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.756 18.831 apply_single 129 13.6 0.001 0.001 18.755 18.831 multiply_cannon_multrec 10140 15.6 10.459 12.390 18.323 18.431 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.407 17.418 calculate_rho_elec 129 8.7 0.483 0.484 17.406 17.417 sum_up_and_integrate 129 10.3 0.002 0.002 16.555 16.652 integrate_v_rspace 129 11.3 0.005 0.005 16.491 16.587 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.368 16.369 cp_fm_diag_elpa_base 84 14.4 11.919 13.595 16.364 16.365 multiply_cannon_sync_h2d 10140 15.6 14.416 14.484 14.416 14.484 pw_transfer 1559 11.6 0.096 0.096 13.184 13.192 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.942 12.947 hybrid_alltoall_any 5257 16.5 1.331 3.113 10.648 12.933 make_images_data 5070 15.6 0.064 0.070 10.368 12.516 init_scf_run 11 5.9 0.000 0.001 12.095 12.096 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.095 12.096 fft_wrap_pw1pw2_140 527 13.2 3.246 3.316 11.535 11.551 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.784 9.788 dbcsr_mm_accdrv_process 20958 16.1 4.321 6.245 7.611 9.528 wfi_extrapolate 11 7.9 0.001 0.001 8.971 8.971 grid_integrate_task_list 129 12.3 8.621 8.815 8.621 8.815 cp_fm_cholesky_decompose 22 10.9 8.756 8.770 8.756 8.770 density_rs2pw 129 9.7 0.005 0.005 8.370 8.414 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.161 8.162 fft3d_ps 1301 14.7 4.009 4.022 7.345 7.390 mp_alltoall_d11v 2429 14.1 6.867 7.007 6.867 7.007 calculate_dm_sparse 129 9.5 0.001 0.001 6.715 6.768 grid_collocate_task_list 129 9.7 6.467 6.542 6.467 6.542 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.331 6.358 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.110 6.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.881000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2858.545455, yerr=149.329631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 1.251979E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.059 85.628 85.629 qs_energies 1 2.0 0.000 0.000 85.020 85.027 ls_scf 1 3.0 0.000 0.000 83.377 83.384 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.478 72.643 multiply_cannon 111 7.7 0.017 0.021 55.883 56.989 multiply_cannon_loop 111 8.7 0.230 0.241 52.447 53.632 ls_scf_main 1 4.0 0.000 0.000 52.203 52.214 density_matrix_trs4 2 5.0 0.002 0.003 46.679 46.773 ls_scf_init_scf 1 4.0 0.000 0.000 28.117 28.118 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.991 27.047 mp_waitall_1 11031 10.9 22.431 25.519 22.431 25.519 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.907 24.924 multiply_cannon_multrec 2664 9.7 8.119 8.844 15.581 17.281 multiply_cannon_sync_h2d 2664 9.7 13.494 15.457 13.494 15.457 make_m2s 222 7.7 0.009 0.011 13.028 13.597 make_images 222 8.7 0.098 0.108 13.006 13.577 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.654 12.075 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.490 8.550 make_images_data 222 9.7 0.004 0.005 7.625 8.168 dbcsr_mm_accdrv_process 4760 10.4 0.597 0.720 7.081 8.062 hybrid_alltoall_any 227 10.6 0.216 1.833 6.494 7.790 dbcsr_mm_accdrv_process_sort 4760 11.4 6.285 7.169 6.285 7.169 calculate_norms 4752 9.8 5.507 6.129 5.507 6.129 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.054 5.146 mp_sum_l 887 5.1 3.035 4.414 3.035 4.414 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.056 3.687 mp_irecv_dv 6231 10.9 2.039 3.658 2.039 3.658 make_images_sizes 222 9.7 0.000 0.000 0.695 3.378 mp_alltoall_i44 222 10.7 0.695 3.377 0.695 3.377 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.261 3.269 arnoldi_extremal 4 6.8 0.000 0.000 3.199 3.223 arnoldi_normal_ev 4 7.8 0.001 0.004 3.199 3.223 build_subspace 16 8.4 0.009 0.012 3.110 3.112 ls_scf_post 1 4.0 0.000 0.000 3.058 