=== This is the CP2K Performance-Test === Updating 443130275..c074904eb Fast-forward data/BASIS_ADMM_ae | 957 ++++++++++++++++++++++++++ src/response_solver.F | 28 +- tests/QS/regtest-tddfpt-force-gapw/TEST_FILES | 4 + tests/QS/regtest-tddfpt-force-gapw/bnt1.inp | 94 +++ tests/QS/regtest-tddfpt-force-gapw/bnt2.inp | 95 +++ tests/QS/regtest-tddfpt-force-gapw/bnt3.inp | 94 +++ 6 files changed, 1260 insertions(+), 12 deletions(-) create mode 100644 data/BASIS_ADMM_ae create mode 100644 tests/QS/regtest-tddfpt-force-gapw/bnt1.inp create mode 100644 tests/QS/regtest-tddfpt-force-gapw/bnt2.inp create mode 100644 tests/QS/regtest-tddfpt-force-gapw/bnt3.inp Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: c074904eb170f1b1b0b6589717b6dd1755d9e84e ################# 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 # # 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: 04.10.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/01 job id: 50043143 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/02 job id: 50043144 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/03 job id: 50043145 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/04 job id: 50043146 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/05 job id: 50043147 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/06 job id: 50043148 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/07 job id: 50043149 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/08 job id: 50043151 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/09 job id: 50043152 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/10 job id: 50043153 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/11 job id: 50043154 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/12 job id: 50043156 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/13 job id: 50043157 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/14 job id: 50043158 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/15 job id: 50043159 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/16 job id: 50043160 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/17 job id: 50043161 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/18 job id: 50043162 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/19 job id: 50043163 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/20 job id: 50043164 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/21 job id: 50043165 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/22 job id: 50043166 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/23 job id: 50043167 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/24 job id: 50043168 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/25 job id: 50043169 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/26 job id: 50043170 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/27 job id: 50043171 --- 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/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.034 138.034 138.034 farming_run 1 2.0 137.512 137.515 138.004 138.007 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.503183E+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 0.006 0.026 114.295 114.295 qs_energies 1 2.0 0.000 0.000 114.085 114.087 mp2_main 1 3.0 0.000 0.000 111.973 111.975 mp2_gpw_main 1 4.0 0.028 0.034 110.935 110.938 mp2_ri_gpw_compute_in 1 5.0 0.172 0.184 91.940 92.285 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.717 55.063 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.163 41.123 46.163 get_2c_integrals 1 6.0 0.008 0.009 36.481 37.050 integrate_v_rspace 273 8.0 0.436 0.452 24.706 29.552 pw_transfer 6555 10.6 0.379 0.391 26.812 27.290 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.466 25.992 grid_integrate_task_list 273 9.0 20.604 25.939 20.604 25.939 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.029 23.004 23.521 rpa_ri_compute_en 1 5.0 0.019 0.021 18.892 19.152 compute_2c_integrals 1 7.0 0.002 0.002 18.964 18.965 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.525 18.696 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.426 18.523 18.695 cp_fm_cholesky_decompose 12 8.2 17.465 18.011 17.465 18.011 cholesky_decomp 1 7.0 0.000 0.000 16.363 16.920 fft3d_s 5443 13.4 16.096 16.455 16.117 16.475 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.328 10.771 15.328 calculate_wavefunction 272 8.0 5.400 5.557 12.342 13.040 rpa_num_int 1 6.0 0.001 0.011 10.660 10.669 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.613 10.647 calc_mat_Q 8 8.0 0.000 0.000 9.464 9.585 contract_S_to_Q 8 9.0 0.000 0.000 8.889 9.008 calc_potential_gpw 544 9.5 0.004 0.005 8.211 8.640 parallel_gemm_fm 14 9.1 0.000 0.000 8.499 8.610 parallel_gemm_fm_cosma 14 10.1 8.499 8.610 8.499 8.610 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.041 8.282 potential_pw2rs 545 10.0 0.106 0.108 7.467 8.071 create_integ_mat 1 6.0 0.015 0.029 7.793 7.803 collocate_single_gaussian 272 10.0 0.039 0.041 7.289 7.569 array2fm 1 7.0 0.000 0.000 6.690 7.211 pw_scatter_s 2720 13.7 4.375 4.542 4.375 4.542 pw_gather_s 2722 13.2 3.447 3.853 3.447 3.853 array2fm_buffer_send 1 8.0 2.988 3.162 2.988 3.162 pw_poisson_solve 545 10.5 1.114 1.182 2.165 2.335 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.937793, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2821.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 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 0.025 0.037 428.037 428.039 farming_run 1 2.0 427.065 427.069 427.997 428.000 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.241784E+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 22521586. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 216.511 216.511 qs_energies 1 2.0 0.002 0.015 216.274 216.292 scf_env_do_scf 1 3.0 0.000 0.000 114.123 114.124 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.233 113.240 rebuild_ks_matrix 4 6.0 0.000 0.000 113.231 113.239 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.231 113.239 hfx_ks_matrix 4 8.0 0.001 0.001 112.887 112.892 integrate_four_center 4 9.0 0.152 0.471 112.887 112.891 integrate_four_center_main 4 10.0 0.121 0.619 101.921 104.768 integrate_four_center_bin 263 11.0 101.800 104.716 101.800 104.716 mp2_main 1 3.0 0.005 0.030 101.857 101.876 mp2_gpw_main 1 4.0 0.049 0.071 100.942 100.958 init_scf_loop 1 4.0 0.000 0.000 96.338 96.338 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.919 74.977 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.768 54.826 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.157 41.507 46.502 integrate_v_rspace 95 8.0 0.394 0.556 27.991 32.820 pw_transfer 2240 10.6 0.145 0.167 29.481 29.858 mp2_ri_gpw_compute_en 1 5.0 0.059 0.077 26.845 28.979 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.497 28.895 grid_integrate_task_list 95 9.0 23.418 28.449 23.418 28.449 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.240 26.582 ao_to_mo_and_store_B_mult_1 91 7.0 10.593 26.175 10.593 26.175 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.824 1.895 25.129 25.140 get_2c_integrals 1 6.0 0.000 0.000 20.063 20.088 compute_2c_integrals 1 7.0 0.002 0.003 19.051 19.057 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.606 18.925 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.869 18.605 18.924 fft3d_s 1823 13.4 18.471 18.778 18.484 18.791 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.784 17.784 calculate_wavefunction 91 8.0 2.007 2.042 9.608 9.841 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.594 8.810 9.305 potential_pw2rs 186 10.0 0.033 0.034 8.412 8.902 local_gemm 172 8.0 8.251 8.735 8.251 8.735 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.512 8.118 8.626 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.139 8.532 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.087 collocate_single_gaussian 91 10.0 0.017 0.020 7.740 8.006 integrate_four_center_load 4 10.0 0.000 0.000 6.752 6.756 hfx_load_balance 1 11.0 0.000 0.000 6.752 6.756 mp_sendrecv_dm3 2068 8.0 6.155 6.688 6.155 6.688 mp2_ri_gpw_compute_en_ener 172 7.0 6.333 6.416 6.333 6.416 mp_sync 37 10.5 2.946 5.728 2.946 5.728 pw_gather_s 912 13.2 4.473 4.917 4.473 4.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.931543, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 452.874240E+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 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.150 0.218 62.370 62.372 qs_mol_dyn_low 1 2.0 0.003 0.005 61.689 61.718 qs_forces 11 3.9 0.002 0.003 61.595 61.597 qs_energies 11 4.9 0.001 0.002 60.120 60.132 scf_env_do_scf 11 5.9 0.001 0.001 45.927 45.927 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.766 43.766 dbcsr_multiply_generic 2286 12.5 0.093 0.096 41.486 42.038 multiply_cannon 2286 13.5 0.188 0.202 34.195 35.784 multiply_cannon_loop 2286 14.5 1.814 1.931 33.518 35.082 qs_scf_new_mos 108 7.5 0.000 0.000 33.703 33.988 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.702 33.988 ot_scf_mini 108 9.5 0.002 0.002 32.015 32.198 velocity_verlet 10 3.0 0.001 0.001 26.726 26.727 mp_waitall_1 245248 16.5 12.101 19.531 12.101 19.531 ot_mini 108 10.5 0.001 0.001 19.143 19.402 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 9.502 18.134 qs_ot_get_derivative 108 11.5 0.001 0.001 16.160 16.357 init_scf_run 11 5.9 0.000 0.001 12.927 12.927 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.927 12.927 multiply_cannon_multrec 54864 15.5 3.661 5.863 7.647 11.205 calculate_first_density_matrix 1 7.0 0.000 0.002 10.640 10.650 qs_ot_get_p 119 10.4 0.001 0.001 8.151 8.431 rebuild_ks_matrix 119 8.3 0.000 0.000 7.887 8.094 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.886 8.093 acc_transpose_blocks 54864 15.5 0.223 0.245 6.538 7.647 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 7.184 7.186 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.950 7.134 mp_sum_l 7287 12.8 5.311 7.116 5.311 7.116 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.865 6.936 multiply_cannon_sync_h2d 54864 15.5 5.170 6.872 5.170 6.872 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.870 6.321 calculate_dm_sparse 119 9.5 0.000 0.000 6.146 6.293 acc_transpose_blocks_kernels 54864 16.5 0.247 0.382 5.072 5.927 dbcsr_mm_accdrv_process 76910 16.1 1.824 2.876 3.900 5.615 jit_kernel_transpose 5 15.5 4.826 5.552 4.826 5.552 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.210 5.244 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.063 5.185 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.929 5.133 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 4.620 4.651 sum_up_and_integrate 119 10.3 0.001 0.002 4.472 4.478 integrate_v_rspace 119 11.3 0.002 0.002 4.461 4.468 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.454 4.455 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.187 4.286 calculate_rho_elec 119 8.7 0.011 0.016 4.187 4.285 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.243 4.244 cp_fm_redistribute_end 50 14.0 2.165 4.216 2.172 4.220 cp_fm_diag_elpa_base 50 14.0 2.041 4.106 2.045 4.117 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.952 3.150 apply_single 119 13.6 0.000 0.000 2.951 3.150 ot_diis_step 108 11.5 0.006 0.006 2.750 2.750 jit_kernel_multiply 13 15.8 2.012 2.708 2.012 2.708 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.531 2.611 density_rs2pw 119 9.7 0.004 0.004 2.171 2.323 wfi_extrapolate 11 7.9 0.001 0.001 2.209 2.209 init_scf_loop 11 6.9 0.000 0.000 2.144 2.144 grid_integrate_task_list 119 12.3 2.016 2.111 2.016 2.111 mp_sum_d 4137 12.0 1.313 1.913 1.313 1.913 potential_pw2rs 119 12.3 0.004 0.004 1.840 1.856 pw_transfer 1439 11.6 0.051 0.055 1.712 1.784 make_m2s 4572 13.5 0.053 0.055 1.708 1.765 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.637 1.712 make_images 4572 14.5 0.133 0.139 1.626 1.682 transfer_rs2pw 487 10.6 0.005 0.006 1.361 1.533 mp_alltoall_d11v 2130 13.8 1.364 1.514 1.364 1.514 acc_transpose_blocks_sync 164592 16.5 1.197 1.443 1.197 1.443 grid_collocate_task_list 119 9.7 1.351 1.423 1.351 1.423 fft3d_ps 1201 14.6 0.371 0.479 1.349 1.423 mp_waitany 12084 13.8 1.249 1.418 1.249 1.418 transfer_pw2rs 487 13.2 0.005 0.006 1.331 1.346 fft_wrap_pw1pw2_140 487 13.2 0.108 0.115 1.261 1.336 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=62.372000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 488.960000E+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 7458. 