=== This is the CP2K Performance-Test === Updating f751b4c36..682b821dd Fast-forward src/emd/rt_delta_pulse.F | 102 +++++++++++++++++++-------------------- src/emd/rt_make_propagators.F | 90 ++++++++++++++++------------------ src/emd/rt_propagation_methods.F | 57 +++++++++++----------- src/emd/rt_propagation_output.F | 61 ++++++++++++----------- src/emd/rt_propagation_utils.F | 55 ++++++++++----------- src/emd/rt_propagator_init.F | 5 +- src/matrix_exp.F | 72 +++++++++++++-------------- src/motion/rt_propagation.F | 11 ++--- src/qs_mo_io.F | 17 +++---- src/qs_moments.F | 19 ++++---- src/qs_wf_history_methods.F | 42 ++++++++-------- src/qs_wf_history_types.F | 9 ++-- src/rpa_main.F | 1 + src/rt_propagation_forces.F | 27 +++++------ src/rt_propagation_types.F | 48 +++++++----------- src/rtp_admm_methods.F | 33 ++++++------- src/xas_env_types.F | 15 +++--- src/xas_methods.F | 38 ++++++++------- src/xas_tdp_correction.F | 59 +++++++++++----------- src/xas_tdp_methods.F | 5 +- src/xas_tdp_types.F | 8 ++- src/xas_tdp_utils.F | 4 +- 22 files changed, 367 insertions(+), 411 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 682b821dde03a8c13c17ad5faba6833238f81c82 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/01 job id: 43332328 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/02 job id: 43332331 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/03 job id: 43332333 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/04 job id: 43332336 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/05 job id: 43332338 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/06 job id: 43332339 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/07 job id: 43332340 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/08 job id: 43332341 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/09 job id: 43332344 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/10 job id: 43332345 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/11 job id: 43332346 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/12 job id: 43332347 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/13 job id: 43332348 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/14 job id: 43332349 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/15 job id: 43332350 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/16 job id: 43332351 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/17 job id: 43332352 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/18 job id: 43332353 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/19 job id: 43332355 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/20 job id: 43332356 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/21 job id: 43332358 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/22 job id: 43332359 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/23 job id: 43332361 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/24 job id: 43332362 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/25 job id: 43332363 --- 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/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/26 job id: 43332365 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.034 135.231 135.234 farming_run 1 2.0 134.517 134.520 135.204 135.210 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.466479E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.036 116.916 116.918 qs_energies 1 2.0 0.000 0.000 116.623 116.626 mp2_main 1 3.0 0.000 0.000 114.654 114.657 mp2_gpw_main 1 4.0 0.333 0.390 113.730 113.733 mp2_ri_gpw_compute_in 1 5.0 0.188 0.205 94.177 95.096 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.545 56.152 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.171 41.855 47.398 get_2c_integrals 1 6.0 0.001 0.007 37.620 38.470 integrate_v_rspace 273 8.0 0.434 0.451 25.219 30.397 pw_transfer 6555 10.6 0.376 0.383 27.455 28.151 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.113 26.779 grid_integrate_task_list 273 9.0 21.017 26.675 21.017 26.675 fft_wrap_pw1pw2_100 2178 12.4 1.198 1.435 23.666 24.334 compute_2c_integrals 1 7.0 0.018 0.020 19.731 20.060 compute_2c_integrals_loop_lm 1 8.0 0.004 0.005 18.924 19.371 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.430 18.920 19.366 rpa_ri_compute_en 1 5.0 0.000 0.000 19.132 19.353 cp_fm_cholesky_decompose 12 8.2 17.855 18.655 17.855 18.655 cholesky_decomp 1 7.0 0.000 0.000 16.744 17.564 fft3d_s 5443 13.4 16.200 16.793 16.221 16.814 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.562 10.851 15.562 calculate_wavefunction 272 8.0 5.403 5.570 12.547 13.129 rpa_num_int 1 6.0 0.005 0.085 10.729 10.729 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.545 10.728 calc_mat_Q 8 8.0 0.000 0.000 9.388 9.634 contract_S_to_Q 8 9.0 0.000 0.000 8.811 9.054 calc_potential_gpw 544 9.5 0.005 0.006 8.249 8.671 parallel_gemm_fm 14 9.1 0.000 0.000 8.393 8.583 parallel_gemm_fm_cosma 14 10.1 8.393 8.583 8.393 8.583 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.236 8.549 potential_pw2rs 545 10.0 0.107 0.109 7.739 8.321 collocate_single_gaussian 272 10.0 0.040 0.043 7.485 7.717 create_integ_mat 1 6.0 0.022 0.027 7.703 7.703 array2fm 1 7.0 0.000 0.000 6.567 7.123 pw_scatter_s 2720 13.7 4.424 4.633 4.424 4.633 pw_gather_s 2722 13.2 3.913 4.312 3.913 4.312 array2fm_buffer_send 1 8.0 2.838 3.088 2.838 3.088 pw_poisson_solve 545 10.5 1.123 1.175 2.170 2.395 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.730455, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2738.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.037 397.501 397.502 farming_run 1 2.0 396.325 396.331 397.460 397.462 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223131E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.048 209.722 209.723 qs_energies 1 2.0 0.000 0.000 209.492 209.507 scf_env_do_scf 1 3.0 0.000 0.000 106.306 106.306 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.414 105.425 rebuild_ks_matrix 4 6.0 0.000 0.000 105.266 105.274 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.265 105.274 hfx_ks_matrix 4 8.0 0.001 0.001 104.891 104.894 integrate_four_center 4 9.0 0.144 0.447 104.890 104.894 mp2_main 1 3.0 0.000 0.000 102.901 102.916 mp2_gpw_main 1 4.0 0.032 0.047 102.037 102.048 integrate_four_center_main 4 10.0 0.099 0.544 96.948 99.468 integrate_four_center_bin 263 11.0 96.849 99.449 96.849 99.449 init_scf_loop 1 4.0 0.000 0.000 91.963 91.963 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.084 76.372 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.707 55.994 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.159 42.249 47.401 integrate_v_rspace 95 8.0 0.397 0.565 28.620 33.803 pw_transfer 2240 10.6 0.144 0.161 29.908 30.368 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.915 29.410 grid_integrate_task_list 95 9.0 23.908 29.265 23.908 29.265 mp2_ri_gpw_compute_en 1 5.0 0.057 0.070 26.804 28.736 ao_to_mo_and_store_B_mult_1 91 7.0 10.775 28.696 10.775 28.696 fft_wrap_pw1pw2_100 730 12.4 1.289 1.450 26.640 27.074 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.841 1.896 24.879 24.888 get_2c_integrals 1 6.0 0.000 0.000 20.275 20.313 compute_2c_integrals 1 7.0 0.002 0.003 19.258 19.265 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.885 19.135 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.863 18.884 19.134 fft3d_s 1823 13.4 18.425 18.725 18.438 18.737 scf_env_do_scf_inner_loop 4 4.0 0.003 0.150 14.341 14.341 calculate_wavefunction 91 8.0 2.011 2.054 9.739 9.994 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.588 8.702 9.234 potential_pw2rs 186 10.0 0.034 0.035 8.645 9.158 local_gemm 172 8.0 8.145 8.659 8.145 8.659 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.255 8.621 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.519 7.949 8.410 calc_potential_gpw 182 9.5 0.002 0.002 7.921 8.187 collocate_single_gaussian 91 10.0 0.017 0.020 7.884 8.144 mp_sendrecv_dm3 2068 8.0 5.994 6.475 5.994 6.475 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.429 6.342 6.429 pw_gather_s 912 13.2 4.908 5.347 4.908 5.347 mp_sync 38 10.4 2.586 4.761 2.586 4.761 pw_scatter_s 910 13.7 3.927 4.265 3.927 4.265 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.043434, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 451.276800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.045 50.177 50.178 qs_mol_dyn_low 1 2.0 0.003 0.004 49.914 49.923 qs_forces 11 3.9 0.001 0.002 49.838 49.845 qs_energies 11 4.9 0.001 0.001 48.374 48.391 scf_env_do_scf 11 5.9 0.000 0.001 42.164 42.165 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 40.076 40.077 dbcsr_multiply_generic 2286 12.5 0.096 0.103 31.933 32.335 qs_scf_new_mos 108 7.5 0.000 0.001 29.941 30.207 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.940 30.207 ot_scf_mini 108 9.5 0.002 0.002 28.408 28.600 multiply_cannon 2286 13.5 0.188 0.199 25.460 26.744 multiply_cannon_loop 2286 14.5 1.511 1.602 24.680 25.986 velocity_verlet 10 3.0 0.001 0.001 24.065 24.065 ot_mini 108 10.5 0.001 0.001 17.924 18.174 qs_ot_get_derivative 108 11.5 0.001 0.001 15.117 15.312 mp_waitall_1 267858 16.1 7.691 13.702 7.691 13.702 multiply_cannon_metrocomm3 54864 15.5 0.070 0.076 5.287 12.171 multiply_cannon_multrec 54864 15.5 4.325 6.757 8.041 11.663 rebuild_ks_matrix 119 8.3 0.000 0.000 8.000 8.124 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.999 8.123 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.064 7.180 multiply_cannon_sync_h2d 54864 15.5 6.008 6.663 6.008 6.663 qs_ot_get_p 119 10.4 0.001 0.001 6.075 6.325 mp_sum_l 7207 12.9 4.333 5.996 4.333 5.996 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.257 5.371 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.829 5.286 dbcsr_mm_accdrv_process 76910 16.1 1.149 1.844 3.638 5.095 init_scf_run 11 5.9 0.000 0.001 4.946 4.947 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.946 4.946 sum_up_and_integrate 119 10.3 0.012 0.014 4.632 4.639 integrate_v_rspace 119 11.3 0.002 0.002 4.620 4.628 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.091 4.172 calculate_rho_elec 119 8.7 0.011 0.016 4.090 4.172 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.437 3.485 jit_kernel_multiply 13 15.8 2.428 3.169 2.428 3.169 calculate_dm_sparse 119 9.5 0.000 0.000 2.928 3.076 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.883 