3.065 ls_scf_store_result 1 5.0 0.000 0.000 2.868 2.919 dbcsr_special_finalize 555 9.7 0.005 0.006 2.354 2.749 dbcsr_merge_single_wm 555 10.7 0.461 0.598 2.346 2.740 make_images_pack 222 9.7 2.207 2.627 2.209 2.629 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.310 2.574 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.481 2.070 2.483 dbcsr_sort_data 658 11.4 2.142 2.461 2.142 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.272 2.351 buffer_matrices_ensure_size 222 8.7 1.744 2.055 1.744 2.055 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.755 1.756 rebuild_ks_matrix 3 7.3 0.000 0.000 1.746 1.747 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.746 1.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.629000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 2.121986E+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 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.036 89.822 89.823 qs_energies 1 2.0 0.000 0.000 89.373 89.377 ls_scf 1 3.0 0.000 0.000 87.961 87.966 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.970 74.375 multiply_cannon 111 7.7 0.028 0.040 52.579 56.423 ls_scf_main 1 4.0 0.000 0.000 54.311 54.317 multiply_cannon_loop 111 8.7 0.136 0.145 49.842 52.983 density_matrix_trs4 2 5.0 0.002 0.003 48.622 48.856 ls_scf_init_scf 1 4.0 0.000 0.000 29.886 29.887 mp_waitall_1 9105 10.9 20.628 29.789 20.628 29.789 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.703 28.796 multiply_cannon_multrec 1332 9.7 13.336 17.488 22.691 27.973 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.340 26.350 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.375 20.440 make_m2s 222 7.7 0.006 0.007 15.052 15.618 make_images 222 8.7 1.374 1.697 15.021 15.587 dbcsr_mm_accdrv_process 4041 10.4 0.313 0.499 8.952 10.505 dbcsr_mm_accdrv_process_sort 4041 11.4 8.467 9.977 8.467 9.977 make_images_data 222 9.7 0.004 0.005 8.722 9.550 hybrid_alltoall_any 227 10.6 0.542 2.560 8.095 9.406 mp_sum_l 887 5.1 4.866 8.178 4.866 8.178 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.186 7.866 mp_irecv_dv 3311 11.0 3.165 7.813 3.165 7.813 calculate_norms 2376 9.8 6.062 6.763 6.062 6.763 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.594 6.337 multiply_cannon_sync_h2d 1332 9.7 4.859 6.165 4.859 6.165 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.928 5.148 arnoldi_extremal 4 6.8 0.000 0.000 4.730 4.743 arnoldi_normal_ev 4 7.8 0.001 0.005 4.730 4.743 build_subspace 16 8.4 0.014 0.021 4.481 4.484 ls_scf_post 1 4.0 0.000 0.000 3.765 3.770 ls_scf_store_result 1 5.0 0.000 0.000 3.447 3.591 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.144 3.389 dbcsr_matrix_vector_mult_local 304 10.0 2.728 3.209 2.731 3.210 mp_allgather_i34 111 8.7 0.720 2.905 0.720 2.905 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.618 2.746 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.176 2.612 dbcsr_data_new 4174 10.1 2.119 2.408 2.119 2.408 make_images_pack 222 9.7 1.821 2.139 1.824 2.141 dbcsr_sort_data 436 11.2 1.818 2.064 1.818 2.064 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.847 1.849 rebuild_ks_matrix 3 7.3 0.000 0.000 1.834 1.837 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.834 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.823000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1766.