=== 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: 75a566c04f83ccf946c3d618f248ad2bb7ee4910 ################# 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, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/01 job id: 51039280 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/02 job id: 51039281 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/03 job id: 51039282 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/04 job id: 51039283 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/05 job id: 51039284 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/06 job id: 51039285 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/07 job id: 51039286 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/08 job id: 51039287 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/09 job id: 51039288 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/10 job id: 51039289 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/11 job id: 51039290 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/12 job id: 51039291 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/13 job id: 51039292 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/14 job id: 51039293 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/15 job id: 51039294 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/16 job id: 51039295 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/17 job id: 51039296 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/18 job id: 51039297 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/19 job id: 51039298 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/20 job id: 51039299 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/21 job id: 51039300 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/22 job id: 51039301 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/23 job id: 51039302 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/24 job id: 51039303 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/25 job id: 51039304 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/26 job id: 51039305 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/27 job id: 51039306 --- 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/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 0.034 133.818 133.819 farming_run 1 2.0 133.122 133.125 133.789 133.792 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.489396E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 114.742 114.742 qs_energies 1 2.0 0.000 0.000 114.528 114.530 mp2_main 1 3.0 0.000 0.000 112.425 112.427 mp2_gpw_main 1 4.0 0.027 0.033 111.413 111.416 mp2_ri_gpw_compute_in 1 5.0 0.172 0.176 92.187 92.690 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.575 55.077 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 40.986 46.149 get_2c_integrals 1 6.0 0.008 0.009 36.913 37.438 integrate_v_rspace 273 8.0 0.441 0.454 24.602 29.480 pw_transfer 6555 10.6 0.384 0.395 26.784 27.505 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.411 26.191 grid_integrate_task_list 273 9.0 20.506 25.823 20.506 25.823 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 22.950 23.700 rpa_ri_compute_en 1 5.0 0.019 0.022 19.121 19.269 compute_2c_integrals 1 7.0 0.002 0.002 19.005 19.007 compute_2c_integrals_loop_lm 1 8.0 0.001 0.004 18.541 18.719 mp2_eri_2c_integrate_gpw 1 9.0 2.383 2.423 18.540 18.718 cp_fm_cholesky_decompose 12 8.2 17.889 18.396 17.889 18.396 cholesky_decomp 1 7.0 0.000 0.000 16.747 17.248 fft3d_s 5443 13.4 16.103 16.325 16.124 16.346 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.320 10.764 15.320 calculate_wavefunction 272 8.0 5.370 5.525 12.295 12.963 rpa_num_int 1 6.0 0.002 0.023 10.717 10.718 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.665 10.694 calc_mat_Q 8 8.0 0.000 0.000 9.487 9.580 contract_S_to_Q 8 9.0 0.000 0.000 8.909 9.001 calc_potential_gpw 544 9.5 0.004 0.004 8.202 8.665 parallel_gemm_fm 14 9.1 0.000 0.000 8.488 8.606 parallel_gemm_fm_cosma 14 10.1 8.488 8.606 8.488 8.606 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.046 8.304 potential_pw2rs 545 10.0 0.106 0.108 7.487 8.199 create_integ_mat 1 6.0 0.024 0.030 7.809 7.810 collocate_single_gaussian 272 10.0 0.038 0.041 7.308 7.587 array2fm 1 7.0 0.000 0.000 6.689 7.225 pw_scatter_s 2720 13.7 4.339 4.448 4.339 4.448 pw_gather_s 2722 13.2 3.439 3.828 3.439 3.828 array2fm_buffer_send 1 8.0 2.899 3.044 2.899 3.044 pw_poisson_solve 545 10.5 1.120 1.190 2.194 2.360 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.415847, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 0.024 0.036 419.822 419.825 farming_run 1 2.0 418.724 418.732 419.779 419.781 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.245118E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 217.042 217.043 qs_energies 1 2.0 0.000 0.000 216.813 216.820 scf_env_do_scf 1 3.0 0.000 0.000 115.051 115.052 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.147 114.156 rebuild_ks_matrix 4 6.0 0.000 0.000 114.146 114.155 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 114.146 114.155 hfx_ks_matrix 4 8.0 0.001 0.001 113.802 113.806 integrate_four_center 4 9.0 0.152 0.457 113.801 113.806 integrate_four_center_main 4 10.0 0.118 0.803 101.928 105.151 integrate_four_center_bin 266 11.0 101.810 105.137 101.810 105.137 mp2_main 1 3.0 0.000 0.001 101.471 101.478 mp2_gpw_main 1 4.0 0.049 0.070 100.614 100.623 init_scf_loop 1 4.0 0.000 0.000 97.069 97.069 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 73.859 74.859 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.768 54.767 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.158 41.271 46.500 integrate_v_rspace 95 8.0 0.395 0.565 27.737 32.812 pw_transfer 2240 10.6 0.147 0.166 29.494 29.954 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.507 28.990 mp2_ri_gpw_compute_en 1 5.0 0.056 0.064 26.608 28.694 grid_integrate_task_list 95 9.0 23.158 28.442 23.158 28.442 ao_to_mo_and_store_B_mult_1 91 7.0 10.825 27.527 10.825 27.527 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.255 26.671 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.895 24.957 24.967 get_2c_integrals 1 6.0 0.000 0.000 20.007 20.028 compute_2c_integrals 1 7.0 0.002 0.003 18.982 18.987 fft3d_s 1823 13.4 18.492 18.886 18.505 18.899 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.592 18.859 mp2_eri_2c_integrate_gpw 1 9.0 1.735 1.825 18.591 18.858 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.981 17.981 calculate_wavefunction 91 8.0 2.017 2.047 9.626 9.817 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.590 8.761 9.304 potential_pw2rs 186 10.0 0.033 0.034 8.382 8.909 local_gemm 172 8.0 8.203 8.735 8.203 8.735 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.134 8.425 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.511 7.975 8.387 calc_potential_gpw 182 9.5 0.002 0.002 7.921 8.153 collocate_single_gaussian 91 10.0 0.016 0.021 7.737 7.960 integrate_four_center_load 4 10.0 0.000 0.000 6.761 6.764 hfx_load_balance 1 11.0 0.000 0.000 6.761 6.764 mp_sync 37 10.5 3.847 6.627 3.847 6.627 mp_sendrecv_dm3 2068 8.0 6.015 6.439 6.015 6.439 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.396 6.338 6.396 pw_gather_s 912 13.2 4.475 4.930 4.475 4.930 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.598753, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 453.468160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 1213312. 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 0.006 0.026 53.727 53.728 qs_mol_dyn_low 1 2.0 0.003 0.003 53.522 53.531 qs_forces 11 3.9 0.002 0.002 53.433 53.434 qs_energies 11 4.9 0.001 0.001 51.962 51.975 scf_env_do_scf 11 5.9 0.000 0.001 45.893 45.893 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.893 43.894 qs_scf_new_mos 108 7.5 0.000 0.001 34.049 34.339 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.048 34.338 dbcsr_multiply_generic 2286 12.5 0.094 0.099 33.869 34.254 ot_scf_mini 108 9.5 0.002 0.002 32.335 32.551 multiply_cannon 2286 13.5 0.189 0.195 26.385 27.956 multiply_cannon_loop 2286 14.5 1.784 1.910 25.746 27.348 velocity_verlet 10 3.0 0.003 0.008 25.967 25.968 ot_mini 108 10.5 0.001 0.001 19.511 19.790 qs_ot_get_derivative 108 11.5 0.001 0.001 16.505 16.722 mp_waitall_1 245248 16.5 8.461 14.509 8.461 14.509 multiply_cannon_metrocomm3 54864 15.5 0.072 0.076 6.104 12.989 multiply_cannon_multrec 54864 15.5 3.703 5.816 7.802 11.252 qs_ot_get_p 119 10.4 0.001 0.001 8.149 8.400 rebuild_ks_matrix 119 8.3 0.000 0.000 7.707 7.854 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.707 7.854 mp_sum_l 7287 12.8 5.601 7.371 5.601 7.371 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.786 6.917 multiply_cannon_sync_h2d 54864 15.5 5.215 6.621 5.215 6.621 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.596 6.046 dbcsr_mm_accdrv_process 76910 16.1 1.773 2.861 4.013 5.776 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.605 5.738 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.245 5.290 init_scf_run 11 5.9 0.000 0.001 4.822 4.822 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.822 4.822 cp_dbcsr_syevd 50 12.0 0.003 0.004 4.471 4.471 sum_up_and_integrate 119 10.3 0.001 0.001 4.349 4.356 integrate_v_rspace 119 11.3 0.002 0.002 4.338 4.347 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.283 4.284 cp_fm_redistribute_end 50 14.0 2.185 4.259 2.191 4.262 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.046 4.183 calculate_rho_elec 119 8.7 0.011 0.016 4.046 4.183 cp_fm_diag_elpa_base 50 14.0 2.066 4.152 2.070 4.161 multiply_cannon_metrocomm1 54864 15.5 0.056 0.063 1.749 3.547 calculate_dm_sparse 119 9.5 0.000 0.000 3.035 3.182 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.913 3.122 apply_single 119 13.6 0.000 0.000 2.913 3.121 jit_kernel_multiply 13 15.8 2.178 2.824 2.178 2.824 acc_transpose_blocks 54864 15.5 0.220 0.247 2.169 2.734 ot_diis_step 108 11.5 0.006 0.006 2.725 2.726 calculate_first_density_matrix 1 7.0 0.000 0.000 2.647 2.653 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.453 2.513 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.362 2.364 density_rs2pw 119 9.7 0.004 0.004 2.058 2.131 grid_integrate_task_list 119 12.3 2.020 2.121 2.020 2.121 wfi_extrapolate 11 7.9 0.001 0.001 2.104 2.104 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.039 2.104 init_scf_loop 11 6.9 0.000 0.000 1.983 1.983 mp_sum_d 4137 12.0 1.368 1.965 1.368 1.965 potential_pw2rs 119 12.3 0.004 0.004 1.743 1.754 pw_transfer 1439 11.6 0.052 0.056 1.591 1.657 make_m2s 4572 13.5 0.053 0.055 1.563 1.606 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.516 1.584 make_images 4572 14.5 0.133 0.139 1.481 1.524 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.484 1.513 mp_alltoall_d11v 2130 13.8 1.271 1.458 1.271 1.458 acc_transpose_blocks_sync 164592 16.5 1.163 1.430 1.163 1.430 grid_collocate_task_list 119 9.7 1.354 1.426 1.354 1.426 mp_waitany 12084 13.8 1.216 1.404 1.216 1.404 transfer_rs2pw 487 10.6 0.005 0.006 1.301 1.391 fft3d_ps 1201 14.6 0.368 0.472 1.227 1.292 transfer_pw2rs 487 13.2 0.006 0.006 1.278 1.284 fft_wrap_pw1pw2_140 487 13.2 0.108 0.116 1.177 1.247 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.745 1.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.728000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.000000, yerr=1.348400 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 489.504768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1757705. 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 0.014 0.031 38.671 38.673 qs_mol_dyn_low 1 2.0 0.003 0.005 38.382 38.390 qs_forces 11 3.9 0.004 0.015 38.307 38.309 qs_energies 11 4.9 0.004 0.017 36.645 36.650 scf_env_do_scf 11 5.9 0.002 0.005 31.412 31.413 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 29.011 29.013 dbcsr_multiply_generic 2286 12.5 0.101 0.105 21.875 22.233 qs_scf_new_mos 108 7.5 0.001 0.001 20.492 20.722 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.491 20.721 ot_scf_mini 108 9.5 0.003 0.003 19.577 19.746 multiply_cannon 2286 13.5 0.207 0.216 16.748 18.178 velocity_verlet 10 3.0 0.001 0.002 18.171 18.173 multiply_cannon_loop 2286 14.5 1.179 1.244 15.601 17.125 ot_mini 108 10.5 0.001 0.002 12.076 12.311 mp_waitall_1 200699 16.5 5.411 10.525 5.411 10.525 qs_ot_get_derivative 108 11.5 0.001 0.001 9.729 9.898 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 3.957 9.333 multiply_cannon_multrec 27432 15.5 1.831 4.059 6.134 9.041 rebuild_ks_matrix 119 8.3 0.000 0.000 6.706 6.851 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.024 6.705 6.850 dbcsr_mm_accdrv_process 47894 16.0 3.485 5.931 4.223 6.282 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.902 6.032 qs_ot_get_p 119 10.4 0.001 0.002 4.585 4.811 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.628 4.534 mp_sum_l 7287 12.8 2.096 4.088 2.096 4.088 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.052 4.030 apply_single 119 13.6 0.000 0.000 3.052 4.030 init_scf_run 11 5.9 0.000 0.001 3.982 3.983 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.982 3.983 sum_up_and_integrate 119 10.3 0.001 0.003 3.640 3.646 integrate_v_rspace 119 11.3 0.002 0.003 3.626 3.631 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.574 3.598 calculate_rho_elec 119 8.7 0.021 0.024 3.573 3.598 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.955 2.974 make_m2s 4572 13.5 0.052 0.053 2.584 2.913 make_images 4572 14.5 0.206 0.244 2.494 2.827 cp_dbcsr_syevd 50 12.0 0.004 0.008 2.562 2.563 calculate_first_density_matrix 1 7.0 0.001 0.003 2.476 2.478 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.304 2.392 init_scf_loop 11 6.9 0.001 0.004 2.380 2.381 ot_diis_step 108 11.5 0.011 0.012 2.299 2.299 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.247 2.248 multiply_cannon_sync_h2d 27432 15.5 1.688 2.225 1.688 2.225 cp_fm_redistribute_end 50 14.0 1.137 2.219 1.140 2.222 cp_fm_diag_elpa_base 50 14.0 1.049 2.128 1.077 2.170 calculate_dm_sparse 119 9.5 0.000 0.001 2.053 2.133 acc_transpose_blocks 27432 15.5 0.115 0.121 1.611 1.956 density_rs2pw 119 9.7 0.004 0.004 1.878 1.951 grid_integrate_task_list 119 12.3 1.833 1.916 1.833 1.