=== 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: aeba16654204b33cadbf84abc85e7a1388c87784 ################# 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 2.3.1, LIBGRPP 20231215, # SPFFT 1.0.6, SPLA 1.5.5 # # 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: 28.03.2024 # # \ 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; \ rm -rf build; \ ./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 --with-spfft --with-spla; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 2.3.1 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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 ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(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_f90cstub.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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/01 job id: 53235823 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/02 job id: 53235824 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/03 job id: 53235825 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/04 job id: 53235826 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/05 job id: 53235827 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/06 job id: 53235829 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/07 job id: 53235830 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/08 job id: 53235831 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/09 job id: 53235833 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/10 job id: 53235834 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/11 job id: 53235835 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/12 job id: 53235836 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/13 job id: 53235837 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/14 job id: 53235839 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/15 job id: 53235840 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/16 job id: 53235841 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/17 job id: 53235842 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/18 job id: 53235844 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/19 job id: 53235845 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/20 job id: 53235846 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/21 job id: 53235850 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/22 job id: 53235852 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/23 job id: 53235853 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/24 job id: 53235854 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/25 job id: 53235856 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/26 job id: 53235857 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/27 job id: 53235858 --- 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/aeba16654204b33cadbf84abc85e7a1388c87784_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 0.018 0.033 130.730 130.730 farming_run 1 2.0 130.108 130.121 130.701 130.704 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.533473E+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 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1131389. 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 0.006 0.027 111.605 111.606 qs_energies 1 2.0 0.001 0.001 111.413 111.415 mp2_main 1 3.0 0.000 0.000 109.482 109.484 mp2_gpw_main 1 4.0 0.029 0.037 108.544 108.546 mp2_ri_gpw_compute_in 1 5.0 0.176 0.182 89.617 89.878 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.320 57.583 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 43.762 48.769 get_2c_integrals 1 6.0 0.008 0.011 31.816 32.121 integrate_v_rspace 273 8.0 0.442 0.457 25.604 30.098 fft_wrap_pw1pw2 5465 10.4 0.064 0.069 28.136 28.709 grid_integrate_task_list 273 9.0 20.830 26.285 20.830 26.285 fft_wrap_pw1pw2_100 2178 11.4 1.299 1.418 25.587 26.151 compute_2c_integrals 1 7.0 0.002 0.002 22.426 22.427 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 21.097 22.160 mp2_eri_2c_integrate_gpw 1 9.0 2.379 2.425 21.095 22.157 rpa_ri_compute_en 1 5.0 0.019 0.027 18.805 19.041 fft3d_s 5443 12.4 16.960 17.510 16.981 17.532 ao_to_mo_and_store_B_mult_1 272 7.0 10.769 15.302 10.769 15.302 calculate_wavefunction 272 8.0 5.295 5.453 12.784 13.701 calc_potential_gpw 544 9.5 0.004 0.004 11.059 12.525 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 10.039 11.256 rpa_num_int 1 6.0 0.009 0.140 10.647 10.657 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.483 10.490 potential_pw2rs 545 10.0 0.106 0.107 8.580 10.412 cp_fm_cholesky_decompose 12 8.2 9.543 9.857 9.543 9.857 calc_mat_Q 8 8.0 0.000 0.000 9.465 9.549 contract_S_to_Q 8 9.0 0.000 0.000 8.885 8.969 cholesky_decomp 1 7.0 0.000 0.000 8.578 8.895 collocate_single_gaussian 272 10.0 0.039 0.041 7.868 8.623 parallel_gemm_fm 14 9.1 0.000 0.000 8.512 8.615 parallel_gemm_fm_cosma 14 10.1 8.512 8.615 8.512 8.615 create_integ_mat 1 6.0 0.024 0.036 7.800 7.809 array2fm 1 7.0 0.000 0.000 6.780 7.255 pw_poisson_solve 545 10.5 0.010 0.012 4.909 6.208 pw_gather_s 2722 12.2 3.995 5.024 3.995 5.024 pw_scatter_s 2720 12.7 4.404 4.888 4.404 4.888 pw_poisson_set 548 11.5 0.017 0.019 3.514 4.530 pw_copy 4911 11.6 2.754 3.468 2.754 3.468 array2fm_buffer_send 1 8.0 3.004 3.185 3.004 3.185 pw_derive 1641 12.5 2.086 2.656 2.086 2.656 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.081 1.239 2.023 2.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=108.543834, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 0.039 0.099 429.633 429.663 farming_run 1 2.0 428.047 428.090 429.322 429.324 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.280774E+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 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 480173. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 12571716. MP_SendRecv 2876 2171443. 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 0.012 0.034 223.922 223.923 qs_energies 1 2.0 0.004 0.015 223.415 223.422 scf_env_do_scf 1 3.0 0.000 0.000 115.186 115.187 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.319 114.326 rebuild_ks_matrix 4 6.0 0.000 0.000 114.318 114.325 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.318 114.325 hfx_ks_matrix 4 8.0 0.001 0.001 113.971 113.974 integrate_four_center 4 9.0 0.153 0.462 113.970 113.973 mp2_main 1 3.0 0.000 0.000 107.937 107.948 mp2_gpw_main 1 4.0 0.049 0.070 107.086 107.101 integrate_four_center_main 4 10.0 0.119 0.613 101.884 105.859 integrate_four_center_bin 270 11.0 101.764 105.847 101.764 105.847 init_scf_loop 1 4.0 0.000 0.000 97.319 97.319 mp2_ri_gpw_compute_in 1 5.0 0.067 0.075 80.027 81.413 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.932 58.322 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.158 44.382 49.392 integrate_v_rspace 95 8.0 0.396 0.560 28.046 32.818 fft_wrap_pw1pw2 1868 10.4 0.029 0.033 30.415 31.187 mp2_ri_gpw_compute_en 1 5.0 0.054 0.062 26.906 29.270 fft_wrap_pw1pw2_100 730 11.4 0.633 0.749 28.119 28.866 grid_integrate_task_list 95 9.0 23.346 28.343 23.346 28.343 ao_to_mo_and_store_B_mult_1 91 7.0 10.877 27.542 10.877 27.542 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.908 24.872 24.883 get_2c_integrals 1 6.0 0.000 0.000 23.005 23.026 compute_2c_integrals 1 7.0 0.002 0.003 22.172 22.180 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 21.771 22.042 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.856 21.770 22.041 fft3d_s 1823 12.4 19.172 19.722 19.186 19.735 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.866 17.866 calc_potential_gpw 182 9.5 0.002 0.002 12.485 13.290 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.776 11.410 calculate_wavefunction 91 8.0 1.986 2.026 10.007 10.675 potential_pw2rs 186 10.0 0.033 0.034 8.994 10.552 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.577 8.803 9.239 local_gemm 172 8.0 8.248 8.675 8.248 8.675 collocate_single_gaussian 91 10.0 0.017 0.024 8.267 8.668 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.528 7.853 8.259 mp_sync 37 10.5 4.064 8.156 4.064 8.156 integrate_four_center_load 4 10.0 0.000 0.000 6.757 6.760 hfx_load_balance 1 11.0 0.000 0.000 6.757 6.760 mp2_ri_gpw_compute_en_ener 172 7.0 6.333 6.427 6.333 6.427 pw_poisson_solve 186 10.4 0.005 0.006 5.640 6.346 mp_sendrecv_dm3 2068 8.0 5.887 6.297 5.887 6.297 pw_gather_s 912 12.2 4.737 5.432 4.737 5.432 pw_poisson_set 189 11.4 0.009 0.011 4.178 4.843 pw_scatter_s 910 12.7 4.118 4.798 4.118 4.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=107.083873, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1538.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 445.673472E+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 3690 72628. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. 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 0.025 0.053 55.238 55.240 qs_mol_dyn_low 1 2.0 0.025 0.060 54.707 54.749 qs_forces 11 3.9 0.010 0.063 54.114 54.140 qs_energies 11 4.9 0.028 0.066 52.562 52.603 scf_env_do_scf 11 5.9 0.000 0.001 46.394 46.394 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.452 44.452 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.544 35.115 qs_scf_new_mos 108 7.5 0.000 0.001 34.518 34.821 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.517 34.821 ot_scf_mini 108 9.5 0.002 0.002 32.820 33.013 multiply_cannon 2286 13.5 0.188 0.193 26.774 28.284 multiply_cannon_loop 2286 14.5 1.816 1.926 26.119 27.662 velocity_verlet 10 3.0 0.081 0.203 26.859 26.896 ot_mini 108 10.5 0.001 0.001 20.015 20.258 qs_ot_get_derivative 108 11.5 0.001 0.001 17.023 17.214 mp_waitall_1 245248 16.5 8.912 14.910 8.912 14.910 multiply_cannon_metrocomm3 54864 15.5 0.071 0.075 6.319 13.158 multiply_cannon_multrec 54864 15.5 3.637 5.614 7.651 11.110 qs_ot_get_p 119 10.4 0.001 0.001 8.147 8.461 rebuild_ks_matrix 119 8.3 0.000 0.001 7.896 8.066 qs_ks_build_kohn_sham_matrix 119 9.3 0.021 0.062 7.896 8.065 mp_sum_l 7287 12.8 5.825 7.637 5.825 7.637 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.920 7.076 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.114 6.550 multiply_cannon_sync_h2d 54864 15.5 5.144 6.147 5.144 6.147 dbcsr_mm_accdrv_process 76910 16.1 1.829 2.869 3.928 5.662 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.383 5.495 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.133 5.168 init_scf_run 11 5.9 0.000 0.001 4.662 4.662 scf_env_initial_rho_setup 11 6.9 0.003 0.013 4.661 4.662 sum_up_and_integrate 119 10.3 0.001 0.002 4.413 4.444 integrate_v_rspace 119 11.3 0.002 0.003 4.402 4.435 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.316 4.316 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.041 4.137 calculate_rho_elec 119 8.7 0.011 0.018 4.041 4.136 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.121 4.122 cp_fm_redistribute_end 50 14.0 2.104 4.097 2.109 4.100 cp_fm_diag_elpa_base 50 14.0 1.985 4.002 1.989 4.008 calculate_dm_sparse 119 9.5 0.000 0.000 3.142 3.325 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.947 3.149 apply_single 119 13.6 0.000 0.000 2.947 3.148 acc_transpose_blocks 54864 15.5 0.218 0.238 2.326 2.896 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.932 2.757 jit_kernel_multiply 13 15.8 2.035 2.712 2.035 2.712 calculate_first_density_matrix 1 7.0 0.000 0.001 2.705 2.712 ot_diis_step 108 11.5 0.006 0.008 2.695 2.696 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.450 2.519 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.380 2.397 density_rs2pw 119 9.7 0.003 0.004 2.047 2.171 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.037 2.125 grid_integrate_task_list 119 12.3 2.000 2.103 2.000 2.103 mp_sum_d 4139 12.0 1.406 2.075 1.406 2.075 init_scf_loop 11 6.9 0.003 0.010 1.926 1.927 wfi_extrapolate 11 7.9 0.003 0.022 1.881 1.881 potential_pw2rs 119 12.3 0.004 0.004 1.792 1.825 make_m2s 4572 13.5 0.054 0.056 1.652 1.702 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.624 1.655 make_images 4572 14.5 0.133 0.139 1.570 1.619 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.523 1.611 mp_alltoall_d11v 2130 13.8 1.337 1.507 1.337 1.507 transfer_rs2pw 487 10.6 0.006 0.007 1.388 1.500 mp_waitany 12084 13.8 1.315 1.458 1.315 1.458 acc_transpose_blocks_sync 164592 16.5 1.193 1.425 1.193 1.425 grid_collocate_task_list 119 9.7 1.349 1.404 1.349 1.404 transfer_pw2rs 487 13.2 0.005 0.006 1.350 1.379 fft3d_ps 1201 13.6 0.369 0.466 1.297 1.371 fft_wrap_pw1pw2_140 487 12.2 0.049 0.053 1.183 1.273 acc_transpose_blocks_kernels 54864 16.5 0.247 0.380 0.870 1.187 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.745 1.166 mp_alltoall_i22 627 13.8 0.857 1.145 0.857 1.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.240000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.818182, yerr=1.028519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 484.118528E+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 3679 72937. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 788952. 