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 2.890863E+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 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.043 91.899 91.900 qs_energies 1 2.0 0.000 0.000 91.338 91.343 ls_scf 1 3.0 0.000 0.000 89.914 89.917 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.696 74.986 multiply_cannon 111 7.7 0.034 0.105 51.936 56.129 ls_scf_main 1 4.0 0.000 0.000 56.012 56.017 multiply_cannon_loop 111 8.7 0.116 0.129 49.204 53.081 density_matrix_trs4 2 5.0 0.002 0.003 50.048 50.279 mp_waitall_1 7281 11.0 23.209 33.162 23.209 33.162 ls_scf_init_scf 1 4.0 0.000 0.000 30.233 30.236 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.048 29.125 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.719 26.735 multiply_cannon_multrec 888 9.7 12.666 15.381 21.308 24.644 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.815 22.691 make_m2s 222 7.7 0.006 0.007 16.073 16.825 make_images 222 8.7 1.586 1.855 16.035 16.785 make_images_data 222 9.7 0.004 0.004 9.373 10.385 hybrid_alltoall_any 227 10.6 0.643 2.954 8.987 9.979 dbcsr_mm_accdrv_process 3754 10.4 0.307 0.494 8.168 9.452 dbcsr_mm_accdrv_process_sort 3754 11.4 7.732 8.958 7.732 8.958 mp_sum_l 887 5.1 4.983 8.190 4.983 8.190 multiply_cannon_sync_h2d 888 9.7 6.055 7.429 6.055 7.429 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.482 7.052 mp_irecv_dv 2335 11.1 2.466 6.991 2.466 6.991 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.667 6.787 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.818 6.467 arnoldi_extremal 4 6.8 0.000 0.000 5.041 5.061 arnoldi_normal_ev 4 7.8 0.001 0.005 5.041 5.061 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.794 5.048 calculate_norms 1584 9.8 4.411 4.867 4.411 4.867 build_subspace 16 8.4 0.014 0.020 4.742 4.747 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.415 3.749 mp_allgather_i34 111 8.7 0.854 3.680 0.854 3.680 ls_scf_post 1 4.0 0.000 0.000 3.668 3.674 dbcsr_matrix_vector_mult_local 304 10.0 3.005 3.582 3.007 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.426 3.496 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.874 2.995 dbcsr_data_new 4116 9.9 2.102 2.449 2.102 2.449 dbcsr_sort_data 325 11.1 1.902 2.144 1.902 2.144 dbcsr_finalize 304 7.8 0.026 0.032 1.606 1.905 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.885 1.887 make_images_pack 222 9.7 1.625 1.883 1.628 1.886 rebuild_ks_matrix 3 7.3 0.000 0.000 1.867 1.869 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.867 1.869 make_images_sizes 222 9.7 0.000 0.000 0.842 1.866 mp_alltoall_i44 222 10.7 0.842 1.866 0.842 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.900000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2206.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 3.347034E+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 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.044 96.997 96.997 qs_energies 1 2.0 0.000 0.000 96.425 96.429 ls_scf 1 3.0 0.000 0.000 94.754 94.757 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.478 78.708 ls_scf_main 1 4.0 0.000 0.000 58.851 58.852 multiply_cannon 111 7.7 0.049 0.083 51.907 56.201 density_matrix_trs4 2 5.0 0.002 0.003 52.679 52.821 multiply_cannon_loop 111 8.7 0.152 0.170 46.860 49.226 ls_scf_init_scf 1 4.0 0.000 0.000 32.665 32.667 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.432 31.517 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.959 28.973 mp_waitall_1 6369 11.0 22.767 28.869 22.767 28.869 multiply_cannon_multrec 1332 9.7 14.162 17.022 22.130 25.087 make_m2s 222 7.7 0.006 0.007 21.044 22.510 make_images 222 8.7 3.145 3.606 20.994 22.462 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.361 17.201 make_images_data 222 9.7 0.004 0.004 11.755 13.332 hybrid_alltoall_any 227 10.6 0.793 3.736 11.057 13.008 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.474 7.613 9.155 dbcsr_mm_accdrv_process_sort 3641 11.4 7.171 8.655 7.171 8.655 mp_sum_l 887 5.1 3.936 7.571 3.936 7.571 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.093 6.295 mp_irecv_dv 3229 10.9 2.068 6.210 2.068 6.210 multiply_cannon_sync_h2d 1332 9.7 