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.891 1.893 pw_transfer 1439 11.6 0.065 0.068 1.826 1.864 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.734 1.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.673 1.713 make_images_data 4572 15.5 0.047 0.054 1.211 1.546 prepare_preconditioner 11 7.9 0.000 0.000 1.484 1.511 make_preconditioner 11 8.9 0.000 0.001 1.484 1.511 jit_kernel_multiply 8 16.4 0.680 1.502 0.680 1.502 wfi_extrapolate 11 7.9 0.001 0.002 1.446 1.446 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.388 1.445 hybrid_alltoall_any 4725 16.4 0.053 0.114 1.076 1.427 fft_wrap_pw1pw2_140 487 13.2 0.127 0.132 1.382 1.420 potential_pw2rs 119 12.3 0.006 0.006 1.402 1.406 fft3d_ps 1201 14.6 0.520 0.576 1.365 1.396 grid_collocate_task_list 119 9.7 1.288 1.355 1.288 1.355 mp_alltoall_d11v 2130 13.8 1.179 1.294 1.179 1.294 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.255 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.200 1.249 transfer_rs2pw 487 10.6 0.005 0.005 1.086 1.178 mp_sum_d 4137 12.0 0.565 1.017 0.565 1.017 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.973 0.976 mp_allgather_i34 2286 14.5 0.572 0.951 0.572 0.951 acc_transpose_blocks_sync 82296 16.5 0.812 0.943 0.812 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.877 transfer_pw2rs 487 13.2 0.004 0.005 0.873 0.876 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.658 0.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.673000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.363636, yerr=1.226431 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 522.129408E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 31.753 31.754 qs_mol_dyn_low 1 2.0 0.003 0.003 31.559 31.568 qs_forces 11 3.9 0.002 0.003 31.501 31.502 qs_energies 11 4.9 0.001 0.001 29.965 29.967 scf_env_do_scf 11 5.9 0.000 0.001 25.228 25.228 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.838 22.839 dbcsr_multiply_generic 2286 12.5 0.096 0.098 16.786 16.870 qs_scf_new_mos 108 7.5 0.001 0.001 15.242 15.262 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.242 15.262 velocity_verlet 10 3.0 0.001 0.001 14.990 14.991 ot_scf_mini 108 9.5 0.002 0.002 14.488 14.505 multiply_cannon 2286 13.5 0.194 0.197 13.624 14.403 multiply_cannon_loop 2286 14.5 0.857 0.893 12.846 13.633 ot_mini 108 10.5 0.001 0.001 8.975 8.995 qs_ot_get_derivative 108 11.5 0.001 0.001 7.481 7.498 multiply_cannon_multrec 18288 15.5 1.882 2.942 6.926 7.284 rebuild_ks_matrix 119 8.3 0.000 0.001 5.906 5.928 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.906 5.928 dbcsr_mm_accdrv_process 38222 16.0 4.887 5.640 4.951 5.704 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.201 5.221 mp_waitall_1 158411 16.6 2.592 3.718 2.592 3.718 init_scf_run 11 5.9 0.000 0.001 3.582 3.582 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.582 3.582 sum_up_and_integrate 119 10.3 0.001 0.001 3.505 3.510 integrate_v_rspace 119 11.3 0.002 0.003 3.492 3.499 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.840 3.401 qs_ot_get_p 119 10.4 0.001 0.001 3.288 3.316 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.230 3.235 calculate_rho_elec 119 8.7 0.031 0.031 3.229 3.235 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.460 2.444 calculate_first_density_matrix 1 7.0 0.000 0.000 2.376 2.377 init_scf_loop 11 6.9 0.000 0.000 2.373 2.374 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.979 2.358 apply_single 119 13.6 0.000 0.000 1.978 2.357 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.194 2.199 make_m2s 4572 13.5 0.044 0.045 1.773 1.909 grid_integrate_task_list 119 12.3 1.798 1.907 1.798 1.907 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.902 1.902 density_rs2pw 119 9.7 0.004 0.004 1.800 1.887 calculate_dm_sparse 119 9.5 0.000 0.001 1.874 1.883 pw_transfer 1439 11.6 0.065 0.069 1.835 1.843 make_images 4572 14.5 0.192 0.204 1.687 1.822 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.742 1.753 acc_transpose_blocks 18288 15.5 0.079 0.081 1.629 1.715 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.688 1.700 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.659 1.670 cp_fm_diag_elpa_base 50 14.0 1.636 1.647 1.657 1.668 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.660 1.662 prepare_preconditioner 11 7.9 0.000 0.000 1.627 1.630 make_preconditioner 11 8.9 0.000 0.000 1.627 1.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.484 1.569 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.513 1.521 ot_diis_step 108 11.5 0.011 0.011 1.480 1.480 fft_wrap_pw1pw2_140 487 13.2 0.179 0.184 1.412 1.423 mp_sum_l 7287 12.8 0.994 1.408 0.994 1.408 grid_collocate_task_list 119 9.7 1.234 1.334 1.234 1.334 potential_pw2rs 119 12.3 0.007 0.008 1.309 1.312 fft3d_ps 1201 14.6 0.545 0.569 1.297 1.310 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.216 1.219 multiply_cannon_sync_h2d 18288 15.5 1.033 1.179 1.033 1.179 wfi_extrapolate 11 7.9 0.001 0.001 1.154 1.154 transfer_rs2pw 487 10.6 0.005 0.005 0.949 1.040 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.940 0.958 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.935 0.936 make_images_data 4572 15.5 0.047 0.050 0.766 0.931 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.663 0.845 acc_transpose_blocks_sync 54864 16.5 0.734 0.816 0.734 0.816 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.795 0.805 transfer_pw2rs 487 13.2 0.004 0.004 0.787 0.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.775 0.778 mp_alltoall_d11v 2130 13.8 0.657 0.767 0.657 0.767 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.672 0.731 cp_fm_cholesky_invert 11 10.9 0.684 0.688 0.684 0.688 mp_alltoall_z22v 1201 16.6 0.604 0.676 0.604 0.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.754000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 558.100480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 36.221 36.222 qs_mol_dyn_low 1 2.0 0.003 0.003 36.064 36.073 qs_forces 11 3.9 0.002 0.002 35.997 35.998 qs_energies 11 4.9 0.001 0.001 34.300 34.305 scf_env_do_scf 11 5.9 0.000 0.001 29.168 29.169 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.821 25.821 dbcsr_multiply_generic 2286 12.5 0.102 0.105 19.812 19.982 velocity_verlet 10 3.0 0.001 0.002 18.261 18.263 qs_scf_new_mos 108 7.5 0.001 0.001 17.816 17.878 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.816 17.877 ot_scf_mini 108 9.5 0.002 0.003 16.792 16.850 multiply_cannon 2286 13.5 0.218 0.224 16.162 16.763 multiply_cannon_loop 2286 14.5 1.532 1.619 15.207 15.610 ot_mini 108 10.5 0.001 0.001 10.432 10.493 multiply_cannon_multrec 27432 15.5 2.451 3.315 8.994 9.520 qs_ot_get_derivative 108 11.5 0.001 0.001 8.574 8.632 dbcsr_mm_accdrv_process 47916 15.9 5.938 7.621 6.440 7.784 rebuild_ks_matrix 119 8.3 0.000 0.000 6.269 6.321 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.269 6.320 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.553 5.597 init_scf_run 11 5.9 0.000 0.001 3.808 3.808 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.808 3.808 qs_ot_get_p 119 10.4 0.001 0.001 3.534 3.614 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.016 3.499 sum_up_and_integrate 119 10.3 0.001 0.001 3.446 3.454 integrate_v_rspace 119 11.3 0.002 0.003 3.435 3.443 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.323 3.355 calculate_rho_elec 119 8.7 0.040 0.046 3.323 3.355 init_scf_loop 11 6.9 0.000 0.000 3.329 3.329 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.174 2.646 apply_single 119 13.6 0.000 0.000 2.173 2.645 prepare_preconditioner 11 7.9 0.000 0.000 2.509 2.517 make_preconditioner 11 8.9 0.000 0.000 2.509 2.517 acc_transpose_blocks 27432 15.5 0.121 0.124 2.409 2.510 calculate_first_density_matrix 1 7.0 0.000 0.000 2.438 2.441 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.108 2.436 make_m2s 4572 13.5 0.054 0.056 2.202 2.335 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.293 2.322 mp_waitall_1 137007 16.6 1.765 2.294 1.765 2.294 make_images 4572 14.5 0.273 0.337 2.094 2.225 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.194 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.140 2.151 pw_transfer 1439 11.6 0.064 0.069 1.962 1.997 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.870 1.909 grid_integrate_task_list 119 12.3 1.833 1.899 1.833 1.899 density_rs2pw 119 9.7 0.004 0.004 1.772 1.853 ot_diis_step 108 11.5 0.012 0.012 1.818 1.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.784 1.785 cp_dbcsr_syevd 50 12.0 0.010 0.012 1.741 1.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.617 1.628 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.561 1.600 mp_sum_l 7287 12.8 1.015 1.545 1.015 1.545 acc_transpose_blocks_sync 82296 16.5 1.443 1.540 1.443 1.540 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.505 1.518 cp_fm_diag_elpa_base 50 14.0 1.469 1.484 1.500 1.507 multiply_cannon_metrocomm3 27432 15.5 0.039 0.041 0.817 1.387 fft3d_ps 1201 14.6 0.591 0.651 1.358 1.383 grid_collocate_task_list 119 9.7 1.248 1.317 1.248 1.317 wfi_extrapolate 11 7.9 0.001 0.001 1.317 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.273 1.288 potential_pw2rs 119 12.3 0.009 0.009 1.258 1.260 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.254 cp_fm_upper_to_full 72 14.2 0.819 1.166 0.819 1.166 jit_kernel_multiply 6 16.1 0.435 1.090 0.435 1.090 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.069 1.069 dbcsr_complete_redistribute 329 12.2 0.120 0.146 0.762 1.039 transfer_rs2pw 487 10.6 0.004 0.005 0.861 0.942 make_images_data 4572 15.5 0.048 0.052 0.808 0.935 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.801 0.877 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.687 0.864 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.566 0.835 acc_transpose_blocks_kernels 27432 16.5 0.271 0.280 0.816 0.827 mp_alltoall_d11v 2130 13.8 0.715 0.811 0.715 0.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.810 multiply_cannon_metrocomm1 27432 15.5 0.035 0.036 0.196 0.788 mp_alltoall_i22 627 13.8 0.428 0.725 0.428 0.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.222000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.909091, yerr=3.232071 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 615.444480E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.048 28.211 28.219 qs_mol_dyn_low 1 2.0 0.003 0.003 28.027 28.042 qs_forces 11 3.9 0.002 0.003 27.920 27.921 qs_energies 11 4.9 0.001 0.001 26.217 26.219 scf_env_do_scf 11 5.9 0.000 0.001 21.455 21.455 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.022 19.022 velocity_verlet 10 3.0 0.001 0.002 14.586 14.592 dbcsr_multiply_generic 2286 12.5 0.095 0.098 12.757 12.817 qs_scf_new_mos 108 7.5 0.001 0.001 11.422 11.444 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.422 11.444 ot_scf_mini 108 9.5 0.002 0.002 10.722 10.747 multiply_cannon 2286 13.5 0.223 0.229 10.260 10.725 multiply_cannon_loop 2286 14.5 0.639 0.658 9.349 9.505 ot_mini 108 10.5 0.001 0.001 6.133 6.162 multiply_cannon_multrec 9144 15.5 1.645 1.834 5.898 6.105 rebuild_ks_matrix 119 8.3 0.000 0.000 5.739 5.763 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.739 5.763 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.104 5.125 qs_ot_get_derivative 108 11.5 0.001 0.001 4.858 4.883 dbcsr_mm_accdrv_process 12550 15.8 3.634 4.173 4.143 4.216 sum_up_and_integrate 119 10.3 0.001 0.001 3.397 3.403 integrate_v_rspace 119 11.3 0.003 0.003 3.387 3.393 init_scf_run 11 5.9 0.000 0.001 3.338 3.338 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.338 3.338 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.331 3.337 calculate_rho_elec 119 8.7 0.060 0.061 3.331 3.336 qs_ot_get_p 119 10.4 0.001 0.001 2.725 2.753 init_scf_loop 11 6.9 0.000 0.000 2.414 2.416 calculate_first_density_matrix 1 7.0 0.000 0.000 2.213 2.215 pw_transfer 1439 11.6 0.065 0.068 2.078 2.087 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.984 1.995 grid_integrate_task_list 119 12.3 1.867 1.931 1.867 1.931 make_m2s 4572 13.5 0.033 0.035 1.738 1.895 density_rs2pw 119 9.7 0.003 0.003 1.716 1.806 make_images 4572 14.5 0.269 0.300 1.649 1.804 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.770 1.773 calculate_dm_sparse 119 9.5 0.000 0.000 1.730 1.746 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.698 1.725 mp_waitall_1 115863 16.7 1.231 1.720 1.231 1.720 prepare_preconditioner 11 7.9 0.000 0.000 1.693 1.696 make_preconditioner 11 8.9 0.000 0.000 1.693 1.696 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.656 1.668 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.585 1.614 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.531 1.531 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.519 1.520 acc_transpose_blocks 9144 15.5 0.043 0.044 1.413 1.448 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.359 1.369 grid_collocate_task_list 119 9.7 1.295 1.366 1.295 1.366 jit_kernel_multiply 8 15.8 0.470 1.356 0.470 1.356 fft3d_ps 1201 14.6 0.649 0.663 1.337 1.348 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.313 1.325 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.260 1.269 cp_fm_diag_elpa_base 50 14.0 1.233 1.249 1.259 1.267 ot_diis_step 108 11.5 0.013 0.013 1.264 1.