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 0.009 0.030 39.273 39.274 qs_mol_dyn_low 1 2.0 0.003 0.003 39.101 39.110 qs_forces 11 3.9 0.002 0.003 39.040 39.040 qs_energies 11 4.9 0.001 0.002 37.342 37.346 scf_env_do_scf 11 5.9 0.001 0.001 32.230 32.230 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.726 29.726 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.142 22.532 qs_scf_new_mos 108 7.5 0.001 0.001 20.807 21.032 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.807 21.032 ot_scf_mini 108 9.5 0.002 0.003 19.879 20.036 velocity_verlet 10 3.0 0.001 0.002 18.995 18.996 multiply_cannon 2286 13.5 0.209 0.218 16.934 18.428 multiply_cannon_loop 2286 14.5 1.185 1.242 15.750 17.272 ot_mini 108 10.5 0.001 0.001 12.281 12.511 mp_waitall_1 200699 16.5 5.671 10.795 5.671 10.795 qs_ot_get_derivative 108 11.5 0.001 0.001 9.906 10.070 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.103 9.485 multiply_cannon_multrec 27432 15.5 1.827 4.074 6.129 9.079 rebuild_ks_matrix 119 8.3 0.000 0.000 6.841 6.994 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.840 6.994 dbcsr_mm_accdrv_process 47894 16.0 3.452 5.762 4.222 6.277 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.002 6.141 qs_ot_get_p 119 10.4 0.001 0.001 4.663 4.875 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.795 4.638 mp_sum_l 7287 12.8 2.085 4.171 2.085 4.171 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.996 4.059 apply_single 119 13.6 0.000 0.000 2.996 4.058 init_scf_run 11 5.9 0.000 0.001 3.903 3.903 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.903 3.903 sum_up_and_integrate 119 10.3 0.001 0.001 3.765 3.769 integrate_v_rspace 119 11.3 0.002 0.002 3.751 3.756 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.709 3.736 calculate_rho_elec 119 8.7 0.021 0.024 3.708 3.735 make_m2s 4572 13.5 0.052 0.054 2.671 3.018 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.991 3.012 make_images 4572 14.5 0.206 0.245 2.581 2.931 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.594 2.594 init_scf_loop 11 6.9 0.000 0.000 2.480 2.481 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.322 2.407 calculate_first_density_matrix 1 7.0 0.000 0.000 2.327 2.329 ot_diis_step 108 11.5 0.011 0.011 2.321 2.322 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.277 2.278 cp_fm_redistribute_end 50 14.0 1.154 2.253 1.157 2.256 cp_fm_diag_elpa_base 50 14.0 1.067 2.164 1.094 2.205 multiply_cannon_sync_h2d 27432 15.5 1.671 2.189 1.671 2.189 calculate_dm_sparse 119 9.5 0.000 0.001 2.070 2.153 density_rs2pw 119 9.7 0.004 0.004 2.005 2.070 acc_transpose_blocks 27432 15.5 0.115 0.121 1.628 1.972 grid_integrate_task_list 119 12.3 1.841 1.949 1.841 1.949 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.938 1.941 pw_transfer 1439 11.6 0.064 0.068 1.890 1.928 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.799 1.839 jit_kernel_multiply 9 16.3 0.711 1.776 0.711 1.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.716 1.759 prepare_preconditioner 11 7.9 0.000 0.000 1.596 1.622 make_preconditioner 11 8.9 0.000 0.000 1.596 1.622 make_images_data 4572 15.5 0.047 0.054 1.252 1.584 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.493 1.549 wfi_extrapolate 11 7.9 0.001 0.001 1.517 1.517 potential_pw2rs 119 12.3 0.006 0.006 1.502 1.509 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.115 1.485 fft3d_ps 1201 14.6 0.523 0.581 1.431 1.464 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.403 1.439 mp_alltoall_d11v 2130 13.8 1.250 1.377 1.250 1.377 grid_collocate_task_list 119 9.7 1.290 1.347 1.290 1.347 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.219 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.262 transfer_rs2pw 487 10.6 0.005 0.005 1.149 1.215 mp_sum_d 4137 12.0 0.550 1.033 0.550 1.033 mp_allgather_i34 2286 14.5 0.606 1.018 0.606 1.018 transfer_pw2rs 487 13.2 0.004 0.005 0.946 0.953 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.936 0.936 acc_transpose_blocks_sync 82296 16.5 0.812 0.933 0.812 0.933 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.895 0.910 acc_transpose_blocks_kernels 27432 16.5 0.189 0.279 0.675 0.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.274000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.454545, yerr=1.776802 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.138176E+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 0.017 0.034 37.751 37.765 qs_mol_dyn_low 1 2.0 0.003 0.006 37.420 37.426 qs_forces 11 3.9 0.004 0.010 37.353 37.354 qs_energies 11 4.9 0.003 0.007 35.786 35.789 scf_env_do_scf 11 5.9 0.001 0.004 25.887 25.888 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.404 23.406 dbcsr_multiply_generic 2286 12.5 0.095 0.097 22.124 22.227 multiply_cannon 2286 13.5 0.195 0.202 18.820 19.617 multiply_cannon_loop 2286 14.5 0.857 0.901 18.033 18.844 qs_scf_new_mos 108 7.5 0.001 0.001 15.653 15.666 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.653 15.666 velocity_verlet 10 3.0 0.002 0.003 15.332 15.334 ot_scf_mini 108 9.5 0.002 0.003 14.885 14.901 ot_mini 108 10.5 0.001 0.001 9.212 9.232 init_scf_run 11 5.9 0.000 0.001 8.690 8.690 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.690 8.690 qs_ot_get_derivative 108 11.5 0.001 0.002 7.645 7.662 calculate_first_density_matrix 1 7.0 0.001 0.004 7.440 7.441 multiply_cannon_multrec 18288 15.5 1.874 2.971 6.943 7.312 acc_transpose_blocks 18288 15.5 0.079 0.081 6.717 6.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.803 6.805 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.640 6.652 rebuild_ks_matrix 119 8.3 0.000 0.000 6.040 6.059 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.017 6.040 6.058 acc_transpose_blocks_kernels 18288 16.5 0.219 0.226 5.885 5.929 dbcsr_mm_accdrv_process 38222 16.0 4.916 5.701 4.976 5.766 jit_kernel_transpose 5 15.6 5.666 5.707 5.666 5.707 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.317 5.333 mp_waitall_1 158411 16.6 2.772 3.952 2.772 3.952 sum_up_and_integrate 119 10.3 0.001 0.002 3.602 3.608 integrate_v_rspace 119 11.3 0.003 0.007 3.589 3.595 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.924 3.482 qs_ot_get_p 119 10.4 0.001 0.002 3.421 3.445 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.306 3.314 calculate_rho_elec 119 8.7 0.031 0.032 3.306 3.314 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.526 2.547 init_scf_loop 11 6.9 0.001 0.004 2.464 2.468 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 1.975 2.358 apply_single 119 13.6 0.000 0.000 1.974 2.358 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.268 2.275 make_m2s 4572 13.5 0.044 0.046 1.852 1.989 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.969 1.970 density_rs2pw 119 9.7 0.004 0.004 1.867 1.948 grid_integrate_task_list 119 12.3 1.795 1.917 1.795 1.917 pw_transfer 1439 11.6 0.065 0.068 1.894 1.908 make_images 4572 14.5 0.192 0.204 1.765 1.899 calculate_dm_sparse 119 9.5 0.000 0.000 1.853 1.863 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.802 1.816 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.733 1.741 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.712 1.724 cp_fm_diag_elpa_base 50 14.0 1.688 1.700 1.710 1.722 prepare_preconditioner 11 7.9 0.000 0.000 1.659 1.662 make_preconditioner 11 8.9 0.000 0.002 1.659 1.662 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.512 1.597 ot_diis_step 108 11.5 0.018 0.064 1.545 1.545 mp_sum_l 7287 12.8 1.097 1.512 1.097 1.512 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.447 1.459 potential_pw2rs 119 12.3 0.007 0.008 1.401 1.409 fft3d_ps 1201 14.6 0.547 0.568 1.357 1.374 grid_collocate_task_list 119 9.7 1.233 1.316 1.233 1.316 wfi_extrapolate 11 7.9 0.001 0.002 1.197 1.197 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.185 1.189 multiply_cannon_sync_h2d 18288 15.5 1.029 1.176 1.029 1.176 transfer_rs2pw 487 10.6 0.005 0.005 1.017 1.109 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.962 0.983 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.967 0.968 make_images_data 4572 15.5 0.047 0.051 0.801 0.961 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.694 0.872 transfer_pw2rs 487 13.2 0.004 0.004 0.850 0.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.797 0.800 mp_alltoall_d11v 2130 13.8 0.704 0.792 0.704 0.792 acc_transpose_blocks_sync 54864 16.5 0.733 0.784 0.733 0.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=37.765000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.909091, yerr=1.083307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 559.067136E+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 0.009 0.027 41.854 41.854 qs_mol_dyn_low 1 2.0 0.003 0.003 41.692 41.700 qs_forces 11 3.9 0.002 0.002 41.632 41.633 qs_energies 11 4.9 0.001 0.001 39.928 39.936 scf_env_do_scf 11 5.9 0.000 0.001 29.858 29.860 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.466 26.466 dbcsr_multiply_generic 2286 12.5 0.101 0.103 25.322 25.449 multiply_cannon 2286 13.5 0.219 0.227 21.557 21.927 multiply_cannon_loop 2286 14.5 1.561 1.621 20.588 21.037 velocity_verlet 10 3.0 0.002 0.002 18.529 18.531 qs_scf_new_mos 108 7.5 0.001 0.001 18.287 18.342 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.286 18.341 ot_scf_mini 108 9.5 0.002 0.003 17.246 17.303 ot_mini 108 10.5 0.001 0.001 10.716 10.777 multiply_cannon_multrec 27432 15.5 2.485 3.361 9.023 9.514 qs_ot_get_derivative 108 11.5 0.001 0.001 8.848 8.906 init_scf_run 11 5.9 0.000 0.001 8.733 8.733 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.733 8.733 dbcsr_mm_accdrv_process 47916 15.9 6.016 7.612 6.435 7.722 acc_transpose_blocks 27432 15.5 0.120 0.122 7.482 7.606 calculate_first_density_matrix 1 7.0 0.000 0.000 7.323 7.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.863 6.864 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.685 6.699 rebuild_ks_matrix 119 8.3 0.000 0.000 6.373 6.420 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.372 6.419 acc_transpose_blocks_kernels 27432 16.5 0.273 0.281 5.855 5.893 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.651 5.691 jit_kernel_transpose 5 15.6 5.583 5.617 5.583 5.617 qs_ot_get_p 119 10.4 0.001 0.002 3.664 3.737 sum_up_and_integrate 119 10.3 0.001 0.001 3.536 3.544 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.152 3.542 integrate_v_rspace 119 11.3 0.003 0.003 3.525 3.533 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.427 3.463 calculate_rho_elec 119 8.7 0.040 0.046 3.426 3.462 init_scf_loop 11 6.9 0.000 0.000 3.373 3.373 mp_waitall_1 137007 16.6 2.137 2.668 2.137 2.668 prepare_preconditioner 11 7.9 0.000 0.000 2.536 2.544 make_preconditioner 11 8.9 0.000 0.000 2.536 2.544 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.143 2.471 make_m2s 4572 13.5 0.054 0.055 2.339 2.443 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.386 2.418 make_images 4572 14.5 0.273 0.336 2.231 2.333 calculate_dm_sparse 119 9.5 0.000 0.000 2.196 2.254 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.236 2.246 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.101 2.219 apply_single 119 13.6 0.000 0.000 2.101 2.219 pw_transfer 1439 11.6 0.064 0.069 2.020 2.057 density_rs2pw 119 9.7 0.004 0.004 1.881 1.973 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.929 1.969 grid_integrate_task_list 119 12.3 1.838 1.917 1.838 1.917 ot_diis_step 108 11.5 0.012 0.012 1.826 1.827 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.823 1.824 fft_wrap_pw1pw2_140 487 13.2 0.227 0.233 1.585 1.625 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.574 1.584 cp_fm_diag_elpa_base 50 14.0 1.540 1.556 1.572 1.581 acc_transpose_blocks_sync 82296 16.5 1.476 1.560 1.476 1.560 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 1.011 1.558 fft3d_ps 1201 14.6 0.591 0.652 1.417 1.443 jit_kernel_multiply 8 16.3 0.352 1.427 0.352 1.427 wfi_extrapolate 11 7.9 0.001 0.001 1.355 1.355 potential_pw2rs 119 12.3 0.009 0.009 1.333 1.337 grid_collocate_task_list 119 9.7 1.246 1.337 1.246 1.337 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.289 1.301 mp_sum_l 7287 12.8 0.997 1.265 0.997 1.265 cp_fm_upper_to_full 72 14.2 0.819 1.166 0.819 1.166 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.069 1.070 dbcsr_complete_redistribute 329 12.2 0.121 0.144 0.787 1.060 transfer_rs2pw 487 10.6 0.004 0.005 0.939 1.034 make_images_data 4572 15.5 0.048 0.052 0.883 0.999 hybrid_alltoall_any 4725 16.4 0.066 0.158 0.765 0.934 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.799 0.875 mp_alltoall_d11v 2130 13.8 0.752 0.872 0.752 0.872 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.579 0.