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 0.006 0.025 39.195 39.196 qs_mol_dyn_low 1 2.0 0.003 0.003 38.956 38.964 qs_forces 11 3.9 0.003 0.004 38.892 38.892 qs_energies 11 4.9 0.002 0.002 37.220 37.225 scf_env_do_scf 11 5.9 0.000 0.001 31.224 31.225 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.000 29.001 dbcsr_multiply_generic 2286 12.5 0.101 0.106 22.499 22.851 qs_scf_new_mos 108 7.5 0.001 0.001 20.485 20.723 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.485 20.722 ot_scf_mini 108 9.5 0.002 0.003 19.584 19.758 multiply_cannon 2286 13.5 0.209 0.216 17.408 18.770 velocity_verlet 10 3.0 0.001 0.002 17.978 17.979 multiply_cannon_loop 2286 14.5 1.186 1.244 16.261 17.723 ot_mini 108 10.5 0.001 0.001 12.119 12.358 mp_waitall_1 200699 16.5 5.732 11.114 5.732 11.114 qs_ot_get_derivative 108 11.5 0.001 0.001 9.774 9.947 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.188 9.893 multiply_cannon_multrec 27432 15.5 1.830 4.205 6.340 9.134 rebuild_ks_matrix 119 8.3 0.000 0.000 6.754 6.898 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.754 6.897 dbcsr_mm_accdrv_process 47894 16.0 3.611 6.052 4.428 6.565 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.934 6.064 qs_ot_get_p 119 10.4 0.001 0.001 4.564 4.792 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.660 4.587 init_scf_run 11 5.9 0.000 0.001 4.452 4.452 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.452 4.452 mp_sum_l 7287 12.8 2.081 4.101 2.081 4.101 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.058 4.070 apply_single 119 13.6 0.000 0.000 3.057 4.069 sum_up_and_integrate 119 10.3 0.001 0.001 3.639 3.647 integrate_v_rspace 119 11.3 0.002 0.003 3.625 3.632 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.517 3.548 calculate_rho_elec 119 8.7 0.021 0.024 3.516 3.547 calculate_first_density_matrix 1 7.0 0.000 0.000 2.976 2.979 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.936 2.954 make_m2s 4572 13.5 0.052 0.054 2.560 2.900 make_images 4572 14.5 0.206 0.246 2.470 2.813 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.541 2.541 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.308 2.398 acc_transpose_blocks 27432 15.5 0.111 0.116 1.737 2.346 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.336 2.338 calculate_dm_sparse 119 9.5 0.000 0.001 2.222 2.302 ot_diis_step 108 11.5 0.011 0.011 2.296 2.297 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.225 2.226 cp_fm_redistribute_end 50 14.0 1.130 2.203 1.133 2.205 init_scf_loop 11 6.9 0.000 0.000 2.204 2.204 multiply_cannon_sync_h2d 27432 15.5 1.711 2.181 1.711 2.181 cp_fm_diag_elpa_base 50 14.0 1.040 2.109 1.068 2.147 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.076 2.116 jit_kernel_multiply 9 16.4 0.758 1.977 0.758 1.977 grid_integrate_task_list 119 12.3 1.840 1.947 1.840 1.947 density_rs2pw 119 9.7 0.003 0.004 1.814 1.893 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.652 1.684 make_images_data 4572 15.5 0.047 0.053 1.216 1.534 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.086 1.459 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.429 1.438 wfi_extrapolate 11 7.9 0.001 0.001 1.417 1.418 fft3d_ps 1201 13.6 0.513 0.567 1.371 1.397 potential_pw2rs 119 12.3 0.006 0.006 1.383 1.391 grid_collocate_task_list 119 9.7 1.291 1.358 1.291 1.358 prepare_preconditioner 11 7.9 0.000 0.000 1.326 1.353 make_preconditioner 11 8.9 0.000 0.000 1.326 1.353 fft_wrap_pw1pw2_140 487 12.2 0.048 0.051 1.318 1.353 mp_alltoall_d11v 2130 13.8 1.204 1.323 1.204 1.323 transfer_rs2pw 487 10.6 0.005 0.005 1.107 1.319 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.789 1.293 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.230 1.282 qs_energies_init_hamiltonians 11 5.9 0.003 0.016 1.260 1.262 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.193 1.239 jit_kernel_transpose 5 15.5 0.601 1.031 0.601 1.031 mp_sum_d 4139 12.0 0.530 0.993 0.530 0.993 mp_allgather_i34 2286 14.5 0.570 0.984 0.570 0.984 acc_transpose_blocks_sync 82296 16.5 0.813 0.936 0.813 0.936 transfer_pw2rs 487 13.2 0.004 0.005 0.910 0.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.876 0.889 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.803 0.836 multiply_cannon_metrocomm1 27432 15.5 0.032 0.035 0.358 0.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.196000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.909091, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 515.227648E+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 3629 73909. 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 0.020 0.047 33.341 33.342 qs_mol_dyn_low 1 2.0 0.003 0.003 32.963 32.998 qs_forces 11 3.9 0.002 0.003 32.440 32.440 qs_energies 11 4.9 0.002 0.002 30.877 30.879 scf_env_do_scf 11 5.9 0.000 0.001 25.815 25.815 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.477 23.478 dbcsr_multiply_generic 2286 12.5 0.095 0.097 17.585 17.677 qs_scf_new_mos 108 7.5 0.001 0.001 15.813 15.835 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.812 15.835 velocity_verlet 10 3.0 0.002 0.002 15.664 15.680 ot_scf_mini 108 9.5 0.002 0.002 15.045 15.065 multiply_cannon 2286 13.5 0.194 0.197 14.095 14.742 multiply_cannon_loop 2286 14.5 0.861 0.897 13.328 14.076 ot_mini 108 10.5 0.001 0.001 9.386 9.409 qs_ot_get_derivative 108 11.5 0.001 0.001 7.884 7.905 multiply_cannon_multrec 18288 15.5 1.876 2.904 7.224 7.613 dbcsr_mm_accdrv_process 38222 16.0 5.190 6.084 5.254 6.150 rebuild_ks_matrix 119 8.3 0.000 0.000 6.074 6.094 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.074 6.093 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.355 5.374 mp_waitall_1 158411 16.6 2.825 4.064 2.825 4.064 init_scf_run 11 5.9 0.000 0.001 3.870 3.870 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.870 3.870 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.011 3.724 sum_up_and_integrate 119 10.3 0.001 0.001 3.550 3.555 integrate_v_rspace 119 11.3 0.003 0.003 3.538 3.544 qs_ot_get_p 119 10.4 0.001 0.001 3.386 3.420 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.219 3.225 calculate_rho_elec 119 8.7 0.031 0.031 3.218 3.224 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.554 2.677 calculate_first_density_matrix 1 7.0 0.000 0.000 2.645 2.646 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.125 2.510 apply_single 119 13.6 0.000 0.000 2.125 2.510 init_scf_loop 11 6.9 0.000 0.000 2.316 2.318 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.256 2.262 make_m2s 4572 13.5 0.044 0.045 1.862 1.993 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.974 1.975 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.959 1.959 calculate_dm_sparse 119 9.5 0.000 0.001 1.919 1.928 density_rs2pw 119 9.7 0.003 0.004 1.783 1.907 make_images 4572 14.5 0.193 0.207 1.776 1.906 grid_integrate_task_list 119 12.3 1.810 1.904 1.810 1.904 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.816 1.823 acc_transpose_blocks 18288 15.5 0.077 0.079 1.683 1.762 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.720 1.728 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.709 1.718 cp_fm_diag_elpa_base 50 14.0 1.686 1.698 1.708 1.717 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.637 1.647 mp_sum_l 7287 12.8 1.246 1.637 1.246 1.637 prepare_preconditioner 11 7.9 0.000 0.000 1.492 1.495 make_preconditioner 11 8.9 0.000 0.000 1.492 1.495 ot_diis_step 108 11.5 0.011 0.011 1.481 1.481 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.374 1.430 potential_pw2rs 119 12.3 0.007 0.008 1.341 1.350 grid_collocate_task_list 119 9.7 1.234 1.343 1.234 1.343 fft3d_ps 1201 13.6 0.534 0.555 1.321 1.331 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.301 1.313 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.255 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 transfer_rs2pw 487 10.6 0.005 0.005 1.016 1.148 multiply_cannon_sync_h2d 18288 15.5 1.023 1.144 1.023 1.144 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.981 0.997 qs_energies_init_hamiltonians 11 5.9 0.007 0.033 0.975 0.975 make_images_data 4572 15.5 0.047 0.051 0.784 0.922 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.848 0.890 transfer_pw2rs 487 13.2 0.004 0.004 0.874 0.882 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.674 0.851 acc_transpose_blocks_sync 54864 16.5 0.739 0.811 0.739 0.811 mp_alltoall_d11v 2130 13.8 0.707 0.809 0.707 0.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.791 0.794 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.678 0.736 mp_alltoall_z22v 1201 15.6 0.636 0.711 0.636 0.711 jit_kernel_transpose 5 15.6 0.632 0.676 0.632 0.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.342000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.454545, yerr=1.616035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.302336E+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 3679 72932. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.072 37.463 37.464 qs_mol_dyn_low 1 2.0 0.003 0.003 37.161 37.175 qs_forces 11 3.9 0.002 0.002 36.801 36.802 qs_energies 11 4.9 0.001 0.001 35.094 35.101 scf_env_do_scf 11 5.9 0.000 0.001 29.772 29.772 scf_env_do_scf_inner_loop 108 6.5 0.007 0.020 26.507 26.508 dbcsr_multiply_generic 2286 12.5 0.101 0.105 20.574 20.710 qs_scf_new_mos 108 7.5 0.001 0.001 18.497 18.546 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.497 18.546 velocity_verlet 10 3.0 0.216 0.309 18.315 18.317 ot_scf_mini 108 9.5 0.002 0.003 17.456 17.510 multiply_cannon 2286 13.5 0.218 0.226 16.673 17.167 multiply_cannon_loop 2286 14.5 1.504 1.625 15.688 16.205 ot_mini 108 10.5 0.001 0.001 10.834 10.901 multiply_cannon_multrec 27432 15.5 2.624 3.347 9.377 9.688 qs_ot_get_derivative 108 11.5 0.001 0.001 8.942 8.997 dbcsr_mm_accdrv_process 47916 15.9 6.201 7.791 6.651 8.121 rebuild_ks_matrix 119 8.3 0.000 0.000 6.644 6.698 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.644 6.697 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.929 5.976 init_scf_run 11 5.9 0.000 0.001 3.952 3.952 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.952 3.952 qs_ot_get_p 119 10.4 0.001 0.001 3.836 3.901 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.194 3.601 sum_up_and_integrate 119 10.3 0.001 0.001 3.454 3.461 integrate_v_rspace 119 11.3 0.003 0.003 3.443 3.450 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.284 3.316 calculate_rho_elec 119 8.7 0.040 0.046 3.283 3.316 init_scf_loop 11 6.9 0.000 0.000 3.243 3.244 acc_transpose_blocks 27432 15.5 0.113 0.116 2.525 2.918 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.149 2.658 apply_single 119 13.6 0.000 0.000 2.149 2.658 calculate_first_density_matrix 1 7.0 0.000 0.000 2.541 2.543 make_m2s 4572 13.5 0.054 0.056 2.346 2.462 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.393 2.420 make_images 4572 14.5 0.273 0.334 2.237 2.352 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.321 2.333 mp_waitall_1 137007 16.6 1.863 2.309 1.863 2.309 calculate_dm_sparse 119 9.5 0.000 0.000 2.162 2.216 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.077 2.078 prepare_preconditioner 11 7.9 0.000 0.000 2.001 2.009 make_preconditioner 11 8.9 0.000 0.000 2.001 2.009 mp_sum_l 7287 12.8 1.135 1.974 1.135 1.974 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.731 1.932 grid_integrate_task_list 119 12.3 1.835 1.922 1.835 1.922 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.895 1.911 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.907 1.907 ot_diis_step 108 11.5 0.012 0.012 1.847 1.847 density_rs2pw 119 9.7 0.003 0.004 1.742 1.842 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.739 1.770 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.617 1.618 cp_fm_redistribute_end 50 14.0 0.809 1.589 0.815 1.591 cp_fm_diag_elpa_base 50 14.0 0.749 1.515 0.774 1.560 acc_transpose_blocks_sync 82296 16.5 1.393 1.558 1.393 1.558 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.422 1.457 fft3d_ps 1201 13.6 0.577 0.640 1.400 1.426 wfi_extrapolate 11 7.9 0.001 0.001 1.357 1.357 grid_collocate_task_list 119 9.7 1.249 1.332 1.249 1.332 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.302 1.321 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.252 1.267 potential_pw2rs 119 12.3 0.008 0.009 1.258 1.263 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.991 1.236 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.792 1.211 jit_kernel_multiply 6 15.9 0.383 1.133 0.383 1.133 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.105 1.106 transfer_rs2pw 487 10.6 0.004 0.005 0.948 1.066 make_images_data 4572 15.5 0.047 0.052 0.865 0.984 jit_kernel_transpose 5 15.6 0.721 0.970 0.721 0.970 mp_sum_d 4137 12.0 0.569 0.958 0.569 0.958 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.777 0.944 mp_alltoall_d11v 2130 13.8 0.824 0.930 0.824 0.930 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.802 0.878 dbcsr_complete_redistribute 329 12.2 0.107 0.122 0.699 0.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.812 transfer_pw2rs 487 13.2 0.003 0.004 0.757 0.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.464000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.909091, yerr=3.396644 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 634.839040E+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 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3629 73907. 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 0.029 0.054 29.627 29.628 qs_mol_dyn_low 1 2.0 0.003 0.003 29.295 29.304 qs_forces 11 3.9 0.002 0.003 28.848 28.848 qs_energies 11 4.9 0.017 0.018 27.149 27.152 scf_env_do_scf 11 5.9 0.000 0.001 21.777 21.777 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.473 19.474 velocity_verlet 10 3.0 0.012 0.012 14.630 14.657 dbcsr_multiply_generic 2286 12.5 0.094 0.098 13.881 13.951 qs_scf_new_mos 108 7.5 0.001 0.001 12.125 12.151 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.124 12.150 multiply_cannon 2286 13.5 0.224 0.234 11.402 11.802 ot_scf_mini 108 9.5 0.002 0.002 11.426 11.450 multiply_cannon_loop 2286 14.5 0.641 0.654 10.526 10.668 multiply_cannon_multrec 9144 15.5 1.967 2.175 6.658 6.868 ot_mini 108 10.5 0.001 0.001 6.816 6.844 rebuild_ks_matrix 119 8.3 0.000 0.000 5.626 5.645 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.626 5.645 qs_ot_get_derivative 108 11.5 0.001 0.001 5.536 5.560 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.000 5.017 dbcsr_mm_accdrv_process 12550 15.8 3.780 4.615 4.580 4.659 init_scf_run 11 5.9 0.000 0.001 3.912 3.912 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.912 3.912 sum_up_and_integrate 119 10.3 0.001 0.001 3.305 3.309 integrate_v_rspace 119 11.3 0.003 0.003 3.295 3.299 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.170 3.177 calculate_rho_elec 119 8.7 0.060 0.061 3.170 3.177 calculate_first_density_matrix 1 7.0 0.000 0.000 2.796 2.797 qs_ot_get_p 119 10.4 0.001 0.001 2.753 2.791 init_scf_loop 11 6.9 0.000 0.000 2.283 2.284 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.043 2.058 calculate_dm_sparse 119 9.5 0.000 0.000 1.911 1.930 grid_integrate_task_list 119 12.3 1.866 1.927 1.866 1.927 make_m2s 4572 13.5 0.034 0.035 1.726 1.851 acc_transpose_blocks 9144 15.5 0.040 0.042 1.798 1.823 mp_waitall_1 115863 16.7 1.252 1.806 1.252 1.806 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.799 1.