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.213 1.250 apply_single 119 13.6 0.000 0.000 1.213 1.250 potential_pw2rs 119 12.3 0.010 0.011 1.215 1.217 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.216 1.217 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.125 1.130 wfi_extrapolate 11 7.9 0.001 0.001 1.072 1.072 hybrid_alltoall_any 4725 16.4 0.064 0.175 0.745 0.968 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.933 make_images_data 4572 15.5 0.042 0.046 0.750 0.922 mp_alltoall_d11v 2130 13.8 0.754 0.843 0.754 0.843 cp_fm_cholesky_invert 11 10.9 0.834 0.837 0.834 0.837 transfer_rs2pw 487 10.6 0.004 0.005 0.738 0.832 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.807 0.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.755 0.757 acc_transpose_blocks_sync 27432 16.5 0.716 0.746 0.716 0.746 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.663 0.707 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.326 0.647 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.638 0.643 mp_allgather_i34 2286 14.5 0.227 0.641 0.227 0.641 transfer_pw2rs 487 13.2 0.003 0.004 0.626 0.628 mp_alltoall_z22v 1201 16.6 0.562 0.600 0.562 0.600 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.219000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=582.636364, yerr=6.270751 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 787.529728E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 42.417 42.418 qs_mol_dyn_low 1 2.0 0.003 0.003 42.222 42.230 qs_forces 11 3.9 0.002 0.002 42.138 42.140 qs_energies 11 4.9 0.001 0.001 40.102 40.107 scf_env_do_scf 11 5.9 0.001 0.001 34.496 34.496 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.803 26.805 velocity_verlet 10 3.0 0.028 0.031 23.959 23.964 dbcsr_multiply_generic 2286 12.5 0.103 0.105 18.827 19.083 qs_scf_new_mos 108 7.5 0.001 0.001 17.134 17.230 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.133 17.229 ot_scf_mini 108 9.5 0.002 0.002 15.984 16.087 multiply_cannon 2286 13.5 0.295 0.302 14.901 15.900 multiply_cannon_loop 2286 14.5 0.863 0.879 13.658 14.697 ot_mini 108 10.5 0.001 0.001 9.758 9.876 multiply_cannon_multrec 9144 15.5 3.382 4.750 8.744 8.818 qs_ot_get_derivative 108 11.5 0.001 0.001 7.717 7.821 init_scf_loop 11 6.9 0.000 0.000 7.667 7.673 rebuild_ks_matrix 119 8.3 0.000 0.000 7.264 7.407 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.263 7.407 prepare_preconditioner 11 7.9 0.000 0.000 6.701 6.714 make_preconditioner 11 8.9 0.000 0.000 6.700 6.714 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.540 6.670 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.296 6.595 dbcsr_mm_accdrv_process 12550 15.8 4.736 6.482 5.226 6.515 cp_fm_upper_to_full 72 14.2 3.144 4.493 3.144 4.493 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.189 4.195 calculate_rho_elec 119 8.7 0.117 0.120 4.189 4.195 sum_up_and_integrate 119 10.3 0.001 0.001 3.844 3.851 integrate_v_rspace 119 11.3 0.003 0.003 3.834 3.840 init_scf_run 11 5.9 0.000 0.001 3.538 3.538 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.538 3.538 qs_ot_get_p 119 10.4 0.001 0.001 3.313 3.456 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.741 3.197 mp_waitall_1 94719 16.7 2.154 3.190 2.154 3.190 pw_transfer 1439 11.6 0.068 0.069 3.017 3.023 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.918 2.925 dbcsr_complete_redistribute 329 12.2 0.286 0.292 1.932 2.727 fft_wrap_pw1pw2_140 487 13.2 0.651 0.657 2.499 2.506 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.237 2.469 apply_single 119 13.6 0.000 0.000 2.237 2.469 make_m2s 4572 13.5 0.037 0.038 2.287 2.435 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.627 2.414 make_images 4572 14.5 0.356 0.388 2.167 2.315 calculate_dm_sparse 119 9.5 0.000 0.000 2.225 2.240 density_rs2pw 119 9.7 0.003 0.003 2.217 2.229 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.268 2.228 mp_alltoall_i22 627 13.8 1.377 2.206 1.377 2.206 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.399 2.179 grid_integrate_task_list 119 12.3 2.075 2.111 2.075 2.111 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.039 2.090 calculate_first_density_matrix 1 7.0 0.000 0.000 2.072 2.075 ot_diis_step 108 11.5 0.014 0.014 2.015 2.015 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.949 1.951 fft3d_ps 1201 14.6 0.870 0.887 1.835 1.844 acc_transpose_blocks 9144 15.5 0.045 0.046 1.814 1.842 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.797 1.797 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.790 1.793 mp_sum_l 7287 12.8 1.090 1.757 1.090 1.757 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.596 1.596 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.543 1.588 grid_collocate_task_list 119 9.7 1.518 1.537 1.518 1.537 cp_fm_cholesky_invert 11 10.9 1.405 1.408 1.405 1.408 wfi_extrapolate 11 7.9 0.001 0.001 1.399 1.399 potential_pw2rs 119 12.3 0.013 0.014 1.353 1.355 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.304 1.325 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.313 1.313 cp_fm_diag_elpa_base 50 14.0 1.171 1.222 1.312 1.312 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.242 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.037 1.214 mp_alltoall_d11v 2130 13.8 1.166 1.201 1.166 1.201 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.117 1.141 make_images_data 4572 15.5 0.046 0.050 0.958 1.140 acc_transpose_blocks_sync 27432 16.5 1.109 1.139 1.109 1.139 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.993 1.019 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.945 qs_create_task_list 11 7.9 0.001 0.001 0.927 0.939 generate_qs_task_list 11 8.9 0.368 0.386 0.927 0.938 mp_alltoall_z22v 1201 16.6 0.828 0.851 0.828 0.851 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.418000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=735.545455, yerr=17.655098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.497280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.028 80.592 80.593 qs_mol_dyn_low 1 2.0 0.003 0.003 80.345 80.355 qs_forces 11 3.9 0.003 0.004 80.262 80.263 qs_energies 11 4.9 0.002 0.002 77.426 77.446 scf_env_do_scf 11 5.9 0.000 0.001 68.692 68.694 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.299 63.299 dbcsr_multiply_generic 2055 12.4 0.105 0.110 50.461 50.832 qs_scf_new_mos 99 7.5 0.000 0.001 46.495 46.647 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.494 46.646 ot_scf_mini 99 9.5 0.002 0.002 44.161 44.289 multiply_cannon 2055 13.4 0.184 0.189 42.278 43.190 velocity_verlet 10 3.0 0.001 0.001 42.570 42.570 multiply_cannon_loop 2055 14.4 1.758 1.803 41.324 42.350 ot_mini 99 10.5 0.001 0.001 26.107 26.211 qs_ot_get_derivative 99 11.5 0.001 0.001 19.257 19.407 multiply_cannon_multrec 49320 15.4 11.509 12.343 17.433 18.198 rebuild_ks_matrix 110 8.3 0.000 0.001 14.351 14.582 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.351 14.582 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.592 12.810 mp_waitall_1 220248 16.4 10.349 11.477 10.349 11.477 multiply_cannon_sync_h2d 49320 15.4 9.737 10.443 9.737 10.443 qs_ot_get_p 110 10.4 0.001 0.001 9.563 9.745 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.244 7.789 multiply_cannon_metrocomm3 49320 15.4 0.081 0.087 6.328 7.773 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.239 7.757 apply_single 110 13.6 0.000 0.001 7.239 7.757 sum_up_and_integrate 110 10.3 0.002 0.003 6.871 6.883 integrate_v_rspace 110 11.3 0.002 0.003 6.846 6.864 init_scf_run 11 5.9 0.000 0.001 6.690 6.691 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.690 6.690 ot_diis_step 99 11.5 0.005 0.006 6.608 6.608 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.399 6.422 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.199 6.346 calculate_rho_elec 110 8.6 0.020 0.024 6.199 6.346 dbcsr_mm_accdrv_process 87628 16.1 2.964 3.113 5.793 6.088 cp_dbcsr_syevd 48 12.0 0.003 0.004 5.499 5.499 init_scf_loop 11 6.9 0.000 0.000 5.367 5.367 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.252 5.308 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.950 4.983 cp_fm_diag_elpa_base 48 14.0 4.936 4.969 4.947 4.980 mp_sum_l 6594 12.7 3.931 4.571 3.931 4.571 wfi_extrapolate 11 7.9 0.001 0.001 4.002 4.002 make_m2s 4110 13.4 0.060 0.065 3.790 3.918 make_images 4110 14.4 0.177 0.191 3.695 3.827 calculate_dm_sparse 110 9.5 0.001 0.001 3.704 3.805 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.529 3.532 density_rs2pw 110 9.6 0.004 0.005 3.211 3.422 grid_integrate_task_list 110 12.3 3.250 3.412 3.250 3.412 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.212 3.288 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.253 3.264 prepare_preconditioner 11 7.9 0.000 0.000 3.217 3.240 make_preconditioner 11 8.9 0.000 0.000 3.217 3.240 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.184 3.217 pw_transfer 1331 11.6 0.055 0.069 2.973 3.049 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.989 3.037 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.885 2.962 acc_transpose_blocks 49320 15.4 0.207 0.218 2.505 2.605 mp_alltoall_d11v 2046 13.8 2.067 2.593 2.067 2.593 calculate_first_density_matrix 1 7.0 0.000 0.000 2.582 2.590 fft_wrap_pw1pw2_140 451 13.1 0.310 0.335 2.486 2.588 jit_kernel_multiply 13 15.9 2.551 2.572 2.551 2.572 potential_pw2rs 110 12.3 0.005 0.006 2.492 2.507 grid_collocate_task_list 110 9.6 2.163 2.298 2.163 2.298 fft3d_ps 1111 14.6 0.796 0.888 2.176 2.227 mp_waitany 14300 13.8 1.785 2.175 1.785 2.175 transfer_rs2pw 451 10.6 0.005 0.006 1.808 2.010 mp_sum_d 3891 11.9 1.395 1.919 1.395 1.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.888 1.908 make_images_data 4110 15.4 0.043 0.048 1.737 1.873 hybrid_alltoall_any 4261 16.3 0.084 0.485 1.523 1.827 cp_fm_cholesky_invert 11 10.9 1.757 1.761 1.757 1.761 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.668 1.695 transfer_pw2rs 451 13.1 0.006 0.007 1.685 1.689 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.593000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.545455, yerr=2.995865 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 592.474112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3120195. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.029 69.133 69.134 qs_mol_dyn_low 1 2.0 0.003 0.003 68.912 68.922 qs_forces 11 3.9 0.003 0.003 68.844 68.845 qs_energies 11 4.9 0.001 0.002 65.536 65.539 scf_env_do_scf 11 5.9 0.000 0.001 57.096 57.100 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.372 49.372 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.943 38.095 velocity_verlet 10 3.0 0.001 0.001 36.241 36.242 qs_scf_new_mos 99 7.5 0.001 0.001 33.491 33.601 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.490 33.600 multiply_cannon 2055 13.4 0.224 0.244 31.258 32.360 ot_scf_mini 99 9.5 0.003 0.003 31.812 31.912 multiply_cannon_loop 2055 14.4 1.155 1.187 30.077 30.915 ot_mini 99 10.5 0.001 0.001 18.490 18.590 multiply_cannon_multrec 24660 15.4 6.983 8.991 14.040 15.775 rebuild_ks_matrix 110 8.3 0.000 0.001 13.534 13.632 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.534 13.632 qs_ot_get_derivative 99 11.5 0.001 0.001 12.709 12.813 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.919 12.001 mp_waitall_1 176588 16.5 7.600 10.357 7.600 10.357 multiply_cannon_metrocomm3 24660 15.4 0.074 0.075 5.149 8.651 init_scf_loop 11 6.9 0.000 0.000 7.689 7.690 multiply_cannon_sync_h2d 24660 15.4 6.412 7.437 6.412 7.437 dbcsr_mm_accdrv_process 52282 16.1 5.432 6.345 6.886 7.215 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.476 7.102 apply_single 110 13.6 0.000 0.001 6.476 7.102 qs_ot_get_p 110 10.4 0.001 0.001 6.654 6.781 sum_up_and_integrate 110 10.3 0.001 0.002 6.406 6.418 integrate_v_rspace 110 11.3 0.002 0.003 6.379 6.392 init_scf_run 11 5.9 0.000 0.001 6.071 6.071 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.071 6.071 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.911 5.921 calculate_rho_elec 110 8.6 0.039 0.047 5.910 5.920 ot_diis_step 99 11.5 0.010 0.010 5.733 5.733 prepare_preconditioner 11 7.9 0.000 0.000 5.674 5.696 make_preconditioner 11 8.9 0.000 0.000 5.674 5.696 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.753 5.492 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.251 