3.072 apply_single 119 13.6 0.000 0.000 2.882 3.071 rs_pw_transfer 974 11.9 0.011 0.012 2.910 3.018 calculate_first_density_matrix 1 7.0 0.000 0.001 2.787 2.790 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.651 2.651 ot_diis_step 108 11.5 0.006 0.006 2.634 2.635 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.403 2.404 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.396 2.397 cp_fm_redistribute_end 50 14.0 2.178 2.377 2.182 2.379 density_rs2pw 119 9.7 0.004 0.004 2.278 2.375 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.372 2.364 cp_fm_diag_elpa_base 50 14.0 0.194 2.334 0.195 2.343 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.186 2.232 acc_transpose_blocks 54864 15.5 0.227 0.251 1.741 2.214 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.124 2.173 grid_integrate_task_list 119 12.3 2.009 2.111 2.009 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.099 2.099 init_scf_loop 11 6.9 0.000 0.000 2.066 2.066 potential_pw2rs 119 12.3 0.004 0.004 2.009 2.023 make_m2s 4572 13.5 0.054 0.056 1.841 1.884 pw_transfer 1439 11.6 0.051 0.055 1.789 1.863 make_images 4572 14.5 0.132 0.138 1.758 1.801 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.714 1.789 mp_sum_d 4125 12.0 1.100 1.669 1.100 1.669 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.566 1.585 fft3d_ps 1201 14.6 0.358 0.459 1.495 1.564 mp_alltoall_d11v 2130 13.8 1.296 1.420 1.296 1.420 fft_wrap_pw1pw2_140 487 13.2 0.079 0.092 1.301 1.375 grid_collocate_task_list 119 9.7 1.287 1.350 1.287 1.350 mp_waitany 12084 13.8 1.235 1.342 1.235 1.342 acc_transpose_blocks_kernels 54864 16.5 0.249 0.390 0.766 1.045 prepare_preconditioner 11 7.9 0.000 0.000 1.015 1.045 make_preconditioner 11 8.9 0.000 0.000 1.015 1.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=50.178000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 486.674432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 866346. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.090 0.232 39.605 39.617 qs_mol_dyn_low 1 2.0 0.005 0.015 38.726 38.733 qs_forces 11 3.9 0.002 0.002 37.981 37.985 qs_energies 11 4.9 0.003 0.006 36.295 36.301 scf_env_do_scf 11 5.9 0.001 0.002 31.245 31.246 scf_env_do_scf_inner_loop 108 6.5 0.004 0.016 28.726 28.729 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.460 21.819 qs_scf_new_mos 108 7.5 0.001 0.001 19.941 20.191 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.940 20.190 ot_scf_mini 108 9.5 0.003 0.004 19.045 19.221 velocity_verlet 10 3.0 0.001 0.001 18.731 18.746 multiply_cannon 2286 13.5 0.209 0.218 16.381 17.909 multiply_cannon_loop 2286 14.5 0.911 0.984 15.241 16.873 ot_mini 108 10.5 0.001 0.002 11.958 12.193 mp_waitall_1 217478 16.2 5.987 11.113 5.987 11.113 qs_ot_get_derivative 108 11.5 0.001 0.001 9.481 9.658 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.210 9.453 multiply_cannon_multrec 27432 15.5 1.972 4.422 5.864 8.660 rebuild_ks_matrix 119 8.3 0.000 0.000 7.038 7.174 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.020 7.037 7.173 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.212 6.336 dbcsr_mm_accdrv_process 47894 16.0 2.996 5.079 3.823 5.972 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.688 4.609 qs_ot_get_p 119 10.4 0.001 0.001 4.179 4.403 mp_sum_l 7207 12.9 2.277 4.396 2.277 4.396 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.127 4.296 apply_single 119 13.6 0.000 0.000 3.127 4.296 sum_up_and_integrate 119 10.3 0.025 0.028 4.092 4.101 integrate_v_rspace 119 11.3 0.002 0.002 4.067 4.077 init_scf_run 11 5.9 0.000 0.001 3.804 3.805 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.804 3.805 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.513 3.533 calculate_rho_elec 119 8.7 0.021 0.024 3.513 3.532 multiply_cannon_sync_h2d 27432 15.5 2.179 2.830 2.179 2.830 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.725 2.744 make_m2s 4572 13.5 0.053 0.055 2.500 2.730 make_images 4572 14.5 0.200 0.238 2.411 2.638 rs_pw_transfer 974 11.9 0.010 0.011 2.460 2.582 init_scf_loop 11 6.9 0.001 0.003 2.493 2.497 ot_diis_step 108 11.5 0.010 0.011 2.423 2.424 calculate_first_density_matrix 1 7.0 0.002 0.005 2.318 2.320 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.290 2.291 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.155 2.244 calculate_dm_sparse 119 9.5 0.001 0.001 2.032 2.107 density_rs2pw 119 9.7 0.004 0.004 1.866 1.955 jit_kernel_multiply 10 16.0 0.775 1.936 0.775 1.936 grid_integrate_task_list 119 12.3 1.836 1.933 1.836 1.933 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.928 1.930 cp_fm_redistribute_end 50 14.0 1.588 1.907 1.592 1.908 cp_fm_diag_elpa_base 50 14.0 0.306 1.850 0.313 1.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.872 1.874 potential_pw2rs 119 12.3 0.006 0.006 1.840 1.850 pw_transfer 1439 11.6 0.063 0.066 1.678 1.708 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.664 1.702 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.588 1.619 prepare_preconditioner 11 7.9 0.000 0.000 1.574 1.600 make_preconditioner 11 8.9 0.001 0.001 1.574 1.600 make_images_data 4572 15.5 0.045 0.052 1.135 1.574 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.468 1.524 acc_transpose_blocks 27432 15.5 0.110 0.116 1.202 1.502 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.980 1.456 wfi_extrapolate 11 7.9 0.001 0.001 1.439 1.439 mp_allgather_i34 2286 14.5 0.582 1.335 0.582 1.335 fft3d_ps 1201 14.6 0.494 0.545 1.296 1.325 mp_alltoall_d11v 2130 13.8 1.203 1.316 1.203 1.316 grid_collocate_task_list 119 9.7 1.240 1.296 1.240 1.296 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.254 fft_wrap_pw1pw2_140 487 13.2 0.075 0.083 1.224 1.254 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.239 mp_sum_d 4125 12.0 0.609 1.071 0.609 1.071 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.957 0.961 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.882 0.894 acc_transpose_blocks_kernels 27432 16.5 0.184 0.273 0.660 0.872 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.617000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.272727, yerr=1.135454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.513856E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.081 33.023 33.025 qs_mol_dyn_low 1 2.0 0.003 0.004 32.713 32.720 qs_forces 11 3.9 0.005 0.017 32.631 32.634 qs_energies 11 4.9 0.003 0.017 30.908 30.911 scf_env_do_scf 11 5.9 0.001 0.012 25.783 25.783 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 23.141 23.142 dbcsr_multiply_generic 2286 12.5 0.095 0.099 17.030 17.148 velocity_verlet 10 3.0 0.001 0.001 15.298 15.300 qs_scf_new_mos 108 7.5 0.001 0.001 15.130 15.142 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.130 15.141 ot_scf_mini 108 9.5 0.003 0.004 14.412 14.425 multiply_cannon 2286 13.5 0.197 0.205 13.461 14.191 multiply_cannon_loop 2286 14.5 0.642 0.673 12.672 13.433 ot_mini 108 10.5 0.023 0.183 9.102 9.116 multiply_cannon_multrec 18288 15.5 1.970 2.892 7.170 7.434 qs_ot_get_derivative 108 11.5 0.001 0.002 7.362 7.377 rebuild_ks_matrix 119 8.3 0.000 0.000 6.481 6.495 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.480 6.495 dbcsr_mm_accdrv_process 38222 16.0 4.298 5.944 5.115 6.000 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.615 5.628 sum_up_and_integrate 119 10.3 0.030 0.031 4.082 4.101 mp_waitall_1 169478 16.3 2.975 4.095 2.975 4.095 integrate_v_rspace 119 11.3 0.002 0.002 4.052 4.074 init_scf_run 11 5.9 0.000 0.001 3.834 3.834 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.834 3.834 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.773 3.415 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.269 3.274 calculate_rho_elec 119 8.7 0.030 0.031 3.268 3.273 qs_ot_get_p 119 10.4 0.001 0.001 3.142 3.159 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.251 2.630 apply_single 119 13.6 0.000 0.000 2.251 2.630 init_scf_loop 11 6.9 0.001 0.003 2.615 2.619 calculate_first_density_matrix 1 7.0 0.001 0.003 2.617 2.618 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.409 2.448 rs_pw_transfer 974 11.9 0.009 0.010 2.331 2.395 calculate_dm_sparse 119 9.5 0.001 0.001 2.092 2.100 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.037 2.042 make_m2s 4572 13.5 0.045 0.046 1.874 2.021 density_rs2pw 119 9.7 0.004 0.004 1.851 1.950 make_images 4572 14.5 0.191 0.203 1.788 1.932 grid_integrate_task_list 119 12.3 1.803 1.887 1.803 1.887 potential_pw2rs 119 12.3 0.007 0.008 1.860 1.883 pw_transfer 1439 11.6 0.063 0.066 1.834 1.857 prepare_preconditioner 11 7.9 0.000 0.000 1.817 1.819 make_preconditioner 11 8.9 0.000 0.001 1.817 1.819 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.743 1.766 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.762 1.763 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.755 1.757 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.661 1.746 ot_diis_step 108 11.5 0.018 0.064 1.698 1.721 mp_sum_l 7207 12.9 1.327 1.680 1.327 1.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.609 1.615 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.588 1.592 multiply_cannon_sync_h2d 18288 15.5 1.390 1.559 1.390 1.559 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.523 1.524 cp_fm_redistribute_end 50 14.0 1.128 1.500 1.129 1.501 cp_fm_diag_elpa_base 50 14.0 0.355 1.442 0.370 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.469 1.472 fft3d_ps 1201 14.6 0.508 0.526 1.429 1.453 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.397 1.421 jit_kernel_multiply 6 15.9 0.766 1.361 0.766 1.361 acc_transpose_blocks 18288 15.5 0.076 0.078 1.300 1.322 grid_collocate_task_list 119 9.7 1.205 1.313 1.205 1.313 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 make_images_data 4572 15.5 0.045 0.050 0.838 1.010 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.967 0.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.936 0.938 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.719 0.917 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.878 0.895 acc_transpose_blocks_kernels 18288 16.5 0.212 0.220 0.851 0.862 mp_alltoall_z22v 1201 16.6 0.764 0.835 0.764 0.835 mp_alltoall_d11v 2130 13.8 0.717 0.802 0.717 0.802 cp_fm_cholesky_invert 11 10.9 0.769 0.774 0.769 0.774 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.655 0.