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.788 1.790 make_images 4572 14.5 0.270 0.302 1.637 1.759 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.729 1.736 density_rs2pw 119 9.7 0.003 0.003 1.577 1.682 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.630 1.637 prepare_preconditioner 11 7.9 0.000 0.000 1.563 1.568 make_preconditioner 11 8.9 0.000 0.000 1.563 1.568 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.560 1.560 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.463 1.488 jit_kernel_multiply 7 15.4 0.761 1.484 0.761 1.484 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.432 1.440 grid_collocate_task_list 119 9.7 1.295 1.385 1.295 1.385 fft3d_ps 1201 13.6 0.636 0.652 1.342 1.350 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.317 1.332 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.310 1.315 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.275 1.283 cp_fm_diag_elpa_base 50 14.0 1.248 1.265 1.274 1.282 ot_diis_step 108 11.5 0.013 0.013 1.269 1.269 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.226 1.260 apply_single 119 13.6 0.000 0.000 1.226 1.260 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.237 1.238 potential_pw2rs 119 12.3 0.010 0.010 1.120 1.123 wfi_extrapolate 11 7.9 0.001 0.001 1.064 1.064 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.026 1.033 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.937 jit_kernel_transpose 5 15.6 0.909 0.915 0.909 0.915 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.736 0.912 mp_alltoall_d11v 2130 13.8 0.766 0.875 0.766 0.875 make_images_data 4572 15.5 0.042 0.047 0.731 0.872 transfer_rs2pw 487 10.6 0.004 0.005 0.775 0.869 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.809 0.816 cp_fm_cholesky_invert 11 10.9 0.766 0.771 0.766 0.771 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.749 acc_transpose_blocks_sync 27432 16.5 0.716 0.737 0.716 0.737 qs_env_update_s_mstruct 11 6.9 0.005 0.007 0.689 0.736 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.333 0.685 transfer_pw2rs 487 13.2 0.003 0.004 0.636 0.637 mp_alltoall_z22v 1201 15.6 0.581 0.614 0.581 0.614 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.628000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=598.727273, yerr=8.781084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 773.701632E+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 3629 74142. 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 0.020 0.047 43.740 43.740 qs_mol_dyn_low 1 2.0 0.003 0.003 43.259 43.276 qs_forces 11 3.9 0.003 0.005 43.178 43.180 qs_energies 11 4.9 0.001 0.001 41.118 41.122 scf_env_do_scf 11 5.9 0.001 0.001 34.072 34.072 scf_env_do_scf_inner_loop 108 6.5 0.012 0.025 27.987 27.989 velocity_verlet 10 3.0 0.002 0.002 23.325 23.333 dbcsr_multiply_generic 2286 12.5 0.104 0.106 20.305 20.449 qs_scf_new_mos 108 7.5 0.001 0.001 18.260 18.367 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.259 18.367 ot_scf_mini 108 9.5 0.002 0.002 17.070 17.181 multiply_cannon 2286 13.5 0.299 0.308 15.906 16.735 multiply_cannon_loop 2286 14.5 0.867 0.895 14.608 15.484 ot_mini 108 10.5 0.001 0.001 10.397 10.522 multiply_cannon_multrec 9144 15.5 3.476 4.841 9.468 9.685 qs_ot_get_derivative 108 11.5 0.001 0.001 8.307 8.417 rebuild_ks_matrix 119 8.3 0.000 0.000 7.360 7.503 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.360 7.502 dbcsr_mm_accdrv_process 12550 15.8 5.037 7.019 5.854 7.233 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.624 6.754 init_scf_loop 11 6.9 0.000 0.000 6.044 6.045 prepare_preconditioner 11 7.9 0.000 0.000 5.032 5.045 make_preconditioner 11 8.9 0.000 0.000 5.032 5.045 init_scf_run 11 5.9 0.000 0.001 4.938 4.938 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.937 4.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.046 4.928 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.190 4.194 calculate_rho_elec 119 8.7 0.118 0.122 4.189 4.193 sum_up_and_integrate 119 10.3 0.001 0.001 3.986 3.992 integrate_v_rspace 119 11.3 0.004 0.004 3.976 3.982 qs_ot_get_p 119 10.4 0.001 0.001 3.661 3.801 mp_waitall_1 94719 16.7 2.608 3.578 2.608 3.578 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.014 3.403 calculate_first_density_matrix 1 7.0 0.000 0.000 3.306 3.308 cp_fm_upper_to_full 72 14.2 2.079 2.969 2.079 2.969 make_m2s 4572 13.5 0.038 0.038 2.691 2.850 calculate_dm_sparse 119 9.5 0.000 0.000 2.774 2.786 make_images 4572 14.5 0.355 0.387 2.571 2.730 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.641 2.648 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.215 2.551 apply_single 119 13.6 0.000 0.000 2.215 2.551 dbcsr_complete_redistribute 329 12.2 0.261 0.269 1.701 2.258 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.174 2.225 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.393 2.221 density_rs2pw 119 9.7 0.003 0.003 2.185 2.218 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.182 2.188 grid_integrate_task_list 119 12.3 2.108 2.169 2.108 2.169 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.143 2.146 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.130 2.130 fft3d_ps 1201 13.6 0.861 0.880 2.093 2.100 ot_diis_step 108 11.5 0.014 0.015 2.061 2.061 mp_sum_l 7287 12.8 1.171 1.982 1.171 1.982 acc_transpose_blocks 9144 15.5 0.043 0.044 1.883 1.946 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.811 1.855 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.316 1.851 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.789 1.793 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.772 1.773 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.761 1.766 grid_collocate_task_list 119 9.7 1.541 1.578 1.541 1.578 wfi_extrapolate 11 7.9 0.001 0.001 1.559 1.559 mp_alltoall_i22 627 13.8 0.975 1.527 0.975 1.527 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.981 1.514 hybrid_alltoall_any 4725 16.4 0.091 0.155 1.168 1.430 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.429 1.429 cp_fm_diag_elpa_base 50 14.0 1.340 1.356 1.427 1.427 potential_pw2rs 119 12.3 0.013 0.014 1.410 1.412 mp_alltoall_d11v 2130 13.8 1.363 1.410 1.363 1.410 make_images_data 4572 15.5 0.046 0.049 1.156 1.404 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.357 1.376 cp_fm_cholesky_invert 11 10.9 1.171 1.178 1.171 1.178 acc_transpose_blocks_sync 27432 16.5 1.111 1.165 1.111 1.165 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.116 1.131 mp_alltoall_z22v 1201 15.6 1.097 1.129 1.097 1.129 jit_kernel_multiply 5 15.4 0.788 1.112 0.788 1.112 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.000 1.028 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.952 0.965 qs_create_task_list 11 7.9 0.000 0.001 0.943 0.954 generate_qs_task_list 11 8.9 0.369 0.389 0.942 0.953 transfer_rs2pw 487 10.6 0.005 0.005 0.831 0.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.740000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.181818, yerr=11.869121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 496.160768E+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 3528 76079. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.055 84.829 84.830 qs_mol_dyn_low 1 2.0 0.005 0.023 84.455 84.473 qs_forces 11 3.9 0.003 0.003 84.309 84.311 qs_energies 11 4.9 0.016 0.042 81.436 81.450 scf_env_do_scf 11 5.9 0.000 0.001 72.133 72.133 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.904 66.904 dbcsr_multiply_generic 2055 12.4 0.108 0.128 53.703 53.903 qs_scf_new_mos 99 7.5 0.000 0.001 50.102 50.250 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.101 50.249 ot_scf_mini 99 9.5 0.002 0.002 47.579 47.707 multiply_cannon 2055 13.4 0.182 0.187 43.919 44.955 multiply_cannon_loop 2055 14.4 1.812 1.880 42.950 43.909 velocity_verlet 10 3.0 0.011 0.062 42.832 42.835 ot_mini 99 10.5 0.001 0.001 28.618 28.726 qs_ot_get_derivative 99 11.5 0.001 0.001 21.770 21.867 multiply_cannon_multrec 49320 15.4 11.105 11.866 17.599 18.373 rebuild_ks_matrix 110 8.3 0.000 0.000 14.379 14.477 qs_ks_build_kohn_sham_matrix 110 9.3 0.010 0.011 14.378 14.477 mp_waitall_1 220248 16.4 12.004 13.236 12.004 13.236 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.611 12.700 qs_ot_get_p 110 10.4 0.001 0.001 10.229 10.350 multiply_cannon_sync_h2d 49320 15.4 9.290 9.873 9.290 9.873 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 7.395 8.828 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.132 8.655 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.188 7.793 apply_single 110 13.6 0.000 0.001 7.188 7.793 init_scf_run 11 5.9 0.000 0.001 7.242 7.242 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.242 7.242 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.954 6.988 sum_up_and_integrate 110 10.3 0.002 0.003 6.846 6.860 integrate_v_rspace 110 11.3 0.003 0.003 6.821 6.840 dbcsr_mm_accdrv_process 87628 16.1 3.175 3.380 6.367 6.790 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.536 6.620 ot_diis_step 99 11.5 0.006 0.006 6.590 6.590 mp_sum_l 6594 12.7 5.424 6.442 5.424 6.442 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.147 6.304 calculate_rho_elec 110 8.6 0.020 0.024 6.146 6.304 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.970 5.970 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.483 5.484 cp_fm_redistribute_end 48 14.0 2.781 5.453 2.793 5.457 cp_fm_diag_elpa_base 48 14.0 2.653 5.328 2.662 5.346 init_scf_loop 11 6.9 0.000 0.000 5.196 5.196 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.057 4.060 calculate_dm_sparse 110 9.5 0.001 0.001 3.958 4.058 make_m2s 4110 13.4 0.061 0.066 3.904 4.023 wfi_extrapolate 11 7.9 0.001 0.001 3.986 3.986 make_images 4110 14.4 0.178 0.190 3.808 3.932 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.786 3.891 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.538 3.590 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.465 3.534 grid_integrate_task_list 110 12.3 3.252 3.396 3.252 3.396 density_rs2pw 110 9.6 0.004 0.004 3.133 3.356 calculate_first_density_matrix 1 7.0 0.000 0.000 3.146 3.158 acc_transpose_blocks 49320 15.4 0.211 0.226 2.766 3.105 prepare_preconditioner 11 7.9 0.000 0.000 3.034 3.055 make_preconditioner 11 8.9 0.000 0.000 3.034 3.055 jit_kernel_multiply 13 15.9 2.909 2.957 2.909 2.957 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.790 2.832 mp_alltoall_d11v 2046 13.8 2.218 2.769 2.218 2.769 fft_wrap_pw1pw2 1111 11.6 0.013 0.018 2.648 2.705 potential_pw2rs 110 12.3 0.006 0.006 2.461 2.477 grid_collocate_task_list 110 9.6 2.156 2.308 2.156 2.308 fft_wrap_pw1pw2_140 451 12.1 0.092 0.098 2.244 2.305 fft3d_ps 1111 13.6 0.733 0.817 2.187 2.242 mp_waitany 14300 13.8 1.812 2.194 1.812 2.194 transfer_rs2pw 451 10.6 0.005 0.006 1.853 2.044 mp_sum_d 3893 11.9 1.451 1.929 1.451 1.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.906 1.922 make_images_data 4110 15.4 0.043 0.047 1.743 1.869 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.515 1.782 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.722 1.752 transfer_pw2rs 451 13.1 0.005 0.007 1.740 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.830000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=471.454545, yerr=2.641062 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 588.996608E+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 3528 76294. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 3488539. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.055 71.415 71.416 qs_mol_dyn_low 1 2.0 0.003 0.003 71.103 71.123 qs_forces 11 3.9 0.017 0.021 70.795 70.797 qs_energies 11 4.9 0.009 0.013 67.450 67.467 scf_env_do_scf 11 5.9 0.000 0.001 56.439 56.440 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.462 49.463 dbcsr_multiply_generic 2055 12.4 0.114 0.117 40.693 40.865 velocity_verlet 10 3.0 0.007 0.021 35.464 35.470 multiply_cannon 2055 13.4 0.225 0.243 33.908 35.153 qs_scf_new_mos 99 7.5 0.001 0.001 33.709 33.850 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.708 33.850 multiply_cannon_loop 2055 14.4 1.157 1.182 32.663 33.515 ot_scf_mini 99 9.5 0.003 0.003 32.038 32.179 ot_mini 99 10.5 0.001 0.001 18.810 18.952 multiply_cannon_multrec 24660 15.4 6.965 8.618 16.215 18.007 rebuild_ks_matrix 110 8.3 0.000 0.001 13.690 13.788 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.690 13.788 qs_ot_get_derivative 99 11.5 0.001 0.001 13.061 13.201 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.063 12.155 mp_waitall_1 176588 16.5 7.986 10.549 7.986 10.549 dbcsr_mm_accdrv_process 52282 16.1 6.574 8.425 9.079 9.458 init_scf_run 11 5.9 0.000 0.001 8.583 8.584 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.583 8.584 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.416 8.301 multiply_cannon_sync_h2d 24660 15.4 6.346 7.451 6.346 7.451 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.448 7.111 apply_single 110 13.6 0.000 0.001 6.448 7.111 init_scf_loop 11 6.9 0.000 0.000 6.939 6.940 qs_ot_get_p 110 10.4 0.001 0.001 6.580 6.703 sum_up_and_integrate 110 10.3 0.001 0.003 6.300 6.312 integrate_v_rspace 110 11.3 0.002 0.003 6.273 6.285 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.051 5.757 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.721 5.731 calculate_rho_elec 110 8.6 0.039 0.047 5.721 5.730 ot_diis_step 99 11.5 0.010 0.010 5.703 5.