5.402 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.736 4.756 make_m2s 4110 13.4 0.057 0.059 4.254 4.592 make_images 4110 14.4 0.407 0.466 4.142 4.477 cp_dbcsr_syevd 48 12.0 0.013 0.033 4.264 4.264 pw_transfer 1331 11.6 0.066 0.072 3.605 3.743 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.666 3.694 cp_fm_diag_elpa_base 48 14.0 3.599 3.618 3.644 3.661 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.499 3.637 wfi_extrapolate 11 7.9 0.001 0.001 3.504 3.504 density_rs2pw 110 9.6 0.004 0.004 3.170 3.377 grid_integrate_task_list 110 12.3 3.142 3.367 3.142 3.367 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.316 3.367 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.226 3.227 fft_wrap_pw1pw2_140 451 13.1 0.357 0.376 2.964 3.108 calculate_dm_sparse 110 9.5 0.001 0.001 2.985 3.015 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.879 2.900 fft3d_ps 1111 14.6 1.115 1.350 2.591 2.749 make_images_data 4110 15.4 0.048 0.053 2.311 2.679 cp_fm_cholesky_invert 11 10.9 2.637 2.644 2.637 2.644 hybrid_alltoall_any 4261 16.3 0.104 0.451 2.017 2.605 calculate_first_density_matrix 1 7.0 0.000 0.000 2.467 2.471 mp_sum_l 6594 12.7 1.798 2.400 1.798 2.400 potential_pw2rs 110 12.3 0.008 0.009 2.297 2.308 grid_collocate_task_list 110 9.6 2.166 2.299 2.166 2.299 mp_alltoall_d11v 2046 13.8 1.811 2.206 1.811 2.206 jit_kernel_multiply 11 16.3 1.088 2.145 1.088 2.145 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.985 2.006 acc_transpose_blocks 24660 15.4 0.117 0.119 1.960 2.002 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.859 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.771 1.783 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.769 1.779 transfer_rs2pw 451 10.6 0.006 0.007 1.441 1.672 mp_irecv_dv 57340 16.2 0.646 1.564 0.646 1.564 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.543 1.555 mp_allgather_i34 2055 14.4 0.489 1.514 0.489 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.491 cp_fm_cholesky_decompose 22 10.9 1.393 1.400 1.393 1.400 dbcsr_complete_redistribute 325 12.2 0.231 0.291 1.130 1.397 mp_waitany 10164 13.8 1.158 1.390 1.158 1.390 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.134000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.636364, yerr=6.498569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 669.507584E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.029 59.156 59.157 qs_mol_dyn_low 1 2.0 0.003 0.003 58.833 58.843 qs_forces 11 3.9 0.003 0.003 58.768 58.768 qs_energies 11 4.9 0.001 0.002 55.601 55.605 scf_env_do_scf 11 5.9 0.001 0.001 47.764 47.764 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.329 39.330 velocity_verlet 10 3.0 0.001 0.002 32.216 32.217 dbcsr_multiply_generic 2055 12.4 0.109 0.112 28.525 28.769 qs_scf_new_mos 99 7.5 0.001 0.001 24.819 24.909 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.818 24.909 ot_scf_mini 99 9.5 0.002 0.003 23.576 23.683 multiply_cannon 2055 13.4 0.212 0.222 22.430 23.493 multiply_cannon_loop 2055 14.4 0.812 0.848 21.289 22.356 ot_mini 99 10.5 0.001 0.001 13.570 13.676 rebuild_ks_matrix 110 8.3 0.000 0.000 12.045 12.177 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.045 12.177 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.584 10.702 multiply_cannon_multrec 16440 15.4 3.677 4.656 9.846 10.699 mp_waitall_1 139946 16.5 6.744 10.029 6.744 10.029 qs_ot_get_derivative 99 11.5 0.001 0.001 9.157 9.265 init_scf_loop 11 6.9 0.000 0.000 8.403 8.404 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.290 7.388 prepare_preconditioner 11 7.9 0.000 0.000 6.665 6.684 make_preconditioner 11 8.9 0.000 0.000 6.665 6.684 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.006 6.354 sum_up_and_integrate 110 10.3 0.001 0.002 6.221 6.238 integrate_v_rspace 110 11.3 0.003 0.003 6.195 6.211 dbcsr_mm_accdrv_process 34862 16.1 5.369 5.637 6.013 6.208 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.699 5.705 calculate_rho_elec 110 8.6 0.059 0.059 5.698 5.705 init_scf_run 11 5.9 0.000 0.001 5.454 5.454 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.454 5.454 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.894 5.357 apply_single 110 13.6 0.000 0.000 4.894 5.357 qs_ot_get_p 110 10.4 0.001 0.001 5.203 5.356 make_m2s 4110 13.4 0.049 0.051 4.069 4.434 ot_diis_step 99 11.5 0.011 0.011 4.385 4.385 make_images 4110 14.4 0.400 0.523 3.952 4.317 multiply_cannon_sync_h2d 16440 15.4 3.262 4.011 3.262 4.011 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.148 3.781 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.608 3.613 pw_transfer 1331 11.6 0.065 0.070 3.577 3.586 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.470 3.482 grid_integrate_task_list 110 12.3 3.203 3.361 3.203 3.361 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.244 3.244 density_rs2pw 110 9.6 0.004 0.004 2.927 3.161 fft_wrap_pw1pw2_140 451 13.1 0.460 0.469 3.021 3.034 wfi_extrapolate 11 7.9 0.001 0.001 2.937 2.937 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.810 2.811 make_images_data 4110 15.4 0.046 0.051 2.279 2.742 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.721 2.730 cp_fm_diag_elpa_base 48 14.0 2.655 2.684 2.719 2.729 hybrid_alltoall_any 4261 16.3 0.108 0.378 2.040 2.702 cp_fm_cholesky_invert 11 10.9 2.558 2.564 2.558 2.564 calculate_dm_sparse 110 9.5 0.001 0.001 2.512 2.547 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.489 2.536 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.403 2.460 calculate_first_density_matrix 1 7.0 0.000 0.000 2.426 2.428 fft3d_ps 1111 14.6 1.119 1.133 2.387 2.400 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.877 2.393 grid_collocate_task_list 110 9.6 2.220 2.389 2.220 2.389 mp_irecv_dv 48980 15.7 0.803 2.259 0.803 2.259 mp_alltoall_d11v 2046 13.8 1.757 2.188 1.757 2.188 potential_pw2rs 110 12.3 0.011 0.011 2.066 2.073 mp_sum_l 6594 12.7 1.399 1.989 1.399 1.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.955 1.956 dbcsr_complete_redistribute 325 12.2 0.330 0.358 1.420 1.866 cp_fm_upper_to_full 70 14.2 1.397 1.764 1.397 1.764 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.652 1.670 acc_transpose_blocks 16440 15.4 0.079 0.083 1.600 1.660 cp_fm_cholesky_decompose 22 10.9 1.555 1.578 1.555 1.578 transfer_rs2pw 451 10.6 0.005 0.006 1.240 1.474 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.452 1.464 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.358 1.455 mp_allgather_i34 2055 14.4 0.452 1.434 0.452 1.434 copy_fm_to_dbcsr 174 11.2 0.001 0.002 0.951 1.387 mp_waitany 17072 13.8 1.115 1.324 1.115 1.324 rs_gather_matrices 110 12.3 0.137 0.150 0.875 1.296 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.265 1.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.157000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=634.000000, yerr=9.496411 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 735.248384E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.031 65.094 65.096 qs_mol_dyn_low 1 2.0 0.003 0.003 64.871 64.882 qs_forces 11 3.9 0.003 0.003 64.804 64.805 qs_energies 11 4.9 0.001 0.002 61.410 61.414 scf_env_do_scf 11 5.9 0.000 0.001 53.220 53.222 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.663 41.664 velocity_verlet 10 3.0 0.001 0.001 37.107 37.109 dbcsr_multiply_generic 2055 12.4 0.116 0.123 30.448 30.641 qs_scf_new_mos 99 7.5 0.001 0.001 27.077 27.189 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.076 27.189 ot_scf_mini 99 9.5 0.002 0.003 25.401 25.522 multiply_cannon 2055 13.4 0.238 0.256 23.363 24.272 multiply_cannon_loop 2055 14.4 1.413 1.476 22.039 22.642 ot_mini 99 10.5 0.001 0.001 14.658 14.802 multiply_cannon_multrec 24660 15.4 4.101 6.744 12.997 14.217 rebuild_ks_matrix 110 8.3 0.000 0.000 11.964 12.067 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.023 11.963 12.067 init_scf_loop 11 6.9 0.000 0.000 11.519 11.519 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.557 10.647 qs_ot_get_derivative 99 11.5 0.001 0.001 10.498 10.626 prepare_preconditioner 11 7.9 0.000 0.000 9.765 9.781 make_preconditioner 11 8.9 0.000 0.000 9.765 9.781 dbcsr_mm_accdrv_process 52304 16.0 7.673 8.927 8.738 9.651 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.982 9.458 sum_up_and_integrate 110 10.3 0.001 0.002 6.207 6.219 integrate_v_rspace 110 11.3 0.003 0.003 6.181 6.193 mp_waitall_1 121746 16.5 4.223 6.095 4.223 6.095 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.840 5.847 calculate_rho_elec 110 8.6 0.077 0.081 5.839 5.847 qs_ot_get_p 110 10.4 0.001 0.001 5.641 5.810 make_m2s 4110 13.4 0.059 0.061 5.493 5.793 make_images 4110 14.4 0.575 0.696 5.353 5.648 init_scf_run 11 5.9 0.000 0.001 5.466 5.466 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.465 5.466 cp_fm_upper_to_full 70 14.2 3.378 4.898 3.378 4.898 ot_diis_step 99 11.5 0.011 0.011 4.122 4.122 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.042 4.118 apply_single 110 13.6 0.000 0.000 4.042 4.118 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.787 3.802 dbcsr_complete_redistribute 325 12.2 0.422 0.467 2.633 3.792 pw_transfer 1331 11.6 0.065 0.073 3.743 3.776 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.637 3.672 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.476 3.541 grid_integrate_task_list 110 12.3 3.272 3.437 3.272 3.437 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.305 3.373 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.130 3.264 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.232 3.232 fft_wrap_pw1pw2_140 451 13.1 0.531 0.544 3.135 3.173 density_rs2pw 110 9.6 0.004 0.004 2.951 3.147 make_images_data 4110 15.4 0.048 0.051 2.782 3.111 hybrid_alltoall_any 4261 16.3 0.122 0.460 2.338 3.087 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 3.004 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.312 2.999 wfi_extrapolate 11 7.9 0.001 0.001 2.992 2.992 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.774 2.894 mp_alltoall_i22 605 13.7 1.659 2.857 1.659 2.857 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.782 2.784 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.712 2.723 cp_fm_diag_elpa_base 48 14.0 2.566 2.628 2.710 2.721 cp_fm_cholesky_invert 11 10.9 2.600 2.608 2.600 2.608 multiply_cannon_sync_h2d 24660 15.4 2.366 2.544 2.366 2.544 acc_transpose_blocks 24660 15.4 0.113 0.115 2.413 2.514 fft3d_ps 1111 14.6 1.150 1.184 2.463 2.476 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.416 2.451 grid_collocate_task_list 110 9.6 2.261 2.398 2.261 2.398 calculate_first_density_matrix 1 7.0 0.000 0.000 2.377 2.381 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.234 2.235 potential_pw2rs 110 12.3 0.012 0.013 2.025 2.029 mp_alltoall_d11v 2046 13.8 1.769 1.972 1.769 1.972 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.781 1.817 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.645 1.658 cp_fm_cholesky_decompose 22 10.9 1.605 1.649 1.605 1.649 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.557 multiply_cannon_metrocomm4 20550 15.4 0.058 0.062 0.849 1.540 acc_transpose_blocks_sync 73980 16.4 1.409 1.508 1.409 1.508 mp_sum_l 6594 12.7 0.900 1.463 0.900 1.463 mp_irecv_dv 62702 16.1 0.746 1.461 0.746 1.461 mp_allgather_i34 2055 14.4 0.417 1.412 0.417 1.412 transfer_rs2pw 451 10.6 0.005 0.006 1.157 1.382 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.096000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.636364, yerr=8.445254 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 851.390464E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 54.711 54.713 qs_mol_dyn_low 1 2.0 0.003 0.003 54.492 54.502 qs_forces 11 3.9 0.003 0.003 54.426 54.427 qs_energies 11 4.9 0.001 0.002 50.791 50.797 scf_env_do_scf 11 5.9 0.000 0.001 42.525 42.525 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.908 34.909 velocity_verlet 10 3.0 0.001 0.001 30.894 30.897 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.136 23.307 qs_scf_new_mos 99 7.5 0.001 0.001 20.438 20.515 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.438 20.514 ot_scf_mini 99 9.5 0.002 0.002 19.202 19.257 multiply_cannon 2055 13.4 0.239 0.254 17.724 18.843 multiply_cannon_loop 2055 14.4 0.601 0.626 16.504 16.844 rebuild_ks_matrix 110 8.3 0.000 0.000 11.569 11.641 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.569 11.641 ot_mini 99 10.5 0.001 0.001 10.732 10.789 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.288 10.352 multiply_cannon_multrec 8220 15.4 3.158 4.781 7.561 8.816 mp_waitall_1 103326 16.6 5.890 7.590 5.890 7.590 init_scf_loop 11 6.9 0.000 0.000 7.572 7.573 qs_ot_get_derivative 99 11.5 0.001 0.001 6.964 7.020 sum_up_and_integrate 110 10.3 0.001 0.002 6.173 6.183 integrate_v_rspace 110 11.3 0.003 0.003 6.146 6.157 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.949 5.965 calculate_rho_elec 110 8.6 0.112 0.113 5.949 5.964 prepare_preconditioner 11 7.9 0.000 0.000 5.940 5.945 make_preconditioner 11 8.9 0.000 0.000 5.940 5.