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=41.854000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.000000, yerr=2.923261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 638.238720E+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 0.011 0.028 33.998 33.999 qs_mol_dyn_low 1 2.0 0.003 0.003 33.829 33.839 qs_forces 11 3.9 0.002 0.002 33.768 33.769 qs_energies 11 4.9 0.001 0.001 32.055 32.058 scf_env_do_scf 11 5.9 0.002 0.014 22.287 22.287 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.766 19.766 dbcsr_multiply_generic 2286 12.5 0.094 0.096 18.262 18.367 multiply_cannon 2286 13.5 0.223 0.231 15.670 16.246 multiply_cannon_loop 2286 14.5 0.640 0.668 14.749 14.933 velocity_verlet 10 3.0 0.002 0.002 14.755 14.758 qs_scf_new_mos 108 7.5 0.001 0.001 12.011 12.040 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.010 12.039 ot_scf_mini 108 9.5 0.002 0.002 11.299 11.327 init_scf_run 11 5.9 0.000 0.001 8.333 8.333 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.333 8.333 calculate_first_density_matrix 1 7.0 0.000 0.001 7.183 7.183 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.587 6.588 ot_mini 108 10.5 0.001 0.001 6.538 6.570 acc_transpose_blocks 9144 15.5 0.043 0.044 6.453 6.552 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.410 6.421 multiply_cannon_multrec 9144 15.5 1.792 1.975 6.113 6.291 rebuild_ks_matrix 119 8.3 0.000 0.000 5.850 5.876 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.850 5.875 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 5.678 5.732 jit_kernel_transpose 5 15.6 5.560 5.612 5.560 5.612 qs_ot_get_derivative 108 11.5 0.001 0.001 5.253 5.280 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.203 5.225 dbcsr_mm_accdrv_process 12550 15.8 3.556 4.252 4.210 4.302 sum_up_and_integrate 119 10.3 0.001 0.001 3.483 3.487 integrate_v_rspace 119 11.3 0.003 0.003 3.473 3.477 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.398 3.403 calculate_rho_elec 119 8.7 0.060 0.061 3.398 3.403 qs_ot_get_p 119 10.4 0.001 0.001 2.879 2.918 init_scf_loop 11 6.9 0.000 0.000 2.500 2.502 pw_transfer 1439 11.6 0.065 0.067 2.143 2.151 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.050 2.059 mp_waitall_1 115863 16.7 1.475 2.051 1.475 2.051 make_m2s 4572 13.5 0.033 0.034 1.815 1.977 grid_integrate_task_list 119 12.3 1.867 1.923 1.867 1.923 density_rs2pw 119 9.7 0.003 0.003 1.789 1.887 make_images 4572 14.5 0.269 0.300 1.725 1.885 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.879 1.882 calculate_dm_sparse 119 9.5 0.000 0.000 1.794 1.811 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.763 1.784 prepare_preconditioner 11 7.9 0.000 0.000 1.753 1.757 make_preconditioner 11 8.9 0.000 0.000 1.753 1.757 fft_wrap_pw1pw2_140 487 13.2 0.324 0.331 1.692 1.703 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.641 1.667 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.636 1.636 fft3d_ps 1201 14.6 0.651 0.666 1.402 1.414 grid_collocate_task_list 119 9.7 1.296 1.393 1.296 1.393 jit_kernel_multiply 8 15.6 0.615 1.383 0.615 1.383 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.352 1.366 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.346 1.355 cp_fm_diag_elpa_base 50 14.0 1.318 1.336 1.345 1.353 potential_pw2rs 119 12.3 0.010 0.010 1.295 1.297 ot_diis_step 108 11.5 0.013 0.013 1.269 1.269 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.206 1.239 apply_single 119 13.6 0.000 0.000 1.206 1.239 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.223 1.223 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.099 1.099 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.786 1.025 make_images_data 4572 15.5 0.042 0.046 0.792 0.976 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.922 cp_fm_cholesky_invert 11 10.9 0.882 0.885 0.882 0.885 transfer_rs2pw 487 10.6 0.004 0.004 0.789 0.872 mp_alltoall_d11v 2130 13.8 0.792 0.858 0.792 0.858 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.815 0.822 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.767 0.770 acc_transpose_blocks_sync 27432 16.5 0.716 0.762 0.716 0.762 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.672 0.720 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.390 0.685 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=33.999000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=603.545455, yerr=8.413881 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 783.458304E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 47.850 47.851 qs_mol_dyn_low 1 2.0 0.003 0.003 47.589 47.598 qs_forces 11 3.9 0.002 0.003 47.529 47.530 qs_energies 11 4.9 0.001 0.001 45.513 45.516 scf_env_do_scf 11 5.9 0.001 0.001 34.927 34.927 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.053 27.054 velocity_verlet 10 3.0 0.002 0.002 24.264 24.269 dbcsr_multiply_generic 2286 12.5 0.102 0.104 23.820 24.060 multiply_cannon 2286 13.5 0.300 0.304 19.886 20.805 multiply_cannon_loop 2286 14.5 0.865 0.891 18.635 19.585 qs_scf_new_mos 108 7.5 0.001 0.001 17.274 17.370 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.273 17.370 ot_scf_mini 108 9.5 0.002 0.002 16.107 16.205 ot_mini 108 10.5 0.001 0.001 9.805 9.919 multiply_cannon_multrec 9144 15.5 3.301 4.594 8.781 8.894 init_scf_run 11 5.9 0.000 0.001 8.534 8.535 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.534 8.534 qs_ot_get_derivative 108 11.5 0.001 0.001 7.763 7.861 init_scf_loop 11 6.9 0.000 0.000 7.847 7.850 rebuild_ks_matrix 119 8.3 0.000 0.000 7.330 7.471 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.329 7.470 calculate_first_density_matrix 1 7.0 0.000 0.000 7.031 7.032 prepare_preconditioner 11 7.9 0.000 0.000 6.868 6.882 make_preconditioner 11 8.9 0.000 0.000 6.868 6.882 acc_transpose_blocks 9144 15.5 0.044 0.045 6.727 6.793 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.382 6.763 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.737 6.737 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.587 6.715 dbcsr_mm_accdrv_process 12550 15.8 4.811 6.541 5.341 6.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.468 6.510 acc_transpose_blocks_kernels 9144 16.5 0.120 0.121 5.558 5.585 jit_kernel_transpose 5 15.6 5.438 5.464 5.438 5.464 cp_fm_upper_to_full 72 14.2 3.214 4.646 3.214 4.646 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.283 4.290 calculate_rho_elec 119 8.7 0.117 0.120 4.282 4.290 sum_up_and_integrate 119 10.3 0.001 0.001 3.913 3.920 integrate_v_rspace 119 11.3 0.003 0.003 3.902 3.909 qs_ot_get_p 119 10.4 0.001 0.001 3.385 3.518 mp_waitall_1 94719 16.7 2.263 3.281 2.263 3.281 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.834 3.234 pw_transfer 1439 11.6 0.068 0.069 3.050 3.055 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.952 2.956 dbcsr_complete_redistribute 329 12.2 0.294 0.299 2.032 2.917 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.727 2.593 make_m2s 4572 13.5 0.037 0.038 2.384 2.530 fft_wrap_pw1pw2_140 487 13.2 0.649 0.651 2.509 2.514 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.189 2.502 apply_single 119 13.6 0.000 0.000 2.189 2.502 make_images 4572 14.5 0.355 0.387 2.264 2.409 mp_alltoall_i22 627 13.8 1.460 2.366 1.460 2.366 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.481 2.340 density_rs2pw 119 9.7 0.003 0.004 2.296 2.315 calculate_dm_sparse 119 9.5 0.000 0.000 2.267 2.283 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.286 2.204 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.082 2.134 grid_integrate_task_list 119 12.3 2.076 2.112 2.076 2.112 ot_diis_step 108 11.5 0.014 0.014 2.016 2.017 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.997 1.999 fft3d_ps 1201 14.6 0.866 0.885 1.873 1.878 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.771 1.772 mp_sum_l 7287 12.8 1.003 1.748 1.003 1.748 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.633 1.635 grid_collocate_task_list 119 9.7 1.519 1.553 1.519 1.553 wfi_extrapolate 11 7.9 0.001 0.001 1.440 1.440 potential_pw2rs 119 12.3 0.014 0.014 1.415 1.418 cp_fm_cholesky_invert 11 10.9 1.410 1.413 1.410 1.413 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.350 1.350 cp_fm_diag_elpa_base 50 14.0 1.203 1.258 1.349 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.302 1.320 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.086 1.310 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.263 1.271 mp_alltoall_d11v 2130 13.8 1.205 1.233 1.205 1.233 make_images_data 4572 15.5 0.046 0.049 1.008 1.200 acc_transpose_blocks_sync 27432 16.5 1.108 1.145 1.108 1.145 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.099 1.115 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.990 1.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.953 0.966 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.851000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=731.636364, yerr=20.042104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 502.874112E+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 0.009 0.029 88.186 88.186 qs_mol_dyn_low 1 2.0 0.003 0.003 87.929 87.939 qs_forces 11 3.9 0.003 0.003 87.855 87.856 qs_energies 11 4.9 0.001 0.002 84.999 85.017 scf_env_do_scf 11 5.9 0.001 0.001 70.977 70.980 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.446 65.447 dbcsr_multiply_generic 2055 12.4 0.105 0.108 56.928 57.103 multiply_cannon 2055 13.4 0.182 0.187 48.129 49.089 qs_scf_new_mos 99 7.5 0.000 0.001 48.201 48.319 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.201 48.319 multiply_cannon_loop 2055 14.4 1.787 1.834 47.125 48.080 ot_scf_mini 99 9.5 0.002 0.002 45.736 45.820 velocity_verlet 10 3.0 0.001 0.002 44.129 44.129 ot_mini 99 10.5 0.001 0.001 27.085 27.165 qs_ot_get_derivative 99 11.5 0.001 0.001 20.303 20.400 multiply_cannon_multrec 49320 15.4 11.327 11.995 17.348 18.003 rebuild_ks_matrix 110 8.3 0.000 0.000 14.530 14.643 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.529 14.643 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.745 12.854 mp_waitall_1 220248 16.4 11.365 12.109 11.365 12.109 init_scf_run 11 5.9 0.000 0.001 11.980 11.980 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.980 11.980 qs_ot_get_p 110 10.4 0.001 0.001 10.053 10.160 multiply_cannon_sync_h2d 49320 15.4 9.544 10.089 9.544 10.089 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 8.777 8.781 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.445 8.503 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.884 8.447 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 6.801 7.963 acc_transpose_blocks 49320 15.4 0.211 0.218 7.689 7.845 calculate_first_density_matrix 1 7.0 0.000 0.001 7.811 7.820 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.225 7.673 apply_single 110 13.6 0.000 0.000 7.224 7.673 sum_up_and_integrate 110 10.3 0.001 0.003 7.046 7.062 integrate_v_rspace 110 11.3 0.002 0.003 7.021 7.042 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.716 6.756 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.481 6.616 calculate_rho_elec 110 8.6 0.020 0.024 6.481 6.616 ot_diis_step 99 11.5 0.006 0.006 6.574 6.574 acc_transpose_blocks_kernels 49320 16.4 0.425 0.438 6.161 6.288 dbcsr_mm_accdrv_process 87628 16.1 3.065 3.138 5.891 6.170 jit_kernel_transpose 5 15.6 5.736 5.854 5.736 5.854 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.816 5.817 init_scf_loop 11 6.9 0.000 0.000 5.502 5.502 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.380 5.420 mp_sum_l 6594 12.7 4.396 5.247 4.396 5.247 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.222 5.244 cp_fm_diag_elpa_base 48 14.0 5.209 5.232 5.221 5.242 wfi_extrapolate 11 7.9 0.001 0.001 4.063 4.064 make_m2s 4110 13.4 0.061 0.064 3.945 4.056 make_images 4110 14.4 0.178 0.191 3.849 3.963 calculate_dm_sparse 110 9.5 0.001 0.001 3.834 3.945 density_rs2pw 110 9.6 0.004 0.004 3.450 3.651 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.645 3.623 grid_integrate_task_list 110 12.3 3.251 3.429 3.251 3.429 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.320 3.394 prepare_preconditioner 11 7.9 0.000 0.000 3.322 3.342 make_preconditioner 11 8.9 0.000 0.000 3.322 3.342 pw_transfer 1331 11.6 0.055 0.071 3.136 3.212 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.113 3.151 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.048 3.129 mp_alltoall_d11v 2046 13.8 2.145 2.702 2.145 2.702 fft_wrap_pw1pw2_140 451 13.1 0.309 0.332 2.595 2.679 potential_pw2rs 110 12.3 0.005 0.006 2.645 2.665 jit_kernel_multiply 13 15.9 2.546 2.565 2.546 2.565 fft3d_ps 1111 14.6 0.797 0.884 2.342 2.397 grid_collocate_task_list 110 9.6 2.160 2.302 2.160 2.302 transfer_rs2pw 451 10.6 0.005 0.006 1.948 2.083 mp_waitany 14300 13.8 1.827 2.076 1.827 2.076 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.913 1.928 make_images_data 4110 15.4 0.043 0.047 1.783 1.921 transfer_pw2rs 451 13.1 0.005 0.007 1.807 1.820 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.542 1.808 cp_fm_cholesky_invert 11 10.9 1.804 1.808 1.804 1.808 mp_sum_d 3891 11.9 1.354 1.801 1.354 1.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=88.186000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.454545, yerr=3.055952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 594.300928E+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 1947637. 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 0.008 0.027 75.648 75.648 qs_mol_dyn_low 1 2.0 0.003 0.003 75.387 75.398 qs_forces 11 3.9 0.003 0.003 75.320 75.321 qs_energies 11 4.9 0.001 0.001 71.965 71.971 scf_env_do_scf 11 5.9 0.000 0.001 58.235 58.238 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.313 50.314 dbcsr_multiply_generic 2055 12.4 0.116 0.119 43.665 43.881 multiply_cannon 2055 13.4 0.225 0.244 36.615 37.726 velocity_verlet 10 3.0 0.001 0.002 36.896 36.897 multiply_cannon_loop 2055 14.4 1.158 1.182 35.264 36.201 qs_scf_new_mos 99 7.5 0.001 0.001 34.247 34.367 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.246 34.367 ot_scf_mini 99 9.5 0.003 0.003 32.536 32.645 ot_mini 99 10.5 0.001 0.001 18.817 18.930 multiply_cannon_multrec 24660 15.4 6.918 8.359 13.965 15.336 rebuild_ks_matrix 110 8.3 0.000 0.000 13.684 13.822 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.683 13.822 qs_ot_get_derivative 99 11.5 0.001 0.001 12.976 13.081 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.018 12.136 init_scf_run 11 5.9 0.000 0.001 11.334 11.335 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.334 11.334 mp_waitall_1 176588 16.5 7.967 10.545 7.967 10.545 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 8.394 8.