5.429 6.134 5.429 6.134 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.065 6.088 arnoldi_extremal 4 6.8 0.000 0.000 5.259 5.272 arnoldi_normal_ev 4 7.8 0.001 0.005 5.259 5.272 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.477 5.088 build_subspace 16 8.4 0.014 0.021 4.928 4.935 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.577 4.714 calculate_norms 2376 9.8 4.217 4.607 4.217 4.607 mp_allgather_i34 111 8.7 2.159 4.474 2.159 4.474 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.594 3.915 dbcsr_matrix_vector_mult_local 304 10.0 3.174 3.677 3.176 3.679 dbcsr_sort_data 658 11.4 3.091 3.445 3.091 3.445 ls_scf_post 1 4.0 0.000 0.000 3.238 3.243 dbcsr_special_finalize 555 9.7 0.006 0.007 2.845 3.208 dbcsr_merge_single_wm 555 10.7 0.543 0.658 2.837 3.200 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.025 3.073 ls_scf_store_result 1 5.0 0.000 0.000 2.981 3.038 dbcsr_data_release 10477 10.7 1.575 2.365 1.575 2.365 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.975 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.997000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2793.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 4.764479E+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 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.052 92.093 92.093 qs_energies 1 2.0 0.000 0.000 91.343 91.347 ls_scf 1 3.0 0.000 0.000 89.260 89.262 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.572 70.807 ls_scf_main 1 4.0 0.000 0.000 56.352 56.352 multiply_cannon 111 7.7 0.065 0.122 52.587 55.485 multiply_cannon_loop 111 8.7 0.088 0.094 50.011 51.090 density_matrix_trs4 2 5.0 0.002 0.003 49.370 49.457 ls_scf_init_scf 1 4.0 0.000 0.000 29.515 29.516 mp_waitall_1 5436 11.0 24.190 28.684 24.190 28.684 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.252 28.293 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.903 25.915 multiply_cannon_multrec 444 9.7 13.716 16.533 20.866 22.137 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.906 15.371 make_m2s 222 7.7 0.005 0.005 13.364 14.304 make_images 222 8.7 2.042 2.481 13.296 14.233 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.984 13.444 hybrid_alltoall_any 227 10.6 0.801 3.830 8.176 9.693 make_images_data 222 9.7 0.003 0.004 8.115 9.640 dbcsr_mm_accdrv_process 3003 10.4 0.352 0.401 6.834 7.950 multiply_cannon_sync_h2d 444 9.7 6.781 7.945 6.781 7.945 dbcsr_mm_accdrv_process_sort 3003 11.4 6.471 7.551 6.471 7.551 arnoldi_extremal 4 6.8 0.000 0.000 5.788 5.801 arnoldi_normal_ev 4 7.8 0.002 0.005 5.788 5.801 build_subspace 16 8.4 0.015 0.020 5.387 5.397 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.392 4.641 mp_sum_l 887 5.1 2.619 4.388 2.619 4.388 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.152 4.339 dbcsr_matrix_vector_mult_local 304 10.0 3.669 4.141 3.671 4.143 calculate_norms 792 9.8 3.619 3.713 3.619 3.713 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.867 3.672 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.515 3.658 mp_irecv_dv 1241 11.2 1.503 3.630 1.503 3.630 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.550 3.598 mp_allgather_i34 111 8.7 1.144 3.500 1.144 3.500 ls_scf_post 1 4.0 0.000 0.000 3.393 3.398 make_images_sizes 222 9.7 0.000 0.000 0.853 3.239 mp_alltoall_i44 222 10.7 0.853 3.239 0.853 3.239 ls_scf_store_result 1 5.0 0.000 0.000 3.188 3.220 dbcsr_data_new 4608 9.7 1.819 2.332 1.819 2.332 dbcsr_finalize 304 7.8 0.062 0.076 2.199 2.269 dbcsr_merge_all 275 8.9 0.480 0.522 2.056 2.116 qs_energies_init_hamiltonians 1 3.0 0.010 0.058 2.067 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.045 2.046 rebuild_ks_matrix 3 7.3 0.000 0.000 2.013 2.014 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.013 