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.548 5.550 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.208 5.249 calculate_first_density_matrix 1 7.0 0.000 0.000 5.036 5.038 prepare_preconditioner 11 7.9 0.000 0.000 4.818 4.844 make_preconditioner 11 8.9 0.000 0.000 4.818 4.844 make_m2s 4110 13.4 0.057 0.060 4.389 4.832 make_images 4110 14.4 0.409 0.470 4.278 4.714 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.671 4.691 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.451 4.551 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.209 4.210 jit_kernel_multiply 10 16.3 2.135 3.967 2.135 3.967 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.671 3.672 cp_fm_redistribute_end 48 14.0 1.850 3.643 1.856 3.645 cp_fm_diag_elpa_base 48 14.0 1.725 3.501 1.782 3.580 wfi_extrapolate 11 7.9 0.001 0.001 3.451 3.451 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.289 3.353 grid_integrate_task_list 110 12.3 3.162 3.334 3.162 3.334 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.139 3.296 density_rs2pw 110 9.6 0.004 0.004 2.961 3.164 calculate_dm_sparse 110 9.5 0.001 0.001 3.038 3.065 make_images_data 4110 15.4 0.048 0.053 2.416 2.839 fft_wrap_pw1pw2_140 451 12.1 0.110 0.115 2.672 2.816 hybrid_alltoall_any 4261 16.3 0.106 0.460 2.054 2.747 fft3d_ps 1111 13.6 1.079 1.309 2.518 2.662 mp_sum_l 6594 12.7 1.798 2.387 1.798 2.387 acc_transpose_blocks 24660 15.4 0.113 0.116 2.099 2.382 cp_fm_cholesky_invert 11 10.9 2.360 2.375 2.360 2.375 grid_collocate_task_list 110 9.6 2.170 2.345 2.170 2.345 potential_pw2rs 110 12.3 0.008 0.008 2.178 2.189 mp_alltoall_d11v 2046 13.8 1.788 1.990 1.788 1.990 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.958 1.977 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.914 1.928 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.788 1.800 transfer_rs2pw 451 10.6 0.007 0.008 1.464 1.734 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.799 1.668 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.604 1.619 mp_allgather_i34 2055 14.4 0.551 1.598 0.551 1.598 mp_irecv_dv 57340 16.2 0.667 1.543 0.667 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.497 mp_waitany 10164 13.8 1.211 1.428 1.211 1.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.416000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.000000, yerr=6.045284 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 666.738688E+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 3480 77281. 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 0.019 0.038 60.351 60.351 qs_mol_dyn_low 1 2.0 0.003 0.003 59.887 59.897 qs_forces 11 3.9 0.012 0.013 59.656 59.657 qs_energies 11 4.9 0.001 0.002 56.487 56.490 scf_env_do_scf 11 5.9 0.001 0.001 47.023 47.023 scf_env_do_scf_inner_loop 99 6.5 0.004 0.013 39.337 39.338 velocity_verlet 10 3.0 0.090 0.145 31.316 31.318 dbcsr_multiply_generic 2055 12.4 0.108 0.111 29.303 29.517 qs_scf_new_mos 99 7.5 0.001 0.001 24.988 25.098 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.988 25.097 multiply_cannon 2055 13.4 0.232 0.248 23.180 24.303 ot_scf_mini 99 9.5 0.002 0.003 23.741 23.860 multiply_cannon_loop 2055 14.4 0.817 0.843 21.993 23.027 ot_mini 99 10.5 0.001 0.001 13.670 13.790 rebuild_ks_matrix 110 8.3 0.000 0.000 12.313 12.428 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.313 12.427 multiply_cannon_multrec 16440 15.4 3.710 5.075 10.520 11.785 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.870 10.977 mp_waitall_1 139946 16.5 6.765 9.566 6.765 9.566 qs_ot_get_derivative 99 11.5 0.001 0.001 9.280 9.400 init_scf_loop 11 6.9 0.000 0.000 7.650 7.651 init_scf_run 11 5.9 0.000 0.001 7.099 7.099 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.099 7.099 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.288 6.907 dbcsr_mm_accdrv_process 34862 16.1 5.943 6.340 6.654 6.834 sum_up_and_integrate 110 10.3 0.001 0.002 6.146 6.163 integrate_v_rspace 110 11.3 0.003 0.003 6.120 6.137 prepare_preconditioner 11 7.9 0.000 0.000 5.652 5.667 make_preconditioner 11 8.9 0.000 0.000 5.652 5.667 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.470 5.477 calculate_rho_elec 110 8.6 0.058 0.058 5.470 5.477 qs_ot_get_p 110 10.4 0.001 0.001 5.269 5.404 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.120 5.359 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.842 5.330 apply_single 110 13.6 0.000 0.000 4.841 5.329 make_m2s 4110 13.4 0.049 0.051 4.130 4.523 make_images 4110 14.4 0.400 0.525 4.014 4.406 ot_diis_step 99 11.5 0.010 0.011 4.362 4.362 calculate_first_density_matrix 1 7.0 0.000 0.000 4.144 4.145 multiply_cannon_sync_h2d 16440 15.4 3.265 3.991 3.265 3.991 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.799 3.800 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.227 3.785 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.648 3.652 grid_integrate_task_list 110 12.3 3.197 3.371 3.197 3.371 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.281 3.282 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.054 3.064 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.973 3.013 density_rs2pw 110 9.6 0.004 0.004 2.704 2.886 wfi_extrapolate 11 7.9 0.001 0.001 2.862 2.862 make_images_data 4110 15.4 0.045 0.050 2.278 2.778 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.734 2.746 cp_fm_diag_elpa_base 48 14.0 2.693 2.710 2.732 2.744 calculate_dm_sparse 110 9.5 0.001 0.001 2.674 2.709 hybrid_alltoall_any 4261 16.3 0.109 0.377 2.052 2.652 fft_wrap_pw1pw2_140 451 12.1 0.127 0.131 2.630 2.644 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.424 2.487 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.872 2.417 fft3d_ps 1111 13.6 1.089 1.099 2.365 2.380 grid_collocate_task_list 110 9.6 2.221 2.365 2.221 2.365 cp_fm_cholesky_invert 11 10.9 2.339 2.351 2.339 2.351 mp_irecv_dv 48980 15.7 0.796 2.278 0.796 2.278 dbcsr_complete_redistribute 325 12.2 0.684 0.836 1.854 2.163 mp_alltoall_d11v 2046 13.8 1.854 2.067 1.854 2.067 mp_sum_l 6594 12.7 1.366 2.031 1.366 2.031 potential_pw2rs 110 12.3 0.010 0.011 1.981 1.986 qs_energies_init_hamiltonians 11 5.9 0.007 0.011 1.951 1.952 acc_transpose_blocks 16440 15.4 0.074 0.077 1.631 1.694 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.359 1.657 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.657 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.616 1.630 mp_allgather_i34 2055 14.4 0.479 1.465 0.479 1.465 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.454 transfer_rs2pw 451 10.6 0.005 0.006 1.235 1.438 cp_fm_cholesky_decompose 22 10.9 1.328 1.353 1.328 1.353 mp_waitany 17072 13.8 1.106 1.331 1.106 1.331 make_basis_sm 11 9.8 0.000 0.001 1.320 1.323 jit_kernel_multiply 6 16.8 0.298 1.321 0.298 1.321 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.269 1.280 parallel_gemm_fm 81 9.0 0.000 0.000 1.241 1.242 parallel_gemm_fm_cosma 81 10.0 1.241 1.242 1.241 1.242 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.351000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=628.818182, yerr=8.705465 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 743.829504E+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 3480 77279. 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 0.029 0.055 65.767 65.768 qs_mol_dyn_low 1 2.0 0.003 0.003 65.223 65.240 qs_forces 11 3.9 0.015 0.052 65.141 65.141 qs_energies 11 4.9 0.001 0.001 61.739 61.754 scf_env_do_scf 11 5.9 0.000 0.001 51.700 51.705 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.251 42.251 velocity_verlet 10 3.0 0.008 0.019 34.738 34.740 dbcsr_multiply_generic 2055 12.4 0.115 0.121 32.075 32.312 qs_scf_new_mos 99 7.5 0.001 0.001 27.901 28.002 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.901 28.002 ot_scf_mini 99 9.5 0.002 0.003 26.219 26.315 multiply_cannon 2055 13.4 0.238 0.259 24.807 26.023 multiply_cannon_loop 2055 14.4 1.389 1.494 23.400 24.122 ot_mini 99 10.5 0.001 0.001 15.441 15.555 multiply_cannon_multrec 24660 15.4 4.229 7.228 13.986 15.205 rebuild_ks_matrix 110 8.3 0.000 0.000 11.998 12.111 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.997 12.111 qs_ot_get_derivative 99 11.5 0.001 0.001 11.272 11.366 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.601 10.704 dbcsr_mm_accdrv_process 52304 16.0 8.348 10.070 9.598 10.653 init_scf_loop 11 6.9 0.000 0.000 9.408 9.409 prepare_preconditioner 11 7.9 0.000 0.000 7.401 7.418 make_preconditioner 11 8.9 0.000 0.000 7.401 7.418 init_scf_run 11 5.9 0.000 0.001 7.290 7.291 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.290 7.291 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.275 7.080 sum_up_and_integrate 110 10.3 0.001 0.002 6.100 6.111 integrate_v_rspace 110 11.3 0.003 0.003 6.074 6.086 qs_ot_get_p 110 10.4 0.001 0.001 5.851 5.998 mp_waitall_1 121746 16.5 4.296 5.955 4.296 5.955 make_m2s 4110 13.4 0.060 0.063 5.550 5.892 make_images 4110 14.4 0.577 0.694 5.408 5.745 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.567 5.579 calculate_rho_elec 110 8.6 0.077 0.081 5.567 5.579 calculate_first_density_matrix 1 7.0 0.000 0.000 4.265 4.267 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.082 4.180 apply_single 110 13.6 0.000 0.000 4.081 4.180 ot_diis_step 99 11.5 0.011 0.013 4.127 4.127 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.772 3.791 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.635 3.712 grid_integrate_task_list 110 12.3 3.291 3.415 3.291 3.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.342 3.396 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.305 3.307 calculate_dm_sparse 110 9.5 0.001 0.001 3.230 3.268 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.197 3.223 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.209 3.209 make_images_data 4110 15.4 0.048 0.052 2.777 3.125 hybrid_alltoall_any 4261 16.3 0.122 0.461 2.292 3.086 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.436 2.995 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.904 2.942 wfi_extrapolate 11 7.9 0.001 0.001 2.929 2.929 cp_fm_upper_to_full 70 14.2 2.056 2.901 2.056 2.901 acc_transpose_blocks 24660 15.4 0.106 0.110 2.699 2.899 density_rs2pw 110 9.6 0.004 0.004 2.690 2.856 fft_wrap_pw1pw2_140 451 12.1 0.146 0.149 2.729 2.757 dbcsr_complete_redistribute 325 12.2 0.396 0.457 2.068 2.748 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.672 2.682 cp_fm_diag_elpa_base 48 14.0 2.607 2.630 2.669 2.680 multiply_cannon_sync_h2d 24660 15.4 2.380 2.593 2.380 2.593 fft3d_ps 1111 13.6 1.126 1.167 2.465 2.486 grid_collocate_task_list 110 9.6 2.264 2.414 2.264 2.414 cp_fm_cholesky_invert 11 10.9 2.383 2.393 2.383 2.393 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.267 2.279 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.526 2.198 mp_alltoall_d11v 2046 13.8 1.813 1.963 1.813 1.963 make_basis_sm 11 9.8 0.052 0.185 1.907 1.908 potential_pw2rs 110 12.3 0.012 0.012 1.903 1.907 jit_kernel_multiply 8 15.7 0.903 1.859 0.903 1.859 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.788 1.821 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.796 1.814 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.118 1.771 mp_alltoall_i22 605 13.7 1.024 1.732 1.024 1.732 mp_sum_l 6594 12.7 1.034 1.725 1.034 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.710 mp_allgather_i34 2055 14.4 0.497 1.670 0.497 1.670 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.642 1.654 parallel_gemm_fm 81 9.0 0.000 0.000 1.581 1.590 parallel_gemm_fm_cosma 81 10.0 1.581 1.590 1.581 1.590 cp_fm_cholesky_decompose 22 10.9 1.476 1.581 1.476 1.581 acc_transpose_blocks_sync 73980 16.4 1.384 1.569 1.384 1.569 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.856 1.510 mp_irecv_dv 62702 16.1 0.752 1.427 0.752 1.427 transfer_rs2pw 451 10.6 0.005 0.006 1.188 1.384 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.211 1.318 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.768000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=702.090909, yerr=11.098358 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 856.797184E+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 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3480 77277. 