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.526 5.601 dbcsr_mm_accdrv_process 17442 15.9 3.121 3.978 4.264 5.204 init_scf_run 11 5.9 0.000 0.001 5.144 5.144 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.144 5.144 qs_ot_get_p 110 10.4 0.001 0.001 4.547 4.614 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.177 4.528 make_m2s 4110 13.4 0.038 0.039 4.071 4.266 make_images 4110 14.4 0.646 0.707 3.940 4.132 pw_transfer 1331 11.6 0.066 0.070 4.058 4.068 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.950 3.963 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.793 3.854 apply_single 110 13.6 0.000 0.000 3.793 3.854 ot_diis_step 99 11.5 0.012 0.014 3.747 3.747 grid_integrate_task_list 110 12.3 3.377 3.530 3.377 3.530 fft_wrap_pw1pw2_140 451 13.1 0.719 0.734 3.439 3.455 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.210 3.213 density_rs2pw 110 9.6 0.004 0.004 2.986 3.137 cp_dbcsr_syevd 48 12.0 0.005 0.005 2.876 2.877 cp_fm_cholesky_invert 11 10.9 2.819 2.823 2.819 2.823 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.675 2.675 wfi_extrapolate 11 7.9 0.001 0.001 2.671 2.671 grid_collocate_task_list 110 9.6 2.363 2.532 2.363 2.532 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.514 2.516 fft3d_ps 1111 14.6 1.291 1.300 2.493 2.510 calculate_dm_sparse 110 9.5 0.001 0.001 2.472 2.507 hybrid_alltoall_any 4261 16.3 0.200 0.848 2.139 2.501 multiply_cannon_sync_h2d 8220 15.4 2.357 2.494 2.357 2.494 make_images_data 4110 15.4 0.040 0.046 2.158 2.419 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.375 2.382 cp_fm_diag_elpa_base 48 14.0 2.320 2.342 2.373 2.381 calculate_first_density_matrix 1 7.0 0.000 0.000 2.376 2.378 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.148 2.174 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.981 2.027 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.791 1.996 potential_pw2rs 110 12.3 0.015 0.015 1.984 1.988 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.879 1.903 mp_alltoall_d11v 2046 13.8 1.596 1.769 1.596 1.769 cp_fm_cholesky_decompose 22 10.9 1.662 1.680 1.662 1.680 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.496 1.616 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.608 1.615 dbcsr_complete_redistribute 325 12.2 0.569 0.591 1.421 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.437 1.449 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.419 mp_allgather_i34 2055 14.4 0.382 1.405 0.382 1.405 acc_transpose_blocks 8220 15.4 0.041 0.042 1.361 1.400 mp_irecv_dv 24056 15.7 0.455 1.373 0.455 1.373 qs_create_task_list 11 7.9 0.001 0.001 1.205 1.307 generate_qs_task_list 11 8.9 0.374 0.442 1.205 1.307 jit_kernel_multiply 7 15.5 0.828 1.267 0.828 1.267 transfer_rs2pw 451 10.6 0.005 0.005 1.086 1.260 mp_waitany 9240 13.8 1.036 1.222 1.036 1.222 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.150 1.192 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.750 1.105 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.713000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.363636, yerr=9.138060 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 1.364242E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.033 88.051 88.053 qs_mol_dyn_low 1 2.0 0.003 0.003 87.757 87.768 qs_forces 11 3.9 0.003 0.003 87.685 87.687 qs_energies 11 4.9 0.001 0.001 83.498 83.501 scf_env_do_scf 11 5.9 0.000 0.001 73.507 73.508 velocity_verlet 10 3.0 0.002 0.003 56.222 56.229 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.195 45.196 dbcsr_multiply_generic 2055 12.4 0.122 0.125 29.650 29.678 init_scf_loop 11 6.9 0.000 0.000 28.239 28.241 qs_scf_new_mos 99 7.5 0.001 0.001 27.154 27.190 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.153 27.189 prepare_preconditioner 11 7.9 0.000 0.000 26.212 26.219 make_preconditioner 11 8.9 0.000 0.000 26.211 26.219 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.408 25.695 ot_scf_mini 99 9.5 0.002 0.002 25.339 25.371 multiply_cannon 2055 13.4 0.331 0.354 22.428 23.173 multiply_cannon_loop 2055 14.4 0.826 0.841 20.632 21.065 cp_fm_upper_to_full 70 14.2 12.750 18.285 12.750 18.285 ot_mini 99 10.5 0.001 0.001 14.241 14.266 rebuild_ks_matrix 110 8.3 0.000 0.000 14.041 14.076 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.040 14.075 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.734 12.765 dbcsr_complete_redistribute 325 12.2 1.009 1.052 7.341 10.613 multiply_cannon_multrec 8220 15.4 4.058 4.274 9.735 9.807 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.348 9.634 qs_ot_get_derivative 99 11.5 0.001 0.001 9.568 9.600 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.788 9.037 mp_alltoall_i22 605 13.7 5.407 8.705 5.407 8.705 mp_waitall_1 84994 16.7 7.659 8.372 7.659 8.372 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.531 7.565 calculate_rho_elec 110 8.6 0.222 0.222 7.531 7.565 sum_up_and_integrate 110 10.3 0.002 0.002 6.957 6.971 integrate_v_rspace 110 11.3 0.003 0.003 6.929 6.943 make_m2s 4110 13.4 0.043 0.043 5.484 5.972 pw_transfer 1331 11.6 0.074 0.075 5.853 5.859 make_images 4110 14.4 0.880 0.920 5.295 5.782 qs_ot_get_p 110 10.4 0.001 0.001 5.749 5.780 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.736 5.741 dbcsr_mm_accdrv_process 11614 15.7 3.850 4.123 5.528 5.730 init_scf_run 11 5.9 0.000 0.001 5.693 5.693 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.693 5.693 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.861 5.324 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.810 5.249 apply_single 110 13.6 0.000 0.000 4.810 5.249 cp_fm_cholesky_invert 11 10.9 5.217 5.221 5.217 5.221 fft_wrap_pw1pw2_140 451 13.1 1.366 1.370 5.026 5.035 ot_diis_step 99 11.5 0.015 0.016 4.648 4.648 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.165 4.174 density_rs2pw 110 9.6 0.004 0.004 3.969 4.001 grid_integrate_task_list 110 12.3 3.708 3.780 3.708 3.780 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.681 3.683 hybrid_alltoall_any 4261 16.3 0.266 0.570 2.943 3.680 cp_dbcsr_syevd 48 12.0 0.054 0.067 3.647 3.648 make_images_data 4110 15.4 0.045 0.048 2.959 3.585 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.031 3.489 fft3d_ps 1111 14.6 1.883 1.896 3.441 3.448 wfi_extrapolate 11 7.9 0.001 0.001 3.339 3.339 calculate_dm_sparse 110 9.5 0.001 0.001 3.178 3.204 multiply_cannon_sync_h2d 8220 15.4 3.122 3.152 3.122 3.152 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.004 3.013 cp_fm_diag_elpa_base 48 14.0 2.456 2.657 2.989 2.989 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.866 2.870 grid_collocate_task_list 110 9.6 2.692 2.706 2.692 2.706 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.654 2.669 potential_pw2rs 110 12.3 0.021 0.021 2.459 2.463 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.373 2.382 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.197 2.265 calculate_first_density_matrix 1 7.0 0.000 0.000 2.235 2.235 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.215 mp_alltoall_d11v 2046 13.8 2.020 2.077 2.020 2.077 cp_fm_cholesky_decompose 22 10.9 1.956 1.976 1.956 1.976 qs_create_task_list 11 7.9 0.001 0.001 1.878 1.923 generate_qs_task_list 11 8.9 0.730 0.784 1.878 1.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.906 1.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.053000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1239.909091, yerr=49.424791 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 632.651776E+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 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529087. 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 0.009 0.029 202.258 202.260 qs_mol_dyn_low 1 2.0 0.003 0.003 201.839 201.856 qs_forces 11 3.9 0.005 0.005 201.723 201.723 qs_energies 11 4.9 0.003 0.009 196.136 196.155 scf_env_do_scf 11 5.9 0.001 0.001 179.749 179.754 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.428 159.430 dbcsr_multiply_generic 2507 12.6 0.177 0.183 122.861 123.305 velocity_verlet 10 3.0 0.001 0.002 121.964 121.965 qs_scf_new_mos 117 7.6 0.001 0.001 120.261 120.492 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.260 120.491 ot_scf_mini 117 9.6 0.003 0.003 113.700 113.954 multiply_cannon 2507 13.6 0.238 0.249 100.889 102.362 multiply_cannon_loop 2507 14.6 2.314 2.422 98.709 100.173 ot_mini 117 10.6 0.001 0.001 64.033 64.276 multiply_cannon_multrec 60168 15.6 32.509 35.304 41.967 43.902 qs_ot_get_derivative 117 11.6 0.001 0.001 38.974 39.217 rebuild_ks_matrix 128 8.3 0.001 0.001 33.585 34.116 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.584 34.115 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.155 30.618 mp_waitall_1 267128 16.5 27.041 30.195 27.041 30.195 multiply_cannon_sync_h2d 60168 15.6 27.313 30.027 27.313 30.027 qs_ot_get_p 128 10.4 0.001 0.001 27.675 28.049 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.537 25.229 apply_single 128 13.6 0.001 0.001 24.537 25.229 ot_diis_step 117 11.6 0.007 0.008 24.830 24.832 qs_ot_p2m_diag 83 11.4 0.079 0.092 21.239 21.351 init_scf_loop 11 6.9 0.000 0.000 20.247 20.248 cp_dbcsr_syevd 83 12.4 0.005 0.006 18.845 18.845 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 17.752 17.984 multiply_cannon_metrocomm3 60168 15.6 0.120 0.126 14.938 17.172 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.926 15.954 cp_fm_diag_elpa_base 83 14.4 15.861 15.890 15.922 15.951 prepare_preconditioner 11 7.9 0.000 0.000 15.670 15.708 make_preconditioner 11 8.9 0.000 0.000 15.670 15.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.929 15.115 make_m2s 5014 13.6 0.104 0.114 13.662 14.100 make_images 5014 14.6 0.396 0.415 13.478 13.929 sum_up_and_integrate 128 10.3 0.002 0.004 13.892 13.905 integrate_v_rspace 128 11.3 0.003 0.004 13.833 13.849 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.924 13.081 calculate_rho_elec 128 8.7 0.045 0.064 12.923 13.080 init_scf_run 11 5.9 0.000 0.001 12.245 12.246 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.245 12.245 dbcsr_mm_accdrv_process 124484 16.2 4.405 4.760 8.994 9.611 wfi_extrapolate 11 7.9 0.001 0.001 9.065 9.065 cp_fm_cholesky_invert 11 10.9 8.973 8.980 8.973 8.980 mp_sum_l 7950 12.9 7.031 8.614 7.031 8.614 calculate_dm_sparse 128 9.5 0.001 0.001 8.372 8.462 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.112 8.255 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.743 7.836 make_images_data 5014 15.6 0.066 0.072 6.695 7.696 pw_transfer 1547 11.6 0.073 0.090 7.499 7.676 multiply_cannon_metrocomm1 60168 15.6 0.098 0.103 5.736 7.644 grid_integrate_task_list 128 12.3 7.041 7.499 7.041 7.499 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.297 7.464 density_rs2pw 128 9.7 0.006 0.007 6.636 7.306 hybrid_alltoall_any 5200 16.5 0.293 2.271 5.846 7.113 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.656 6.667 fft_wrap_pw1pw2_140 523 13.2 0.859 0.898 6.445 6.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.806 5.952 fft3d_ps 1291 14.7 2.192 2.855 5.299 5.633 mp_alltoall_d11v 2415 14.1 4.196 5.308 4.196 5.308 grid_collocate_task_list 128 9.7 4.842 5.296 4.842 5.296 cp_fm_cholesky_decompose 22 10.9 4.531 4.543 4.531 4.543 potential_pw2rs 128 12.3 0.009 0.010 4.411 4.428 mp_sum_d 4472 12.1 3.398 4.426 3.398 4.426 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.260000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.818182, yerr=6.952020 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 846.913536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5051614. 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 0.014 0.032 187.318 187.322 qs_mol_dyn_low 1 2.0 0.003 0.006 186.933 186.947 qs_forces 11 3.9 0.004 0.005 186.844 186.848 qs_energies 11 4.9 0.002 0.009 180.120 180.133 scf_env_do_scf 11 5.9 0.001 0.001 163.690 163.700 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.229 131.232 velocity_verlet 10 3.0 0.001 0.002 117.715 117.717 dbcsr_multiply_generic 2507 12.6 0.188 0.194 96.644 97.924 qs_scf_new_mos 117 7.6 0.001 0.001 93.111 93.696 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.111 93.695 ot_scf_mini 117 9.6 0.004 0.004 88.325 88.978 multiply_cannon 2507 13.6 0.501 0.557 76.708 80.891 multiply_cannon_loop 2507 14.6 1.560 1.632 73.850 76.799 ot_mini 117 10.6 0.001 0.001 49.808 50.409 mp_waitall_1 214728 16.6 23.697 36.982 23.697 36.982 multiply_cannon_multrec 30084 15.6 21.040 24.921 31.624 35.833 rebuild_ks_matrix 128 8.3 0.001 0.001 31.980 32.604 qs_ks_build_kohn_sham_matrix 128 9.3 0.049 0.092 31.979 32.603 init_scf_loop 11 6.9 0.001 0.003 32.373 32.374 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.757 29.319 multiply_cannon_metrocomm3 30084 15.6 0.099 0.104 15.522 28.851 qs_ot_get_derivative 117 11.6 0.001 0.002 27.902 28.532 prepare_preconditioner 11 7.9 0.000 0.000 28.059 28.142 make_preconditioner 11 8.9 0.000 0.000 28.059 28.142 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.734 27.271 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.032 23.205 apply_single 128 13.6 0.001 0.001 22.032 23.205 ot_diis_step 117 11.6 0.014 0.015 21.728 21.730 qs_ot_get_p 128 10.4 0.001 0.001 20.730 21.483 multiply_cannon_sync_h2d 30084 15.6 18.081 21.011 18.081 21.011 cp_fm_cholesky_invert 11 10.9 16.290 16.302 16.290 16.302 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.918 15.951 make_m2s 5014 13.6 0.087 0.094 14.060 15.146 make_images 5014 14.6 1.166 1.374 13.844 14.932 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.726 14.726 sum_up_and_integrate 128 10.3 0.002 0.004 13.864 13.892 integrate_v_rspace 128 11.3 0.003 0.003 13.804 13.834 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.237 13.269 calculate_rho_elec 128 8.7 0.087 0.103 13.236 13.269 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.587 11.627 cp_fm_diag_elpa_base 83 14.4 11.337 11.432 11.581 11.616 init_scf_run 11 5.9 0.000 0.001 11.605 11.607 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.605 11.607 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.041 11.517 multiply_cannon_metrocomm4 27577 15.6 0.098 0.112 3.829 10.885 dbcsr_mm_accdrv_process 62242 16.2 5.393 6.053 10.034 10.624 mp_irecv_dv 69486 16.3 3.630 10.477 3.630 10.477 make_images_data 5014 15.6 0.067 0.077 8.205 9.622 hybrid_alltoall_any 5200 16.5 0.348 1.545 6.995 8.955 pw_transfer 1547 11.6 0.085 0.094 8.493 8.555 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.271 8.335 wfi_extrapolate 11 7.9 0.001 0.001 8.280 8.280 grid_integrate_task_list 128 12.3 7.177 7.575 7.177 7.575 fft_wrap_pw1pw2_140 523 13.2 0.934 0.965 7.351 7.422 density_rs2pw 128 9.7 0.006 0.006 6.951 7.318 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.242 7.024 cp_fm_cholesky_decompose 22 10.9 6.802 6.897 6.802 6.897 calculate_dm_sparse 128 9.5 0.001 0.001 6.439 6.588 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.260 6.269 mp_sum_l 7950 12.9 4.087 6.235 4.087 6.235 fft3d_ps 1291 14.7 2.828 3.003 5.826 5.887 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.360 5.532 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.353 5.424 grid_collocate_task_list 128 9.7 5.007 5.362 5.007 5.362 mp_alltoall_d11v 2415 14.1 4.108 4.603 4.108 4.603 potential_pw2rs 128 12.3 0.015 0.018 4.387 4.404 mp_allgather_i34 2507 14.6 1.376 4.368 1.376 4.368 mp_sum_d 4467 12.1 2.624 3.985 2.624 3.985 dbcsr_complete_redistribute 395 12.7 0.779 0.857 3.145 3.