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.033 8.070 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.341 7.938 init_scf_loop 11 6.9 0.000 0.000 7.879 7.879 calculate_first_density_matrix 1 7.0 0.000 0.000 7.623 7.625 multiply_cannon_sync_h2d 24660 15.4 6.357 7.259 6.357 7.259 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.511 7.210 apply_single 110 13.6 0.000 0.001 6.511 7.210 dbcsr_mm_accdrv_process 52282 16.1 5.449 6.081 6.878 7.204 acc_transpose_blocks 24660 15.4 0.117 0.120 7.059 7.122 qs_ot_get_p 110 10.4 0.001 0.001 6.942 7.071 sum_up_and_integrate 110 10.3 0.001 0.002 6.475 6.485 integrate_v_rspace 110 11.3 0.002 0.003 6.448 6.461 acc_transpose_blocks_kernels 24660 16.4 0.321 0.329 5.993 6.047 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.033 6.046 calculate_rho_elec 110 8.6 0.039 0.047 6.032 6.045 prepare_preconditioner 11 7.9 0.000 0.000 5.840 5.860 make_preconditioner 11 8.9 0.000 0.000 5.840 5.860 ot_diis_step 99 11.5 0.010 0.010 5.783 5.783 jit_kernel_transpose 5 15.6 5.672 5.718 5.672 5.718 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.394 5.549 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.826 5.533 make_m2s 4110 13.4 0.057 0.060 4.471 4.917 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.881 4.903 make_images 4110 14.4 0.406 0.463 4.358 4.799 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.409 4.410 pw_transfer 1331 11.6 0.065 0.072 3.687 3.825 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.801 3.817 cp_fm_diag_elpa_base 48 14.0 3.754 3.770 3.799 3.814 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.581 3.721 wfi_extrapolate 11 7.9 0.001 0.001 3.605 3.605 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.439 3.496 density_rs2pw 110 9.6 0.004 0.004 3.246 3.445 grid_integrate_task_list 110 12.3 3.172 3.321 3.172 3.321 fft_wrap_pw1pw2_140 451 13.1 0.358 0.379 3.017 3.161 calculate_dm_sparse 110 9.5 0.001 0.001 3.015 3.050 make_images_data 4110 15.4 0.048 0.053 2.470 2.935 hybrid_alltoall_any 4261 16.3 0.105 0.459 2.160 2.928 fft3d_ps 1111 14.6 1.117 1.353 2.673 2.831 cp_fm_cholesky_invert 11 10.9 2.771 2.779 2.771 2.779 mp_sum_l 6594 12.7 1.944 2.617 1.944 2.617 potential_pw2rs 110 12.3 0.008 0.009 2.327 2.341 grid_collocate_task_list 110 9.6 2.174 2.329 2.174 2.329 jit_kernel_multiply 11 16.2 1.062 2.108 1.062 2.108 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.040 2.055 mp_alltoall_d11v 2046 13.8 1.869 2.054 1.869 2.054 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.868 1.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.823 1.841 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.781 1.798 mp_allgather_i34 2055 14.4 0.659 1.762 0.659 1.762 transfer_rs2pw 451 10.6 0.006 0.007 1.469 1.654 mp_irecv_dv 57340 16.2 0.656 1.578 0.656 1.578 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.545 1.557 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=75.648000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.818182, yerr=6.176468 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 664.346624E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.042 65.804 65.806 qs_mol_dyn_low 1 2.0 0.003 0.003 65.208 65.220 qs_forces 11 3.9 0.003 0.003 65.142 65.143 qs_energies 11 4.9 0.001 0.002 61.947 61.951 scf_env_do_scf 11 5.9 0.000 0.001 49.013 49.013 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.434 40.434 dbcsr_multiply_generic 2055 12.4 0.109 0.112 33.802 34.051 velocity_verlet 10 3.0 0.002 0.002 32.823 32.824 multiply_cannon 2055 13.4 0.212 0.220 27.443 28.655 multiply_cannon_loop 2055 14.4 0.813 0.841 26.264 27.292 qs_scf_new_mos 99 7.5 0.001 0.001 25.546 25.669 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.545 25.668 ot_scf_mini 99 9.5 0.002 0.003 24.285 24.419 ot_mini 99 10.5 0.001 0.001 13.812 13.944 rebuild_ks_matrix 110 8.3 0.000 0.000 12.270 12.404 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.270 12.403 multiply_cannon_multrec 16440 15.4 3.694 4.920 10.045 11.180 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.802 10.920 init_scf_run 11 5.9 0.000 0.001 10.521 10.522 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.521 10.521 mp_waitall_1 139946 16.5 6.973 10.000 6.973 10.000 qs_ot_get_derivative 99 11.5 0.001 0.001 9.412 9.544 init_scf_loop 11 6.9 0.000 0.000 8.546 8.546 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.706 7.707 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.369 7.423 calculate_first_density_matrix 1 7.0 0.000 0.000 7.392 7.393 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.248 7.146 prepare_preconditioner 11 7.9 0.000 0.000 6.796 6.814 make_preconditioner 11 8.9 0.000 0.000 6.796 6.814 acc_transpose_blocks 16440 15.4 0.079 0.082 6.410 6.506 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.118 6.481 sum_up_and_integrate 110 10.3 0.001 0.002 6.398 6.411 dbcsr_mm_accdrv_process 34862 16.1 5.504 5.813 6.196 6.386 integrate_v_rspace 110 11.3 0.003 0.003 6.372 6.385 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.891 5.898 calculate_rho_elec 110 8.6 0.059 0.059 5.890 5.897 qs_ot_get_p 110 10.4 0.001 0.001 5.628 5.766 acc_transpose_blocks_kernels 16440 16.4 0.218 0.221 5.620 5.653 jit_kernel_transpose 5 15.6 5.402 5.433 5.402 5.433 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.862 5.344 apply_single 110 13.6 0.000 0.000 4.862 5.344 make_m2s 4110 13.4 0.049 0.050 4.320 4.691 make_images 4110 14.4 0.396 0.517 4.202 4.570 ot_diis_step 99 11.5 0.011 0.011 4.370 4.371 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.970 3.975 multiply_cannon_sync_h2d 16440 15.4 3.239 3.893 3.239 3.893 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.241 3.846 pw_transfer 1331 11.6 0.065 0.072 3.744 3.754 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.637 3.651 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.585 3.586 grid_integrate_task_list 110 12.3 3.206 3.394 3.206 3.394 density_rs2pw 110 9.6 0.004 0.004 3.100 3.291 fft_wrap_pw1pw2_140 451 13.1 0.460 0.471 3.116 3.134 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.031 3.039 cp_fm_diag_elpa_base 48 14.0 2.965 2.991 3.029 3.038 wfi_extrapolate 11 7.9 0.001 0.001 3.037 3.037 make_images_data 4110 15.4 0.046 0.050 2.426 2.930 hybrid_alltoall_any 4261 16.3 0.108 0.378 2.078 2.772 cp_fm_cholesky_invert 11 10.9 2.662 2.669 2.662 2.669 calculate_dm_sparse 110 9.5 0.001 0.001 2.589 2.622 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.529 2.605 fft3d_ps 1111 14.6 1.121 1.137 2.556 2.571 grid_collocate_task_list 110 9.6 2.220 2.426 2.220 2.426 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.876 2.407 mp_irecv_dv 48980 15.7 0.802 2.275 0.802 2.275 potential_pw2rs 110 12.3 0.011 0.011 2.220 2.225 mp_sum_l 6594 12.7 1.415 2.084 1.415 2.084 mp_alltoall_d11v 2046 13.8 1.793 2.011 1.793 2.011 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.976 1.977 dbcsr_complete_redistribute 325 12.2 0.318 0.350 1.460 1.921 cp_fm_upper_to_full 70 14.2 1.409 1.801 1.409 1.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.660 1.677 cp_fm_cholesky_decompose 22 10.9 1.617 1.635 1.617 1.635 mp_allgather_i34 2055 14.4 0.493 1.539 0.493 1.539 transfer_rs2pw 451 10.6 0.005 0.005 1.330 1.533 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.513 1.522 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.374 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.971 1.428 mp_waitany 17072 13.8 1.163 1.342 1.163 1.342 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.806000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.727273, yerr=7.932567 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 739.889152E+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 0.021 0.095 66.339 66.341 qs_mol_dyn_low 1 2.0 0.003 0.005 66.027 66.037 qs_forces 11 3.9 0.003 0.005 65.961 65.962 qs_energies 11 4.9 0.005 0.035 62.555 62.589 scf_env_do_scf 11 5.9 0.001 0.002 53.923 53.926 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.194 42.195 velocity_verlet 10 3.0 0.001 0.002 37.372 37.374 dbcsr_multiply_generic 2055 12.4 0.120 0.141 30.755 30.904 qs_scf_new_mos 99 7.5 0.001 0.001 27.494 27.596 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.493 27.596 ot_scf_mini 99 9.5 0.003 0.003 25.805 25.903 multiply_cannon 2055 13.4 0.238 0.252 23.552 24.483 multiply_cannon_loop 2055 14.4 1.439 1.490 22.204 22.713 ot_mini 99 10.5 0.001 0.001 14.840 14.963 multiply_cannon_multrec 24660 15.4 4.092 6.642 12.985 14.266 rebuild_ks_matrix 110 8.3 0.000 0.000 12.073 12.197 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.072 12.196 init_scf_loop 11 6.9 0.001 0.003 11.690 11.691 qs_ot_get_derivative 99 11.5 0.001 0.001 10.694 10.797 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.681 10.792 prepare_preconditioner 11 7.9 0.000 0.000 9.957 9.971 make_preconditioner 11 8.9 0.000 0.002 9.957 9.971 dbcsr_mm_accdrv_process 52304 16.0 7.702 9.003 8.735 9.702 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.169 9.650 sum_up_and_integrate 110 10.3 0.001 0.002 6.326 6.339 integrate_v_rspace 110 11.3 0.003 0.003 6.300 6.312 mp_waitall_1 121746 16.5 4.286 6.135 4.286 6.135 qs_ot_get_p 110 10.4 0.001 0.002 5.841 5.976 init_scf_run 11 5.9 0.000 0.001 5.863 5.864 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.863 5.863 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.829 5.843 calculate_rho_elec 110 8.6 0.078 0.081 5.828 5.843 make_m2s 4110 13.4 0.059 0.062 5.534 5.803 make_images 4110 14.4 0.574 0.694 5.393 5.657 cp_fm_upper_to_full 70 14.2 3.369 4.896 3.369 4.896 ot_diis_step 99 11.5 0.011 0.011 4.109 4.110 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.025 4.092 apply_single 110 13.6 0.000 0.000 4.025 4.092 dbcsr_complete_redistribute 325 12.2 0.555 0.632 2.852 4.009 qs_ot_p2m_diag 48 11.0 0.055 0.065 3.963 3.979 pw_transfer 1331 11.6 0.065 0.072 3.788 3.822 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.681 3.718 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.538 3.583 copy_fm_to_dbcsr 174 11.2 0.002 0.008 2.338 3.482 grid_integrate_task_list 110 12.3 3.274 3.461 3.274 3.461 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.386 3.447 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.398 3.398 fft_wrap_pw1pw2_140 451 13.1 0.530 0.542 3.158 3.198 calculate_dm_sparse 110 9.5 0.001 0.001 3.042 3.087 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.073 3.075 density_rs2pw 110 9.6 0.004 0.004 2.951 3.066 make_images_data 4110 15.4 0.049 0.053 2.760 3.055 hybrid_alltoall_any 4261 16.3 0.124 0.462 2.356 3.033 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.321 3.006 wfi_extrapolate 11 7.9 0.001 0.001 2.979 2.979 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.779 2.913 mp_alltoall_i22 605 13.7 1.702 2.883 1.702 2.883 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.856 2.864 cp_fm_diag_elpa_base 48 14.0 2.704 2.763 2.853 2.861 calculate_first_density_matrix 1 7.0 0.001 0.005 2.779 2.783 cp_fm_cholesky_invert 11 10.9 2.734 2.742 2.734 2.742 acc_transpose_blocks 24660 15.4 0.113 0.116 2.531 2.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.514 2.559 fft3d_ps 1111 14.6 1.152 1.189 2.508 2.527 multiply_cannon_sync_h2d 24660 15.4 2.369 2.510 2.369 2.510 grid_collocate_task_list 110 9.6 2.262 2.411 2.262 2.411 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.260 2.262 potential_pw2rs 110 12.3 0.013 0.013 2.125 2.127 mp_alltoall_d11v 2046 13.8 1.791 2.010 1.791 2.010 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.802 1.835 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.611 1.709 cp_fm_cholesky_decompose 22 10.9 1.659 1.706 1.659 1.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.631 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.602 1.619 mp_sum_l 6594 12.7 0.988 1.595 0.988 1.595 acc_transpose_blocks_sync 73980 16.4 1.448 1.581 1.448 1.581 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.846 1.528 mp_irecv_dv 62702 16.1 0.743 1.445 0.743 1.445 jit_kernel_multiply 8 15.7 0.689 1.425 0.689 1.425 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.341000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.363636, yerr=11.022891 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 847.224832E+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 0.011 0.031 61.795 61.796 qs_mol_dyn_low 1 2.0 0.003 0.003 61.567 61.577 qs_forces 11 3.9 0.003 0.003 61.230 61.231 qs_energies 11 4.9 0.001 0.001 57.529 57.533 scf_env_do_scf 11 5.9 0.000 0.001 43.974 43.974 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.049 36.051 velocity_verlet 10 3.0 0.002 0.002 32.134 32.154 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.546 28.659 multiply_cannon 2055 13.4 0.240 0.259 22.804 24.214 multiply_cannon_loop 2055 14.4 0.603 0.621 21.512 21.797 qs_scf_new_mos 99 7.5 0.001 0.001 21.145 21.204 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.144 21.203 ot_scf_mini 99 9.5 0.002 0.002 19.879 19.917 rebuild_ks_matrix 110 8.3 0.000 0.000 11.854 11.881 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.853 11.881 ot_mini 99 10.5 0.001 0.001 10.896 10.926 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.537 10.564 init_scf_run 11 5.9 0.000 0.001 10.402 10.402 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.401 10.402 multiply_cannon_multrec 8220 15.4 3.187 4.366 7.575 8.452 init_scf_loop 11 6.9 0.000 0.000 7.877 7.878 mp_waitall_1 103326 16.6 6.051 7.686 6.051 7.686 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.624 7.626 calculate_first_density_matrix 1 7.0 0.000 0.000 7.501 7.503 qs_ot_get_derivative 99 11.5 0.001 0.001 7.207 7.245 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.230 7.242 acc_transpose_blocks 8220 15.4 0.041 0.042 6.437 6.488 sum_up_and_integrate 110 10.3 0.001 0.002 6.353 6.364 integrate_v_rspace 110 11.3 0.003 0.003 6.326 6.338 prepare_preconditioner 11 7.9 0.000 0.000 6.215 6.220 make_preconditioner 11 8.9 0.000 0.000 6.215 6.220 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.172 6.187 calculate_rho_elec 110 8.6 0.113 0.113 6.171 6.186 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.782 5.861 acc_transpose_blocks_kernels 8220 16.4 0.114 0.116 5.714 5.749 jit_kernel_transpose 5 15.6 5.600 5.634 5.600 5.634 dbcsr_mm_accdrv_process 17442 15.9 3.065 4.095 4.249 5.167 qs_ot_get_p 110 10.4 0.002 0.002 4.995 5.017 make_m2s 4110 13.4 0.038 0.039 4.361 4.653 make_images 4110 14.4 0.643 0.706 4.230 4.523 pw_transfer 1331 11.6 0.065 0.069 4.236 4.244 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.946 4.194 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.129 4.140 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.722 3.767 apply_single 110 13.6 0.000 0.000 3.722 3.767 ot_diis_step 99 11.5 0.012 0.012 3.663 3.664 fft_wrap_pw1pw2_140 451 13.1 0.716 0.728 3.561 3.574 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.557 3.561 grid_integrate_task_list 110 12.3 3.376 3.486 3.376 3.486 density_rs2pw 110 9.6 0.004 0.004 3.172 3.294 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.209 3.209 cp_fm_cholesky_invert 11 10.9 2.984 2.988 2.984 2.988 wfi_extrapolate 11 7.9 0.001 0.001 2.802 2.802 hybrid_alltoall_any 4261 16.3 0.200 0.846 2.308 2.761 make_images_data 4110 15.4 0.040 0.046 2.354 2.740 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.680 2.687 cp_fm_diag_elpa_base 48 14.0 2.622 2.645 2.678 2.686 fft3d_ps 1111 14.6 1.293 1.302 2.676 2.685 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.684 2.685 calculate_dm_sparse 110 9.5 0.001 0.001 2.494 2.532 multiply_cannon_sync_h2d 8220 15.4 2.374 2.520 2.374 2.520 grid_collocate_task_list 110 9.6 2.365 2.442 2.365 2.442 potential_pw2rs 110 12.3 0.015 0.016 2.108 2.113 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.072 2.104 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 2.010 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.973 1.993 mp_alltoall_d11v 2046 13.8 1.745 1.910 1.745 1.910 cp_fm_cholesky_decompose 22 10.9 1.770 1.791 1.770 1.