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 0.031 0.046 56.536 56.536 qs_mol_dyn_low 1 2.0 0.003 0.003 56.027 56.044 qs_forces 11 3.9 0.006 0.008 55.894 55.895 qs_energies 11 4.9 0.023 0.025 52.248 52.253 scf_env_do_scf 11 5.9 0.000 0.001 41.837 41.837 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.729 34.730 velocity_verlet 10 3.0 0.017 0.017 30.186 30.193 dbcsr_multiply_generic 2055 12.4 0.106 0.108 24.824 24.976 qs_scf_new_mos 99 7.5 0.001 0.001 20.660 20.707 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.659 20.706 multiply_cannon 2055 13.4 0.265 0.298 19.356 20.479 ot_scf_mini 99 9.5 0.002 0.002 19.409 19.445 multiply_cannon_loop 2055 14.4 0.604 0.623 18.106 18.446 rebuild_ks_matrix 110 8.3 0.000 0.000 11.441 11.491 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 11.441 11.491 ot_mini 99 10.5 0.001 0.001 10.942 10.973 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.174 10.219 multiply_cannon_multrec 8220 15.4 3.381 4.894 8.431 9.878 mp_waitall_1 103326 16.6 5.955 7.773 5.955 7.773 qs_ot_get_derivative 99 11.5 0.001 0.001 7.173 7.209 init_scf_loop 11 6.9 0.000 0.000 7.061 7.063 init_scf_run 11 5.9 0.000 0.001 6.762 6.762 scf_env_initial_rho_setup 11 6.9 0.050 0.051 6.762 6.762 sum_up_and_integrate 110 10.3 0.001 0.002 6.019 6.034 integrate_v_rspace 110 11.3 0.003 0.003 5.993 6.007 dbcsr_mm_accdrv_process 17442 15.9 3.132 4.426 4.909 5.934 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.624 5.641 calculate_rho_elec 110 8.6 0.115 0.116 5.624 5.640 prepare_preconditioner 11 7.9 0.000 0.000 5.429 5.433 make_preconditioner 11 8.9 0.000 0.000 5.429 5.433 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.029 5.100 qs_ot_get_p 110 10.4 0.001 0.001 4.551 4.592 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.065 4.499 make_m2s 4110 13.4 0.038 0.039 4.110 4.305 make_images 4110 14.4 0.651 0.716 3.980 4.170 calculate_first_density_matrix 1 7.0 0.000 0.000 4.090 4.092 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.872 3.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.801 3.860 apply_single 110 13.6 0.000 0.000 3.800 3.860 ot_diis_step 99 11.5 0.012 0.012 3.748 3.748 grid_integrate_task_list 110 12.3 3.379 3.540 3.379 3.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.440 3.459 fft_wrap_pw1pw2 1111 11.6 0.013 0.015 3.308 3.315 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.212 3.216 qs_energies_init_hamiltonians 11 5.9 0.016 0.016 3.093 3.094 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.878 2.879 fft_wrap_pw1pw2_140 451 12.1 0.192 0.195 2.856 2.864 density_rs2pw 110 9.6 0.004 0.004 2.645 2.798 calculate_dm_sparse 110 9.5 0.001 0.001 2.628 2.664 cp_fm_cholesky_invert 11 10.9 2.560 2.567 2.560 2.567 wfi_extrapolate 11 7.9 0.001 0.001 2.528 2.528 hybrid_alltoall_any 4261 16.3 0.201 0.849 2.138 2.518 fft3d_ps 1111 13.6 1.271 1.283 2.468 2.483 grid_collocate_task_list 110 9.6 2.366 2.480 2.366 2.480 multiply_cannon_sync_h2d 8220 15.4 2.359 2.465 2.359 2.465 make_images_data 4110 15.4 0.041 0.047 2.170 2.430 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.368 2.376 cp_fm_diag_elpa_base 48 14.0 2.313 2.338 2.366 2.374 acc_transpose_blocks 8220 15.4 0.038 0.039 2.043 2.126 jit_kernel_multiply 8 15.8 1.455 2.109 1.455 2.109 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.074 2.105 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.797 2.015 qs_env_update_s_mstruct 11 6.9 0.001 0.005 1.874 1.986 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.871 1.887 mp_alltoall_d11v 2046 13.8 1.657 1.841 1.657 1.841 potential_pw2rs 110 12.3 0.015 0.015 1.824 1.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.597 1.605 dbcsr_complete_redistribute 325 12.2 0.531 0.586 1.503 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.582 1.593 cp_fm_cholesky_decompose 22 10.9 1.466 1.493 1.466 1.493 mp_allgather_i34 2055 14.4 0.394 1.419 0.394 1.419 acc_transpose_blocks_kernels 8220 16.4 0.113 0.115 1.325 1.415 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.909 1.309 qs_create_task_list 11 7.9 0.001 0.001 1.208 1.308 generate_qs_task_list 11 8.9 0.375 0.443 1.208 1.308 jit_kernel_transpose 5 15.6 1.212 1.302 1.212 1.302 transfer_rs2pw 451 10.6 0.005 0.006 1.106 1.286 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.481 1.274 mp_waitany 9240 13.8 1.054 1.239 1.054 1.239 mp_irecv_dv 24056 15.7 0.454 1.232 0.454 1.232 copy_dbcsr_to_fm 151 11.3 0.002 0.003 1.163 1.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.536000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=809.454545, yerr=10.773399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 1.354101E+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 3469 77990. 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 0.020 0.040 80.927 80.927 qs_mol_dyn_low 1 2.0 0.003 0.003 80.501 80.519 qs_forces 11 3.9 0.011 0.013 80.371 80.372 qs_energies 11 4.9 0.002 0.003 76.255 76.260 scf_env_do_scf 11 5.9 0.000 0.001 63.953 63.953 velocity_verlet 10 3.0 0.044 0.064 47.965 47.979 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.556 43.557 dbcsr_multiply_generic 2055 12.4 0.121 0.125 32.116 32.154 qs_scf_new_mos 99 7.5 0.001 0.001 26.655 26.687 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.655 26.687 multiply_cannon 2055 13.4 0.334 0.382 24.830 25.569 ot_scf_mini 99 9.5 0.002 0.002 24.853 24.884 multiply_cannon_loop 2055 14.4 0.821 0.835 23.016 23.546 init_scf_loop 11 6.9 0.000 0.000 20.326 20.327 prepare_preconditioner 11 7.9 0.000 0.000 18.291 18.296 make_preconditioner 11 8.9 0.000 0.000 18.291 18.296 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.476 17.789 ot_mini 99 10.5 0.001 0.001 14.301 14.327 rebuild_ks_matrix 110 8.3 0.000 0.001 13.492 13.533 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.491 13.532 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.240 12.276 multiply_cannon_multrec 8220 15.4 4.096 4.254 11.803 11.983 cp_fm_upper_to_full 70 14.2 8.206 11.619 8.206 11.619 qs_ot_get_derivative 99 11.5 0.001 0.001 9.625 9.657 mp_waitall_1 84994 16.7 7.823 8.612 7.823 8.612 init_scf_run 11 5.9 0.000 0.001 7.933 7.933 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.933 7.933 dbcsr_mm_accdrv_process 11614 15.7 3.977 4.253 7.559 7.810 dbcsr_complete_redistribute 325 12.2 0.935 0.948 5.332 7.343 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.808 6.844 calculate_rho_elec 110 8.6 0.227 0.227 6.808 6.844 sum_up_and_integrate 110 10.3 0.002 0.002 6.567 6.581 integrate_v_rspace 110 11.3 0.004 0.004 6.539 6.554 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.339 6.346 make_m2s 4110 13.4 0.043 0.044 5.475 5.987 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.798 5.809 make_images 4110 14.4 0.879 0.922 5.286 5.797 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.967 5.597 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.891 5.473 apply_single 110 13.6 0.000 0.000 4.891 5.473 mp_alltoall_i22 605 13.7 3.412 5.408 3.412 5.408 qs_ot_get_p 110 10.4 0.001 0.001 5.219 5.247 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.977 4.979 ot_diis_step 99 11.5 0.015 0.016 4.652 4.652 calculate_first_density_matrix 1 7.0 0.000 0.000 4.624 4.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.485 4.499 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.450 4.471 fft_wrap_pw1pw2_140 451 12.1 0.363 0.365 3.886 3.907 cp_fm_cholesky_invert 11 10.9 3.825 3.834 3.825 3.834 qs_energies_init_hamiltonians 11 5.9 0.007 0.049 3.788 3.791 grid_integrate_task_list 110 12.3 3.669 3.735 3.669 3.735 hybrid_alltoall_any 4261 16.3 0.262 0.564 2.943 3.733 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.134 3.659 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.644 3.651 make_images_data 4110 15.4 0.044 0.047 2.982 3.617 jit_kernel_multiply 8 15.0 3.383 3.485 3.383 3.485 fft3d_ps 1111 13.6 1.827 1.853 3.310 3.329 calculate_dm_sparse 110 9.5 0.001 0.001 3.300 3.326 density_rs2pw 110 9.6 0.004 0.004 3.290 3.311 wfi_extrapolate 11 7.9 0.001 0.001 3.198 3.198 multiply_cannon_sync_h2d 8220 15.4 3.131 3.147 3.131 3.147 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.126 3.127 grid_collocate_task_list 110 9.6 2.660 2.669 2.660 2.669 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.624 2.639 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.544 2.544 cp_fm_diag_elpa_base 48 14.0 2.248 2.318 2.542 2.543 qs_env_update_s_mstruct 11 6.9 0.002 0.006 2.241 2.316 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.106 2.193 potential_pw2rs 110 12.3 0.021 0.021 2.119 2.124 mp_alltoall_d11v 2046 13.8 2.015 2.072 2.015 2.072 acc_transpose_blocks 8220 15.4 0.039 0.040 1.896 2.009 qs_create_task_list 11 7.9 0.000 0.001 1.882 1.926 generate_qs_task_list 11 8.9 0.733 0.788 1.881 1.926 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.854 1.858 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.798 1.814 cp_fm_cholesky_decompose 22 10.9 1.734 1.748 1.734 1.748 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.693 1.708 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.623 1.646 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=80.927000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.909091, yerr=44.813664 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/15/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 1.094965E+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 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 629.587968E+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 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4009 67203. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2490150. 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 0.043 0.142 214.106 214.107 qs_mol_dyn_low 1 2.0 0.018 0.072 211.144 213.186 qs_forces 11 3.9 0.005 0.005 210.514 210.520 qs_energies 11 4.9 0.002 0.003 204.546 204.566 scf_env_do_scf 11 5.9 0.001 0.001 186.768 186.785 scf_env_do_scf_inner_loop 117 6.6 0.004 0.012 167.235 167.237 dbcsr_multiply_generic 2507 12.6 0.178 0.185 128.016 128.740 qs_scf_new_mos 117 7.6 0.001 0.001 127.161 127.429 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.160 127.429 velocity_verlet 10 3.0 0.117 0.455 124.735 124.747 ot_scf_mini 117 9.6 0.003 0.003 120.462 120.745 multiply_cannon 2507 13.6 0.242 0.259 103.414 105.795 multiply_cannon_loop 2507 14.6 2.379 2.435 101.116 103.382 ot_mini 117 10.6 0.001 0.001 67.334 67.595 multiply_cannon_multrec 60168 15.6 31.987 34.703 42.198 44.244 qs_ot_get_derivative 117 11.6 0.001 0.001 42.309 42.568 rebuild_ks_matrix 128 8.3 0.004 0.015 34.612 34.961 qs_ks_build_kohn_sham_matrix 128 9.3 0.074 0.268 34.608 34.961 mp_waitall_1 267128 16.5 29.708 32.858 29.708 32.858 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.940 31.249 qs_ot_get_p 128 10.4 0.001 0.001 30.747 31.096 multiply_cannon_sync_h2d 60168 15.6 26.432 28.613 26.432 28.613 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.504 25.470 apply_single 128 13.6 0.001 0.001 24.503 25.469 ot_diis_step 117 11.6 0.008 0.008 24.758 24.760 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.893 23.975 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.951 20.952 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.098 20.286 init_scf_loop 11 6.9 0.004 0.017 19.444 19.445 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 16.343 18.765 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.738 17.778 cp_fm_diag_elpa_base 83 14.4 17.679 17.714 17.730 17.768 make_m2s 5014 13.6 0.106 0.114 14.475 14.888 make_images 5014 14.6 0.402 0.422 14.289 14.707 prepare_preconditioner 11 7.9 0.000 0.000 14.578 14.620 make_preconditioner 11 8.9 0.000 0.000 14.578 14.620 sum_up_and_integrate 128 10.3 0.002 0.004 14.246 14.276 integrate_v_rspace 128 11.3 0.004 0.004 14.187 14.219 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.800 13.967 init_scf_run 11 5.9 0.000 0.001 13.426 13.427 scf_env_initial_rho_setup 11 6.9 0.003 0.017 13.426 13.427 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.217 13.385 calculate_rho_elec 128 8.7 0.046 0.068 13.216 13.384 mp_sum_l 7950 12.9 8.959 10.801 8.959 10.801 dbcsr_mm_accdrv_process 124484 16.2 4.692 4.864 9.774 10.292 calculate_dm_sparse 128 9.5 0.001 0.001 9.165 9.286 wfi_extrapolate 11 7.9 0.001 0.001 9.088 9.089 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.557 8.729 multiply_cannon_metrocomm1 60168 15.6 0.090 0.095 6.384 8.317 cp_fm_cholesky_invert 11 10.9 8.256 8.275 8.256 8.275 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.141 8.218 make_images_data 5014 15.6 0.066 0.072 7.041 7.984 density_rs2pw 128 9.7 0.005 0.006 6.772 7.624 grid_integrate_task_list 128 12.3 7.027 7.569 7.027 7.569 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.991 7.322 hybrid_alltoall_any 5200 16.5 0.294 2.283 6.161 7.309 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.174 7.186 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.048 6.157 fft_wrap_pw1pw2_140 523 12.2 0.231 0.243 5.893 6.117 fft3d_ps 1291 13.7 2.089 2.581 5.821 6.093 mp_alltoall_d11v 2415 14.1 4.864 6.006 4.864 6.006 grid_collocate_task_list 128 9.7 4.844 5.255 4.844 5.255 potential_pw2rs 128 12.3 0.009 0.010 4.675 4.734 acc_transpose_blocks 60168 15.6 0.277 0.304 4.174 4.473 mp_sum_d 4474 12.1 3.726 4.341 3.726 4.341 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=214.107000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.636364, yerr=6.314093 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 858.636288E+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 3999 67821. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5242228. 