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.025000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=2.449490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 553.529344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.085 37.177 37.179 qs_mol_dyn_low 1 2.0 0.003 0.004 36.772 36.779 qs_forces 11 3.9 0.002 0.007 36.682 36.682 qs_energies 11 4.9 0.015 0.032 34.784 34.798 scf_env_do_scf 11 5.9 0.001 0.015 28.414 28.416 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.832 24.833 dbcsr_multiply_generic 2286 12.5 0.102 0.105 18.635 18.752 velocity_verlet 10 3.0 0.001 0.001 18.258 18.261 qs_scf_new_mos 108 7.5 0.001 0.001 16.571 16.634 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.571 16.634 ot_scf_mini 108 9.5 0.007 0.019 15.629 15.689 multiply_cannon 2286 13.5 0.271 0.308 14.800 15.173 multiply_cannon_loop 2286 14.5 0.949 0.981 13.758 14.164 ot_mini 108 10.5 0.001 0.001 9.646 9.720 multiply_cannon_multrec 27432 15.5 2.379 3.061 8.748 9.107 qs_ot_get_derivative 108 11.5 0.001 0.001 7.809 7.878 dbcsr_mm_accdrv_process 47916 15.9 5.363 7.033 6.276 7.499 rebuild_ks_matrix 119 8.3 0.000 0.000 6.654 6.710 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.654 6.710 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.802 5.851 init_scf_run 11 5.9 0.000 0.001 4.185 4.186 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.185 4.185 sum_up_and_integrate 119 10.3 0.035 0.037 3.906 3.943 integrate_v_rspace 119 11.3 0.002 0.003 3.871 3.909 init_scf_loop 11 6.9 0.001 0.003 3.546 3.548 qs_ot_get_p 119 10.4 0.001 0.001 3.368 3.443 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.988 3.438 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.357 3.388 calculate_rho_elec 119 8.7 0.040 0.046 3.357 3.388 mp_waitall_1 145218 16.4 2.354 2.947 2.354 2.947 rs_pw_transfer 974 11.9 0.009 0.009 2.618 2.794 calculate_first_density_matrix 1 7.0 0.001 0.004 2.744 2.745 prepare_preconditioner 11 7.9 0.000 0.000 2.662 2.671 make_preconditioner 11 8.9 0.000 0.001 2.662 2.671 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.265 2.595 make_m2s 4572 13.5 0.055 0.057 2.326 2.470 make_images 4572 14.5 0.274 0.334 2.218 2.361 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.100 2.230 apply_single 119 13.6 0.000 0.000 2.100 2.230 calculate_dm_sparse 119 9.5 0.000 0.001 2.078 2.130 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.007 2.017 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.990 1.992 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.932 1.963 density_rs2pw 119 9.7 0.004 0.004 1.854 1.950 grid_integrate_task_list 119 12.3 1.830 1.910 1.830 1.910 pw_transfer 1439 11.6 0.063 0.066 1.844 1.905 qs_energies_init_hamiltonians 11 5.9 0.054 0.341 1.851 1.853 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.754 1.818 ot_diis_step 108 11.5 0.012 0.012 1.792 1.793 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.755 1.767 potential_pw2rs 119 12.3 0.008 0.010 1.693 1.726 jit_kernel_multiply 9 16.1 0.853 1.711 0.853 1.711 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.681 1.681 acc_transpose_blocks 27432 15.5 0.112 0.114 1.513 1.543 fft3d_ps 1201 14.6 0.529 0.581 1.436 1.490 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.862 1.478 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.404 1.469 mp_sum_l 7207 12.9 1.116 1.446 1.116 1.446 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.412 1.413 wfi_extrapolate 11 7.9 0.001 0.001 1.391 1.392 cp_fm_redistribute_end 50 14.0 0.929 1.388 0.931 1.389 cp_fm_diag_elpa_base 50 14.0 0.436 1.329 0.456 1.369 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.279 1.367 grid_collocate_task_list 119 9.7 1.222 1.309 1.222 1.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.261 1.270 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.135 1.156 rs_pw_transfer_RS2PW_140 130 11.5 0.112 0.140 0.958 1.137 mp_waitany 7280 13.7 0.956 1.130 0.956 1.130 dbcsr_complete_redistribute 329 12.2 0.128 0.169 0.847 1.123 cp_fm_upper_to_full 72 13.5 0.807 1.120 0.807 1.120 multiply_cannon_sync_h2d 27432 15.5 0.979 1.048 0.979 1.048 make_images_data 4572 15.5 0.045 0.049 0.879 1.034 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.774 0.971 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.943 0.948 mp_alltoall_d11v 2130 13.8 0.790 0.920 0.790 0.920 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 0.879 0.905 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.618 0.888 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.804 0.870 calculate_rho_core 11 7.9 0.061 0.323 0.790 0.864 mp_alltoall_z22v 1201 16.6 0.790 0.850 0.790 0.850 cp_fm_cholesky_invert 11 10.9 0.831 0.834 0.831 0.834 mp_sum_d 4123 12.0 0.657 0.830 0.657 0.830 mp_alltoall_i22 627 13.8 0.457 0.750 0.457 0.750 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.179000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.181818, yerr=2.166614 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 599.150592E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.082 29.882 29.883 qs_mol_dyn_low 1 2.0 0.010 0.012 29.171 29.178 qs_forces 11 3.9 0.011 0.018 29.072 29.079 qs_energies 11 4.9 0.001 0.001 27.374 27.391 scf_env_do_scf 11 5.9 0.000 0.001 21.809 21.809 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.304 19.305 velocity_verlet 10 3.0 0.001 0.001 14.227 14.232 dbcsr_multiply_generic 2286 12.5 0.093 0.095 13.292 13.372 qs_scf_new_mos 108 7.5 0.001 0.001 11.702 11.730 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.701 11.730 multiply_cannon 2286 13.5 0.236 0.243 10.726 11.211 ot_scf_mini 108 9.5 0.002 0.002 11.028 11.057 multiply_cannon_loop 2286 14.5 0.334 0.343 9.802 9.990 multiply_cannon_multrec 9144 15.5 1.926 2.229 6.681 6.949 ot_mini 108 10.5 0.001 0.001 6.542 6.577 rebuild_ks_matrix 119 8.3 0.000 0.000 5.804 5.826 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.803 5.826 qs_ot_get_derivative 108 11.5 0.001 0.001 5.251 5.278 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.178 5.199 dbcsr_mm_accdrv_process 12550 15.8 3.284 4.019 4.654 4.807 init_scf_run 11 5.9 0.000 0.001 4.071 4.071 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.071 4.071 sum_up_and_integrate 119 10.3 0.037 0.041 3.557 3.563 integrate_v_rspace 119 11.3 0.002 0.003 3.520 3.525 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.220 3.227 calculate_rho_elec 119 8.7 0.059 0.061 3.219 3.227 calculate_first_density_matrix 1 7.0 0.000 0.000 2.904 2.905 qs_ot_get_p 119 10.4 0.001 0.001 2.693 2.732 init_scf_loop 11 6.9 0.000 0.000 2.484 2.486 mp_waitall_1 121218 16.5 1.921 2.486 1.921 2.486 jit_kernel_multiply 10 15.9 1.333 2.334 1.333 2.334 make_m2s 4572 13.5 0.035 0.036 1.787 1.970 calculate_dm_sparse 119 9.5 0.000 0.000 1.921 1.939 grid_integrate_task_list 119 12.3 1.842 1.934 1.842 1.934 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.910 1.927 make_images 4572 14.5 0.268 0.301 1.697 1.878 rs_pw_transfer 974 11.9 0.008 0.008 1.707 1.838 prepare_preconditioner 11 7.9 0.000 0.000 1.761 1.766 make_preconditioner 11 8.9 0.000 0.000 1.761 1.766 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.739 1.742 density_rs2pw 119 9.7 0.003 0.004 1.644 1.739 pw_transfer 1439 11.6 0.064 0.065 1.726 1.733 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.706 1.707 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.648 1.676 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.635 1.641 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.540 1.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.480 1.493 grid_collocate_task_list 119 9.7 1.272 1.363 1.272 1.363 potential_pw2rs 119 12.3 0.010 0.010 1.344 1.346 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.339 1.345 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.326 1.334 fft3d_ps 1201 14.6 0.539 0.552 1.302 1.309 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.277 1.293 ot_diis_step 108 11.5 0.012 0.012 1.275 1.276 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.264 1.265 cp_fm_redistribute_end 50 14.0 0.626 1.242 0.627 1.243 cp_fm_diag_elpa_base 50 14.0 0.570 1.172 0.614 1.227 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.160 1.174 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.147 1.172 apply_single 119 13.6 0.000 0.000 1.146 1.172 wfi_extrapolate 11 7.9 0.001 0.001 1.122 1.123 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.796 1.083 make_images_data 4572 15.5 0.039 0.043 0.840 1.076 acc_transpose_blocks 9144 15.5 0.039 0.040 0.963 0.983 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.919 cp_fm_cholesky_invert 11 10.9 0.899 0.902 0.899 0.902 mp_alltoall_d11v 2130 13.8 0.788 0.898 0.788 0.898 parallel_gemm_fm 81 9.0 0.000 0.000 0.803 0.804 parallel_gemm_fm_cosma 81 10.0 0.803 0.804 0.803 0.804 multiply_cannon_sync_h2d 9144 15.5 0.701 0.774 0.701 0.774 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.750 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.361 0.749 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.701 0.748 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.734 0.742 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.717 0.736 make_basis_sm 11 9.8 0.001 0.001 0.731 0.732 multiply_cannon_metrocomm1 9144 15.5 0.022 0.022 0.475 0.692 mp_alltoall_z22v 1201 16.6 0.637 0.669 0.637 0.669 mp_allgather_i34 2286 14.5 0.253 0.667 0.253 0.667 jit_kernel_transpose 5 15.6 0.600 0.620 0.600 0.620 dbcsr_complete_redistribute 329 12.2 0.178 0.215 0.572 0.603 mp_waitany 5200 13.7 0.474 0.599 0.474 0.599 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.883000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.727273, yerr=5.344341 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 755.892224E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.048 43.282 43.283 qs_mol_dyn_low 1 2.0 0.003 0.003 42.600 42.607 qs_forces 11 3.9 0.002 0.002 42.539 42.540 qs_energies 11 4.9 0.001 0.001 40.592 40.595 scf_env_do_scf 11 5.9 0.001 0.003 33.385 33.385 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.428 25.430 velocity_verlet 10 3.0 0.001 0.001 23.226 23.232 dbcsr_multiply_generic 2286 12.5 0.101 0.102 19.021 19.282 qs_scf_new_mos 108 7.5 0.001 0.001 16.408 16.497 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.408 16.496 multiply_cannon 2286 13.5 0.302 0.310 14.969 15.921 ot_scf_mini 108 9.5 0.002 0.002 15.321 15.413 multiply_cannon_loop 2286 14.5 0.348 0.352 