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.649 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.514 1.626 dbcsr_complete_redistribute 325 12.2 0.559 0.584 1.490 1.602 mp_allgather_i34 2055 14.4 0.465 1.598 0.465 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.437 1.449 qs_create_task_list 11 7.9 0.000 0.000 1.220 1.317 generate_qs_task_list 11 8.9 0.375 0.441 1.220 1.317 transfer_rs2pw 451 10.6 0.005 0.005 1.189 1.313 mp_waitany 9240 13.8 1.094 1.246 1.094 1.246 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=61.796000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=802.272727, yerr=7.072236 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 1.350738E+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 0.034 0.083 93.668 93.674 qs_mol_dyn_low 1 2.0 0.003 0.003 93.269 93.293 qs_forces 11 3.9 0.003 0.004 92.346 92.347 qs_energies 11 4.9 0.003 0.005 88.180 88.182 scf_env_do_scf 11 5.9 0.000 0.001 72.943 72.943 velocity_verlet 10 3.0 0.002 0.002 56.728 56.841 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.622 44.623 dbcsr_multiply_generic 2055 12.4 0.121 0.126 34.328 34.420 init_scf_loop 11 6.9 0.000 0.000 28.251 28.252 multiply_cannon 2055 13.4 0.331 0.366 27.352 28.144 qs_scf_new_mos 99 7.5 0.001 0.001 26.650 26.686 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.649 26.685 prepare_preconditioner 11 7.9 0.000 0.000 26.233 26.241 make_preconditioner 11 8.9 0.000 0.000 26.233 26.241 multiply_cannon_loop 2055 14.4 0.831 0.845 25.570 25.809 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.418 25.724 ot_scf_mini 99 9.5 0.002 0.002 24.857 24.883 cp_fm_upper_to_full 70 14.2 12.695 18.250 12.695 18.250 ot_mini 99 10.5 0.001 0.001 13.989 14.015 rebuild_ks_matrix 110 8.3 0.000 0.001 13.978 14.011 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.977 14.010 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.671 12.701 init_scf_run 11 5.9 0.000 0.001 10.947 10.947 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.946 10.946 dbcsr_complete_redistribute 325 12.2 1.022 1.037 7.399 10.622 multiply_cannon_multrec 8220 15.4 4.040 4.223 9.699 9.847 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.374 9.603 qs_ot_get_derivative 99 11.5 0.001 0.001 9.425 9.449 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.800 8.991 mp_alltoall_i22 605 13.7 5.440 8.693 5.440 8.693 mp_waitall_1 84994 16.7 7.404 8.246 7.404 8.246 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.950 7.952 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.509 7.544 calculate_rho_elec 110 8.6 0.222 0.223 7.509 7.543 calculate_first_density_matrix 1 7.0 0.000 0.000 7.494 7.497 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.439 7.456 sum_up_and_integrate 110 10.3 0.002 0.002 6.947 6.962 integrate_v_rspace 110 11.3 0.003 0.004 6.919 6.934 acc_transpose_blocks 8220 15.4 0.041 0.042 6.802 6.829 pw_transfer 1331 11.6 0.074 0.075 5.848 5.855 make_m2s 4110 13.4 0.043 0.043 5.305 5.795 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.731 5.738 acc_transpose_blocks_kernels 8220 16.4 0.117 0.118 5.715 5.729 dbcsr_mm_accdrv_process 11614 15.7 3.841 4.100 5.512 5.728 qs_ot_get_p 110 10.4 0.003 0.004 5.595 5.625 jit_kernel_transpose 5 15.6 5.598 5.612 5.598 5.612 make_images 4110 14.4 0.881 0.935 5.115 5.604 cp_fm_cholesky_invert 11 10.9 5.202 5.206 5.202 5.206 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.699 5.165 apply_single 110 13.6 0.000 0.000 4.699 5.164 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.764 5.087 fft_wrap_pw1pw2_140 451 13.1 1.366 1.370 5.025 5.034 ot_diis_step 99 11.5 0.015 0.016 4.546 4.546 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.038 4.046 density_rs2pw 110 9.6 0.004 0.004 3.967 3.988 grid_integrate_task_list 110 12.3 3.680 3.760 3.680 3.760 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.683 3.683 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.522 3.522 hybrid_alltoall_any 4261 16.3 0.264 0.564 2.785 3.515 fft3d_ps 1111 14.6 1.882 1.887 3.437 3.444 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.992 3.438 make_images_data 4110 15.4 0.045 0.048 2.784 3.392 wfi_extrapolate 11 7.9 0.001 0.001 3.332 3.332 calculate_dm_sparse 110 9.5 0.001 0.001 3.163 3.186 multiply_cannon_sync_h2d 8220 15.4 3.112 3.136 3.112 3.136 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.936 2.936 cp_fm_diag_elpa_base 48 14.0 2.394 2.596 2.934 2.934 grid_collocate_task_list 110 9.6 2.671 2.690 2.671 2.690 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.630 2.644 potential_pw2rs 110 12.3 0.021 0.021 2.472 2.475 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.202 2.256 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.169 mp_alltoall_d11v 2046 13.8 2.007 2.055 2.007 2.055 cp_fm_cholesky_decompose 22 10.9 2.010 2.024 2.010 2.024 qs_create_task_list 11 7.9 0.000 0.000 1.877 1.924 generate_qs_task_list 11 8.9 0.735 0.786 1.877 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.903 1.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=93.674000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1235.727273, yerr=45.463454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/15/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 1.094965E+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 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 631.001088E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2246334. 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 0.014 0.035 207.317 207.318 qs_mol_dyn_low 1 2.0 0.003 0.003 206.770 206.785 qs_forces 11 3.9 0.033 0.044 206.664 206.665 qs_energies 11 4.9 0.001 0.002 201.043 201.085 scf_env_do_scf 11 5.9 0.001 0.002 184.639 184.643 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.226 164.228 dbcsr_multiply_generic 2507 12.6 0.177 0.182 126.272 126.630 qs_scf_new_mos 117 7.6 0.001 0.001 124.887 125.103 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.887 125.103 velocity_verlet 10 3.0 0.001 0.002 123.788 123.789 ot_scf_mini 117 9.6 0.003 0.003 118.217 118.430 multiply_cannon 2507 13.6 0.237 0.246 101.995 103.457 multiply_cannon_loop 2507 14.6 2.397 2.451 99.856 101.309 ot_mini 117 10.6 0.001 0.001 66.802 67.004 multiply_cannon_multrec 60168 15.6 31.883 33.720 41.711 43.715 qs_ot_get_derivative 117 11.6 0.001 0.001 41.908 42.151 rebuild_ks_matrix 128 8.3 0.001 0.001 33.548 33.970 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.547 33.970 mp_waitall_1 267128 16.5 29.216 32.076 29.216 32.076 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.114 30.512 qs_ot_get_p 128 10.4 0.001 0.001 29.158 29.429 multiply_cannon_sync_h2d 60168 15.6 26.307 28.497 26.307 28.497 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.386 24.938 apply_single 128 13.6 0.001 0.001 24.385 24.937 ot_diis_step 117 11.6 0.008 0.008 24.612 24.613 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.608 22.676 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.245 20.430 init_scf_loop 11 6.9 0.000 0.000 20.339 20.340 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.720 19.720 multiply_cannon_metrocomm3 60168 15.6 0.116 0.121 16.431 18.741 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.774 16.808 cp_fm_diag_elpa_base 83 14.4 16.699 16.741 16.770 16.805 prepare_preconditioner 11 7.9 0.000 0.000 15.747 15.785 make_preconditioner 11 8.9 0.000 0.000 15.747 15.785 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.980 15.154 make_m2s 5014 13.6 0.105 0.111 13.821 14.244 make_images 5014 14.6 0.397 0.415 13.634 14.063 sum_up_and_integrate 128 10.3 0.002 0.004 13.939 13.954 integrate_v_rspace 128 11.3 0.003 0.004 13.880 13.899 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.087 13.273 calculate_rho_elec 128 8.7 0.045 0.064 13.087 13.273 init_scf_run 11 5.9 0.000 0.001 12.266 12.267 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.266 12.267 mp_sum_l 7950 12.9 9.201 10.313 9.201 10.313 dbcsr_mm_accdrv_process 124484 16.2 4.734 4.867 9.399 9.916 wfi_extrapolate 11 7.9 0.001 0.001 9.026 9.027 cp_fm_cholesky_invert 11 10.9 8.908 8.916 8.908 8.916 calculate_dm_sparse 128 9.5 0.001 0.001 8.480 8.554 multiply_cannon_metrocomm1 60168 15.6 0.094 0.097 6.340 8.437 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.198 8.335 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.088 8.164 pw_transfer 1547 11.6 0.073 0.088 7.575 7.751 make_images_data 5014 15.6 0.066 0.072 6.746 7.712 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.373 7.547 grid_integrate_task_list 128 12.3 7.042 7.477 7.042 7.477 density_rs2pw 128 9.7 0.006 0.007 6.711 7.423 hybrid_alltoall_any 5200 16.5 0.293 2.256 5.933 7.139 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.685 6.698 fft_wrap_pw1pw2_140 523 13.2 0.855 0.891 6.483 6.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.804 5.900 fft3d_ps 1291 14.7 2.195 2.857 5.384 5.730 mp_alltoall_d11v 2415 14.1 4.355 5.394 4.355 5.394 grid_collocate_task_list 128 9.7 4.843 5.204 4.843 5.204 cp_fm_cholesky_decompose 22 10.9 4.656 4.673 4.656 4.673 potential_pw2rs 128 12.3 0.009 0.010 4.446 4.467 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.318000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.636364, yerr=5.973772 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 845.815808E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 6491806. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.100 192.771 192.774 qs_mol_dyn_low 1 2.0 0.009 0.043 192.223 192.308 qs_forces 11 3.9 0.006 0.020 192.018 192.021 qs_energies 11 4.9 0.003 0.016 185.199 185.217 scf_env_do_scf 11 5.9 0.001 0.001 166.295 166.304 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.877 132.879 velocity_verlet 10 3.0 0.005 0.035 119.104 119.110 dbcsr_multiply_generic 2485 12.5 0.186 0.193 99.294 100.557 qs_scf_new_mos 116 7.6 0.001 0.001 94.482 95.015 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.482 95.014 ot_scf_mini 116 9.6 0.004 0.004 89.698 90.283 multiply_cannon 2485 13.5 0.498 0.552 78.910 84.290 multiply_cannon_loop 2485 14.5 1.550 1.621 75.902 78.727 ot_mini 116 10.6 0.001 0.002 50.091 50.648 mp_waitall_1 212858 16.6 24.392 38.384 24.392 38.384 multiply_cannon_multrec 29820 15.5 20.666 25.451 31.175 36.465 init_scf_loop 11 6.9 0.000 0.000 33.329 33.331 rebuild_ks_matrix 127 8.3 0.001 0.001 32.170 32.903 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.042 32.169 32.902 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.872 29.555 prepare_preconditioner 11 7.9 0.000 0.000 29.021 29.098 make_preconditioner 11 8.9 0.000 0.000 29.021 29.098 qs_ot_get_derivative 116 11.6 0.001 0.002 28.244 28.826 multiply_cannon_metrocomm3 29820 15.5 0.097 0.105 15.828 28.632 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.723 28.277 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 21.975 23.090 apply_single 127 13.6 0.001 0.001 21.975 23.089 qs_ot_get_p 127 10.4 0.001 0.002 21.779 22.523 ot_diis_step 116 11.6 0.016 0.030 21.663 21.674 multiply_cannon_sync_h2d 29820 15.5 17.767 20.644 17.767 20.644 cp_fm_cholesky_invert 11 10.9 16.983 16.995 16.983 16.995 qs_ot_p2m_diag 82 11.4 0.186 0.213 16.804 16.844 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.596 15.597 make_m2s 4970 13.5 0.086 0.091 14.284 15.589 make_images 4970 14.5 1.152 1.345 14.068 15.374 sum_up_and_integrate 127 10.3 0.002 0.006 14.060 14.096 integrate_v_rspace 127 11.3 0.003 0.013 14.000 14.040 init_scf_run 11 5.9 0.000 0.001 14.018 14.019 scf_env_initial_rho_setup 11 6.9 0.001 0.002 14.017 14.019 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.377 13.406 calculate_rho_elec 127 8.7 0.086 0.103 13.376 13.406 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.422 12.459 cp_fm_diag_elpa_base 82 14.4 12.170 12.263 12.416 12.449 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.178 11.600 multiply_cannon_metrocomm4 27335 15.5 0.099 0.112 3.857 10.698 dbcsr_mm_accdrv_process 61726 16.2 5.418 6.024 9.970 10.523 mp_irecv_dv 68888 16.3 3.658 10.303 3.658 10.303 make_images_data 4970 15.5 0.065 0.075 8.459 10.038 