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 0.018 0.056 188.479 188.480 qs_mol_dyn_low 1 2.0 0.003 0.003 187.904 187.919 qs_forces 11 3.9 0.008 0.026 187.793 187.795 qs_energies 11 4.9 0.001 0.002 180.949 180.965 scf_env_do_scf 11 5.9 0.001 0.001 163.182 163.198 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 136.505 136.508 velocity_verlet 10 3.0 0.015 0.028 115.119 115.120 dbcsr_multiply_generic 2507 12.6 0.187 0.192 99.454 100.599 qs_scf_new_mos 117 7.6 0.001 0.001 97.426 97.930 qs_scf_loop_do_ot 117 8.6 0.001 0.001 97.425 97.929 ot_scf_mini 117 9.6 0.004 0.004 92.597 93.153 multiply_cannon 2507 13.6 0.505 0.559 78.767 83.822 multiply_cannon_loop 2507 14.6 1.561 1.623 75.253 77.714 ot_mini 117 10.6 0.001 0.001 50.913 51.445 mp_waitall_1 214728 16.6 24.397 39.005 24.397 39.005 multiply_cannon_multrec 30084 15.6 21.042 25.172 32.141 36.951 rebuild_ks_matrix 128 8.3 0.001 0.001 33.005 33.710 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 33.004 33.709 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.722 30.378 qs_ot_get_derivative 117 11.6 0.001 0.002 28.989 29.554 multiply_cannon_metrocomm3 30084 15.6 0.095 0.101 15.546 28.541 init_scf_loop 11 6.9 0.000 0.000 26.573 26.575 qs_ot_get_p 128 10.4 0.001 0.001 23.535 24.157 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.073 23.236 apply_single 128 13.6 0.001 0.001 22.072 23.236 prepare_preconditioner 11 7.9 0.000 0.000 22.167 22.241 make_preconditioner 11 8.9 0.000 0.000 22.167 22.240 ot_diis_step 117 11.6 0.014 0.015 21.743 21.746 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.042 21.491 multiply_cannon_sync_h2d 30084 15.6 18.184 20.438 18.184 20.438 qs_ot_p2m_diag 83 11.4 0.188 0.217 18.418 18.454 cp_dbcsr_syevd 83 12.4 0.006 0.006 17.174 17.176 make_m2s 5014 13.6 0.087 0.092 14.643 16.374 make_images 5014 14.6 1.173 1.379 14.427 16.155 sum_up_and_integrate 128 10.3 0.002 0.004 14.119 14.146 integrate_v_rspace 128 11.3 0.003 0.004 14.059 14.087 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.572 13.606 cp_fm_diag_elpa_base 83 14.4 13.365 13.427 13.560 13.591 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.237 13.261 calculate_rho_elec 128 8.7 0.088 0.106 13.237 13.260 init_scf_run 11 5.9 0.000 0.001 12.639 12.640 scf_env_initial_rho_setup 11 6.9 0.012 0.020 12.638 12.640 cp_fm_cholesky_invert 11 10.9 12.233 12.272 12.233 12.272 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.437 11.835 dbcsr_mm_accdrv_process 62242 16.2 5.653 6.860 10.537 11.235 multiply_cannon_metrocomm4 27577 15.6 0.107 0.118 3.811 10.644 make_images_data 5014 15.6 0.065 0.076 8.526 10.626 mp_irecv_dv 69486 16.3 3.600 10.232 3.600 10.232 hybrid_alltoall_any 5200 16.5 0.350 1.534 7.270 10.042 wfi_extrapolate 11 7.9 0.001 0.001 8.376 8.376 fft_wrap_pw1pw2 1291 11.7 0.019 0.025 7.739 7.823 grid_integrate_task_list 128 12.3 7.177 7.611 7.177 7.611 density_rs2pw 128 9.7 0.006 0.006 6.883 7.227 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.181 7.192 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.357 7.136 calculate_dm_sparse 128 9.5 0.001 0.001 6.603 6.761 fft_wrap_pw1pw2_140 523 12.2 0.248 0.257 6.644 6.717 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.133 6.314 fft3d_ps 1291 13.7 2.762 2.940 6.198 6.257 mp_sum_l 7950 12.9 4.258 6.166 4.258 6.166 cp_fm_cholesky_decompose 22 10.9 5.729 5.798 5.729 5.798 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.423 5.501 grid_collocate_task_list 128 9.7 5.004 5.364 5.004 5.364 mp_alltoall_d11v 2415 14.1 4.672 5.240 4.672 5.240 mp_allgather_i34 2507 14.6 2.039 5.237 2.039 5.237 potential_pw2rs 128 12.3 0.015 0.017 4.546 4.566 dbcsr_complete_redistribute 395 12.7 0.739 0.896 3.531 4.226 calculate_first_density_matrix 1 7.0 0.000 0.000 4.057 4.074 mp_sum_d 4469 12.1 2.632 3.951 2.632 3.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.480000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=814.272727, yerr=5.847935 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.031176E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4008 67664. MP_Allreduce 11080 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.099 178.888 178.889 qs_mol_dyn_low 1 2.0 0.003 0.003 177.582 177.599 qs_forces 11 3.9 0.004 0.004 176.473 176.474 qs_energies 11 4.9 0.002 0.002 169.784 169.797 scf_env_do_scf 11 5.9 0.001 0.001 153.410 153.410 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 123.310 123.311 velocity_verlet 10 3.0 0.008 0.010 110.870 110.904 qs_scf_new_mos 116 7.6 0.001 0.001 85.950 86.262 qs_scf_loop_do_ot 116 8.6 0.001 0.001 85.949 86.261 dbcsr_multiply_generic 2485 12.5 0.182 0.187 83.164 84.287 ot_scf_mini 116 9.6 0.003 0.004 81.731 82.110 multiply_cannon 2485 13.5 0.500 0.521 62.989 68.222 multiply_cannon_loop 2485 14.5 1.119 1.179 60.033 63.001 ot_mini 116 10.6 0.001 0.001 43.393 43.745 mp_waitall_1 169034 16.6 25.828 35.086 25.828 35.086 rebuild_ks_matrix 127 8.3 0.001 0.001 30.733 31.185 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.016 30.732 31.184 init_scf_loop 11 6.9 0.000 0.000 29.970 29.972 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.693 28.103 multiply_cannon_multrec 19880 15.5 12.778 16.128 22.728 26.081 prepare_preconditioner 11 7.9 0.000 0.000 25.898 25.948 make_preconditioner 11 8.9 0.000 0.000 25.897 25.948 multiply_cannon_metrocomm3 19880 15.5 0.059 0.063 16.012 25.203 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.869 24.841 qs_ot_get_derivative 116 11.6 0.001 0.002 23.767 24.139 qs_ot_get_p 127 10.4 0.001 0.001 23.601 24.076 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.725 20.671 apply_single 127 13.6 0.001 0.001 19.725 20.670 ot_diis_step 116 11.6 0.017 0.018 19.509 19.510 qs_ot_p2m_diag 82 11.4 0.262 0.269 19.067 19.089 cp_dbcsr_syevd 82 12.4 0.005 0.005 17.962 17.963 make_m2s 4970 13.5 0.078 0.082 14.953 15.969 make_images 4970 14.5 1.138 1.217 14.722 15.735 multiply_cannon_sync_h2d 19880 15.5 13.373 15.644 13.373 15.644 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.180 14.202 cp_fm_diag_elpa_base 82 14.4 13.906 13.994 14.175 14.197 sum_up_and_integrate 127 10.3 0.002 0.003 14.079 14.107 integrate_v_rspace 127 11.3 0.003 0.004 14.020 14.044 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.469 13.501 calculate_rho_elec 127 8.7 0.131 0.145 13.468 13.501 cp_fm_cholesky_invert 11 10.9 12.686 12.709 12.686 12.709 init_scf_run 11 5.9 0.000 0.001 11.275 11.275 scf_env_initial_rho_setup 11 6.9 0.008 0.027 11.275 11.275 make_images_data 4970 15.5 0.061 0.070 8.929 10.352 hybrid_alltoall_any 5155 16.4 0.449 2.037 7.674 9.769 multiply_cannon_metrocomm4 17395 15.5 0.066 0.078 3.523 9.617 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.311 9.589 dbcsr_mm_accdrv_process 41158 16.2 6.278 6.603 9.425 9.553 mp_irecv_dv 49801 16.2 3.392 9.357 3.392 9.357 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 7.924 8.040 wfi_extrapolate 11 7.9 0.001 0.001 7.691 7.691 grid_integrate_task_list 127 12.3 7.236 7.685 7.236 7.685 density_rs2pw 127 9.7 0.005 0.006 6.937 7.403 fft_wrap_pw1pw2_140 519 12.2 0.281 0.289 6.836 6.953 fft3d_ps 1281 13.7 2.744 2.963 6.246 6.333 cp_fm_cholesky_decompose 22 10.9 6.177 6.238 6.177 6.238 dbcsr_complete_redistribute 393 12.7 1.054 1.134 4.910 6.118 calculate_dm_sparse 127 9.5 0.001 0.001 5.981 6.055 mp_alltoall_d11v 2401 14.1 5.071 6.032 5.071 6.032 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.914 5.918 grid_collocate_task_list 127 9.7 5.105 5.496 5.105 5.496 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.724 5.379 cp_fm_upper_to_full 104 14.8 3.973 5.016 3.973 5.016 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.840 4.971 mp_sum_l 7884 12.9 3.417 4.778 3.417 4.778 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.496 4.690 mp_allgather_i34 2485 14.5 1.475 4.633 1.475 4.633 potential_pw2rs 127 12.3 0.019 0.021 4.507 4.526 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.046 4.086 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 3.749 3.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.549 3.593 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.889000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=980.363636, yerr=4.597879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 4.353792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.096188E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379572320 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4027 67348. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.134 189.811 189.812 qs_mol_dyn_low 1 2.0 0.009 0.053 188.579 188.622 qs_forces 11 3.9 0.019 0.021 188.403 188.410 qs_energies 11 4.9 0.002 0.002 181.164 181.175 scf_env_do_scf 11 5.9 0.001 0.001 163.499 163.521 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 124.682 124.683 velocity_verlet 10 3.0 0.010 0.022 121.687 121.696 qs_scf_new_mos 117 7.6 0.001 0.001 87.800 88.161 qs_scf_loop_do_ot 117 8.6 0.001 0.001 87.799 88.161 ot_scf_mini 117 9.6 0.003 0.003 83.189 83.510 dbcsr_multiply_generic 2507 12.6 0.190 0.195 82.342 82.995 multiply_cannon 2507 13.6 0.540 0.570 56.339 59.579 multiply_cannon_loop 2507 14.6 1.802 1.924 52.600 54.192 ot_mini 117 10.6 0.001 0.001 44.206 44.539 init_scf_loop 11 6.9 0.000 0.000 38.674 38.675 prepare_preconditioner 11 7.9 0.000 0.000 34.487 34.520 make_preconditioner 11 8.9 0.000 0.000 34.487 34.520 make_full_inverse_cholesky 11 9.9 0.018 0.023 29.647 33.248 multiply_cannon_multrec 30084 15.6 13.446 18.466 26.558 31.242 rebuild_ks_matrix 128 8.3 0.001 0.001 29.978 30.226 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.978 30.226 mp_waitall_1 147882 16.7 18.256 27.994 18.256 27.994 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.050 27.283 qs_ot_get_p 128 10.4 0.001 0.001 24.865 25.268 qs_ot_get_derivative 117 11.6 0.001 0.002 24.218 24.551 make_m2s 5014 13.6 0.094 0.098 21.242 22.117 make_images 5014 14.6 2.004 2.271 20.929 21.797 qs_ot_p2m_diag 83 11.4 0.343 0.391 20.280 20.335 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.418 20.060 apply_single 128 13.6 0.001 0.001 19.418 20.059 ot_diis_step 117 11.6 0.018 0.020 19.856 19.858 cp_dbcsr_syevd 83 12.4 0.005 0.006 18.794 18.796 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.695 15.255 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.037 15.055 cp_fm_diag_elpa_base 83 14.4 14.480 14.644 15.029 15.041 sum_up_and_integrate 128 10.3 0.002 0.003 14.173 14.199 integrate_v_rspace 128 11.3 0.003 0.004 14.113 14.142 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.804 13.857 calculate_rho_elec 128 8.7 0.174 0.190 13.803 13.857 dbcsr_mm_accdrv_process 62264 16.2 8.335 9.355 12.683 13.169 cp_fm_cholesky_invert 11 10.9 13.049 13.075 13.049 13.075 make_images_data 5014 15.6 0.064 0.070 11.373 12.702 cp_fm_upper_to_full 105 14.8 8.901 12.651 8.901 12.651 hybrid_alltoall_any 5200 16.5 0.531 2.214 10.235 11.681 init_scf_run 11 5.9 0.000 0.001 11.458 11.459 scf_env_initial_rho_setup 11 6.9 0.018 0.028 11.458 11.459 multiply_cannon_sync_h2d 30084 15.6 10.463 11.085 10.463 11.085 dbcsr_complete_redistribute 395 12.7 1.417 1.516 8.066 10.899 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.246 10.490 copy_fm_to_dbcsr 209 11.7 0.001 0.002 6.466 9.282 fft_wrap_pw1pw2 1291 11.7 0.018 0.024 8.223 8.329 grid_integrate_task_list 128 12.3 7.527 7.944 7.527 7.944 wfi_extrapolate 11 7.9 0.001 0.001 7.792 7.792 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.856 7.589 transfer_fm_to_dbcsr 11 9.9 0.002 0.011 4.817 7.579 mp_irecv_dv 76098 16.2 2.702 7.299 2.702 7.299 density_rs2pw 128 9.7 0.005 0.005 6.818 7.272 fft_wrap_pw1pw2_140 523 12.2 0.326 0.337 7.130 7.263 mp_alltoall_i22 716 14.1 4.207 7.059 4.207 7.059 cp_fm_cholesky_decompose 22 10.9 6.479 6.532 6.479 6.532 fft3d_ps 1291 13.7 2.941 3.008 6.420 6.492 calculate_dm_sparse 128 9.5 0.001 0.001 6.372 6.463 mp_alltoall_d11v 2415 14.1 5.435 6.407 5.435 6.407 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.754 5.795 grid_collocate_task_list 128 9.7 5.277 5.739 5.277 5.739 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.598 4.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.485 4.602 qs_energies_init_hamiltonians 11 5.9 0.023 0.041 4.601 4.602 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.371 4.442 potential_pw2rs 128 12.3 0.022 0.022 4.367 4.396 mp_sum_l 7950 12.9 3.035 3.972 3.035 3.972 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.812000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1039.090909, yerr=13.221570 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.420751E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 755071. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67810. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838186. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.068 166.664 166.664 qs_mol_dyn_low 1 2.0 0.003 0.003 165.450 165.545 qs_forces 11 3.9 0.005 0.007 165.282 165.290 qs_energies 11 4.9 0.001 0.002 157.834 157.843 scf_env_do_scf 11 5.9 0.001 0.001 139.050 139.063 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 108.187 108.188 velocity_verlet 10 3.0 0.039 0.075 106.432 106.435 dbcsr_multiply_generic 2507 12.6 0.183 0.193 73.340 73.904 qs_scf_new_mos 117 7.6 0.001 0.001 71.944 72.069 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.943 72.069 ot_scf_mini 117 9.6 0.003 0.003 67.530 67.648 multiply_cannon 2507 13.6 0.571 0.591 54.877 57.883 multiply_cannon_loop 2507 14.6 0.810 0.838 51.809 53.053 ot_mini 117 10.6 0.001 0.001 37.796 37.916 mp_waitall_1 125778 16.7 24.735 30.994 24.735 30.994 init_scf_loop 11 6.9 0.000 0.000 30.706 30.708 rebuild_ks_matrix 128 8.3 0.001 0.001 29.024 29.192 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 29.023 29.191 prepare_preconditioner 11 7.9 0.000 0.000 26.849 26.861 make_preconditioner 11 8.9 0.000 0.000 26.849 26.861 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.476 26.636 make_full_inverse_cholesky 11 9.9 0.022 0.028 24.955 25.220 multiply_cannon_multrec 10028 15.6 10.326 16.448 18.431 22.784 qs_ot_get_derivative 117 11.6 0.002 0.002 20.601 20.718 multiply_cannon_metrocomm3 10028 15.6 0.024 0.026 12.894 20.287 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.254 17.546 apply_single 128 13.6 0.001 0.001 17.254 17.545 qs_ot_get_p 128 10.4 0.001 0.001 17.049 17.194 ot_diis_step 117 11.6 0.020 0.020 17.120 17.121 make_m2s 5014 13.6 0.064 0.069 14.548 15.416 make_images 5014 14.6 2.171 2.583 14.240 15.102 cp_fm_cholesky_invert 11 10.9 14.810 14.821 14.810 14.821 sum_up_and_integrate 128 10.3 0.002 0.002 13.785 13.830 integrate_v_rspace 128 11.3 0.004 0.004 13.724 13.770 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.443 13.472 calculate_rho_elec 128 8.7 0.257 0.268 13.443 13.471 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.304 13.320 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.142 12.143 init_scf_run 11 5.9 0.000 0.001 11.888 11.888 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.887 11.888 multiply_cannon_sync_h2d 10028 15.6 10.700 11.217 10.700 11.217 make_images_data 5014 15.6 0.055 0.068 8.254 9.558 hybrid_alltoall_any 5200 16.5 0.846 3.777 8.167 9.342 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.222 9.233 cp_fm_diag_elpa_base 83 14.4 8.985 9.059 9.218 9.230 dbcsr_mm_accdrv_process 20762 16.1 3.283 4.687 7.760 8.595 fft_wrap_pw1pw2 1291 11.7 0.017 0.018 8.098 8.134 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.041 8.133 grid_integrate_task_list 128 12.3 7.797 8.080 7.797 8.080 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.344 7.804 fft_wrap_pw1pw2_140 523 12.2 0.433 0.439 7.146 7.181 wfi_extrapolate 11 7.9 0.001 0.001 7.090 7.090 cp_fm_cholesky_decompose 22 10.9 6.519 6.596 6.519 6.596 density_rs2pw 128 9.7 0.005 0.005 6.159 6.433 calculate_dm_sparse 128 9.5 0.001 0.001 6.250 6.329 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.281 6.295 multiply_cannon_metrocomm4 7521 15.6 0.025 0.028 1.882 6.127 fft3d_ps 1291 13.7 3.130 3.253 6.054 6.086 mp_irecv_dv 28860 15.9 1.842 6.025 1.842 6.025 grid_collocate_task_list 128 9.7 5.568 5.801 5.568 5.801 dbcsr_complete_redistribute 395 12.7 2.010 2.132 5.179 5.467 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 5.345 5.347 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.063 5.110 mp_alltoall_d11v 2415 14.1 4.585 4.920 4.585 4.920 calculate_first_density_matrix 1 7.0 0.000 0.000 4.563 4.565 mp_allgather_i34 2507 14.6 1.239 4.461 1.239 4.461 potential_pw2rs 128 12.3 0.026 0.027 3.943 3.949 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.609 3.947 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.659 3.747 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.623 3.672 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.318 3.621 copy_dbcsr_to_fm 186 11.8 0.003 0.004 3.483 3.576 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.455 3.472 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=166.664000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1342.090909, yerr=19.317701 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.964406E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3998 68758. MP_Allreduce 11055 1504. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.080 247.667 247.668 qs_mol_dyn_low 1 2.0 0.003 0.003 246.280 246.340 qs_forces 11 3.9 0.050 0.062 245.695 245.698 qs_energies 11 4.9 0.007 0.013 236.890 236.907 scf_env_do_scf 11 5.9 0.001 0.001 212.672 212.680 velocity_verlet 10 3.0 0.006 0.007 169.747 169.815 scf_env_do_scf_inner_loop 116 6.6 0.010 0.015 129.432 129.435 qs_scf_new_mos 116 7.6 0.001 0.001 87.216 87.276 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.215 87.275 dbcsr_multiply_generic 2485 12.5 0.214 0.225 83.613 84.012 init_scf_loop 11 6.9 0.000 0.000 82.943 82.949 ot_scf_mini 116 9.6 0.003 0.004 82.243 82.282 prepare_preconditioner 11 7.9 0.000 0.000 78.245 78.252 make_preconditioner 11 8.9 0.000 0.000 78.245 78.252 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.853 75.486 multiply_cannon 2485 13.5 0.690 0.748 60.862 61.798 multiply_cannon_loop 2485 14.5 1.039 1.052 56.909 58.113 cp_fm_upper_to_full 104 14.8 34.452 49.194 34.452 49.194 ot_mini 116 10.6 0.001 0.001 44.257 44.301 rebuild_ks_matrix 127 8.3 0.001 0.001 33.345 33.369 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.344 33.369 mp_waitall_1 102768 16.8 28.550 32.841 28.550 32.841 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.997 31.031 dbcsr_complete_redistribute 393 12.7 3.918 4.048 22.481 31.005 copy_fm_to_dbcsr 208 11.6 0.001 0.001 18.855 27.383 qs_ot_get_derivative 116 11.6 0.002 0.002 24.767 24.805 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 16.349 24.779 qs_ot_get_p 127 10.4 0.001 0.001 23.045 23.091 mp_alltoall_i22 712 14.1 14.076 22.780 14.076 22.780 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 19.768 21.013 make_m2s 4970 13.5 0.074 0.076 18.128 19.568 ot_diis_step 116 11.6 0.022 0.022 19.461 19.461 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.978 19.133 apply_single 127 13.6 0.001 0.001 18.978 19.132 qs_ot_p2m_diag 82 11.4 0.868 0.873 19.081 19.111 make_images 4970 14.5 3.077 3.265 17.651 19.092 multiply_cannon_multrec 9940 15.5 10.242 12.135 18.533 18.976 cp_fm_cholesky_invert 11 10.9 18.286 18.298 18.286 18.298 cp_dbcsr_syevd 82 12.4 0.006 0.006 17.311 17.314 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.671 15.684 calculate_rho_elec 127 8.7 0.479 0.479 15.670 15.683 sum_up_and_integrate 127 10.3 0.002 0.002 15.570 15.651 integrate_v_rspace 127 11.3 0.005 0.005 15.508 15.589 multiply_cannon_sync_h2d 9940 15.5 14.196 14.228 14.196 14.228 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.167 14.167 cp_fm_diag_elpa_base 82 14.4 11.355 12.361 14.162 14.162 init_scf_run 11 5.9 0.000 0.001 13.765 13.766 scf_env_initial_rho_setup 11 6.9 0.008 0.016 13.765 13.765 hybrid_alltoall_any 5155 16.4 1.301 3.035 10.283 12.390 make_images_data 4970 15.5 0.062 0.067 10.036 12.153 dbcsr_mm_accdrv_process 20590 16.0 4.402 6.272 8.042 10.169 fft_wrap_pw1pw2 1281 11.7 0.020 0.020 9.930 9.948 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.575 9.609 wfi_extrapolate 11 7.9 0.001 0.001 8.873 8.873 fft_wrap_pw1pw2_140 519 12.2 0.817 0.819 8.798 8.814 grid_integrate_task_list 127 12.3 8.497 8.691 8.497 8.691 qs_energies_init_hamiltonians 11 5.9 0.017 0.025 8.245 8.256 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.946 7.995 cp_fm_cholesky_decompose 22 10.9 7.355 7.423 7.355 7.423 fft3d_ps 1281 13.7 3.938 3.958 7.234 7.257 mp_alltoall_d11v 2401 14.1 7.081 7.195 7.081 7.195 calculate_dm_sparse 127 9.5 0.001 0.001 6.953 7.004 density_rs2pw 127 9.7 0.005 0.005 6.809 6.878 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.335 6.448 grid_collocate_task_list 127 9.7 6.359 6.431 6.359 6.431 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.814 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.058 5.603 qs_env_update_s_mstruct 11 6.9 0.006 0.014 5.005 5.254 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=247.668000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2652.909091, yerr=154.642964 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 1.259151E+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 1033 290947. 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 0.025 0.069 85.837 85.837 qs_energies 1 2.0 0.000 0.000 85.141 85.155 ls_scf 1 3.0 0.000 0.000 84.100 84.114 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.681 72.845 multiply_cannon 111 7.7 0.028 0.065 55.961 57.269 multiply_cannon_loop 111 8.7 0.228 0.247 52.542 53.980 ls_scf_main 1 4.0 0.000 0.000 52.246 52.247 density_matrix_trs4 2 5.0 0.002 0.003 46.687 46.753 ls_scf_init_scf 1 4.0 0.000 0.000 28.797 28.798 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.647 27.707 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.350 25.370 mp_waitall_1 11031 10.9 22.587 25.156 22.587 25.156 multiply_cannon_multrec 2664 9.7 8.134 8.868 15.622 17.335 multiply_cannon_sync_h2d 2664 9.7 13.414 15.217 13.414 15.217 make_m2s 222 7.7 0.009 0.011 13.138 13.681 make_images 222 8.7 0.099 0.109 13.116 13.662 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.747 12.305 make_images_data 222 9.7 0.004 0.005 7.713 8.397 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.470 8.271 dbcsr_mm_accdrv_process 4760 10.4 0.593 0.705 7.106 8.096 hybrid_alltoall_any 227 10.6 0.215 1.826 6.672 7.807 dbcsr_mm_accdrv_process_sort 4760 11.4 6.312 7.217 6.312 7.217 calculate_norms 4752 9.8 5.503 6.056 5.503 6.056 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.067 5.190 mp_sum_l 887 5.1 3.103 4.589 3.103 4.589 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.065 3.603 mp_irecv_dv 6231 10.9 2.048 3.577 2.048 3.577 make_images_sizes 222 9.7 0.000 0.000 0.709 3.411 mp_alltoall_i44 222 10.7 0.709 3.411 0.709 3.411 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.296 3.372 arnoldi_extremal 4 6.8 0.000 0.000 3.252 3.278 arnoldi_normal_ev 4 7.8 0.020 0.084 3.251 3.278 build_subspace 16 8.4 0.009 0.012 3.085 3.087 ls_scf_post 1 4.0 0.000 0.000 3.057 3.071 ls_scf_store_result 1 5.0 0.000 0.000 2.871 2.915 dbcsr_special_finalize 555 9.7 0.005 0.006 2.359 2.749 dbcsr_merge_single_wm 555 10.7 0.460 0.612 2.351 2.741 make_images_pack 222 9.7 2.209 2.626 2.211 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.309 2.550 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.458 2.069 2.460 dbcsr_sort_data 658 11.4 2.148 2.459 2.148 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.295 2.371 buffer_matrices_ensure_size 222 8.7 1.747 2.113 1.747 2.113 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.808 1.809 rebuild_ks_matrix 3 7.3 0.000 0.000 1.799 1.800 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.799 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.837000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1136.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 2.214818E+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 1033 301899. 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 0.046 0.068 90.365 90.366 qs_energies 1 2.0 0.000 0.000 89.744 89.749 ls_scf 1 3.0 0.000 0.000 88.302 88.308 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.323 74.653 multiply_cannon 111 7.7 0.028 0.041 52.993 56.978 ls_scf_main 1 4.0 0.000 0.000 54.466 54.470 multiply_cannon_loop 111 8.7 0.135 0.146 50.159 53.065 density_matrix_trs4 2 5.0 0.002 0.003 48.802 49.033 ls_scf_init_scf 1 4.0 0.000 0.000 29.963 29.964 mp_waitall_1 9105 10.9 20.784 29.878 20.784 29.878 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.785 28.867 multiply_cannon_multrec 1332 9.7 13.333 17.012 22.729 27.521 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.408 26.420 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.519 20.294 make_m2s 222 7.7 0.006 0.007 15.019 15.604 make_images 222 8.7 1.373 1.695 14.988 15.575 dbcsr_mm_accdrv_process 4041 10.4 0.329 0.564 8.987 10.570 dbcsr_mm_accdrv_process_sort 4041 11.4 8.495 10.011 8.495 10.011 make_images_data 222 9.7 0.004 0.005 8.729 9.676 hybrid_alltoall_any 227 10.6 0.539 2.574 8.088 9.483 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.189 7.801 mp_irecv_dv 3311 11.0 3.168 7.743 3.168 7.743 mp_sum_l 887 5.1 4.840 7.560 4.840 7.560 calculate_norms 2376 9.8 6.063 6.742 6.063 6.742 multiply_cannon_sync_h2d 1332 9.7 4.911 6.011 4.911 6.011 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.594 5.729 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.954 5.173 arnoldi_extremal 4 6.8 0.000 0.000 4.608 4.625 arnoldi_normal_ev 4 7.8 0.001 0.005 4.608 4.624 build_subspace 16 8.4 0.014 0.021 4.354 4.358 ls_scf_post 1 4.0 0.000 0.000 3.873 3.878 ls_scf_store_result 1 5.0 0.000 0.000 3.381 3.453 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.130 3.367 dbcsr_matrix_vector_mult_local 304 10.0 2.752 3.238 2.754 3.240 mp_allgather_i34 111 8.7 0.812 2.990 0.812 2.990 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.184 2.773 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.590 2.666 dbcsr_data_new 4174 10.1 2.126 2.411 2.126 2.411 make_images_pack 222 9.7 1.822 2.127 1.825 2.130 dbcsr_sort_data 436 11.2 1.811 2.011 1.811 2.011 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.882 1.885 rebuild_ks_matrix 3 7.3 0.000 0.000 1.869 1.871 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.869 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.366000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1855.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 2.956059E+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 1033 300682. 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 0.070 0.085 93.301 93.303 qs_energies 1 2.0 0.003 0.022 92.532 92.558 ls_scf 1 3.0 0.001 0.005 90.688 90.708 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.050 75.383 ls_scf_main 1 4.0 0.023 0.204 56.051 56.059 multiply_cannon 111 7.7 0.034 0.076 52.284 55.885 multiply_cannon_loop 111 8.7 0.117 0.131 49.532 54.081 density_matrix_trs4 2 5.0 0.003 0.017 50.102 50.332 mp_waitall_1 7281 11.0 23.026 33.688 23.026 33.688 ls_scf_init_scf 1 4.0 0.000 0.002 30.995 31.019 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.660 29.746 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.193 27.206 multiply_cannon_multrec 888 9.7 12.707 15.511 21.387 24.834 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.817 23.236 make_m2s 222 7.7 0.006 0.006 15.891 16.647 make_images 222 8.7 1.591 1.860 15.853 16.608 make_images_data 222 9.7 0.004 0.005 9.233 10.327 hybrid_alltoall_any 227 10.6 0.642 2.952 8.920 10.026 dbcsr_mm_accdrv_process 3754 10.4 0.322 0.511 8.209 9.495 dbcsr_mm_accdrv_process_sort 3754 11.4 7.767 8.984 7.767 8.984 mp_sum_l 887 5.1 5.191 8.405 5.191 8.405 multiply_cannon_sync_h2d 888 9.7 6.009 7.391 6.009 7.391 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.473 7.296 mp_irecv_dv 2335 11.1 2.457 7.246 2.457 7.246 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.622 6.733 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.918 6.518 arnoldi_extremal 4 6.8 0.000 0.000 5.104 5.121 arnoldi_normal_ev 4 7.8 0.010 0.037 5.104 5.121 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.773 5.011 calculate_norms 1584 9.8 4.397 4.906 4.397 4.906 build_subspace 16 8.4 0.014 0.020 4.779 4.786 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.442 3.776 ls_scf_post 1 4.0 0.001 0.005 3.642 3.662 dbcsr_matrix_vector_mult_local 304 10.0 3.035 3.616 3.037 3.618 mp_allgather_i34 111 8.7 0.849 3.599 0.849 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.450 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.848 2.964 dbcsr_data_new 4116 9.9 2.138 2.485 2.138 2.485 dbcsr_sort_data 325 11.1 1.884 2.089 1.884 2.089 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.063 2.088 rebuild_ks_matrix 3 7.3 0.000 0.000 2.045 2.070 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.028 2.045 2.070 make_images_pack 222 9.7 1.630 1.886 1.633 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.303000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2283.