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.322000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=805.909091, yerr=3.527929 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 936.431616E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 173.598 173.600 qs_mol_dyn_low 1 2.0 0.003 0.003 173.171 173.225 qs_forces 11 3.9 0.004 0.005 173.061 173.063 qs_energies 11 4.9 0.006 0.007 166.535 166.542 scf_env_do_scf 11 5.9 0.001 0.001 151.175 151.178 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.430 116.432 velocity_verlet 10 3.0 0.001 0.002 111.546 111.549 dbcsr_multiply_generic 2507 12.6 0.185 0.189 81.532 82.548 qs_scf_new_mos 117 7.6 0.001 0.001 79.588 79.945 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.587 79.944 ot_scf_mini 117 9.6 0.003 0.004 75.365 75.774 multiply_cannon 2507 13.6 0.504 0.528 61.832 66.342 multiply_cannon_loop 2507 14.6 1.126 1.184 59.148 61.466 ot_mini 117 10.6 0.001 0.001 42.410 42.822 init_scf_loop 11 6.9 0.000 0.000 34.647 34.649 mp_waitall_1 170520 16.6 24.800 34.534 24.800 34.534 prepare_preconditioner 11 7.9 0.000 0.000 30.670 30.717 make_preconditioner 11 8.9 0.000 0.000 30.670 30.717 rebuild_ks_matrix 128 8.3 0.001 0.001 30.028 30.543 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.027 30.543 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.283 29.725 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.069 27.545 multiply_cannon_multrec 20056 15.6 13.075 16.349 22.389 25.538 multiply_cannon_metrocomm3 20056 15.6 0.062 0.067 15.447 24.717 qs_ot_get_derivative 117 11.6 0.002 0.002 22.764 23.167 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.795 20.692 apply_single 128 13.6 0.001 0.001 19.794 20.692 ot_diis_step 117 11.6 0.018 0.018 19.539 19.540 qs_ot_get_p 128 10.4 0.001 0.001 18.356 18.840 make_m2s 5014 13.6 0.078 0.084 14.754 15.478 make_images 5014 14.6 1.149 1.246 14.519 15.245 multiply_cannon_sync_h2d 20056 15.6 13.555 15.169 13.555 15.169 cp_fm_cholesky_invert 11 10.9 14.224 14.233 14.224 14.233 qs_ot_p2m_diag 83 11.4 0.265 0.271 14.170 14.180 sum_up_and_integrate 128 10.3 0.002 0.003 13.879 13.906 integrate_v_rspace 128 11.3 0.003 0.003 13.819 13.846 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.478 13.514 calculate_rho_elec 128 8.7 0.129 0.143 13.477 13.513 cp_dbcsr_syevd 83 12.4 0.006 0.007 13.124 13.125 init_scf_run 11 5.9 0.000 0.001 10.358 10.359 scf_env_initial_rho_setup 11 6.9 0.002 0.002 10.358 10.358 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.956 9.978 make_images_data 5014 15.6 0.064 0.074 8.897 9.975 cp_fm_diag_elpa_base 83 14.4 9.544 9.714 9.952 9.973 multiply_cannon_metrocomm4 17549 15.6 0.064 0.073 3.487 9.594 hybrid_alltoall_any 5200 16.5 0.449 2.039 7.614 9.519 mp_irecv_dv 50230 16.2 3.358 9.325 3.358 9.325 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.976 9.284 dbcsr_mm_accdrv_process 41502 16.2 5.625 5.918 8.768 8.918 pw_transfer 1547 11.6 0.084 0.103 8.715 8.813 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.492 8.597 grid_integrate_task_list 128 12.3 7.292 7.803 7.292 7.803 fft_wrap_pw1pw2_140 523 13.2 1.057 1.104 7.554 7.669 wfi_extrapolate 11 7.9 0.001 0.001 7.388 7.388 cp_fm_upper_to_full 105 14.8 5.736 7.306 5.736 7.306 density_rs2pw 128 9.7 0.005 0.006 6.868 7.295 cp_fm_cholesky_decompose 22 10.9 7.257 7.280 7.257 7.280 dbcsr_complete_redistribute 395 12.7 1.165 1.192 4.564 6.305 calculate_dm_sparse 128 9.5 0.001 0.001 5.810 5.915 fft3d_ps 1291 14.7 2.814 3.053 5.811 5.907 grid_collocate_task_list 128 9.7 5.147 5.530 5.147 5.530 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.409 5.413 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.541 5.231 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.417 5.159 mp_alltoall_d11v 2415 14.1 4.250 4.774 4.250 4.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.537 4.668 mp_sum_l 7950 12.9 3.109 4.410 3.109 4.410 potential_pw2rs 128 12.3 0.020 0.022 4.323 4.333 mp_allgather_i34 2507 14.6 1.183 4.290 1.183 4.290 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.367 4.098 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.966 3.994 mp_alltoall_i22 716 14.1 1.960 3.860 1.960 3.860 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.732 3.733 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.456 3.495 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.600000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.909091, yerr=4.925897 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.154015E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. 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 0.010 0.032 186.927 186.929 qs_mol_dyn_low 1 2.0 0.003 0.003 186.562 186.576 qs_forces 11 3.9 0.004 0.004 186.431 186.438 qs_energies 11 4.9 0.001 0.002 179.338 179.347 scf_env_do_scf 11 5.9 0.001 0.001 162.784 162.796 velocity_verlet 10 3.0 0.004 0.006 124.225 124.228 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.441 116.441 dbcsr_multiply_generic 2507 12.6 0.193 0.198 80.243 80.927 qs_scf_new_mos 117 7.6 0.001 0.001 80.300 80.644 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.300 80.643 ot_scf_mini 117 9.6 0.003 0.004 75.762 76.094 multiply_cannon 2507 13.6 0.550 0.580 55.870 58.692 multiply_cannon_loop 2507 14.6 1.838 1.909 52.155 53.777 init_scf_loop 11 6.9 0.000 0.000 46.222 46.223 ot_mini 117 10.6 0.001 0.001 42.858 43.198 prepare_preconditioner 11 7.9 0.000 0.000 42.142 42.165 make_preconditioner 11 8.9 0.000 0.000 42.142 42.165 make_full_inverse_cholesky 11 9.9 0.011 0.022 35.652 40.741 multiply_cannon_multrec 30084 15.6 13.478 18.937 26.355 31.530 rebuild_ks_matrix 128 8.3 0.001 0.001 29.197 29.461 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.197 29.460 mp_waitall_1 147882 16.7 17.015 26.842 17.015 26.842 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.349 26.581 qs_ot_get_derivative 117 11.6 0.001 0.002 23.161 23.498 make_m2s 5014 13.6 0.092 0.097 19.990 21.163 make_images 5014 14.6 1.944 2.228 19.681 20.856 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.124 19.726 apply_single 128 13.6 0.001 0.001 19.123 19.726 ot_diis_step 117 11.6 0.017 0.018 19.578 19.579 qs_ot_get_p 128 10.4 0.001 0.001 19.064 19.432 cp_fm_upper_to_full 105 14.8 11.594 17.083 11.594 17.083 cp_fm_cholesky_invert 11 10.9 15.935 15.944 15.935 15.944 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.493 15.115 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.909 14.960 sum_up_and_integrate 128 10.3 0.002 0.003 13.945 13.973 integrate_v_rspace 128 11.3 0.003 0.003 13.885 13.915 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.684 13.736 calculate_rho_elec 128 8.7 0.170 0.186 13.683 13.735 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.513 13.515 dbcsr_complete_redistribute 395 12.7 1.498 1.622 9.204 13.037 dbcsr_mm_accdrv_process 62264 16.2 8.387 9.242 12.445 12.911 make_images_data 5014 15.6 0.065 0.072 10.655 12.411 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.840 11.641 multiply_cannon_sync_h2d 30084 15.6 10.465 11.499 10.465 11.499 hybrid_alltoall_any 5200 16.5 0.530 2.213 9.679 11.253 init_scf_run 11 5.9 0.000 0.001 10.667 10.668 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.666 10.668 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.302 10.316 cp_fm_diag_elpa_base 83 14.4 9.332 9.667 10.296 10.308 transfer_fm_to_dbcsr 11 9.9 0.002 0.016 6.468 10.152 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.713 9.961 mp_alltoall_i22 716 14.1 5.681 9.405 5.681 9.405 pw_transfer 1547 11.6 0.085 0.102 9.141 9.220 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.916 9.004 fft_wrap_pw1pw2_140 523 13.2 1.246 1.296 7.950 8.033 grid_integrate_task_list 128 12.3 7.529 7.772 7.529 7.772 wfi_extrapolate 11 7.9 0.001 0.001 7.612 7.612 cp_fm_cholesky_decompose 22 10.9 7.430 7.512 7.430 7.512 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.851 7.480 mp_irecv_dv 76098 16.2 2.699 7.199 2.699 7.199 density_rs2pw 128 9.7 0.005 0.006 6.773 7.079 calculate_dm_sparse 128 9.5 0.001 0.001 6.216 6.290 fft3d_ps 1291 14.7 2.992 3.091 5.968 6.060 grid_collocate_task_list 128 9.7 5.283 5.727 5.283 5.727 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.426 5.468 mp_alltoall_d11v 2415 14.1 4.884 5.367 4.884 5.367 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.494 4.599 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.345 4.420 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.411 4.412 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.231 4.314 potential_pw2rs 128 12.3 0.022 0.023 4.260 4.269 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.929000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1089.454545, yerr=19.592966 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.548034E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3981 58503. MP_Allreduce 11030 1168. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 168.493 168.494 qs_mol_dyn_low 1 2.0 0.003 0.003 168.110 168.125 qs_forces 11 3.9 0.004 0.004 168.018 168.020 qs_energies 11 4.9 0.002 0.002 160.571 160.579 scf_env_do_scf 11 5.9 0.001 0.001 143.465 143.481 velocity_verlet 10 3.0 0.002 0.002 109.952 109.956 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 107.926 107.928 dbcsr_multiply_generic 2485 12.5 0.182 0.189 71.666 72.110 qs_scf_new_mos 116 7.6 0.001 0.001 71.207 71.308 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.206 71.308 ot_scf_mini 116 9.6 0.003 0.003 66.848 66.934 multiply_cannon 2485 13.5 0.556 0.577 53.362 56.767 multiply_cannon_loop 2485 14.5 0.799 0.831 50.485 51.352 ot_mini 116 10.6 0.001 0.001 37.635 37.721 init_scf_loop 11 6.9 0.000 0.000 35.389 35.391 prepare_preconditioner 11 7.9 0.000 0.000 31.487 31.515 make_preconditioner 11 8.9 0.000 0.000 31.487 31.515 mp_waitall_1 124680 16.7 24.317 30.511 24.317 30.511 make_full_inverse_cholesky 11 9.9 0.014 0.025 29.370 29.649 rebuild_ks_matrix 127 8.3 0.001 0.001 28.976 29.062 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.975 29.061 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.381 26.462 multiply_cannon_multrec 9940 15.5 10.316 14.260 18.483 21.370 qs_ot_get_derivative 116 11.6 0.002 0.002 20.842 20.942 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.445 19.457 cp_fm_cholesky_invert 11 10.9 17.937 17.943 17.937 17.943 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.844 16.984 apply_single 127 13.6 0.001 0.001 16.844 16.983 qs_ot_get_p 127 10.4 0.001 0.001 16.720 16.828 ot_diis_step 116 11.6 0.019 0.020 16.722 16.723 make_m2s 4970 13.5 0.064 0.068 14.515 15.318 make_images 4970 14.5 2.146 2.573 14.211 15.007 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.113 14.135 calculate_rho_elec 127 8.7 0.252 0.262 14.113 14.134 sum_up_and_integrate 127 10.3 0.002 0.002 14.080 14.129 integrate_v_rspace 127 11.3 0.004 0.004 14.021 14.070 qs_ot_p2m_diag 82 11.4 0.489 0.494 12.976 12.994 cp_dbcsr_syevd 82 12.4 0.006 0.006 11.830 11.830 multiply_cannon_sync_h2d 9940 15.5 10.809 11.180 10.809 11.180 init_scf_run 11 5.9 0.000 0.001 10.257 10.258 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.257 10.257 pw_transfer 1535 11.6 0.084 0.094 9.788 9.818 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.568 9.605 hybrid_alltoall_any 5155 16.4 0.833 3.747 8.244 9.600 make_images_data 