hybrid_alltoall_any 5155 16.4 0.347 1.517 7.153 9.413 pw_transfer 1535 11.6 0.083 0.093 8.695 8.753 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 8.564 8.575 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.474 8.533 wfi_extrapolate 11 7.9 0.004 0.023 8.365 8.365 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.685 7.873 density_rs2pw 127 9.7 0.006 0.006 7.117 7.581 fft_wrap_pw1pw2_140 519 13.2 0.925 0.959 7.461 7.527 grid_integrate_task_list 127 12.3 7.112 7.469 7.112 7.469 cp_fm_cholesky_decompose 22 10.9 7.124 7.194 7.124 7.194 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.291 7.033 calculate_dm_sparse 127 9.5 0.001 0.001 6.471 6.639 mp_sum_l 7884 12.9 4.339 6.555 4.339 6.555 fft3d_ps 1281 14.7 2.803 2.998 6.054 6.113 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.391 5.480 calculate_first_density_matrix 1 7.0 0.000 0.001 5.468 5.473 mp_alltoall_d11v 2401 14.1 4.205 5.313 4.205 5.313 grid_collocate_task_list 127 9.7 4.964 5.305 4.964 5.305 acc_transpose_blocks 29820 15.5 0.150 0.158 4.900 4.951 mp_allgather_i34 2485 14.5 1.542 4.700 1.542 4.700 potential_pw2rs 127 12.3 0.016 0.018 4.600 4.620 mp_sum_d 4456 12.1 2.616 4.036 2.616 4.036 dbcsr_complete_redistribute 393 12.7 0.761 0.850 3.170 4.019 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.774000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.727273, yerr=5.064868 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 955.584512E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58221. MP_Allreduce 11080 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 176.054 176.054 qs_mol_dyn_low 1 2.0 0.003 0.003 175.640 175.653 qs_forces 11 3.9 0.004 0.004 175.548 175.550 qs_energies 11 4.9 0.001 0.002 169.016 169.028 scf_env_do_scf 11 5.9 0.001 0.001 153.293 153.294 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 118.074 118.075 velocity_verlet 10 3.0 0.001 0.002 112.175 112.177 dbcsr_multiply_generic 2485 12.5 0.199 0.202 81.622 82.758 qs_scf_new_mos 116 7.6 0.001 0.001 80.978 81.280 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.977 81.280 ot_scf_mini 116 9.6 0.003 0.004 76.760 77.128 multiply_cannon 2485 13.5 0.500 0.517 61.925 66.914 multiply_cannon_loop 2485 14.5 1.112 1.187 58.939 61.467 ot_mini 116 10.6 0.001 0.001 42.522 42.881 init_scf_loop 11 6.9 0.000 0.000 35.115 35.116 mp_waitall_1 169034 16.6 24.608 34.245 24.608 34.245 prepare_preconditioner 11 7.9 0.000 0.000 31.101 31.146 make_preconditioner 11 8.9 0.000 0.000 31.101 31.146 rebuild_ks_matrix 127 8.3 0.001 0.001 30.076 30.498 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 30.076 30.498 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.713 30.124 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.098 27.490 multiply_cannon_multrec 19880 15.5 12.975 16.092 22.231 25.356 multiply_cannon_metrocomm3 19880 15.5 0.060 0.063 15.170 24.308 qs_ot_get_derivative 116 11.6 0.001 0.002 23.132 23.496 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.463 20.583 apply_single 127 13.6 0.001 0.001 19.463 20.582 qs_ot_get_p 127 10.4 0.001 0.002 19.769 20.267 ot_diis_step 116 11.6 0.018 0.020 19.268 19.269 make_m2s 4970 13.5 0.077 0.084 14.648 15.913 make_images 4970 14.5 1.134 1.223 14.413 15.676 qs_ot_p2m_diag 82 11.4 0.261 0.268 15.365 15.374 multiply_cannon_sync_h2d 19880 15.5 13.536 15.265 13.536 15.265 cp_fm_cholesky_invert 11 10.9 14.521 14.530 14.521 14.530 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.302 14.303 sum_up_and_integrate 127 10.3 0.002 0.003 14.000 14.024 integrate_v_rspace 127 11.3 0.003 0.004 13.941 13.963 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.715 13.744 calculate_rho_elec 127 8.7 0.128 0.142 13.714 13.744 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.076 11.089 cp_fm_diag_elpa_base 82 14.4 10.663 10.828 11.073 11.086 init_scf_run 11 5.9 0.000 0.001 10.730 10.730 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.730 10.730 make_images_data 4970 15.5 0.063 0.073 8.692 10.441 hybrid_alltoall_any 5155 16.4 0.443 2.019 7.427 9.794 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.161 9.434 multiply_cannon_metrocomm4 17395 15.5 0.064 0.072 3.472 9.237 mp_irecv_dv 49801 16.2 3.345 8.987 3.345 8.987 pw_transfer 1535 11.6 0.084 0.101 8.870 8.975 dbcsr_mm_accdrv_process 41158 16.2 5.586 5.953 8.714 8.848 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.649 8.758 fft_wrap_pw1pw2_140 519 13.2 1.045 1.088 7.625 7.741 grid_integrate_task_list 127 12.3 7.257 7.708 7.257 7.708 wfi_extrapolate 11 7.9 0.001 0.001 7.462 7.462 cp_fm_cholesky_decompose 22 10.9 7.370 7.414 7.370 7.414 density_rs2pw 127 9.7 0.005 0.006 7.067 7.349 cp_fm_upper_to_full 104 14.8 5.778 7.332 5.778 7.332 dbcsr_complete_redistribute 393 12.7 1.295 1.339 4.817 6.608 fft3d_ps 1281 14.7 2.797 3.036 5.995 6.079 calculate_dm_sparse 127 9.5 0.001 0.001 5.827 5.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.687 5.694 grid_collocate_task_list 127 9.7 5.113 5.602 5.113 5.602 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.628 5.419 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.729 5.334 mp_alltoall_d11v 2401 14.1 4.455 4.811 4.455 4.811 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.628 4.765 mp_sum_l 7884 12.9 3.218 4.750 3.218 4.750 potential_pw2rs 127 12.3 0.021 0.022 4.455 4.476 mp_allgather_i34 2485 14.5 1.499 4.473 1.499 4.473 transfer_fm_to_dbcsr 11 9.9 0.019 0.022 2.370 4.135 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.896 3.922 mp_alltoall_i22 712 14.1 1.996 3.917 1.996 3.917 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.702 3.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.054000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=906.909091, yerr=8.585977 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/18/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 4.353788E+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 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 1.146307E+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 1.770319E+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 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. 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 0.015 0.033 192.080 192.081 qs_mol_dyn_low 1 2.0 0.003 0.003 191.721 191.736 qs_forces 11 3.9 0.005 0.005 191.617 191.621 qs_energies 11 4.9 0.001 0.002 184.521 184.532 scf_env_do_scf 11 5.9 0.001 0.001 162.906 162.918 velocity_verlet 10 3.0 0.002 0.002 124.216 124.218 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.537 116.538 dbcsr_multiply_generic 2507 12.6 0.192 0.198 85.391 86.033 qs_scf_new_mos 117 7.6 0.001 0.001 80.181 80.491 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.181 80.490 ot_scf_mini 117 9.6 0.003 0.003 75.625 75.958 multiply_cannon 2507 13.6 0.552 0.582 60.879 64.201 multiply_cannon_loop 2507 14.6 1.870 1.931 57.299 58.913 init_scf_loop 11 6.9 0.000 0.000 46.249 46.250 ot_mini 117 10.6 0.001 0.001 42.814 43.140 prepare_preconditioner 11 7.9 0.000 0.000 42.169 42.208 make_preconditioner 11 8.9 0.000 0.000 42.169 42.208 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.847 40.863 multiply_cannon_multrec 30084 15.6 13.612 18.784 26.545 31.447 rebuild_ks_matrix 128 8.3 0.001 0.001 29.372 29.691 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.372 29.690 mp_waitall_1 147882 16.7 17.224 27.023 17.224 27.023 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.490 26.780 qs_ot_get_derivative 117 11.6 0.001 0.002 23.219 23.554 make_m2s 5014 13.6 0.092 0.098 20.188 21.180 make_images 5014 14.6 1.931 2.236 19.878 20.871 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.012 19.585 apply_single 128 13.6 0.001 0.001 19.012 19.585 ot_diis_step 117 11.6 0.018 0.018 19.468 19.469 qs_ot_get_p 128 10.4 0.001 0.002 18.954 19.318 cp_fm_upper_to_full 105 14.8 11.511 16.926 11.511 16.926 cp_fm_cholesky_invert 11 10.9 16.133 16.141 16.133 16.141 init_scf_run 11 5.9 0.000 0.001 15.747 15.748 scf_env_initial_rho_setup 11 6.9 0.001 0.001 15.746 15.748 multiply_cannon_metrocomm3 30084 15.6 0.049 0.053 6.426 15.232 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.848 14.898 sum_up_and_integrate 128 10.3 0.002 0.003 14.027 14.053 integrate_v_rspace 128 11.3 0.003 0.004 13.968 13.996 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.768 13.807 calculate_rho_elec 128 8.7 0.170 0.185 13.767 13.807 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.434 13.436 dbcsr_mm_accdrv_process 62264 16.2 8.451 9.229 12.500 12.989 dbcsr_complete_redistribute 395 12.7 1.486 1.614 9.005 12.842 make_images_data 5014 15.6 0.065 0.071 10.849 12.378 hybrid_alltoall_any 5200 16.5 0.529 2.217 9.743 11.778 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.647 11.449 multiply_cannon_sync_h2d 30084 15.6 10.402 11.169 10.402 11.169 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 10.491 10.542 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.238 10.252 cp_fm_diag_elpa_base 83 14.4 9.271 9.590 10.231 10.243 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.299 10.015 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.728 9.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 9.413 9.509 mp_alltoall_i22 716 14.1 5.563 9.405 5.563 9.405 pw_transfer 1547 11.6 0.085 0.100 9.167 9.229 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.942 9.009 acc_transpose_blocks 30084 15.6 0.149 0.153 8.049 8.228 fft_wrap_pw1pw2_140 523 13.2 1.242 1.278 7.969 8.049 calculate_first_density_matrix 1 7.0 0.000 0.001 7.952 7.955 grid_integrate_task_list 128 12.3 7.531 7.916 7.531 7.916 wfi_extrapolate 11 7.9 0.001 0.001 7.596 7.596 cp_fm_cholesky_decompose 22 10.9 7.495 7.579 7.495 7.579 density_rs2pw 128 9.7 0.005 0.006 6.804 7.116 multiply_cannon_metrocomm4 25070 15.6 0.080 0.088 2.805 6.923 mp_irecv_dv 76098 16.2 2.651 6.658 2.651 6.658 calculate_dm_sparse 128 9.5 0.001 0.001 6.243 6.318 fft3d_ps 1291 14.7 2.995 3.055 5.997 6.055 acc_transpose_blocks_kernels 30084 16.6 0.403 0.413 5.991 6.009 grid_collocate_task_list 128 9.7 5.284 5.779 5.284 5.779 jit_kernel_transpose 5 15.6 5.588 5.605 5.588 5.605 mp_alltoall_d11v 2415 14.1 4.902 5.282 4.902 5.282 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.458 4.556 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.386 4.386 potential_pw2rs 128 12.3 0.023 0.023 4.302 4.312 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.233 4.294 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.081000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.000000, yerr=18.379831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/19/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 5.820059E+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 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 1.546854E+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 716108580288 ------------------------------------------------------------------------------- - - - 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 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828189. 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 0.039 0.099 173.931 173.957 qs_mol_dyn_low 1 2.0 0.003 0.003 173.456 173.508 qs_forces 11 3.9 0.004 0.005 173.352 173.355 qs_energies 11 4.9 0.002 0.002 165.942 165.952 scf_env_do_scf 11 5.9 0.001 0.002 143.849 143.858 velocity_verlet 10 3.0 0.002 0.002 110.798 110.803 scf_env_do_scf_inner_loop 116 6.6 0.005 0.010 108.099 108.100 dbcsr_multiply_generic 2485 12.5 0.183 0.190 76.506 77.002 qs_scf_new_mos 116 7.6 0.001 0.001 71.302 71.389 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.301 71.389 ot_scf_mini 116 9.6 0.003 0.003 66.927 67.025 multiply_cannon 2485 13.5 0.557 0.581 58.124 61.168 multiply_cannon_loop 2485 14.5 0.802 0.844 54.985 56.132 ot_mini 116 10.6 0.001 0.001 37.500 37.597 init_scf_loop 11 6.9 0.000 0.000 35.601 35.602 prepare_preconditioner 11 7.9 0.000 0.000 31.770 31.795 make_preconditioner 11 8.9 0.000 0.000 31.770 31.795 mp_waitall_1 124680 16.7 24.690 31.292 24.690 31.292 make_full_inverse_cholesky 11 9.9 0.018 0.027 29.741 30.002 rebuild_ks_matrix 127 8.3 0.001 0.001 28.988 29.098 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.987 29.098 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.391 26.489 multiply_cannon_multrec 9940 15.5 10.242 15.609 17.763 22.149 qs_ot_get_derivative 116 11.6 0.001 0.002 20.389 20.478 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.809 20.220 cp_fm_cholesky_invert 11 10.9 18.202 18.209 18.202 18.209 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.169 17.353 apply_single 127 13.6 0.001 0.001 17.169 17.353 ot_diis_step 116 11.6 0.019 0.020 17.040 17.040 qs_ot_get_p 127 10.4 0.003 0.003 16.827 16.966 make_m2s 4970 13.5 0.063 0.068 14.549 15.570 init_scf_run 11 5.9 0.000 0.001 15.263 15.263 scf_env_initial_rho_setup 11 6.9 0.001 0.002 15.263 15.263 make_images 4970 14.5 2.138 2.570 14.243 15.258 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.185 14.226 calculate_rho_elec 127 8.7 0.251 0.263 14.184 14.225 sum_up_and_integrate 127 10.3 0.002 0.002 14.131 14.177 integrate_v_rspace 127 11.3 0.004 0.004 14.071 14.118 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.098 13.114 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.945 11.946 multiply_cannon_sync_h2d 9940 15.5 10.624 11.060 10.624 11.060 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 10.047 10.060 pw_transfer 1535 11.6 0.084 0.093 9.873 9.909 make_images_data 4970 15.5 0.054 0.063 8.358 9.810 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.653 9.695 hybrid_alltoall_any 5155 16.4 0.834 3.746 8.192 9.409 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.092 9.102 cp_fm_diag_elpa_base 82 14.4 8.857 8.938 9.088 9.098 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.949 9.000 fft_wrap_pw1pw2_140 519 13.2 1.659 1.696 8.542 8.586 cp_fm_cholesky_decompose 22 10.9 7.938 8.136 7.938 8.136 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.953 8.014 grid_integrate_task_list 127 12.3 7.738 7.998 7.738 7.998 dbcsr_mm_accdrv_process 20590 16.1 3.252 4.406 7.174 7.884 calculate_first_density_matrix 1 7.0 0.000 0.001 7.683 7.685 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.274 7.489 density_rs2pw 127 9.7 0.005 0.005 6.937 7.389 wfi_extrapolate 11 7.9 0.001 0.001 7.341 7.341 acc_transpose_blocks 9940 15.5 0.054 0.056 6.541 6.584 fft3d_ps 1281 14.7 3.115 3.216 6.136 6.171 calculate_dm_sparse 127 9.5 0.001 0.001 6.030 6.117 grid_collocate_task_list 127 9.7 5.519 5.888 5.519 5.888 acc_transpose_blocks_kernels 9940 16.5 0.149 0.153 5.578 5.606 dbcsr_complete_redistribute 393 12.7 2.091 2.144 5.116 5.530 multiply_cannon_metrocomm4 7455 15.5 0.026 0.029 1.856 5.500 jit_kernel_transpose 5 15.6 5.429 5.456 5.429 5.456 mp_irecv_dv 28618 15.9 1.816 5.415 1.816 5.415 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.254 5.255 mp_alltoall_d11v 2401 14.1 4.513 4.808 4.513 4.808 mp_allgather_i34 2485 14.5 1.345 4.614 1.345 4.614 potential_pw2rs 127 12.3 0.026 0.027 4.375 4.379 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.583 3.870 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.606 3.692 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.620 3.670 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.300 3.618 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.491 3.502 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.434 3.