13.680 14.711 multiply_cannon_multrec 9144 15.5 3.502 5.262 9.067 9.704 ot_mini 108 10.5 0.001 0.001 9.155 9.266 init_scf_loop 11 6.9 0.000 0.000 7.926 7.931 qs_ot_get_derivative 108 11.5 0.001 0.001 7.125 7.218 rebuild_ks_matrix 119 8.3 0.000 0.000 6.929 7.065 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.929 7.064 dbcsr_mm_accdrv_process 12550 15.8 4.319 5.862 5.441 7.040 prepare_preconditioner 11 7.9 0.000 0.000 6.982 6.994 make_preconditioner 11 8.9 0.000 0.000 6.982 6.994 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.528 6.868 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.263 6.386 init_scf_run 11 5.9 0.000 0.001 5.191 5.191 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.190 5.190 cp_fm_upper_to_full 72 14.2 3.161 4.553 3.161 4.553 mp_waitall_1 97218 16.6 3.304 4.121 3.304 4.121 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.754 3.766 calculate_rho_elec 119 8.7 0.118 0.121 3.754 3.766 sum_up_and_integrate 119 10.3 0.064 0.066 3.746 3.752 calculate_first_density_matrix 1 7.0 0.000 0.000 3.723 3.724 integrate_v_rspace 119 11.3 0.003 0.003 3.681 3.689 qs_ot_get_p 119 10.4 0.001 0.001 3.367 3.500 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.635 3.052 calculate_dm_sparse 119 9.5 0.000 0.000 3.008 3.031 dbcsr_complete_redistribute 329 12.2 0.288 0.293 2.025 2.834 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.855 2.687 make_m2s 4572 13.5 0.038 0.039 2.428 2.603 mp_alltoall_i22 627 13.8 1.472 2.553 1.472 2.553 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.703 2.511 make_images 4572 14.5 0.350 0.383 2.306 2.480 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.179 2.417 apply_single 119 13.6 0.000 0.000 2.179 2.417 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.277 2.277 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.448 2.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.070 2.080 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.076 2.078 pw_transfer 1439 11.6 0.066 0.067 2.048 2.052 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.963 2.046 grid_integrate_task_list 119 12.3 2.027 2.041 2.027 2.041 ot_diis_step 108 11.5 0.014 0.014 2.002 2.002 jit_kernel_multiply 10 15.7 1.095 1.977 1.095 1.977 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.952 1.957 density_rs2pw 119 9.7 0.003 0.004 1.847 1.880 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.797 1.798 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.744 1.791 mp_sum_l 7207 12.9 1.100 1.784 1.100 1.784 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.750 1.752 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.617 1.622 fft3d_ps 1201 14.6 0.565 0.575 1.588 1.592 cp_fm_cholesky_invert 11 10.9 1.565 1.569 1.565 1.569 rs_pw_transfer 974 11.9 0.009 0.009 1.480 1.505 grid_collocate_task_list 119 9.7 1.464 1.491 1.464 1.491 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.480 1.480 cp_fm_diag_elpa_base 50 14.0 1.335 1.387 1.478 1.478 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.217 1.466 wfi_extrapolate 11 7.9 0.001 0.001 1.407 1.407 make_images_data 4572 15.5 0.042 0.045 1.161 1.376 potential_pw2rs 119 12.3 0.014 0.015 1.252 1.254 acc_transpose_blocks 9144 15.5 0.038 0.039 1.176 1.240 mp_alltoall_d11v 2130 13.8 1.208 1.225 1.208 1.225 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.171 1.191 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.092 1.106 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.771 1.061 multiply_cannon_sync_h2d 9144 15.5 1.032 1.038 1.032 1.038 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.985 1.023 acc_transpose_blocks_kernels 9144 16.5 0.119 0.120 0.919 0.975 qs_create_task_list 11 7.9 0.001 0.001 0.943 0.956 generate_qs_task_list 11 8.9 0.372 0.392 0.943 0.955 mp_alltoall_z22v 1201 16.6 0.888 0.908 0.888 0.908 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.878 0.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.283000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.727273, yerr=13.129363 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.583872E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1391902. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.063 84.132 84.142 qs_mol_dyn_low 1 2.0 0.005 0.019 83.558 83.568 qs_forces 11 3.9 0.002 0.002 83.462 83.463 qs_energies 11 4.9 0.001 0.006 80.675 80.684 scf_env_do_scf 11 5.9 0.000 0.001 71.190 71.193 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.617 65.617 dbcsr_multiply_generic 2055 12.4 0.108 0.112 53.273 53.510 qs_scf_new_mos 99 7.5 0.000 0.001 48.770 48.903 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.769 48.902 ot_scf_mini 99 9.5 0.002 0.002 46.344 46.432 multiply_cannon 2055 13.4 0.179 0.187 43.508 44.419 multiply_cannon_loop 2055 14.4 1.576 1.621 42.525 43.465 velocity_verlet 10 3.0 0.001 0.001 41.780 41.780 ot_mini 99 10.5 0.001 0.001 28.315 28.429 qs_ot_get_derivative 99 11.5 0.001 0.001 21.474 21.583 multiply_cannon_multrec 49320 15.4 12.093 12.737 17.624 18.322 rebuild_ks_matrix 110 8.3 0.000 0.000 14.341 14.492 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.341 14.491 mp_waitall_1 241148 16.1 12.073 13.657 12.073 13.657 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.611 12.753 multiply_cannon_sync_h2d 49320 15.4 9.925 10.459 9.925 10.459 qs_ot_get_p 110 10.4 0.001 0.001 9.422 9.591 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 7.127 8.595 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.421 8.010 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.271 7.752 apply_single 110 13.6 0.000 0.000 7.271 7.752 init_scf_run 11 5.9 0.000 0.001 7.345 7.345 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.344 7.345 sum_up_and_integrate 110 10.3 0.036 0.043 6.947 6.959 integrate_v_rspace 110 11.3 0.002 0.003 6.911 6.931 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.822 6.888 ot_diis_step 99 11.5 0.005 0.006 6.531 6.531 mp_sum_l 6514 12.8 5.467 6.386 5.467 6.386 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.176 6.305 calculate_rho_elec 110 8.6 0.020 0.024 6.176 6.304 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.136 6.174 dbcsr_mm_accdrv_process 87628 16.1 2.098 2.200 5.411 5.857 init_scf_loop 11 6.9 0.000 0.000 5.545 5.545 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.085 5.085 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.588 4.589 cp_fm_redistribute_end 48 14.0 3.992 4.561 3.995 4.563 cp_fm_diag_elpa_base 48 14.0 0.561 4.494 0.565 4.519 calculate_dm_sparse 110 9.5 0.001 0.001 4.044 4.144 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.880 4.061 make_m2s 4110 13.4 0.061 0.065 3.925 4.044 wfi_extrapolate 11 7.9 0.001 0.001 3.989 3.989 make_images 4110 14.4 0.176 0.190 3.830 3.951 rs_pw_transfer 902 11.9 0.011 0.013 3.695 3.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.821 3.825 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.491 3.550 prepare_preconditioner 11 7.9 0.000 0.000 3.400 3.418 make_preconditioner 11 8.9 0.000 0.000 3.400 3.418 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.347 3.404 density_rs2pw 110 9.6 0.004 0.004 3.207 3.389 grid_integrate_task_list 110 12.3 3.244 3.382 3.244 3.382 jit_kernel_multiply 13 15.9 3.028 3.290 3.028 3.290 calculate_first_density_matrix 1 7.0 0.002 0.015 3.269 3.277 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.191 3.229 pw_transfer 1331 11.6 0.054 0.064 2.797 2.861 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.710 2.778 potential_pw2rs 110 12.3 0.006 0.007 2.549 2.571 mp_alltoall_d11v 2046 13.8 2.155 2.569 2.155 2.569 acc_transpose_blocks 49320 15.4 0.225 0.235 2.314 2.469 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.284 2.355 fft3d_ps 1111 14.6 0.751 0.838 2.248 2.306 grid_collocate_task_list 110 9.6 2.087 2.211 2.087 2.211 mp_waitany 14300 13.8 1.834 2.067 1.834 2.067 mp_sum_d 3879 11.9 1.466 2.003 1.466 2.003 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.893 1.921 make_images_data 4110 15.4 0.043 0.046 1.788 1.915 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.870 1.887 hybrid_alltoall_any 4261 16.3 0.083 0.484 1.562 1.852 cp_fm_cholesky_invert 11 10.9 1.829 1.834 1.829 1.834 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.142000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.181818, yerr=2.790184 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 587.190272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2661790. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.094 70.713 70.725 qs_mol_dyn_low 1 2.0 0.003 0.004 70.111 70.122 qs_forces 11 3.9 0.002 0.002 70.041 70.042 qs_energies 11 4.9 0.009 0.042 66.485 66.488 scf_env_do_scf 11 5.9 0.001 0.002 56.561 56.564 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 48.949 48.950 dbcsr_multiply_generic 2055 12.4 0.115 0.120 39.143 39.346 velocity_verlet 10 3.0 0.001 0.001 36.043 36.045 multiply_cannon 2055 13.4 0.224 0.242 32.293 33.489 qs_scf_new_mos 99 7.5 0.001 0.001 33.137 33.269 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.137 33.268 multiply_cannon_loop 2055 14.4 0.934 0.959 30.991 31.864 ot_scf_mini 99 9.5 0.003 0.005 31.480 31.602 ot_mini 99 10.5 0.001 0.001 18.782 18.902 multiply_cannon_multrec 24660 15.4 7.608 9.457 14.170 16.029 rebuild_ks_matrix 110 8.3 0.000 0.000 13.871 13.953 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.870 13.953 qs_ot_get_derivative 99 11.5 0.001 0.001 12.922 13.044 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.010 12.081 mp_waitall_1 186928 16.3 8.325 10.815 8.325 10.815 multiply_cannon_sync_h2d 24660 15.4 6.998 8.229 6.998 8.229 multiply_cannon_metrocomm3 24660 15.4 0.069 0.070 5.548 8.164 init_scf_loop 11 6.9 0.001 0.004 7.575 7.582 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.703 7.467 apply_single 110 13.6 0.000 0.001 6.703 7.467 init_scf_run 11 5.9 0.000 0.001 7.332 7.333 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.332 7.333 sum_up_and_integrate 110 10.3 0.053 0.059 6.731 6.775 integrate_v_rspace 110 11.3 0.002 0.002 6.678 6.725 dbcsr_mm_accdrv_process 52282 16.1 5.053 5.970 6.403 6.647 qs_ot_get_p 110 10.4 0.001 0.001 6.055 6.204 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.990 5.849 ot_diis_step 99 11.5 0.010 0.010 5.807 5.808 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.731 5.747 calculate_rho_elec 110 8.6 0.039 0.047 5.731 5.747 