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 3.387814E+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 1033 301899. 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 0.032 0.044 97.752 97.753 qs_energies 1 2.0 0.000 0.000 96.373 96.378 ls_scf 1 3.0 0.000 0.000 94.696 94.699 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.348 78.582 ls_scf_main 1 4.0 0.000 0.000 58.650 58.651 multiply_cannon 111 7.7 0.050 0.123 51.587 56.156 density_matrix_trs4 2 5.0 0.002 0.003 52.511 52.647 multiply_cannon_loop 111 8.7 0.153 0.170 46.517 48.934 ls_scf_init_scf 1 4.0 0.000 0.000 32.814 32.816 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.594 31.676 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.956 28.969 mp_waitall_1 6369 11.0 22.459 28.754 22.459 28.754 multiply_cannon_multrec 1332 9.7 14.157 16.883 22.203 25.252 make_m2s 222 7.7 0.006 0.008 21.098 22.576 make_images 222 8.7 3.149 3.600 21.047 22.528 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.032 17.109 make_images_data 222 9.7 0.004 0.004 11.764 13.353 hybrid_alltoall_any 227 10.6 0.797 3.788 11.159 13.025 dbcsr_mm_accdrv_process 3641 10.4 0.296 0.482 7.684 9.201 dbcsr_mm_accdrv_process_sort 3641 11.4 7.221 8.707 7.221 8.707 mp_sum_l 887 5.1 4.083 7.640 4.083 7.640 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.082 6.167 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.107 6.094 mp_irecv_dv 3229 10.9 2.055 6.079 2.055 6.079 multiply_cannon_sync_h2d 1332 9.7 5.427 6.036 5.427 6.036 arnoldi_extremal 4 6.8 0.000 0.000 5.243 5.254 arnoldi_normal_ev 4 7.8 0.001 0.004 5.242 5.254 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.492 5.079 build_subspace 16 8.4 0.014 0.021 4.910 4.922 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.632 4.815 mp_allgather_i34 111 8.7 2.184 4.698 2.184 4.698 calculate_norms 2376 9.8 4.167 4.455 4.167 4.455 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.590 3.898 dbcsr_matrix_vector_mult_local 304 10.0 3.205 3.707 3.208 3.709 dbcsr_sort_data 658 11.4 3.096 3.433 3.096 3.433 dbcsr_special_finalize 555 9.7 0.006 0.007 2.860 3.241 ls_scf_post 1 4.0 0.000 0.000 3.231 3.236 dbcsr_merge_single_wm 555 10.7 0.546 0.668 2.851 3.232 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.998 3.048 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.037 dbcsr_data_release 10477 10.7 1.574 2.385 1.574 2.385 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.995 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.957 1.958 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.753000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 4.841632E+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 1033 300769. 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 0.077 0.101 93.206 93.207 qs_energies 1 2.0 0.000 0.000 91.652 91.687 ls_scf 1 3.0 0.000 0.000 89.676 89.712 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.738 70.900 ls_scf_main 1 4.0 0.000 0.000 56.301 56.301 multiply_cannon 111 7.7 0.087 0.136 52.603 55.790 multiply_cannon_loop 111 8.7 0.088 0.094 50.054 51.286 density_matrix_trs4 2 5.0 0.002 0.003 49.328 49.404 ls_scf_init_scf 1 4.0 0.000 0.000 29.943 29.948 mp_waitall_1 5436 11.0 24.134 28.743 24.134 28.743 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.478 28.515 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.305 26.315 multiply_cannon_multrec 444 9.7 13.801 16.573 20.967 22.292 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.319 15.553 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.514 14.794 make_m2s 222 7.7 0.005 0.005 13.396 14.318 make_images 222 8.7 2.042 2.488 13.328 14.248 make_images_data 222 9.7 0.003 0.004 8.118 9.640 hybrid_alltoall_any 227 10.6 0.801 3.823 8.103 9.498 dbcsr_mm_accdrv_process 3003 10.4 0.351 0.401 6.862 8.049 multiply_cannon_sync_h2d 444 9.7 6.732 7.838 6.732 7.838 dbcsr_mm_accdrv_process_sort 3003 11.4 6.499 7.648 6.499 7.648 arnoldi_extremal 4 6.8 0.000 0.000 5.875 5.887 arnoldi_normal_ev 4 7.8 0.002 0.005 5.875 5.887 build_subspace 16 8.4 0.015 0.020 5.479 5.492 mp_sum_l 887 5.1 2.790 4.739 2.790 4.739 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.429 4.626 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.198 4.395 dbcsr_matrix_vector_mult_local 304 10.0 3.731 4.205 3.733 4.207 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.984 3.853 calculate_norms 792 9.8 3.620 3.746 3.620 3.746 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.503 3.629 mp_irecv_dv 1241 11.2 1.490 3.606 1.490 3.606 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.504 3.589 mp_allgather_i34 111 8.7 1.122 3.490 1.122 3.490 ls_scf_post 1 4.0 0.000 0.000 3.432 3.467 make_images_sizes 222 9.7 0.000 0.000 0.880 3.283 mp_alltoall_i44 222 10.7 0.880 3.283 0.880 3.283 ls_scf_store_result 1 5.0 0.000 0.000 3.196 3.229 dbcsr_data_new 4608 9.7 1.792 2.271 1.792 2.271 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.267 2.268 dbcsr_finalize 304 7.8 0.062 0.077 2.193 2.261 rebuild_ks_matrix 3 7.3 0.000 0.000 2.234 2.235 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.234 2.235 dbcsr_merge_all 275 8.9 0.478 0.521 2.049 2.094 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.960 1.960 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.207000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3802.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 9.048433E+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 1033 319197. 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 0.097 0.118 113.062 113.063 qs_energies 1 2.0 0.000 0.000 110.950 110.957 ls_scf 1 3.0 0.000 0.000 107.948 107.955 dbcsr_multiply_generic 111 6.7 0.024 0.028 81.521 81.688 ls_scf_main 1 4.0 0.000 0.000 63.690 63.691 density_matrix_trs4 2 5.0 0.002 0.003 54.903 54.978 multiply_cannon 111 7.7 0.164 0.242 49.334 54.652 multiply_cannon_loop 111 8.7 0.098 0.100 46.334 52.763 ls_scf_init_scf 1 4.0 0.000 0.000 40.399 40.400 ls_scf_init_matrix_S 1 5.0 0.000 0.000 38.745 38.755 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 36.039 36.048 mp_waitall_1 4527 11.1 21.872 25.852 21.872 25.852 make_m2s 222 7.7 0.005 0.005 22.667 23.773 make_images 222 8.7 3.577 3.865 22.560 23.664 multiply_cannon_multrec 444 9.7 17.806 18.436 22.523 23.102 hybrid_alltoall_any 227 10.6 1.653 3.625 12.795 15.773 make_images_data 222 9.7 0.003 0.004 13.025 15.185 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.252 10.827 multiply_cannon_sync_h2d 444 9.7 8.782 8.850 8.782 8.850 mp_sum_l 887 5.1 7.118 8.782 7.118 8.782 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 6.537 7.787 arnoldi_extremal 4 6.8 0.000 0.000 7.520 7.531 arnoldi_normal_ev 4 7.8 0.005 0.011 7.520 7.531 compute_matrix_preconditioner 1 6.0 0.002 0.002 7.405 7.411 build_subspace 16 8.4 0.026 0.036 6.960 6.980 acc_transpose_blocks 444 9.7 0.003 0.003 1.048 6.864 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 0.965 6.782 jit_kernel_transpose 1 13.0 0.960 6.776 0.960 6.776 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.561 5.719 dbcsr_matrix_vector_mult_local 304 10.0 5.118 5.435 5.120 5.437 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.889 5.151 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.910 5.003 dbcsr_mm_accdrv_process 1814 10.4 0.314 0.357 4.527 4.648 dbcsr_mm_accdrv_process_sort 1814 11.4 4.188 4.318 4.188 4.318 ls_scf_post 1 4.0 0.000 0.000 3.859 3.866 make_images_sizes 222 9.7 0.000 0.000 1.500 3.674 mp_alltoall_i44 222 10.7 1.499 3.674 1.499 3.674 mp_allgather_i34 111 8.7 1.115 3.658 1.115 3.658 ls_scf_store_result 1 5.0 0.000 0.000 3.518 3.558 calculate_norms 792 9.8 3.240 3.327 3.240 3.327 dbcsr_finalize 304 7.8 0.082 0.089 3.077 3.151 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.971 2.971 dbcsr_merge_all 275 8.9 0.883 0.907 2.861 2.930 dbcsr_complete_redistribute 5 7.6 1.426 1.471 2.742 2.848 matrix_ls_to_qs 2 6.0 0.000 0.000 2.395 2.499 dbcsr_sort_data 325 11.1 2.439 2.494 2.439 2.494 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.451 2.453 rebuild_ks_matrix 3 7.3 0.000 0.000 2.385 2.387 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.384 2.386 dbcsr_data_new 6591 9.6 1.838 2.368 1.838 2.368 dbcsr_new_transposed 4 7.5 0.259 0.285 2.330 2.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=113.063000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7152.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/aeba16654204b33cadbf84abc85e7a1388c87784_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 585.003008E+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 77937. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 993468628. 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 0.048 0.123 230.345 230.346 qs_mol_dyn_low 1 2.0 0.020 0.188 228.614 228.636 qs_forces 5 3.8 0.027 0.066 228.316 228.321 qs_energies 5 4.8 0.002 0.042 225.178 225.225 scf_env_do_scf 5 5.8 0.000 0.011 206.539 206.540 scf_env_do_scf_inner_loop 105 6.6 0.007 0.262 180.939 181.149 qs_scf_new_mos 105 7.6 0.000 0.001 141.270 141.383 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.269 141.383 dbcsr_multiply_generic 1445 12.2 0.126 0.135 137.074 137.455 ot_scf_mini 105 9.6 0.003 0.004 131.386 131.529 multiply_cannon 1445 13.2 0.275 0.286 118.133 120.136 multiply_cannon_loop 1445 14.2 2.857 3.021 116.402 117.582 velocity_verlet 4 3.0 0.012 0.067 103.281 103.287 ot_mini 105 10.6 0.001 0.001 59.743 59.859 multiply_cannon_multrec 69360 15.2 29.722 34.764 40.502 46.055 qs_ot_get_p 112 10.4 0.001 0.001 41.717 41.959 mp_waitall_1 488190 16.1 34.920 41.653 34.920 41.653 qs_ot_get_derivative 55 11.6 0.001 0.001 38.027 38.144 multiply_cannon_metrocomm3 69360 15.2 0.199 0.212 25.539 33.177 multiply_cannon_sync_h2d 69360 15.2 29.292 33.076 29.292 33.076 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.695 30.778 rebuild_ks_matrix 110 8.4 0.000 0.000 29.042 29.200 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.098 29.042 29.200 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.518 27.519 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.583 26.746 init_scf_loop 7 6.6 0.001 0.019 25.566 25.570 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.011 23.266 apply_single 62 13.6 0.000 0.000 23.011 23.266 cp_fm_syevd 40 13.0 0.000 0.000 22.283 22.433 ot_new_cg_direction 55 11.6 0.001 0.001 21.031 21.032 prepare_preconditioner 7 7.6 0.000 0.000 20.405 20.438 make_preconditioner 7 8.6 0.000 0.020 20.405 20.438 cp_fm_redistribute_end 40 14.0 8.789 17.503 8.803 17.509 cp_fm_syevd_base 40 14.0 8.695 17.410 8.695 17.410 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.592 16.855 calculate_rho_elec 110 8.6 0.029 0.039 16.591 16.854 init_scf_run 5 5.8 0.000 0.001 16.120 16.121 scf_env_initial_rho_setup 5 6.8 0.001 0.003 16.120 16.121 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.696 14.897 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.579 13.681 mp_sum_l 4764 12.2 11.920 12.819 11.920 12.819 calculate_dm_sparse 110 9.5 0.000 0.001 12.486 12.685 make_full_inverse_cholesky 7 9.6 0.000 0.000 12.448 12.522 calculate_first_density_matrix 1 7.0 0.000 0.014 11.792 11.814 dbcsr_mm_accdrv_process 154766 15.8 6.239 6.444 10.647 11.683 density_rs2pw 110 9.6 0.004 0.005 10.885 11.338 fft_wrap_pw1pw2 1425 12.5 0.018 0.024 11.050 11.277 acc_transpose_blocks 69360 15.2 0.342 0.357 10.333 11.122 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 10.730 10.749 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 10.190 10.239 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.164 10.228 qs_vxc_create 110 10.4 0.002 0.002 10.095 10.120 fft_wrap_pw1pw2_240 915 14.0 0.198 0.222 9.702 9.870 check_diag 80 13.5 8.651 8.924 9.508 9.651 multiply_cannon_metrocomm1 69360 15.2 0.093 0.100 4.785 8.899 fft3d_pb 915 15.0 2.245 2.456 8.667 8.867 sum_up_and_integrate 60 10.3 0.001 0.032 8.580 8.590 integrate_v_rspace 60 11.3 0.002 0.002 8.562 8.574 transfer_rs2pw 445 10.6 0.008 0.009 7.834 8.317 acc_transpose_blocks_kernels 69360 16.2 0.851 0.896 7.598 8.314 cp_fm_cholesky_invert 7 10.6 8.258 8.275 8.258 8.275 make_full_single_inverse 7 9.6 0.001 0.031 7.671 7.704 jit_kernel_transpose 5 15.0 6.747 7.442 6.747 7.442 xc_rho_set_and_dset_create 110 12.4 0.072 0.092 7.142 7.370 make_m2s 2890 13.2 0.078 0.088 6.599 7.159 make_images 2890 14.2 0.239 0.259 6.492 7.053 xc_vxc_pw_create 60 11.3 0.036 0.039 6.765 6.794 mp_alltoall_z22v 2340 16.7 5.853 6.048 5.853 6.048 xc_pw_derive 510 13.4 0.005 0.006 5.947 6.000 mp_waitany 7680 13.5 4.453 5.025 4.453 5.025 multiply_cannon_metrocomm4 67915 15.2 0.181 0.197 2.018 4.970 potential_pw2rs 60 12.3 0.002 0.003 4.768 4.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.346000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=556.800000, yerr=1.600000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: aeba16654204b33cadbf84abc85e7a1388c87784 Summary: empty Status: OK