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.957000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1472.454545, yerr=2.840091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 11.785519E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.183239E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591807936 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4040 58600. MP_Allreduce 11177 1500. MP_Sync 87 MP_Alltoall 1724 36993717. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.115 297.942 297.960 qs_mol_dyn_low 1 2.0 0.003 0.003 297.289 297.320 qs_forces 11 3.9 0.005 0.005 297.189 297.192 qs_energies 11 4.9 0.002 0.002 288.353 288.359 scf_env_do_scf 11 5.9 0.001 0.001 260.879 260.887 velocity_verlet 10 3.0 0.002 0.002 212.236 212.244 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 135.925 135.928 init_scf_loop 11 6.9 0.000 0.000 124.681 124.684 prepare_preconditioner 11 7.9 0.000 0.000 119.799 119.819 make_preconditioner 11 8.9 0.000 0.000 119.799 119.819 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.492 116.948 qs_scf_new_mos 118 7.6 0.001 0.001 90.516 90.589 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.515 90.588 dbcsr_multiply_generic 2527 12.6 0.218 0.228 87.579 88.121 ot_scf_mini 118 9.6 0.004 0.004 85.638 85.663 cp_fm_upper_to_full 105 14.8 52.794 76.126 52.794 76.126 multiply_cannon 2527 13.6 0.688 0.736 64.136 65.457 multiply_cannon_loop 2527 14.6 1.061 1.078 60.041 61.491 ot_mini 118 10.6 0.001 0.001 44.955 44.995 dbcsr_complete_redistribute 395 12.7 3.983 4.034 30.260 43.089 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.776 39.592 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.264 36.933 rebuild_ks_matrix 129 8.3 0.001 0.001 35.102 35.142 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 35.102 35.141 mp_alltoall_i22 718 14.1 22.041 34.976 22.041 34.976 cp_fm_cholesky_invert 11 10.9 33.637 33.643 33.637 33.643 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.609 32.668 mp_waitall_1 104500 16.8 28.232 32.463 28.232 32.463 qs_ot_get_p 129 10.4 0.002 0.002 25.740 25.817 qs_ot_get_derivative 118 11.6 0.002 0.002 24.803 24.829 qs_ot_p2m_diag 83 11.4 0.880 0.885 21.632 21.663 make_m2s 5054 13.6 0.075 0.078 18.845 20.385 multiply_cannon_metrocomm3 10108 15.6 0.025 0.025 18.826 20.243 ot_diis_step 118 11.6 0.022 0.022 20.119 20.119 make_images 5054 14.6 3.084 3.278 18.356 19.894 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.818 19.820 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.430 19.556 apply_single 129 13.6 0.001 0.001 19.430 19.556 multiply_cannon_multrec 10108 15.6 10.332 12.173 18.144 18.185 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.587 17.606 calculate_rho_elec 129 8.7 0.482 0.482 17.586 17.606 init_scf_run 11 5.9 0.000 0.001 17.187 17.188 scf_env_initial_rho_setup 11 6.9 0.001 0.001 17.186 17.188 sum_up_and_integrate 129 10.3 0.002 0.002 16.648 16.740 integrate_v_rspace 129 11.3 0.004 0.005 16.585 16.677 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.640 16.641 cp_fm_diag_elpa_base 83 14.4 12.205 13.832 16.636 16.636 multiply_cannon_sync_h2d 10108 15.6 14.405 14.434 14.405 14.434 pw_transfer 1559 11.6 0.095 0.096 13.387 13.402 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 13.147 13.161 hybrid_alltoall_any 5240 16.5 1.326 3.086 10.874 13.084 make_images_data 5054 15.6 0.063 0.068 10.702 12.940 fft_wrap_pw1pw2_140 527 13.2 3.239 3.298 11.684 11.698 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 11.376 11.409 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.718 9.736 dbcsr_mm_accdrv_process 20926 16.1 4.297 6.019 7.560 9.443 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 9.392 9.433 wfi_extrapolate 11 7.9 0.001 0.001 8.977 8.977 cp_fm_cholesky_decompose 22 10.9 8.920 8.937 8.920 8.937 grid_integrate_task_list 129 12.3 8.613 8.807 8.613 8.807 density_rs2pw 129 9.7 0.005 0.005 8.506 8.551 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.156 8.157 calculate_first_density_matrix 1 7.0 0.000 0.000 7.943 7.944 fft3d_ps 1301 14.7 4.006 4.025 7.571 7.629 mp_alltoall_d11v 2423 14.1 7.045 7.173 7.045 7.173 acc_transpose_blocks 10108 15.6 0.054 0.055 7.122 7.146 calculate_dm_sparse 129 9.5 0.001 0.001 6.745 6.773 grid_collocate_task_list 129 9.7 6.464 6.539 6.464 6.539 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.125 6.227 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.960000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2900.727273, yerr=151.038882 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 1.261265E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.033 84.691 84.692 qs_energies 1 2.0 0.000 0.000 84.248 84.256 ls_scf 1 3.0 0.000 0.000 83.351 83.359 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.364 72.532 multiply_cannon 111 7.7 0.017 0.020 55.821 57.104 multiply_cannon_loop 111 8.7 0.228 0.245 52.450 53.841 ls_scf_main 1 4.0 0.000 0.000 52.131 52.144 density_matrix_trs4 2 5.0 0.002 0.003 46.578 46.662 ls_scf_init_scf 1 4.0 0.000 0.000 28.149 28.150 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.021 27.084 mp_waitall_1 11031 10.9 22.223 25.751 22.223 25.751 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.903 24.916 multiply_cannon_multrec 2664 9.7 8.163 8.974 15.557 17.362 multiply_cannon_sync_h2d 2664 9.7 13.615 15.937 13.615 15.937 make_m2s 222 7.7 0.009 0.012 12.946 13.406 make_images 222 8.7 0.098 0.108 12.924 13.386 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.641 12.261 make_images_data 222 9.7 0.004 0.005 7.524 8.083 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.698 7.011 7.986 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.390 7.636 hybrid_alltoall_any 227 10.6 0.215 1.837 6.515 7.618 dbcsr_mm_accdrv_process_sort 4760 11.4 6.220 7.126 6.220 7.126 calculate_norms 4752 9.8 5.519 6.264 5.519 6.264 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.049 5.230 mp_sum_l 887 5.1 3.056 4.195 3.056 4.195 make_images_sizes 222 9.7 0.000 0.000 0.694 3.655 mp_alltoall_i44 222 10.7 0.694 3.655 0.694 3.655 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.069 3.333 mp_irecv_dv 6231 10.9 2.052 3.305 2.052 3.305 arnoldi_extremal 4 6.8 0.000 0.000 3.216 3.236 arnoldi_normal_ev 4 7.8 0.001 0.003 3.216 3.236 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.266 3.158 build_subspace 16 8.4 0.009 0.012 3.123 3.124 ls_scf_post 1 4.0 0.000 0.000 3.071 3.079 ls_scf_store_result 1 5.0 0.000 0.000 2.877 2.924 dbcsr_special_finalize 555 9.7 0.005 0.006 2.375 2.697 dbcsr_merge_single_wm 555 10.7 0.451 0.594 2.367 2.689 make_images_pack 222 9.7 2.207 2.626 2.209 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.327 2.564 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.452 2.068 2.454 dbcsr_sort_data 658 11.4 2.174 2.446 2.174 2.446 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.312 2.390 buffer_matrices_ensure_size 222 8.7 1.751 2.146 1.751 2.146 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.764 1.766 rebuild_ks_matrix 3 7.3 0.000 0.000 1.755 1.756 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.755 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.692000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 2.165948E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.037 98.112 98.113 qs_energies 1 2.0 0.000 0.000 97.623 97.629 ls_scf 1 3.0 0.000 0.000 96.284 96.289 dbcsr_multiply_generic 111 6.7 0.015 0.016 82.334 82.669 multiply_cannon 111 7.7 0.028 0.043 58.694 63.619 multiply_cannon_loop 111 8.7 0.135 0.145 55.886 61.502 ls_scf_main 1 4.0 0.000 0.000 54.746 54.750 density_matrix_trs4 2 5.0 0.002 0.003 48.980 49.250 ls_scf_init_scf 1 4.0 0.000 0.000 37.804 37.806 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.600 36.696 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 34.217 34.232 mp_waitall_1 9105 10.9 22.185 32.398 22.185 32.398 multiply_cannon_multrec 1332 9.7 13.331 16.679 22.637 27.147 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.405 23.831 make_m2s 222 7.7 0.006 0.008 15.363 15.898 make_images 222 8.7 1.365 1.689 15.332 15.868 mp_sum_l 887 5.1 6.793 10.846 6.793 10.846 dbcsr_mm_accdrv_process 4041 10.4 0.348 0.539 8.900 10.454 dbcsr_mm_accdrv_process_sort 4041 11.4 8.417 9.916 8.417 9.916 make_images_data 222 9.7 0.004 0.005 8.952 9.877 hybrid_alltoall_any 227 10.6 0.541 2.556 8.225 9.742 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 5.462 9.211 compute_matrix_preconditioner 1 6.0 0.000 0.001 8.593 8.598 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.188 7.771 mp_irecv_dv 3311 11.0 3.167 7.710 3.167 7.710 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.473 7.288 calculate_norms 2376 9.8 6.071 6.648 6.071 6.648 multiply_cannon_sync_h2d 1332 9.7 4.888 6.407 4.888 6.407 acc_transpose_blocks 1332 9.7 0.008 0.009 4.960 5.269 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.018 5.235 acc_transpose_blocks_kernels 1332 10.7 0.020 0.022 4.883 5.190 jit_kernel_transpose 1 13.0 4.863 5.169 4.863 5.169 arnoldi_extremal 4 6.8 0.000 0.000 4.687 4.705 arnoldi_normal_ev 4 7.8 0.001 0.005 4.687 4.705 build_subspace 16 8.4 0.014 0.021 4.431 4.433 ls_scf_post 1 4.0 0.000 0.000 3.735 3.741 ls_scf_store_result 1 5.0 0.000 0.000 3.444 3.558 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.195 3.408 dbcsr_matrix_vector_mult_local 304 10.0 2.790 3.261 2.792 3.263 mp_allgather_i34 111 8.7 0.800 3.092 0.800 3.092 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.615 2.708 dbcsr_data_new 4174 10.1 2.115 2.406 2.115 2.406 make_images_pack 222 9.7 1.815 2.120 1.818 2.122 dbcsr_sort_data 436 11.2 1.841 2.116 1.841 2.116 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=98.113000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 2.826060E+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 0.026 0.044 93.044 93.045 qs_energies 1 2.0 0.000 0.000 92.516 92.523 ls_scf 1 3.0 0.000 0.000 91.091 91.097 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.688 76.013 multiply_cannon 111 7.7 0.037 0.120 53.144 56.782 ls_scf_main 1 4.0 0.000 0.000 55.897 55.901 multiply_cannon_loop 111 8.7 0.116 0.131 50.393 54.558 density_matrix_trs4 2 5.0 0.002 0.003 50.004 50.160 mp_waitall_1 7281 11.0 23.167 32.791 23.167 32.791 ls_scf_init_scf 1 4.0 0.000 0.000 31.488 31.490 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.291 30.371 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.989 28.005 multiply_cannon_multrec 888 9.7 12.656 15.196 21.240 24.390 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.957 22.485 make_m2s 222 7.7 0.006 0.007 15.949 16.728 make_images 222 8.7 1.580 1.849 15.911 16.688 make_images_data 222 9.7 0.004 0.004 9.261 10.361 hybrid_alltoall_any 227 10.6 0.640 2.939 8.899 9.979 dbcsr_mm_accdrv_process 3754 10.4 0.316 0.488 8.110 9.350 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.862 7.676 8.862 mp_sum_l 887 5.1 4.897 7.975 4.897 7.975 multiply_cannon_sync_h2d 888 9.7 5.993 7.306 5.993 7.306 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.484 7.032 mp_irecv_dv 2335 11.1 2.468 6.983 2.468 6.983 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.592 6.880 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.787 6.340 arnoldi_extremal 4 6.8 0.000 0.000 5.132 5.149 arnoldi_normal_ev 4 7.8 0.001 0.005 5.132 5.149 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.707 4.913 calculate_norms 1584 9.8 4.399 4.857 4.399 4.857 build_subspace 16 8.4 0.014 0.020 4.827 4.834 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.504 3.814 ls_scf_post 1 4.0 0.000 0.000 3.707 3.713 dbcsr_matrix_vector_mult_local 304 10.0 3.080 3.649 3.082 3.651 mp_allgather_i34 111 8.7 0.871 3.567 0.871 3.567 ls_scf_store_result 1 5.0 0.000 0.000 3.437 3.532 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.872 2.966 dbcsr_data_new 4116 9.9 2.102 2.454 2.102 2.454 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.118 2.123 dbcsr_sort_data 325 11.1 1.877 2.105 1.877 2.105 make_images_sizes 222 9.7 0.000 0.000 0.864 2.052 mp_alltoall_i44 222 10.7 0.863 2.051 0.863 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.891 1.893 make_images_pack 222 9.7 1.622 1.877 1.625 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.873 1.875 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.873 1.874 dbcsr_finalize 304 7.8 0.026 0.032 1.605 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.045000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2203.