prepare_preconditioner 11 7.9 0.000 0.000 5.552 5.570 make_preconditioner 11 8.9 0.000 0.001 5.552 5.570 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.099 5.247 make_m2s 4110 13.4 0.057 0.061 4.243 4.700 make_images 4110 14.4 0.399 0.442 4.134 4.586 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.172 4.174 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.134 4.157 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.752 3.775 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.697 3.698 pw_transfer 1331 11.6 0.066 0.071 3.518 3.689 calculate_first_density_matrix 1 7.0 0.003 0.018 3.628 3.630 wfi_extrapolate 11 7.9 0.001 0.001 3.621 3.621 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.411 3.584 grid_integrate_task_list 110 12.3 3.145 3.339 3.145 3.339 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.200 3.260 rs_pw_transfer 902 11.9 0.012 0.013 3.000 3.251 density_rs2pw 110 9.6 0.004 0.004 3.017 3.217 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.141 3.142 cp_fm_redistribute_end 48 14.0 2.338 3.114 2.340 3.115 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.919 3.091 cp_fm_diag_elpa_base 48 14.0 0.741 3.005 0.772 3.090 calculate_dm_sparse 110 9.5 0.001 0.001 3.058 3.084 fft3d_ps 1111 14.6 1.073 1.267 2.768 2.926 make_images_data 4110 15.4 0.047 0.051 2.416 2.873 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.135 2.778 mp_sum_l 6514 12.8 2.077 2.772 2.077 2.772 potential_pw2rs 110 12.3 0.008 0.009 2.596 2.645 cp_fm_cholesky_invert 11 10.9 2.593 2.600 2.593 2.600 grid_collocate_task_list 110 9.6 2.117 2.270 2.117 2.270 mp_alltoall_d11v 2046 13.8 1.842 2.270 1.842 2.270 acc_transpose_blocks 24660 15.4 0.111 0.114 2.054 2.217 jit_kernel_multiply 9 16.4 1.001 2.191 1.001 2.191 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.028 2.040 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.024 2.025 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.956 1.980 mp_allgather_i34 2055 14.4 0.628 1.718 0.628 1.718 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.635 1.645 multiply_cannon_metrocomm4 22605 15.4 0.076 0.080 0.788 1.609 acc_transpose_blocks_kernels 24660 16.4 0.313 0.323 1.426 1.584 mp_alltoall_z22v 1111 16.6 1.441 1.542 1.441 1.542 dbcsr_complete_redistribute 325 12.2 0.256 0.330 1.233 1.507 mp_waitany 10164 13.8 1.222 1.494 1.222 1.494 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.482 mp_irecv_dv 57340 16.2 0.663 1.446 0.663 1.446 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.725000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.090909, yerr=6.815151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 662.126592E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.100 62.599 62.606 qs_mol_dyn_low 1 2.0 0.011 0.022 61.419 61.429 qs_forces 11 3.9 0.009 0.015 61.344 61.351 qs_energies 11 4.9 0.002 0.009 58.191 58.197 scf_env_do_scf 11 5.9 0.001 0.001 49.084 49.085 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.427 40.427 velocity_verlet 10 3.0 0.001 0.001 32.595 32.596 dbcsr_multiply_generic 2055 12.4 0.116 0.120 29.869 30.171 qs_scf_new_mos 99 7.5 0.001 0.001 25.890 25.994 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.889 25.994 ot_scf_mini 99 9.5 0.003 0.003 24.638 24.758 multiply_cannon 2055 13.4 0.213 0.227 23.177 24.288 multiply_cannon_loop 2055 14.4 0.621 0.633 21.948 23.066 ot_mini 99 10.5 0.001 0.001 14.517 14.643 rebuild_ks_matrix 110 8.3 0.000 0.000 12.195 12.365 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.026 12.194 12.365 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.759 10.913 mp_waitall_1 146766 16.3 7.885 10.899 7.885 10.899 multiply_cannon_multrec 16440 15.4 3.755 4.625 9.784 10.738 qs_ot_get_derivative 99 11.5 0.001 0.001 9.948 10.074 init_scf_loop 11 6.9 0.001 0.002 8.621 8.622 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 5.043 8.184 prepare_preconditioner 11 7.9 0.000 0.000 6.830 6.850 make_preconditioner 11 8.9 0.000 0.001 6.830 6.850 init_scf_run 11 5.9 0.000 0.001 6.691 6.691 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.691 6.691 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.163 6.514 sum_up_and_integrate 110 10.3 0.061 0.061 6.398 6.414 integrate_v_rspace 110 11.3 0.002 0.003 6.337 6.352 dbcsr_mm_accdrv_process 34862 16.1 4.784 5.534 5.885 6.202 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.308 5.911 apply_single 110 13.6 0.000 0.000 5.307 5.910 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.552 5.564 calculate_rho_elec 110 8.6 0.058 0.058 5.551 5.564 qs_ot_get_p 110 10.4 0.001 0.001 5.254 5.408 make_m2s 4110 13.4 0.050 0.052 4.238 4.610 ot_diis_step 99 11.5 0.010 0.011 4.537 4.538 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.618 4.525 make_images 4110 14.4 0.392 0.507 4.123 4.496 multiply_cannon_sync_h2d 16440 15.4 3.599 4.105 3.599 4.105 calculate_first_density_matrix 1 7.0 0.010 0.076 3.652 3.654 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.645 3.649 grid_integrate_task_list 110 12.3 3.199 3.465 3.199 3.465 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.306 3.307 pw_transfer 1331 11.6 0.065 0.071 3.193 3.200 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.170 3.171 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.087 3.095 density_rs2pw 110 9.6 0.004 0.004 2.793 3.027 wfi_extrapolate 11 7.9 0.001 0.001 2.950 2.950 make_images_data 4110 15.4 0.043 0.047 2.472 2.932 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.227 2.910 rs_pw_transfer 902 11.9 0.010 0.011 2.621 2.866 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.786 2.842 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.786 2.788 cp_fm_redistribute_end 48 14.0 1.729 2.761 1.731 2.762 cp_fm_diag_elpa_base 48 14.0 0.969 2.635 1.026 2.740 cp_fm_cholesky_invert 11 10.9 2.699 2.705 2.699 2.705 mp_sum_l 6514 12.8 1.866 2.652 1.866 2.652 calculate_dm_sparse 110 9.5 0.001 0.001 2.612 2.648 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.632 2.640 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.936 2.543 jit_kernel_multiply 12 16.3 0.714 2.524 0.714 2.524 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.373 2.445 mp_irecv_dv 48980 15.7 0.866 2.418 0.866 2.418 fft3d_ps 1111 14.6 1.065 1.077 2.402 2.415 grid_collocate_task_list 110 9.6 2.180 2.295 2.180 2.295 potential_pw2rs 110 12.3 0.010 0.010 2.201 2.210 mp_alltoall_d11v 2046 13.8 1.863 2.168 1.863 2.168 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 1.976 1.979 dbcsr_complete_redistribute 325 12.2 0.357 0.396 1.484 1.944 cp_fm_upper_to_full 70 13.6 1.399 1.860 1.399 1.860 acc_transpose_blocks 16440 15.4 0.072 0.076 1.470 1.746 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.651 mp_allgather_i34 2055 14.4 0.552 1.634 0.552 1.634 cp_fm_cholesky_decompose 22 10.9 1.600 1.623 1.600 1.623 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.553 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.026 1.471 mp_sum_d 3877 11.9 1.070 1.458 1.070 1.458 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.349 1.453 mp_waitany 17072 13.8 1.141 1.365 1.141 1.365 dbcsr_dot_sd 1091 11.9 0.302 0.308 1.033 1.346 acc_transpose_blocks_kernels 16440 16.4 0.210 0.215 1.040 1.303 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.283 1.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.606000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=9.703028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 724.598784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.176 66.122 66.123 qs_mol_dyn_low 1 2.0 0.003 0.003 65.580 65.591 qs_forces 11 3.9 0.002 0.003 65.482 65.483 qs_energies 11 4.9 0.004 0.019 62.133 62.136 scf_env_do_scf 11 5.9 0.001 0.002 52.753 52.756 scf_env_do_scf_inner_loop 99 6.5 0.015 0.055 40.974 40.975 velocity_verlet 10 3.0 0.001 0.001 36.667 36.670 dbcsr_multiply_generic 2055 12.4 0.116 0.118 30.611 30.842 qs_scf_new_mos 99 7.5 0.001 0.001 26.578 26.710 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.578 26.709 ot_scf_mini 99 9.5 0.003 0.004 24.940 25.047 multiply_cannon 2055 13.4 0.245 0.268 23.632 24.799 multiply_cannon_loop 2055 14.4 0.893 0.916 22.208 22.845 ot_mini 99 10.5 0.001 0.001 14.370 14.503 multiply_cannon_multrec 24660 15.4 4.216 6.891 13.081 14.266 rebuild_ks_matrix 110 8.3 0.000 0.000 11.882 11.978 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.027 11.881 11.977 init_scf_loop 11 6.9 0.002 0.006 11.738 11.739 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.523 10.606 qs_ot_get_derivative 99 11.5 0.001 0.001 10.222 10.336 prepare_preconditioner 11 7.9 0.000 0.000 9.998 10.013 make_preconditioner 11 8.9 0.000 0.001 9.998 10.013 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.224 9.658 dbcsr_mm_accdrv_process 52304 16.0 6.876 8.673 8.719 9.588 mp_waitall_1 126806 16.4 4.779 6.732 4.779 6.732 init_scf_run 11 5.9 0.000 0.002 6.488 6.488 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.487 6.488 sum_up_and_integrate 110 10.3 0.068 0.071 6.296 6.310 integrate_v_rspace 110 11.3 0.002 0.005 6.228 6.242 qs_ot_get_p 110 10.4 0.001 0.001 5.548 5.702 make_m2s 4110 13.4 0.060 0.063 5.353 5.684 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.615 5.644 calculate_rho_elec 110 8.6 0.077 0.081 5.615 5.644 make_images 4110 14.4 0.575 0.696 5.212 5.540 cp_fm_upper_to_full 70 13.8 3.268 4.642 3.268 4.642 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.049 4.177 apply_single 110 13.6 0.000 0.000 4.049 4.177 ot_diis_step 99 11.5 0.011 0.011 4.113 4.113 dbcsr_complete_redistribute 325 12.2 0.418 0.478 2.701 3.819 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.772 3.797 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.560 3.562 calculate_first_density_matrix 1 7.0 0.008 0.033 3.461 3.463 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.382 3.455 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.661 3.431 grid_integrate_task_list 110 12.3 3.295 3.410 3.295 3.410 multiply_cannon_sync_h2d 24660 15.4 3.188 3.336 3.188 3.336 pw_transfer 1331 11.6 0.065 0.071 3.253 3.278 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.164 3.274 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.254 3.254 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.173 3.204 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.148 3.176 calculate_dm_sparse 110 9.5 0.001 0.001 3.053 3.082 