4970 15.5 0.054 0.065 8.351 9.592 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.019 9.031 cp_fm_diag_elpa_base 82 14.4 8.786 8.860 9.016 9.028 dbcsr_mm_accdrv_process 20590 16.1 3.667 5.073 7.817 8.540 fft_wrap_pw1pw2_140 519 13.2 1.655 1.688 8.493 8.531 grid_integrate_task_list 127 12.3 7.717 8.104 7.717 8.104 cp_fm_cholesky_decompose 22 10.9 7.862 7.981 7.862 7.981 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.868 7.939 wfi_extrapolate 11 7.9 0.001 0.001 7.285 7.285 density_rs2pw 127 9.7 0.005 0.005 6.874 7.279 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.291 6.996 fft3d_ps 1281 14.7 3.112 3.212 6.054 6.084 calculate_dm_sparse 127 9.5 0.001 0.001 6.004 6.070 grid_collocate_task_list 127 9.7 5.527 5.801 5.527 5.801 dbcsr_complete_redistribute 393 12.7 2.108 2.171 5.193 5.477 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.284 5.285 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.083 5.097 mp_alltoall_d11v 2401 14.1 4.597 4.969 4.597 4.969 mp_allgather_i34 2485 14.5 1.079 4.475 1.079 4.475 potential_pw2rs 127 12.3 0.025 0.026 4.314 4.321 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.238 4.300 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.020 4.081 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.599 3.920 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.371 3.696 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.522 3.552 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.703 3.512 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.486 3.495 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.408 3.489 mp_irecv_dv 28618 15.9 1.665 3.434 1.665 3.434 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.494000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1461.909091, yerr=32.038200 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/20/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 11.606413E+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 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 3.166966E+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 11.342275E+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 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59121. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. 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 0.019 0.040 289.916 289.918 qs_mol_dyn_low 1 2.0 0.003 0.003 289.397 289.410 qs_forces 11 3.9 0.005 0.005 289.298 289.305 qs_energies 11 4.9 0.045 0.049 280.391 280.398 scf_env_do_scf 11 5.9 0.001 0.001 257.926 257.935 velocity_verlet 10 3.0 0.002 0.002 209.860 209.868 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.673 132.675 init_scf_loop 11 6.9 0.000 0.000 124.986 124.990 prepare_preconditioner 11 7.9 0.000 0.000 120.120 120.141 make_preconditioner 11 8.9 0.000 0.000 120.120 120.141 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.071 117.290 qs_scf_new_mos 116 7.6 0.001 0.001 88.250 88.290 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.249 88.289 ot_scf_mini 116 9.6 0.004 0.004 83.437 83.441 dbcsr_multiply_generic 2485 12.5 0.215 0.226 80.849 81.311 cp_fm_upper_to_full 104 14.8 53.143 77.260 53.143 77.260 multiply_cannon 2485 13.5 0.670 0.708 58.201 59.723 multiply_cannon_loop 2485 14.5 1.039 1.075 54.211 55.398 dbcsr_complete_redistribute 393 12.7 3.986 4.026 30.879 44.608 ot_mini 116 10.6 0.001 0.001 43.422 43.427 copy_fm_to_dbcsr 208 11.6 0.001 0.002 27.433 41.154 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 25.007 38.538 mp_alltoall_i22 712 14.1 22.776 36.633 22.776 36.633 rebuild_ks_matrix 127 8.3 0.001 0.001 34.498 34.505 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 34.497 34.505 cp_fm_cholesky_invert 11 10.9 32.942 32.949 32.942 32.949 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.984 32.005 mp_waitall_1 102768 16.8 27.437 31.847 27.437 31.847 qs_ot_get_p 127 10.4 0.001 0.001 25.058 25.100 qs_ot_get_derivative 116 11.6 0.002 0.002 24.302 24.306 qs_ot_p2m_diag 82 11.4 0.869 0.873 21.072 21.100 make_m2s 4970 13.5 0.073 0.075 18.233 19.708 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 18.530 19.694 cp_dbcsr_syevd 82 12.4 0.006 0.007 19.302 19.303 make_images 4970 14.5 3.047 3.253 17.757 19.231 ot_diis_step 116 11.6 0.022 0.022 19.095 19.096 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.457 18.536 apply_single 127 13.6 0.001 0.001 18.456 18.535 multiply_cannon_multrec 9940 15.5 10.176 12.000 17.890 17.953 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.159 17.172 calculate_rho_elec 127 8.7 0.475 0.476 17.158 17.172 sum_up_and_integrate 127 10.3 0.002 0.002 16.360 16.453 integrate_v_rspace 127 11.3 0.004 0.004 16.298 16.392 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.190 16.190 cp_fm_diag_elpa_base 82 14.4 11.802 13.484 16.186 16.186 multiply_cannon_sync_h2d 9940 15.5 14.189 14.260 14.189 14.260 pw_transfer 1535 11.6 0.094 0.095 13.014 13.020 hybrid_alltoall_any 5155 16.4 1.306 3.028 10.504 12.792 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.778 12.785 make_images_data 4970 15.5 0.063 0.069 10.228 12.370 init_scf_run 11 5.9 0.000 0.001 12.077 12.080 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.077 12.080 fft_wrap_pw1pw2_140 519 13.2 3.185 3.227 11.381 11.397 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.545 9.548 dbcsr_mm_accdrv_process 20590 16.0 4.192 5.988 7.466 9.300 wfi_extrapolate 11 7.9 0.001 0.001 8.907 8.907 cp_fm_cholesky_decompose 22 10.9 8.781 8.810 8.781 8.810 grid_integrate_task_list 127 12.3 8.499 8.698 8.499 8.698 density_rs2pw 127 9.7 0.005 0.005 8.245 8.295 qs_energies_init_hamiltonians 11 5.9 0.038 0.050 8.204 8.205 fft3d_ps 1281 14.7 3.935 3.945 7.292 7.357 mp_alltoall_d11v 2401 14.1 6.838 6.983 6.838 6.983 calculate_dm_sparse 127 9.5 0.001 0.001 6.665 6.719 grid_collocate_task_list 127 9.7 6.367 6.442 6.367 6.442 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.289 6.340 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.064 6.154 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.918000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2879.909091, yerr=150.292717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 1.260827E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 12289157. 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 0.011 0.029 84.905 84.906 qs_energies 1 2.0 0.000 0.000 84.468 84.476 ls_scf 1 3.0 0.000 0.000 83.560 83.567 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.486 72.645 multiply_cannon 111 7.7 0.017 0.020 55.850 57.194 multiply_cannon_loop 111 8.7 0.223 0.240 52.404 53.730 ls_scf_main 1 4.0 0.000 0.000 52.172 52.172 density_matrix_trs4 2 5.0 0.002 0.003 46.671 46.754 ls_scf_init_scf 1 4.0 0.000 0.001 28.364 28.366 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.262 27.303 mp_waitall_1 11031 10.9 22.349 25.487 22.349 25.487 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.020 25.040 multiply_cannon_multrec 2664 9.7 8.158 8.893 15.526 17.244 multiply_cannon_sync_h2d 2664 9.7 13.512 15.088 13.512 15.088 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.620 13.549 make_m2s 222 7.7 0.009 0.011 13.002 13.481 make_images 222 8.7 0.097 0.108 12.980 13.462 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.527 8.900 make_images_data 222 9.7 0.004 0.005 7.536 8.083 dbcsr_mm_accdrv_process 4760 10.4 0.587 0.705 6.987 7.957 hybrid_alltoall_any 227 10.6 0.216 1.852 6.554 7.952 dbcsr_mm_accdrv_process_sort 4760 11.4 6.202 7.081 6.202 7.081 calculate_norms 4752 9.8 5.507 6.128 5.507 6.128 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.998 5.120 mp_sum_l 887 5.1 3.091 4.565 3.091 4.565 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.326 3.636 make_images_sizes 222 9.7 0.000 0.000 0.764 3.588 mp_alltoall_i44 222 10.7 0.764 3.588 0.764 3.588 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.491 mp_irecv_dv 6231 10.9 2.041 3.470 2.041 3.470 arnoldi_extremal 4 6.8 0.000 0.000 3.259 3.280 arnoldi_normal_ev 4 7.8 0.001 0.004 3.259 3.280 build_subspace 16 8.4 0.009 0.012 3.169 3.171 ls_scf_post 1 4.0 0.000 0.000 3.024 3.031 ls_scf_store_result 1 5.0 0.000 0.000 2.849 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.348 2.860 dbcsr_merge_single_wm 555 10.7 0.469 0.603 2.340 2.851 make_images_pack 222 9.7 2.209 2.629 2.211 2.631 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.335 2.591 dbcsr_sort_data 658 11.4 2.127 2.580 2.127 2.580 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.477 2.073 2.479 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.292 2.368 buffer_matrices_ensure_size 222 8.7 1.792 2.160 1.792 2.160 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.764 1.765 rebuild_ks_matrix 3 7.3 0.000 0.000 1.755 1.756 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.755 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.906000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 2.173202E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.038 90.088 90.089 qs_energies 1 2.0 0.000 0.000 89.614 89.619 ls_scf 1 3.0 0.000 0.000 88.246 88.250 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.408 74.729 multiply_cannon 111 7.7 0.027 0.038 52.752 57.148 ls_scf_main 1 4.0 0.000 0.000 54.763 54.767 multiply_cannon_loop 111 8.7 0.135 0.145 49.901 53.472 density_matrix_trs4 2 5.0 0.002 0.003 49.141 49.359 mp_waitall_1 9105 10.9 20.752 30.923 20.752 30.923 ls_scf_init_scf 1 4.0 0.000 0.000 29.856 29.857 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.634 28.752 multiply_cannon_multrec 1332 9.7 13.305 17.728 22.560 28.070 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.236 26.250 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.570 21.324 make_m2s 222 7.7 0.006 0.007 15.074 15.676 make_images 222 8.7 1.369 1.699 15.043 15.647 dbcsr_mm_accdrv_process 4041 10.4 0.361 0.526 8.851 10.388 dbcsr_mm_accdrv_process_sort 4041 11.4 8.371 9.862 8.371 9.862 make_images_data 222 9.7 0.004 0.005 8.719 9.681 hybrid_alltoall_any 227 10.6 0.541 2.551 8.066 9.487 mp_sum_l 887 5.1 5.095 8.302 5.095 8.302 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.183 8.126 mp_irecv_dv 3311 11.0 3.164 8.068 3.164 8.068 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.938 6.772 calculate_norms 2376 9.8 6.060 6.660 6.060 6.660 multiply_cannon_sync_h2d 1332 9.7 4.930 6.121 4.930 6.121 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.893 5.107 arnoldi_extremal 4 6.8 0.000 0.000 4.586 4.598 arnoldi_normal_ev 4 7.8 0.002 0.006 4.586 4.598 build_subspace 16 8.4 0.014 0.021 4.343 4.346 ls_scf_post 1 4.0 0.000 0.000 3.628 3.632 ls_scf_store_result 1 5.0 0.000 0.000 3.377 3.451 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.126 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.227 2.746 3.229 mp_allgather_i34 111 8.7 0.837 3.108 0.837 3.108 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.108 2.696 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.585 2.667 dbcsr_data_new 4174 10.1 2.116 2.416 2.116 2.416 make_images_pack 222 9.7 1.822 2.113 1.825 2.116 dbcsr_sort_data 436 11.2 1.841 2.063 1.841 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.843 1.845 rebuild_ks_matrix 3 7.3 0.000 0.000 1.830 1.832 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.830 1.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.089000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 2.884502E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.047 93.575 93.578 qs_energies 1 2.0 0.000 0.003 92.968 92.974 ls_scf 1 3.0 0.000 0.003 90.668 90.673 dbcsr_multiply_generic 111 6.7 0.016 0.019 75.456 75.735 ls_scf_main 1 4.0 0.000 0.003 56.470 56.474 multiply_cannon 111 7.7 0.037 0.110 52.094 56.303 multiply_cannon_loop 111 8.7 0.116 0.133 49.333 53.317 density_matrix_trs4 2 5.0 0.002 0.004 50.609 50.759 mp_waitall_1 7281 11.0 23.611 33.291 23.611 33.291 ls_scf_init_scf 1 4.0 0.000 0.002 30.501 30.508 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.317 29.378 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.976 26.984 multiply_cannon_multrec 888 9.7 12.643 15.303 21.192 24.462 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.993 23.030 make_m2s 222 7.7 0.006 0.007 16.438 17.178 make_images 222 8.7 1.586 1.850 16.400 17.138 make_images_data 222 9.7 0.004 0.004 9.449 10.668 hybrid_alltoall_any 227 10.6 0.642 2.956 8.891 10.325 dbcsr_mm_accdrv_process 3754 10.4 0.303 0.486 8.070 9.286 dbcsr_mm_accdrv_process_sort 3754 11.4 7.637 8.800 7.637 8.800 mp_sum_l 887 5.1 5.233 8.657 5.233 8.657 multiply_cannon_sync_h2d 888 9.7 6.039 7.230 6.039 7.230 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.817 7.046 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.081 6.978 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.491 6.915 mp_irecv_dv 2335 11.1 2.476 6.860 2.476 6.860 arnoldi_extremal 4 6.8 0.000 0.000 5.043 5.055 arnoldi_normal_ev 4 7.8 0.001 0.005 5.043 5.055 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.704 4.928 build_subspace 16 8.4 0.014 0.020 4.744 4.749 calculate_norms 1584 9.8 4.340 4.717 4.340 4.717 mp_allgather_i34 111 8.7 0.880 3.765 0.880 3.765 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.420 3.759 ls_scf_post 1 4.0 0.000 0.002 3.697 3.702 dbcsr_matrix_vector_mult_local 304 10.0 3.024 3.604 3.026 3.606 ls_scf_store_result 1 5.0 0.000 0.000 3.433 3.520 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.831 2.929 dbcsr_data_new 4116 9.9 2.096 2.460 2.096 2.460 make_images_sizes 222 9.7 0.000 0.000 1.189 2.384 mp_alltoall_i44 222 10.7 1.189 2.384 1.189 2.384 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.292 2.292 dbcsr_sort_data 325 11.1 1.843 2.144 1.843 2.144 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.894 make_images_pack 222 9.7 1.625 1.883 1.628 1.886 dbcsr_finalize 304 7.8 0.026 0.032 1.604 1.885 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.876 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.012 1.868 1.