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 3.367645E+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 0.032 0.046 103.833 103.833 qs_energies 1 2.0 0.000 0.000 103.249 103.256 ls_scf 1 3.0 0.000 0.000 101.587 101.595 dbcsr_multiply_generic 111 6.7 0.016 0.017 85.164 85.416 multiply_cannon 111 7.7 0.040 0.083 56.673 61.702 ls_scf_main 1 4.0 0.000 0.000 58.757 58.759 multiply_cannon_loop 111 8.7 0.153 0.167 51.540 55.160 density_matrix_trs4 2 5.0 0.002 0.003 52.645 52.765 ls_scf_init_scf 1 4.0 0.000 0.000 39.554 39.556 ls_scf_init_matrix_S 1 5.0 0.000 0.000 38.336 38.418 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 35.872 35.883 mp_waitall_1 6369 11.0 23.310 33.534 23.310 33.534 multiply_cannon_multrec 1332 9.7 14.173 17.509 22.090 24.782 make_m2s 222 7.7 0.007 0.008 21.055 22.557 make_images 222 8.7 3.134 3.600 21.005 22.510 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.841 20.323 make_images_data 222 9.7 0.004 0.004 11.781 13.341 hybrid_alltoall_any 227 10.6 0.798 3.779 11.093 13.002 mp_sum_l 887 5.1 5.864 9.744 5.864 9.744 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.481 7.561 9.090 dbcsr_mm_accdrv_process_sort 3641 11.4 7.122 8.598 7.122 8.598 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.835 8.078 compute_matrix_preconditioner 1 6.0 0.001 0.001 7.706 7.713 multiply_cannon_sync_h2d 1332 9.7 5.370 5.997 5.370 5.997 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.066 5.943 mp_irecv_dv 3229 10.9 2.040 5.875 2.040 5.875 arnoldi_extremal 4 6.8 0.000 0.000 5.350 5.363 arnoldi_normal_ev 4 7.8 0.001 0.005 5.350 5.363 acc_transpose_blocks 1332 9.7 0.008 0.008 4.626 5.099 build_subspace 16 8.4 0.014 0.021 5.015 5.024 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 4.514 4.985 jit_kernel_transpose 1 13.0 4.496 4.967 4.496 4.967 mp_allgather_i34 111 8.7 2.237 4.936 2.237 4.936 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.521 4.835 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.558 4.721 calculate_norms 2376 9.8 4.181 4.529 4.181 4.529 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.683 4.012 dbcsr_matrix_vector_mult_local 304 10.0 3.254 3.762 3.256 3.764 dbcsr_sort_data 658 11.4 3.060 3.530 3.060 3.530 ls_scf_post 1 4.0 0.000 0.000 3.276 3.283 dbcsr_special_finalize 555 9.7 0.006 0.007 2.814 3.245 dbcsr_merge_single_wm 555 10.7 0.541 0.656 2.806 3.237 ls_scf_store_result 1 5.0 0.000 0.000 2.997 3.069 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.002 3.045 dbcsr_data_release 10477 10.7 1.593 2.416 1.593 2.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=103.833000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 4.767343E+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 0.037 0.055 91.696 91.696 qs_energies 1 2.0 0.000 0.000 90.964 90.969 ls_scf 1 3.0 0.000 0.000 89.013 89.018 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.544 70.745 ls_scf_main 1 4.0 0.000 0.000 56.410 56.410 multiply_cannon 111 7.7 0.078 0.163 52.180 55.549 multiply_cannon_loop 111 8.7 0.088 0.095 49.606 51.178 density_matrix_trs4 2 5.0 0.002 0.003 49.375 49.439 ls_scf_init_scf 1 4.0 0.000 0.000 29.173 29.175 mp_waitall_1 5436 11.0 24.083 28.896 24.083 28.896 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.925 27.956 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.841 25.858 multiply_cannon_multrec 444 9.7 13.722 16.422 20.819 22.997 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.528 15.187 make_m2s 222 7.7 0.005 0.005 13.613 14.541 make_images 222 8.7 2.036 2.475 13.546 14.472 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.038 13.775 hybrid_alltoall_any 227 10.6 0.804 3.859 8.115 9.709 make_images_data 222 9.7 0.003 0.004 8.338 9.682 multiply_cannon_sync_h2d 444 9.7 6.712 8.269 6.712 8.269 dbcsr_mm_accdrv_process 3003 10.4 0.362 0.398 6.781 7.895 dbcsr_mm_accdrv_process_sort 3003 11.4 6.419 7.499 6.419 7.499 arnoldi_extremal 4 6.8 0.000 0.000 5.869 5.886 arnoldi_normal_ev 4 7.8 0.002 0.005 5.869 5.886 build_subspace 16 8.4 0.015 0.020 5.474 5.490 mp_sum_l 887 5.1 2.751 4.901 2.751 4.901 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.475 4.683 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.223 4.418 dbcsr_matrix_vector_mult_local 304 10.0 3.753 4.197 3.755 4.200 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.939 3.880 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.563 3.870 mp_irecv_dv 1241 11.2 1.550 3.849 1.550 3.849 calculate_norms 792 9.8 3.610 3.750 3.610 3.750 mp_allgather_i34 111 8.7 1.155 3.647 1.155 3.647 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.547 3.633 ls_scf_post 1 4.0 0.000 0.000 3.430 3.435 make_images_sizes 222 9.7 0.000 0.000 0.895 3.342 mp_alltoall_i44 222 10.7 0.895 3.341 0.895 3.341 ls_scf_store_result 1 5.0 0.000 0.000 3.226 3.257 dbcsr_finalize 304 7.8 0.062 0.076 2.194 2.295 dbcsr_data_new 4608 9.7 1.792 2.251 1.792 2.251 dbcsr_merge_all 275 8.9 0.478 0.521 2.053 2.136 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.029 2.030 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.997 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.996 1.997 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.935 1.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.696000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3759.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 8.819880E+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 0.081 0.169 111.883 111.883 qs_energies 1 2.0 0.000 0.000 110.503 110.508 ls_scf 1 3.0 0.000 0.000 107.503 107.508 dbcsr_multiply_generic 111 6.7 0.024 0.027 80.984 81.124 ls_scf_main 1 4.0 0.000 0.000 63.930 63.931 density_matrix_trs4 2 5.0 0.002 0.003 54.872 54.923 multiply_cannon 111 7.7 0.144 0.206 52.831 54.684 multiply_cannon_loop 111 8.7 0.099 0.100 49.857 52.199 ls_scf_init_scf 1 4.0 0.000 0.000 39.777 39.778 ls_scf_init_matrix_S 1 5.0 0.000 0.000 38.228 38.242 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 35.504 35.514 mp_waitall_1 4527 11.1 22.456 25.917 22.456 25.917 make_m2s 222 7.7 0.005 0.005 22.616 23.693 make_images 222 8.7 3.569 3.864 22.508 23.585 multiply_cannon_multrec 444 9.7 17.878 18.467 22.534 23.177 hybrid_alltoall_any 227 10.6 1.654 3.614 12.799 15.649 make_images_data 222 9.7 0.003 0.004 13.009 15.099 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.853 14.732 multiply_cannon_sync_h2d 444 9.7 8.787 8.848 8.787 8.848 arnoldi_extremal 4 6.8 0.000 0.000 7.494 7.506 arnoldi_normal_ev 4 7.8 0.003 0.009 7.494 7.506 build_subspace 16 8.4 0.026 0.036 6.936 6.952 compute_matrix_preconditioner 1 6.0 0.002 0.002 6.891 6.899 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.588 5.732 dbcsr_matrix_vector_mult_local 304 10.0 5.147 5.432 5.150 5.435 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.196 5.288 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.862 5.094 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.362 4.466 4.586 dbcsr_mm_accdrv_process_sort 1814 11.4 4.130 4.265 4.130 4.265 acc_transpose_blocks 444 9.7 0.003 0.003 3.967 4.194 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 3.883 4.108 jit_kernel_transpose 1 13.0 3.877 4.102 3.877 4.102 mp_sum_l 887 5.1 3.127 3.950 3.127 3.950 ls_scf_post 1 4.0 0.000 0.000 3.796 3.802 make_images_sizes 222 9.7 0.000 0.000 1.478 3.655 mp_alltoall_i44 222 10.7 1.478 3.655 1.478 3.655 mp_allgather_i34 111 8.7 1.088 3.555 1.088 3.555 ls_scf_store_result 1 5.0 0.000 0.000 3.513 3.553 calculate_norms 792 9.8 3.233 3.270 3.233 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.172 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.574 3.041 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.969 2.969 dbcsr_merge_all 275 8.9 0.885 0.922 2.868 2.952 dbcsr_complete_redistribute 5 7.6 1.440 1.498 2.762 2.880 matrix_ls_to_qs 2 6.0 0.000 0.000 2.416 2.542 dbcsr_sort_data 325 11.1 2.438 2.496 2.438 2.496 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.475 2.477 rebuild_ks_matrix 3 7.3 0.000 0.000 2.409 2.411 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.409 2.411 dbcsr_data_new 6591 9.6 1.857 2.333 1.857 2.333 dbcsr_new_transposed 4 7.5 0.241 0.253 2.292 2.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=111.883000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7029.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c074904eb170f1b1b0b6589717b6dd1755d9e84e_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 594.186240E+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 1431471399. 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 0.025 0.082 224.808 224.810 qs_mol_dyn_low 1 2.0 0.005 0.070 223.717 223.755 qs_forces 5 3.8 0.006 0.069 223.495 223.509 qs_energies 5 4.8 0.005 0.051 220.374 220.405 scf_env_do_scf 5 5.8 0.000 0.002 207.236 207.239 scf_env_do_scf_inner_loop 105 6.6 0.003 0.019 180.096 180.111 qs_scf_new_mos 105 7.6 0.000 0.001 140.705 140.847 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.705 140.846 dbcsr_multiply_generic 1445 12.2 0.128 0.137 131.254 131.667 ot_scf_mini 105 9.6 0.003 0.004 130.862 130.990 multiply_cannon 1445 13.2 0.274 0.288 112.453 115.039 multiply_cannon_loop 1445 14.2 2.839 2.985 110.666 111.823 velocity_verlet 4 3.0 0.009 0.043 106.076 106.077 ot_mini 105 10.6 0.001 0.002 59.684 59.836 multiply_cannon_multrec 69360 15.2 29.736 34.873 39.589 44.942 qs_ot_get_p 112 10.4 0.001 0.002 41.139 41.446 mp_waitall_1 488190 16.1 34.256 40.990 34.256 40.990 qs_ot_get_derivative 55 11.6 0.001 0.002 37.912 38.064 multiply_cannon_sync_h2d 69360 15.2 29.049 32.943 29.049 32.943 multiply_cannon_metrocomm3 69360 15.2 0.199 0.208 25.457 32.829 qs_ot_p2m_diag 40 11.0 0.020 0.034 30.166 30.248 rebuild_ks_matrix 110 8.4 0.000 0.000 28.844 29.014 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.041 28.843 29.014 init_scf_loop 7 6.6 0.000 0.004 27.111 27.114 cp_dbcsr_syevd 40 12.0 0.002 0.003 26.999 27.000 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.413 26.572 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.112 23.481 apply_single 62 13.6 0.000 0.000 23.112 23.481 prepare_preconditioner 7 7.6 0.000 0.000 22.276 22.320 make_preconditioner 7 8.6 0.000 0.004 22.276 22.320 cp_fm_syevd 40 13.0 0.000 0.003 21.977 22.128 ot_new_cg_direction 55 11.6 0.001 0.002 21.106 21.107 cp_fm_redistribute_end 40 14.0 8.660 17.280 8.665 17.282 cp_fm_syevd_base 40 14.0 8.608 17.229 8.608 17.229 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.355 16.736 calculate_rho_elec 110 8.6 0.030 0.032 16.354 16.735 make_full_inverse_cholesky 7 9.6 0.000 0.001 15.093 15.170 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.805 14.968 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.649 13.753 mp_sum_l 4764 12.2 11.886 12.538 11.886 12.538 pw_transfer 1645 12.4 0.079 0.100 11.724 11.975 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.586 11.842 calculate_dm_sparse 110 9.5 0.000 0.001 11.409 11.605 density_rs2pw 110 9.6 0.005 0.006 10.679 11.116 qs_vxc_create 110 10.4 0.002 0.005 10.462 10.509 dbcsr_mm_accdrv_process 154766 15.8 6.242 6.413 9.720 10.499 init_scf_run 5 5.8 0.000 0.001 10.452 10.453 scf_env_initial_rho_setup 5 6.8 0.002 0.003 10.452 10.453 cp_fm_cholesky_invert 7 10.6 10.446 10.453 10.446 10.453 fft_wrap_pw1pw2_240 915 15.0 0.828 0.930 10.213 10.444 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.083 10.156 check_diag 80 13.5 8.593 8.885 9.388 9.539 fft3d_pb 915 16.0 2.394 2.701 8.414 8.694 sum_up_and_integrate 60 10.3 0.001 0.003 8.370 8.404 integrate_v_rspace 60 11.3 0.002 0.020 8.352 8.388 transfer_rs2pw 445 10.6 0.007 0.008 7.451 7.880 xc_rho_set_and_dset_create 110 12.4 0.075 0.098 7.309 7.565 make_m2s 2890 13.2 0.078 0.088 6.466 7.092 xc_vxc_pw_create 60 11.3 0.039 0.050 7.027 7.073 make_images 2890 14.2 0.236 0.257 6.358 6.983 make_full_single_inverse 7 9.6 0.001 0.003 6.891 6.936 multiply_cannon_metrocomm1 69360 15.2 0.098 0.106 4.417 6.709 acc_transpose_blocks 69360 15.2 0.359 0.375 6.207 6.506 xc_pw_derive 510 13.4 0.005 0.007 6.226 6.314 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 5.949 5.968 mp_alltoall_z22v 2340 17.7 5.438 5.697 5.438 5.697 calculate_first_density_matrix 1 7.0 0.000 0.003 5.623 5.642 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.522 5.579 mp_waitany 7680 13.5 4.386 5.014 4.386 5.014 potential_pw2rs 60 12.3 0.002 0.003 4.623 4.667 multiply_cannon_metrocomm4 67915 15.2 0.188 0.205 2.017 4.625 wfi_extrapolate 5 7.8 0.000 0.001 4.514 4.515 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=224.810000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.800000, yerr=3.815757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c074904eb170f1b1b0b6589717b6dd1755d9e84e Summary: empty Status: OK