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.010 3.074 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.313 3.053 make_images_data 4110 15.4 0.046 0.050 2.707 3.015 density_rs2pw 110 9.6 0.004 0.004 2.757 2.943 wfi_extrapolate 11 7.9 0.001 0.001 2.940 2.940 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.765 2.860 cp_fm_cholesky_invert 11 10.9 2.823 2.832 2.823 2.832 mp_alltoall_i22 605 13.7 1.629 2.782 1.629 2.782 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.717 2.719 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 2.674 2.704 cp_fm_redistribute_end 48 14.0 1.351 2.689 1.352 2.690 cp_fm_diag_elpa_base 48 14.0 1.256 2.560 1.334 2.671 jit_kernel_multiply 12 15.7 1.514 2.626 1.514 2.626 rs_pw_transfer 902 11.9 0.010 0.011 2.381 2.572 fft3d_ps 1111 14.6 1.064 1.096 2.451 2.471 grid_collocate_task_list 110 9.6 2.224 2.386 2.224 2.386 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.369 2.370 potential_pw2rs 110 12.3 0.013 0.013 2.068 2.077 acc_transpose_blocks 24660 15.4 0.104 0.106 1.892 2.034 mp_alltoall_d11v 2046 13.8 1.812 1.971 1.812 1.971 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.731 1.788 cp_fm_cholesky_decompose 22 10.9 1.668 1.714 1.668 1.714 mp_sum_l 6514 12.8 1.007 1.701 1.007 1.701 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.701 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.651 1.669 mp_allgather_i34 2055 14.4 0.521 1.622 0.521 1.622 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.604 1.616 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.237 1.421 multiply_cannon_metrocomm4 20550 15.4 0.058 0.062 0.847 1.392 acc_transpose_blocks_kernels 24660 16.4 0.306 0.311 1.244 1.371 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.123000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.545455, yerr=7.679532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 827.727872E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.061 56.626 56.626 qs_mol_dyn_low 1 2.0 0.003 0.003 56.003 56.012 qs_forces 11 3.9 0.002 0.002 55.850 55.851 qs_energies 11 4.9 0.001 0.001 52.254 52.258 scf_env_do_scf 11 5.9 0.000 0.001 42.764 42.764 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.726 34.727 velocity_verlet 10 3.0 0.001 0.001 30.622 30.631 dbcsr_multiply_generic 2055 12.4 0.107 0.110 23.866 23.976 qs_scf_new_mos 99 7.5 0.001 0.001 20.644 20.700 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.644 20.699 multiply_cannon 2055 13.4 0.246 0.260 18.408 19.651 ot_scf_mini 99 9.5 0.002 0.002 19.400 19.447 multiply_cannon_loop 2055 14.4 0.325 0.337 17.048 17.276 rebuild_ks_matrix 110 8.3 0.000 0.000 11.436 11.478 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.435 11.477 ot_mini 99 10.5 0.001 0.001 10.568 10.611 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.192 10.229 multiply_cannon_multrec 8220 15.4 3.284 4.504 8.128 9.243 init_scf_loop 11 6.9 0.000 0.000 7.991 7.993 mp_waitall_1 106626 16.5 6.285 7.915 6.285 7.915 qs_ot_get_derivative 99 11.5 0.001 0.001 6.713 6.759 init_scf_run 11 5.9 0.000 0.001 6.401 6.402 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.401 6.401 sum_up_and_integrate 110 10.3 0.080 0.081 6.135 6.152 integrate_v_rspace 110 11.3 0.003 0.003 6.055 6.072 prepare_preconditioner 11 7.9 0.000 0.000 6.029 6.037 make_preconditioner 11 8.9 0.000 0.000 6.029 6.037 dbcsr_mm_accdrv_process 17442 15.9 3.223 4.103 4.713 5.760 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.600 5.673 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.607 5.623 calculate_rho_elec 110 8.6 0.115 0.115 5.607 5.622 qs_ot_get_p 110 10.4 0.001 0.001 5.207 5.258 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.178 4.572 make_m2s 4110 13.4 0.039 0.041 4.182 4.473 make_images 4110 14.4 0.634 0.692 4.052 4.345 ot_diis_step 99 11.5 0.012 0.012 3.834 3.834 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.768 3.817 apply_single 110 13.6 0.000 0.000 3.767 3.817 calculate_first_density_matrix 1 7.0 0.000 0.000 3.634 3.635 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.572 3.575 grid_integrate_task_list 110 12.3 3.360 3.501 3.360 3.501 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.272 3.272 pw_transfer 1331 11.6 0.065 0.070 3.257 3.266 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.150 3.163 calculate_dm_sparse 110 9.5 0.001 0.001 3.027 3.053 multiply_cannon_sync_h2d 8220 15.4 2.907 3.036 2.907 3.036 cp_fm_cholesky_invert 11 10.9 2.915 2.918 2.915 2.918 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.862 2.863 make_images_data 4110 15.4 0.039 0.043 2.409 2.851 density_rs2pw 110 9.6 0.004 0.004 2.652 2.772 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.747 2.748 cp_fm_redistribute_end 48 14.0 0.693 2.720 0.697 2.720 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 2.695 2.704 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.331 2.703 cp_fm_diag_elpa_base 48 14.0 1.840 2.537 2.016 2.692 wfi_extrapolate 11 7.9 0.001 0.001 2.682 2.682 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.643 2.643 grid_collocate_task_list 110 9.6 2.329 2.541 2.329 2.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.445 2.461 fft3d_ps 1111 14.6 1.113 1.143 2.409 2.421 rs_pw_transfer 902 11.9 0.010 0.011 2.108 2.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.984 1.995 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.951 1.992 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.770 1.991 jit_kernel_multiply 9 15.9 1.184 1.941 1.184 1.941 potential_pw2rs 110 12.3 0.015 0.016 1.883 1.888 mp_alltoall_d11v 2046 13.8 1.652 1.865 1.652 1.865 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.770 1.794 cp_fm_cholesky_decompose 22 10.9 1.668 1.684 1.668 1.684 mp_allgather_i34 2055 14.4 0.545 1.666 0.545 1.666 dbcsr_complete_redistribute 325 12.2 0.594 0.640 1.512 1.605 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.485 1.597 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.574 1.579 qs_create_task_list 11 7.9 0.000 0.001 1.220 1.317 generate_qs_task_list 11 8.9 0.378 0.446 1.220 1.317 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.476 1.272 mp_irecv_dv 24056 15.7 0.452 1.230 0.452 1.230 mp_waitany 9240 13.8 1.036 1.202 1.036 1.202 make_basis_sm 11 9.8 0.020 0.023 1.172 1.174 acc_transpose_blocks 8220 15.4 0.036 0.037 1.150 1.169 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.129 1.143 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.626000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.000000, yerr=8.090398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 1.313718E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.059 88.780 88.781 qs_mol_dyn_low 1 2.0 0.003 0.004 87.822 87.830 qs_forces 11 3.9 0.002 0.003 87.753 87.754 qs_energies 11 4.9 0.001 0.001 83.687 83.688 scf_env_do_scf 11 5.9 0.000 0.001 72.728 72.728 velocity_verlet 10 3.0 0.001 0.001 55.429 55.435 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.965 43.966 dbcsr_multiply_generic 2055 12.4 0.121 0.124 30.110 30.187 init_scf_loop 11 6.9 0.000 0.000 28.685 28.688 qs_scf_new_mos 99 7.5 0.001 0.001 27.466 27.526 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.465 27.525 prepare_preconditioner 11 7.9 0.000 0.000 26.730 26.738 make_preconditioner 11 8.9 0.000 0.000 26.730 26.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.701 26.207 ot_scf_mini 99 9.5 0.002 0.002 25.683 25.725 multiply_cannon 2055 13.4 0.348 0.373 22.979 23.803 multiply_cannon_loop 2055 14.4 0.345 0.348 21.170 21.585 cp_fm_upper_to_full 70 14.2 12.740 18.502 12.740 18.502 ot_mini 99 10.5 0.001 0.001 14.530 14.572 rebuild_ks_matrix 110 8.3 0.001 0.001 13.186 13.225 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.185 13.224 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.990 12.025 dbcsr_complete_redistribute 325 12.2 1.018 1.042 7.648 10.942 multiply_cannon_multrec 8220 15.4 4.439 4.625 10.104 10.389 qs_ot_get_derivative 99 11.5 0.001 0.001 9.954 9.996 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.606 9.901 mp_waitall_1 87304 16.6 8.288 9.303 8.288 9.303 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.014 9.287 mp_alltoall_i22 605 13.7 5.646 8.927 5.646 8.927 sum_up_and_integrate 110 10.3 0.151 0.152 6.671 6.685 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.528 6.563 calculate_rho_elec 110 8.6 0.227 0.227 6.528 6.563 init_scf_run 11 5.9 0.027 0.217 6.540 6.540 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.513 6.540 integrate_v_rspace 110 11.3 0.003 0.003 6.520 6.533 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.258 6.015 qs_ot_get_p 110 10.4 0.001 0.001 5.929 5.980 dbcsr_mm_accdrv_process 11614 15.7 3.639 4.107 5.523 5.908 make_m2s 4110 13.4 0.044 0.045 5.254 5.770 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.768 5.648 apply_single 110 13.6 0.000 0.000 4.767 5.648 make_images 4110 14.4 0.877 0.937 5.066 5.581 cp_fm_cholesky_invert 11 10.9 5.380 5.385 5.380 5.385 ot_diis_step 99 11.5 0.015 0.015 4.548 4.548 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.358 4.366 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.590 4.118 multiply_cannon_sync_h2d 8220 15.4 3.942 3.947 3.942 3.947 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.910 3.910 pw_transfer 1331 11.6 0.074 0.074 3.822 3.825 grid_integrate_task_list 110 12.3 3.661 3.715 3.661 3.715 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.705 3.709 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.639 3.639 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.907 3.574 make_images_data 4110 15.4 0.042 0.045 2.850 3.512 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.473 3.477 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.293 3.293 cp_fm_diag_elpa_base 48 14.0 2.762 2.960 3.291 3.291 wfi_extrapolate 11 7.9 0.001 0.001 3.267 3.267 calculate_dm_sparse 110 9.5 0.001 0.001 3.200 3.226 fft_wrap_pw1pw2_140 451 13.1 0.214 0.218 3.187 3.192 calculate_first_density_matrix 1 7.0 0.000 0.000 3.139 3.167 density_rs2pw 110 9.6 0.004 0.004 3.025 3.045 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.948 2.959 fft3d_ps 1111 14.6 1.269 1.275 2.918 2.921 grid_collocate_task_list 110 9.6 2.630 2.653 2.630 2.653 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.429 2.456 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.170 2.244 rs_pw_transfer 902 11.9 0.010 0.011 2.166 2.219 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.096 2.177 jit_kernel_multiply 10 15.2 1.685 2.175 1.685 2.175 mp_alltoall_d11v 2046 13.8 2.074 2.121 2.074 2.121 cp_fm_cholesky_decompose 22 10.9 2.088 2.100 2.088 2.100 potential_pw2rs 110 12.3 0.021 0.022 2.096 2.100 qs_create_task_list 11 7.9 0.000 0.000 1.900 1.945 generate_qs_task_list 11 8.9 0.739 0.791 1.899 1.