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.578000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2243.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 3.370643E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.045 97.045 97.046 qs_energies 1 2.0 0.000 0.000 96.473 96.479 ls_scf 1 3.0 0.000 0.000 94.734 94.739 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.496 78.799 ls_scf_main 1 4.0 0.000 0.000 58.852 58.853 multiply_cannon 111 7.7 0.049 0.114 51.653 55.633 density_matrix_trs4 2 5.0 0.002 0.003 52.734 52.847 multiply_cannon_loop 111 8.7 0.154 0.167 46.687 49.921 ls_scf_init_scf 1 4.0 0.000 0.001 32.633 32.635 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.404 31.492 mp_waitall_1 6369 11.0 22.877 29.715 22.877 29.715 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.915 28.925 multiply_cannon_multrec 1332 9.7 14.140 17.121 22.021 25.060 make_m2s 222 7.7 0.007 0.008 21.225 22.558 make_images 222 8.7 3.140 3.598 21.174 22.510 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.369 16.257 make_images_data 222 9.7 0.004 0.004 11.871 13.462 hybrid_alltoall_any 227 10.6 0.797 3.761 11.285 12.956 dbcsr_mm_accdrv_process 3641 10.4 0.293 0.497 7.525 9.045 dbcsr_mm_accdrv_process_sort 3641 11.4 7.085 8.558 7.085 8.558 mp_sum_l 887 5.1 4.022 7.161 4.022 7.161 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.087 6.150 mp_irecv_dv 3229 10.9 2.061 6.065 2.061 6.065 multiply_cannon_sync_h2d 1332 9.7 5.463 6.049 5.463 6.049 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.111 5.606 arnoldi_extremal 4 6.8 0.000 0.000 5.200 5.226 arnoldi_normal_ev 4 7.8 0.001 0.005 5.200 5.225 build_subspace 16 8.4 0.014 0.021 4.859 4.868 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.471 4.820 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.507 4.683 calculate_norms 2376 9.8 4.172 4.506 4.172 4.506 mp_allgather_i34 111 8.7 2.079 4.141 2.079 4.141 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.574 3.884 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.706 3.192 3.708 dbcsr_sort_data 658 11.4 3.075 3.418 3.075 3.418 ls_scf_post 1 4.0 0.000 0.000 3.249 3.255 dbcsr_special_finalize 555 9.7 0.006 0.008 2.819 3.199 dbcsr_merge_single_wm 555 10.7 0.540 0.666 2.811 3.191 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.026 3.079 ls_scf_store_result 1 5.0 0.000 0.000 2.988 3.052 dbcsr_data_release 10477 10.7 1.596 2.421 1.596 2.421 dbcsr_finalize 304 7.8 0.049 0.061 1.809 1.991 buffer_matrices_ensure_size 222 8.7 1.594 1.966 1.594 1.966 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.046000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 4.754026E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.054 92.509 92.510 qs_energies 1 2.0 0.000 0.000 90.896 90.902 ls_scf 1 3.0 0.000 0.000 88.895 88.901 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.453 70.666 ls_scf_main 1 4.0 0.000 0.000 56.230 56.230 multiply_cannon 111 7.7 0.093 0.163 52.191 55.837 multiply_cannon_loop 111 8.7 0.087 0.094 49.649 51.102 density_matrix_trs4 2 5.0 0.002 0.003 49.215 49.279 ls_scf_init_scf 1 4.0 0.000 0.001 29.269 29.270 mp_waitall_1 5436 11.0 24.201 28.605 24.201 28.605 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.013 28.049 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.911 25.921 multiply_cannon_multrec 444 9.7 13.660 16.264 20.712 22.893 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.756 15.129 make_m2s 222 7.7 0.005 0.005 13.587 14.520 make_images 222 8.7 2.038 2.477 13.520 14.450 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.959 13.882 hybrid_alltoall_any 227 10.6 0.801 3.834 8.085 9.792 make_images_data 222 9.7 0.003 0.004 8.301 9.567 multiply_cannon_sync_h2d 444 9.7 6.763 8.157 6.763 8.157 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.401 6.748 7.886 dbcsr_mm_accdrv_process_sort 3003 11.4 6.387 7.490 6.387 7.490 arnoldi_extremal 4 6.8 0.000 0.000 5.855 5.867 arnoldi_normal_ev 4 7.8 0.002 0.005 5.855 5.867 build_subspace 16 8.4 0.015 0.020 5.446 5.451 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.457 4.662 mp_sum_l 887 5.1 2.668 4.658 2.668 4.658 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.172 4.381 dbcsr_matrix_vector_mult_local 304 10.0 3.700 4.175 3.702 4.177 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.511 3.925 mp_irecv_dv 1241 11.2 1.498 3.890 1.498 3.890 calculate_norms 792 9.8 3.614 3.762 3.614 3.762 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.866 3.698 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.537 3.612 mp_allgather_i34 111 8.7 1.118 3.597 1.118 3.597 ls_scf_post 1 4.0 0.000 0.000 3.396 3.402 make_images_sizes 222 9.7 0.000 0.000 0.902 3.354 mp_alltoall_i44 222 10.7 0.901 3.354 0.901 3.354 ls_scf_store_result 1 5.0 0.000 0.000 3.190 3.220 dbcsr_data_new 4608 9.7 1.787 2.333 1.787 2.333 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.279 dbcsr_merge_all 275 8.9 0.481 0.530 2.055 2.115 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.033 2.034 rebuild_ks_matrix 3 7.3 0.000 0.000 2.001 2.002 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.001 2.002 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.985 1.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.510000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3772.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.937660E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.087 105.674 105.676 qs_energies 1 2.0 0.000 0.000 104.364 104.369 ls_scf 1 3.0 0.000 0.000 101.354 101.358 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.981 75.118 ls_scf_main 1 4.0 0.000 0.000 63.987 63.988 density_matrix_trs4 2 5.0 0.002 0.003 54.939 55.008 multiply_cannon 111 7.7 0.187 0.367 48.623 50.795 multiply_cannon_loop 111 8.7 0.097 0.100 45.623 46.076 ls_scf_init_scf 1 4.0 0.000 0.001 33.585 33.586 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.040 32.048 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.310 29.317 mp_waitall_1 4527 11.1 22.072 25.960 22.072 25.960 make_m2s 222 7.7 0.005 0.005 22.704 23.779 make_images 222 8.7 3.564 3.864 22.596 23.671 multiply_cannon_multrec 444 9.7 17.874 18.432 22.514 23.094 hybrid_alltoall_any 227 10.6 1.651 3.615 12.767 15.761 make_images_data 222 9.7 0.003 0.004 13.086 15.246 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.401 10.912 multiply_cannon_sync_h2d 444 9.7 8.790 8.847 8.790 8.847 arnoldi_extremal 4 6.8 0.000 0.000 7.388 7.398 arnoldi_normal_ev 4 7.8 0.003 0.008 7.388 7.398 build_subspace 16 8.4 0.026 0.036 6.836 6.852 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.493 5.641 dbcsr_matrix_vector_mult_local 304 10.0 5.047 5.356 5.049 5.358 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.175 5.268 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.878 5.134 dbcsr_mm_accdrv_process 1814 10.4 0.294 0.374 4.454 4.582 dbcsr_mm_accdrv_process_sort 1814 11.4 4.112 4.250 4.112 4.250 ls_scf_post 1 4.0 0.000 0.000 3.781 3.786 make_images_sizes 222 9.7 0.000 0.000 1.490 3.785 mp_alltoall_i44 222 10.7 1.489 3.785 1.489 3.785 mp_allgather_i34 111 8.7 1.117 3.596 1.117 3.596 ls_scf_store_result 1 5.0 0.000 0.000 3.499 3.535 calculate_norms 792 9.8 3.230 3.278 3.230 3.278 dbcsr_finalize 304 7.8 0.082 0.089 3.062 3.095 qs_energies_init_hamiltonians 1 3.0 0.002 0.005 2.980 2.980 dbcsr_merge_all 275 8.9 0.886 0.912 2.847 2.883 dbcsr_complete_redistribute 5 7.6 1.439 1.473 2.737 2.859 matrix_ls_to_qs 2 6.0 0.000 0.000 2.391 2.516 dbcsr_sort_data 325 11.1 2.433 2.478 2.433 2.478 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.459 2.460 rebuild_ks_matrix 3 7.3 0.000 0.000 2.393 2.394 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.393 2.394 dbcsr_data_new 6591 9.6 1.813 2.364 1.813 2.364 dbcsr_new_transposed 4 7.5 0.242 0.255 2.292 2.310 dbcsr_frobenius_norm 74 6.6 2.054 2.136 2.189 2.220 dbcsr_add_d 103 6.2 0.000 0.000 2.127 2.204 dbcsr_add_anytype 103 7.2 0.858 0.893 2.127 2.204 dbcsr_data_release 12724 10.6 1.978 2.200 1.978 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.676000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7038.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/75a566c04f83ccf946c3d618f248ad2bb7ee4910_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 589.484032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1320539570. 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 0.019 0.131 223.187 223.188 qs_mol_dyn_low 1 2.0 0.003 0.006 222.165 222.187 qs_forces 5 3.8 0.004 0.006 222.098 222.103 qs_energies 5 4.8 0.002 0.012 219.218 219.253 scf_env_do_scf 5 5.8 0.000 0.001 205.748 205.750 scf_env_do_scf_inner_loop 105 6.6 0.002 0.008 179.406 179.407 qs_scf_new_mos 105 7.6 0.000 0.001 140.407 140.635 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.406 140.635 dbcsr_multiply_generic 1445 12.2 0.129 0.139 131.567 132.070 ot_scf_mini 105 9.6 0.003 0.004 130.518 130.736 multiply_cannon 1445 13.2 0.275 0.286 113.056 114.760 multiply_cannon_loop 1445 14.2 2.774 2.972 111.434 112.690 velocity_verlet 4 3.0 0.001 0.002 105.549 105.549 ot_mini 105 10.6 0.001 0.001 59.880 60.001 multiply_cannon_multrec 69360 15.2 29.864 35.233 39.544 44.717 qs_ot_get_p 112 10.4 0.001 0.002 40.921 41.178 mp_waitall_1 488190 16.1 34.185 40.776 34.185 40.776 qs_ot_get_derivative 55 11.6 0.001 0.001 38.145 38.325 multiply_cannon_sync_h2d 69360 15.2 29.436 33.522 29.436 33.522 multiply_cannon_metrocomm3 69360 15.2 0.201 0.217 25.505 33.013 qs_ot_p2m_diag 40 11.0 0.020 0.031 29.899 30.028 rebuild_ks_matrix 110 8.4 0.000 0.000 28.365 28.514 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.018 28.365 28.514 cp_dbcsr_syevd 40 12.0 0.002 0.005 26.765 26.766 init_scf_loop 7 6.6 0.000 0.011 26.313 26.314 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.069 26.200 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.981 23.200 apply_single 62 13.6 0.000 0.000 22.981 23.199 cp_fm_syevd 40 13.0 0.000 0.005 21.485 21.617 prepare_preconditioner 7 7.6 0.000 0.000 21.502 21.526 make_preconditioner 7 8.6 0.000 0.004 21.502 21.526 ot_new_cg_direction 55 11.6 0.001 0.001 20.970 20.970 cp_fm_redistribute_end 40 14.0 8.397 16.753 8.402 16.754 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.270 16.719 calculate_rho_elec 110 8.6 0.030 0.032 16.269 16.719 cp_fm_syevd_base 40 14.0 8.345 16.703 8.345 16.703 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.580 14.805 make_full_inverse_cholesky 7 9.6 0.000 0.006 14.487 14.561 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.867 13.975 mp_sum_l 4764 12.2 11.916 12.641 11.916 12.641 calculate_dm_sparse 110 9.5 0.000 0.001 11.493 11.645 pw_transfer 1645 12.4 0.079 0.095 11.285 11.454 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.146 11.315 density_rs2pw 110 9.6 0.005 0.006 10.530 10.974 init_scf_run 5 5.8 0.000 0.001 10.851 10.852 scf_env_initial_rho_setup 5 6.8 0.002 0.003 10.851 10.852 dbcsr_mm_accdrv_process 154766 15.8 6.021 6.344 9.547 10.504 qs_vxc_create 110 10.4 0.002 0.004 10.210 10.243 cp_fm_cholesky_invert 7 10.6 10.226 10.233 10.226 10.233 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.133 10.233 fft_wrap_pw1pw2_240 915 15.0 0.839 0.917 9.820 10.012 check_diag 80 13.5 8.686 8.935 9.571 9.701 sum_up_and_integrate 60 10.3 0.001 0.004 8.187 8.197 integrate_v_rspace 60 11.3 0.001 0.002 8.170 8.180 fft3d_pb 915 16.0 2.396 2.587 7.997 8.170 transfer_rs2pw 445 10.6 0.007 0.008 7.374 7.825 xc_rho_set_and_dset_create 110 12.4 0.075 0.099 7.198 7.455 multiply_cannon_metrocomm1 69360 15.2 0.098 0.110 4.524 7.316 acc_transpose_blocks 69360 15.2 0.351 0.377 6.526 7.052 xc_vxc_pw_create 60 11.3 0.039 0.050 6.811 6.844 make_full_single_inverse 7 9.6 0.001 0.005 6.739 6.764 make_m2s 2890 13.2 0.079 0.089 6.124 6.682 make_images 2890 14.2 0.238 0.264 6.016 6.573 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.434 6.450 xc_pw_derive 510 13.4 0.005 0.007 6.011 6.093 calculate_first_density_matrix 1 7.0 0.000 0.003 6.058 6.072 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.014 6.070 mp_alltoall_z22v 2340 17.7 4.978 5.271 4.978 5.271 mp_waitany 7680 13.5 4.329 4.833 4.329 4.833 wfi_extrapolate 5 7.8 0.000 0.001 4.488 4.488 multiply_cannon_metrocomm4 67915 15.2 0.185 0.209 2.016 4.479 potential_pw2rs 60 12.3 0.002 0.003 4.455 4.475 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=223.188000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 75a566c04f83ccf946c3d618f248ad2bb7ee4910 Summary: empty Status: OK