2.014 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.093000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_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 8.867738E+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 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.105 0.296 106.255 106.256 qs_energies 1 2.0 0.000 0.000 104.059 104.063 ls_scf 1 3.0 0.000 0.000 101.076 101.080 dbcsr_multiply_generic 111 6.7 0.023 0.028 74.699 74.835 ls_scf_main 1 4.0 0.000 0.000 63.675 63.675 density_matrix_trs4 2 5.0 0.002 0.003 54.612 54.669 multiply_cannon 111 7.7 0.155 0.196 48.298 50.374 multiply_cannon_loop 111 8.7 0.098 0.099 45.355 46.070 ls_scf_init_scf 1 4.0 0.000 0.000 33.622 33.623 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.058 32.082 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.368 29.376 mp_waitall_1 4527 11.1 21.825 25.358 21.825 25.358 make_m2s 222 7.7 0.005 0.005 22.673 23.710 make_images 222 8.7 3.578 3.874 22.565 23.602 multiply_cannon_multrec 444 9.7 17.809 18.497 22.494 23.182 hybrid_alltoall_any 227 10.6 1.652 3.605 12.893 15.798 make_images_data 222 9.7 0.003 0.004 13.086 15.168 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.155 10.523 multiply_cannon_sync_h2d 444 9.7 8.789 8.850 8.789 8.850 arnoldi_extremal 4 6.8 0.000 0.000 7.318 7.330 arnoldi_normal_ev 4 7.8 0.003 0.009 7.317 7.330 build_subspace 16 8.4 0.026 0.036 6.775 6.788 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.427 5.565 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.190 5.283 dbcsr_matrix_vector_mult_local 304 10.0 4.974 5.264 4.976 5.267 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.855 5.106 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.354 4.500 4.631 dbcsr_mm_accdrv_process_sort 1814 11.4 4.164 4.297 4.164 4.297 ls_scf_post 1 4.0 0.000 0.000 3.779 3.784 make_images_sizes 222 9.7 0.000 0.000 1.450 3.541 mp_alltoall_i44 222 10.7 1.450 3.540 1.450 3.540 ls_scf_store_result 1 5.0 0.000 0.000 3.499 3.539 mp_allgather_i34 111 8.7 1.058 3.367 1.058 3.367 calculate_norms 792 9.8 3.232 3.282 3.232 3.282 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.160 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.952 2.952 dbcsr_merge_all 275 8.9 0.887 0.919 2.863 2.939 dbcsr_complete_redistribute 5 7.6 1.438 1.474 2.744 2.853 matrix_ls_to_qs 2 6.0 0.000 0.000 2.403 2.521 dbcsr_sort_data 325 11.1 2.435 2.491 2.435 2.491 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.463 2.465 rebuild_ks_matrix 3 7.3 0.000 0.000 2.397 2.399 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.006 2.397 2.399 dbcsr_new_transposed 4 7.5 0.242 0.252 2.289 2.296 dbcsr_data_new 6591 9.6 1.841 2.290 1.841 2.290 dbcsr_frobenius_norm 74 6.6 2.057 2.133 2.212 2.246 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.204 dbcsr_add_anytype 103 7.2 0.858 0.891 2.126 2.204 dbcsr_data_release 12724 10.6 1.977 2.173 1.977 2.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.256000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7025.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/99544d10ee311fd730c849cdf817355d5af06832_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 585.351168E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 793765747. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.043 224.878 224.879 qs_mol_dyn_low 1 2.0 0.010 0.063 223.883 223.912 qs_forces 5 3.8 0.005 0.032 223.703 223.706 qs_energies 5 4.8 0.003 0.040 220.692 220.719 scf_env_do_scf 5 5.8 0.000 0.001 207.517 207.519 scf_env_do_scf_inner_loop 105 6.6 0.002 0.007 181.265 181.266 qs_scf_new_mos 105 7.6 0.000 0.001 141.351 141.503 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.350 141.503 ot_scf_mini 105 9.6 0.003 0.003 131.538 131.681 dbcsr_multiply_generic 1445 12.2 0.129 0.138 130.938 131.413 multiply_cannon 1445 13.2 0.274 0.285 112.230 114.334 multiply_cannon_loop 1445 14.2 2.836 2.965 110.500 111.638 velocity_verlet 4 3.0 0.005 0.031 106.228 106.231 ot_mini 105 10.6 0.001 0.001 59.688 59.813 multiply_cannon_multrec 69360 15.2 29.780 34.639 39.652 44.659 qs_ot_get_p 112 10.4 0.001 0.001 41.931 42.266 mp_waitall_1 488190 16.1 34.085 39.957 34.085 39.957 qs_ot_get_derivative 55 11.6 0.001 0.001 37.946 38.058 multiply_cannon_sync_h2d 69360 15.2 29.159 33.229 29.159 33.229 multiply_cannon_metrocomm3 69360 15.2 0.201 0.213 25.278 31.643 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.832 30.926 rebuild_ks_matrix 110 8.4 0.000 0.000 28.999 29.197 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.016 28.998 29.197 cp_dbcsr_syevd 40 12.0 0.002 0.003 27.616 27.617 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.649 26.820 init_scf_loop 7 6.6 0.000 0.000 26.219 26.220 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.064 23.300 apply_single 62 13.6 0.000 0.000 23.064 23.300 cp_fm_syevd 40 13.0 0.000 0.000 22.481 22.612 prepare_preconditioner 7 7.6 0.000 0.000 21.370 21.408 make_preconditioner 7 8.6 0.000 0.000 21.370 21.408 ot_new_cg_direction 55 11.6 0.001 0.001 21.081 21.082 cp_fm_redistribute_end 40 14.0 8.872 17.694 8.877 17.695 cp_fm_syevd_base 40 14.0 8.811 17.636 8.811 17.636 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.749 17.129 calculate_rho_elec 110 8.6 0.029 0.033 16.749 17.129 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.715 14.870 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.290 14.354 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.605 13.703 mp_sum_l 4764 12.2 11.797 12.533 11.797 12.533 pw_transfer 1645 12.4 0.079 0.102 11.775 12.026 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.637 11.894 density_rs2pw 110 9.6 0.005 0.006 10.977 11.548 calculate_dm_sparse 110 9.5 0.000 0.001 11.361 11.520 qs_vxc_create 110 10.4 0.002 0.003 10.520 10.570 dbcsr_mm_accdrv_process 154766 15.8 6.259 6.465 9.740 10.521 fft_wrap_pw1pw2_240 915 15.0 0.823 0.924 10.278 10.506 init_scf_run 5 5.8 0.000 0.001 10.386 10.387 scf_env_initial_rho_setup 5 6.8 0.002 0.003 10.386 10.387 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.309 10.366 cp_fm_cholesky_invert 7 10.6 9.988 9.995 9.988 9.995 check_diag 80 13.5 8.686 8.937 9.550 9.684 fft3d_pb 915 16.0 2.396 2.687 8.483 8.750 sum_up_and_integrate 60 10.3 0.001 0.007 8.477 8.488 integrate_v_rspace 60 11.3 0.002 0.003 8.460 8.471 transfer_rs2pw 445 10.6 0.007 0.008 7.676 8.324 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.370 7.625 xc_vxc_pw_create 60 11.3 0.039 0.049 7.058 7.106 make_m2s 2890 13.2 0.078 0.088 6.411 6.989 make_images 2890 14.2 0.237 0.258 6.303 6.878 make_full_single_inverse 7 9.6 0.001 0.001 6.823 6.861 multiply_cannon_metrocomm1 69360 15.2 0.098 0.104 4.319 6.702 acc_transpose_blocks 69360 15.2 0.357 0.375 6.164 6.445 xc_pw_derive 510 13.4 0.005 0.007 6.274 6.354 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 5.817 5.833 mp_alltoall_z22v 2340 17.7 5.507 5.789 5.507 5.789 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.389 5.454 calculate_first_density_matrix 1 7.0 0.000 0.002 5.418 5.433 mp_waitany 7680 13.5 4.389 5.195 4.389 5.195 potential_pw2rs 60 12.3 0.002 0.003 4.674 4.703 wfi_extrapolate 5 7.8 0.000 0.002 4.678 4.678 multiply_cannon_metrocomm4 67915 15.2 0.190 0.209 2.018 4.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=224.879000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=555.200000, yerr=3.429286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 99544d10ee311fd730c849cdf817355d5af06832 Summary: empty Status: OK