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.799 1.803 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.772 1.789 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.781000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1187.909091, yerr=47.227601 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.057856E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4093 56650. MP_Allreduce 11273 784. MP_Sync 170 MP_Alltoall 2226 3406736. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 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 0.547 0.752 206.431 206.436 qs_mol_dyn_low 1 2.0 0.005 0.010 204.789 204.919 qs_forces 11 3.9 0.005 0.007 204.693 204.695 qs_energies 11 4.9 0.001 0.001 199.182 199.204 scf_env_do_scf 11 5.9 0.001 0.001 181.130 181.134 scf_env_do_scf_inner_loop 117 6.6 0.036 0.272 160.170 160.172 dbcsr_multiply_generic 2507 12.6 0.178 0.182 124.409 125.126 velocity_verlet 10 3.0 0.001 0.001 122.028 122.030 qs_scf_new_mos 117 7.6 0.001 0.001 121.081 121.317 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.080 121.316 ot_scf_mini 117 9.6 0.003 0.003 114.522 114.709 multiply_cannon 2507 13.6 0.236 0.245 100.952 103.196 multiply_cannon_loop 2507 14.6 2.147 2.216 98.748 100.925 ot_mini 117 10.6 0.001 0.001 65.874 66.129 multiply_cannon_multrec 60168 15.6 32.975 35.231 41.564 43.222 qs_ot_get_derivative 117 11.6 0.001 0.001 41.083 41.290 rebuild_ks_matrix 128 8.3 0.001 0.001 33.244 33.469 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.244 33.468 mp_waitall_1 291448 16.2 28.773 32.005 28.773 32.005 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.912 30.104 multiply_cannon_sync_h2d 60168 15.6 27.281 28.523 27.281 28.523 qs_ot_get_p 128 10.4 0.001 0.001 26.990 27.324 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.348 25.100 apply_single 128 13.6 0.001 0.001 24.347 25.100 ot_diis_step 117 11.6 0.007 0.008 24.465 24.466 init_scf_loop 11 6.9 0.000 0.000 20.880 20.882 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.128 20.201 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.897 19.059 multiply_cannon_metrocomm3 60168 15.6 0.118 0.124 15.606 18.054 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.579 17.580 prepare_preconditioner 11 7.9 0.000 0.000 16.160 16.210 make_preconditioner 11 8.9 0.000 0.000 16.160 16.210 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.392 15.558 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.690 14.695 cp_fm_redistribute_end 83 14.4 11.571 14.610 11.583 14.613 cp_fm_diag_elpa_base 83 14.4 2.987 14.375 3.017 14.493 make_m2s 5014 13.6 0.104 0.112 13.931 14.262 make_images 5014 14.6 0.400 0.419 13.750 14.094 sum_up_and_integrate 128 10.3 0.089 0.107 14.042 14.055 integrate_v_rspace 128 11.3 0.003 0.004 13.953 13.971 init_scf_run 11 5.9 0.000 0.002 13.316 13.317 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.316 13.316 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.967 13.080 calculate_rho_elec 128 8.7 0.044 0.063 12.966 13.079 mp_sum_l 7870 13.0 8.410 9.573 8.410 9.573 cp_fm_cholesky_invert 11 10.9 9.261 9.269 9.261 9.269 wfi_extrapolate 11 7.9 0.001 0.001 9.026 9.026 dbcsr_mm_accdrv_process 124484 16.2 3.326 3.471 8.156 8.721 calculate_dm_sparse 128 9.5 0.001 0.001 8.554 8.686 multiply_cannon_metrocomm1 60168 15.6 0.088 0.091 6.069 8.441 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.154 8.350 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.812 7.922 make_images_data 5014 15.6 0.067 0.072 6.881 7.881 grid_integrate_task_list 128 12.3 7.109 7.549 7.109 7.549 hybrid_alltoall_any 5200 16.5 0.293 2.253 6.045 7.251 pw_transfer 1547 11.6 0.075 0.106 6.937 7.221 density_rs2pw 128 9.7 0.006 0.007 6.521 7.053 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.732 6.992 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.795 6.806 rs_pw_transfer 1046 11.9 0.016 0.018 6.012 6.743 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.895 6.000 fft_wrap_pw1pw2_140 523 13.2 0.444 0.510 5.747 5.924 mp_alltoall_d11v 2415 14.1 4.496 5.878 4.496 5.878 fft3d_ps 1291 14.7 2.092 2.560 5.529 5.723 grid_collocate_task_list 128 9.7 4.727 5.064 4.727 5.064 cp_fm_cholesky_decompose 22 10.9 4.698 4.713 4.698 4.713 mp_sum_d 4464 12.2 3.792 4.571 3.792 4.571 potential_pw2rs 128 12.3 0.009 0.010 4.516 4.543 calculate_first_density_matrix 1 7.0 0.000 0.001 4.120 4.130 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.436000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.545455, yerr=7.487731 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 825.520128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 6274936. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.085 0.149 194.545 194.593 qs_mol_dyn_low 1 2.0 0.004 0.011 193.707 193.722 qs_forces 11 3.9 0.004 0.013 193.608 193.610 qs_energies 11 4.9 0.013 0.033 186.780 186.806 scf_env_do_scf 11 5.9 0.001 0.002 169.007 169.017 scf_env_do_scf_inner_loop 117 6.6 0.009 0.052 135.875 135.878 velocity_verlet 10 3.0 0.001 0.001 119.874 119.876 dbcsr_multiply_generic 2507 12.6 0.188 0.194 99.634 100.782 qs_scf_new_mos 117 7.6 0.001 0.001 96.922 97.238 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.921 97.237 ot_scf_mini 117 9.6 0.004 0.005 92.118 92.518 multiply_cannon 2507 13.6 0.481 0.534 78.946 83.187 multiply_cannon_loop 2507 14.6 1.258 1.307 75.698 78.480 ot_mini 117 10.6 0.001 0.001 51.099 51.445 mp_waitall_1 226760 16.4 25.125 38.386 25.125 38.386 multiply_cannon_multrec 30084 15.6 22.196 26.445 32.752 37.461 rebuild_ks_matrix 128 8.3 0.001 0.001 33.005 33.470 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.004 33.469 init_scf_loop 11 6.9 0.017 0.035 33.039 33.041 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.711 30.137 qs_ot_get_derivative 117 11.6 0.001 0.002 29.249 29.637 prepare_preconditioner 11 7.9 0.000 0.000 28.564 28.622 make_preconditioner 11 8.9 0.000 0.002 28.564 28.622 multiply_cannon_metrocomm3 30084 15.6 0.092 0.097 15.648 27.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.234 27.768 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.449 23.886 apply_single 128 13.6 0.001 0.001 22.449 23.886 qs_ot_get_p 128 10.4 0.001 0.002 22.694 23.125 ot_diis_step 117 11.6 0.014 0.015 21.685 21.687 multiply_cannon_sync_h2d 30084 15.6 19.488 21.328 19.488 21.328 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.769 17.804 cp_fm_cholesky_invert 11 10.9 16.694 16.707 16.694 16.707 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.618 16.619 make_m2s 5014 13.6 0.090 0.097 14.507 16.027 make_images 5014 14.6 1.155 1.347 14.296 15.818 sum_up_and_integrate 128 10.3 0.116 0.134 14.475 14.519 integrate_v_rspace 128 11.3 0.003 0.003 14.358 14.409 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.480 13.490 cp_fm_redistribute_end 83 14.4 7.840 13.415 7.852 13.417 cp_fm_diag_elpa_base 83 14.4 5.326 12.958 5.546 13.319 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.101 13.130 calculate_rho_elec 128 8.7 0.087 0.105 13.100 13.129 init_scf_run 11 5.9 0.000 0.001 12.927 12.929 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.927 12.929 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.261 11.538 make_images_data 5014 15.6 0.064 0.073 8.830 10.667 dbcsr_mm_accdrv_process 62242 16.2 4.867 6.051 10.017 10.589 multiply_cannon_metrocomm4 27577 15.6 0.101 0.113 3.708 10.509 mp_irecv_dv 69486 16.3 3.511 10.113 3.511 10.113 hybrid_alltoall_any 5200 16.5 0.344 1.517 7.466 9.929 wfi_extrapolate 11 7.9 0.001 0.001 8.370 8.370 pw_transfer 1547 11.6 0.085 0.105 7.899 7.965 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.675 7.741 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.528 7.730 grid_integrate_task_list 128 12.3 7.178 7.557 7.178 7.557 calculate_dm_sparse 128 9.5 0.002 0.004 7.310 7.444 density_rs2pw 128 9.7 0.006 0.006 6.813 7.306 cp_fm_cholesky_decompose 22 10.9 6.934 7.024 6.934 7.024 fft_wrap_pw1pw2_140 523 13.2 0.468 0.517 6.684 6.752 mp_sum_l 7870 13.0 4.473 6.613 4.473 6.613 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.325 6.334 rs_pw_transfer 1046 11.9 0.014 0.016 5.656 6.236 fft3d_ps 1291 14.7 2.770 2.908 6.055 6.094 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.457 5.572 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.316 5.380 grid_collocate_task_list 128 9.7 4.909 5.289 4.909 5.289 mp_allgather_i34 2507 14.6 1.814 5.137 1.814 5.137 mp_alltoall_d11v 2415 14.1 4.586 5.021 4.586 5.021 potential_pw2rs 128 12.3 0.014 0.016 4.905 4.945 calculate_first_density_matrix 1 7.0 0.001 0.003 4.384 4.392 dbcsr_complete_redistribute 395 12.7 0.768 0.869 3.244 4.085 mp_sum_d 4455 12.2 2.685 4.006 2.685 4.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.593000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.636364, yerr=3.699453 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 938.590208E+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 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 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 0.021 0.036 177.241 177.244 qs_mol_dyn_low 1 2.0 0.003 0.003 176.609 176.623 qs_forces 11 3.9 0.003 0.003 176.509 176.514 qs_energies 11 4.9 0.001 0.005 170.023 170.032 scf_env_do_scf 11 5.9 0.001 0.001 153.331 153.333 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 118.078 118.080 velocity_verlet 10 3.0 0.001 0.001 111.682 111.685 dbcsr_multiply_generic 2485 12.5 0.182 0.186 82.411 83.539 qs_scf_new_mos 116 7.6 0.001 0.001 81.877 82.180 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.876 82.179 ot_scf_mini 116 9.6 0.003 0.004 77.778 78.136 multiply_cannon 2485 13.5 0.497 0.522 62.838 67.120 multiply_cannon_loop 2485 14.5 0.855 0.889 59.684 62.337 ot_mini 116 10.6 0.001 0.001 42.882 43.244 init_scf_loop 11 6.9 0.000 0.000 35.151 35.152 mp_waitall_1 176908 16.5 25.219 34.588 25.219 34.588 prepare_preconditioner 11 7.9 0.000 0.000 31.142 31.193 make_preconditioner 11 8.9 0.000 0.000 31.142 31.193 rebuild_ks_matrix 127 8.3 0.001 0.001 29.796 30.252 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.795 30.251 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.619 30.173 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.895 27.311 multiply_cannon_multrec 19880 15.5 13.420 16.343 23.031 25.776 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 15.122 24.582 qs_ot_get_derivative 116 11.6 0.001 0.002 23.386 23.759 qs_ot_get_p 127 10.4 0.001 0.001 20.386 20.789 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.726 20.697 apply_single 127 13.6 0.001 0.001 19.725 20.697 ot_diis_step 116 11.6 0.018 0.018 19.388 19.389 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.182 16.188 multiply_cannon_sync_h2d 19880 15.5 14.209 15.834 14.209 15.834 make_m2s 4970 13.5 0.081 0.087 14.578 15.700 make_images 4970 14.5 1.164 1.268 14.346 15.461 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.196 15.197 cp_fm_cholesky_invert 11 10.9 14.424 14.432 14.424 14.432 sum_up_and_integrate 127 10.3 0.131 0.144 13.983 14.007 integrate_v_rspace 127 11.3 0.003 0.004 13.852 13.879 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.954 12.994 calculate_rho_elec 127 8.7 0.131 0.145 12.954 12.993 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.086 12.088 cp_fm_redistribute_end 82 14.4 4.540 12.026 4.553 12.028 cp_fm_diag_elpa_base 82 14.4 7.046 11.474 7.454 11.939 init_scf_run 11 5.9 0.000 0.001 11.631 11.631 scf_env_initial_rho_setup 11 6.9 0.002 0.016 11.631 11.631 make_images_data 4970 15.5 0.060 0.066 8.996 10.486 hybrid_alltoall_any 5155 16.4 0.427 1.940 7.752 9.818 dbcsr_mm_accdrv_process 41158 16.2 4.987 6.097 9.067 9.422 multiply_cannon_metrocomm4 17395 15.5 0.060 0.070 3.390 9.098 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.769 9.024 mp_irecv_dv 49801 16.2 3.269 8.860 3.269 8.860 pw_transfer 1535 11.6 0.084 0.102 7.723 7.825 grid_integrate_task_list 127 12.3 7.282 7.678 7.282 7.678 cp_fm_upper_to_full 104 14.5 5.880 7.642 5.880 7.642 fft_wrap_pw1pw2 1281 12.7 0.009 0.010 7.502 7.609 wfi_extrapolate 11 7.9 0.001 0.001 7.391 7.391 cp_fm_cholesky_decompose 22 10.9 7.275 7.304 7.275 7.304 dbcsr_complete_redistribute 393 12.7 1.169 1.195 5.137 7.046 density_rs2pw 127 9.7 0.005 0.006 6.425 6.773 fft_wrap_pw1pw2_140 519 13.2 0.474 0.527 6.594 6.713 calculate_dm_sparse 127 9.5 0.002 0.011 6.282 6.456 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.017 6.029 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.944 5.852 fft3d_ps 1281 14.7 2.673 2.884 5.784 5.849 grid_collocate_task_list 127 9.7 5.043 5.496 5.043 5.496 rs_pw_transfer 1038 11.9 0.013 0.014 4.988 5.398 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.596 5.318 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.102 5.220 mp_alltoall_d11v 2401 14.1 4.252 4.700 4.252 4.700 mp_allgather_i34 2485 14.5 1.686 4.697 1.686 4.697 mp_sum_l 7804 13.0 3.185 4.559 3.185 4.559 potential_pw2rs 127 12.3 0.020 0.022 4.433 4.448 mp_alltoall_i22 712 14.1 2.487 4.443 2.487 4.443 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.503 4.292 calculate_first_density_matrix 1 7.0 0.050 0.400 4.061 4.067 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.963 4.013 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.804 3.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.244000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.636364, yerr=10.645295 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.149014E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57219. MP_Allreduce 11246 1068. MP_Sync 168 MP_Alltoall 1700 12496371. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.120 186.814 186.818 qs_mol_dyn_low 1 2.0 0.003 0.003 186.034 186.047 qs_forces 11 3.9 0.003 0.003 185.925 185.934 qs_energies 11 4.9 0.011 0.077 178.870 178.901 scf_env_do_scf 11 5.9 0.003 0.040 161.678 161.690 velocity_verlet 10 3.0 0.001 0.001 122.661 122.663 scf_env_do_scf_inner_loop 116 6.6 0.004 0.016 115.300 115.301 qs_scf_new_mos 116 7.6 0.001 0.001 80.171 80.471 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.171 80.470 dbcsr_multiply_generic 2485 12.5 0.189 0.193 79.377 80.069 ot_scf_mini 116 9.6 0.004 0.004 75.699 76.055 multiply_cannon 2485 13.5 0.558 0.606 54.996 57.854 multiply_cannon_loop 2485 14.5 1.183 1.213 51.115 52.853 init_scf_loop 11 6.9 0.001 0.003 46.214 46.218 ot_mini 116 10.6 0.001 0.001 42.245 42.570 prepare_preconditioner 11 7.9 0.000 0.000 42.132 42.158 make_preconditioner 11 8.9 0.000 0.001 42.132 42.158 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.843 40.807 multiply_cannon_multrec 29820 15.5 14.108 19.522 26.371 31.131 rebuild_ks_matrix 127 8.3 0.001 0.001 28.773 29.074 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 28.772 29.074 mp_waitall_1 152434 16.5 17.413 27.263 17.413 27.263 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.993 26.268 qs_ot_get_derivative 116 11.6 0.001 0.002 22.730 23.082 make_m2s 4970 13.5 0.095 0.097 20.093 21.227 make_images 4970 14.5 1.946 2.260 19.792 20.926 qs_ot_get_p 127 10.4 0.001 0.001 19.797 20.121 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.902 19.409 apply_single 127 13.6 0.001 0.001 18.902 19.408 ot_diis_step 116 11.6 0.017 0.019 19.385 19.387 cp_fm_upper_to_full 104 14.7 10.986 16.260 10.986 16.260 cp_fm_cholesky_invert 11 10.9 16.161 16.170 16.161 16.170 qs_ot_p2m_diag 82 11.4 0.340 0.385 15.702 15.754 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.359 15.089 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.433 14.435 sum_up_and_integrate 127 10.3 0.139 0.150 13.945 13.970 integrate_v_rspace 127 11.3 0.003 0.003 13.806 13.836 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.956 12.991 calculate_rho_elec 127 8.7 0.173 0.189 12.955 12.990 make_images_data 4970 15.5 0.063 0.067 10.843 12.850 dbcsr_complete_redistribute 393 12.7 1.500 1.623 8.876 12.598 dbcsr_mm_accdrv_process 61748 16.2 7.415 8.596 11.843 12.369 multiply_cannon_sync_h2d 29820 15.5 11.605 12.363 11.605 12.363 hybrid_alltoall_any 5155 16.4 0.522 2.204 9.718 11.401 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.387 11.390 cp_fm_redistribute_end 82 14.4 1.940 11.315 1.955 11.319 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.499 11.215 cp_fm_diag_elpa_base 82 14.4 8.760 10.725 9.331 11.206 init_scf_run 11 5.9 0.001 0.007 11.158 11.159 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.157 11.159 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.266 9.865 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.268 9.533 mp_alltoall_i22 712 14.1 5.575 9.211 5.575 9.211 grid_integrate_task_list 127 12.3 7.464 7.884 7.464 7.884 pw_transfer 1535 11.6 0.084 0.099 7.768 7.839 cp_fm_cholesky_decompose 22 10.9 7.630 7.740 7.630 7.740 fft_wrap_pw1pw2 1281 12.7 0.009 0.010 7.544 7.621 wfi_extrapolate 11 7.9 0.001 0.001 7.618 7.619 multiply_cannon_metrocomm4 24850 15.5 0.076 0.086 2.729 7.173 mp_irecv_dv 75445 16.2 2.585 6.901 2.585 6.901 fft_wrap_pw1pw2_140 519 13.2 0.478 0.486 6.678 6.776 density_rs2pw 127 9.7 0.005 0.006 6.165 6.647 calculate_dm_sparse 127 9.5 0.001 0.001 6.229 6.314 fft3d_ps 1281 14.7 2.777 2.864 5.791 5.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.720 5.772 grid_collocate_task_list 127 9.7 5.161 5.508 5.161 5.508 mp_alltoall_d11v 2401 14.1 4.752 5.395 4.752 5.395 rs_pw_transfer 1038 11.9 0.013 0.014 4.534 5.013 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.638 4.714 qs_energies_init_hamiltonians 11 5.9 0.006 0.020 4.471 4.473 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.379 4.468 potential_pw2rs 127 12.3 0.022 0.022 4.203 4.218 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.130 4.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.818000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.363636, yerr=21.968046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 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 11613089636352 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 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.511834E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57813. MP_Allreduce 11072 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.055 174.508 174.509 qs_mol_dyn_low 1 2.0 0.003 0.003 173.913 173.925 qs_forces 11 3.9 0.003 0.003 173.815 173.819 qs_energies 11 4.9 0.001 0.001 166.583 166.592 scf_env_do_scf 11 5.9 0.002 0.030 148.205 148.217 velocity_verlet 10 3.0 0.001 0.001 112.803 112.807 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 112.634 112.635 qs_scf_new_mos 118 7.6 0.001 0.001 76.580 76.708 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.580 76.707 dbcsr_multiply_generic 2527 12.6 0.186 0.192 75.927 76.407 ot_scf_mini 118 9.6 0.004 0.004 72.117 72.195 multiply_cannon 2527 13.6 0.587 0.613 55.871 60.412 multiply_cannon_loop 2527 14.6 0.455 0.465 51.156 52.013 ot_mini 118 10.6 0.001 0.001 40.259 40.332 init_scf_loop 11 6.9 0.000 0.000 35.388 35.389 mp_waitall_1 130650 16.6 26.544 34.116 26.544 34.116 prepare_preconditioner 11 7.9 0.000 0.000 31.509 31.548 make_preconditioner 11 8.9 0.000 0.000 31.509 31.548 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.298 29.563 rebuild_ks_matrix 129 8.3 0.001 0.001 28.772 28.858 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.772 28.857 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.353 26.429 multiply_cannon_multrec 10108 15.6 10.482 14.628 18.604 21.379 qs_ot_get_derivative 118 11.6 0.001 0.002 20.107 20.177 ot_diis_step 118 11.6 0.020 0.021 20.085 20.086 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.615 19.901 apply_single 129 13.6 0.001 0.001 19.615 19.901 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 12.385 19.609 qs_ot_get_p 129 10.4 0.001 0.001 18.796 18.907 make_m2s 5054 13.6 0.068 0.072 16.439 18.520 make_images 5054 14.6 2.322 2.697 16.129 18.213 cp_fm_cholesky_invert 11 10.9 17.869 17.875 17.869 17.875 qs_ot_p2m_diag 83 11.4 0.496 0.501 14.972 14.988 sum_up_and_integrate 129 10.3 0.183 0.193 14.168 14.222 integrate_v_rspace 129 11.3 0.003 0.004 13.985 14.049 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.820 13.822 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.334 13.372 calculate_rho_elec 129 8.7 0.260 0.271 13.333 13.371 make_images_data 5054 15.6 0.053 0.061 10.055 12.430 hybrid_alltoall_any 5240 16.5 0.824 3.858 9.770 12.267 multiply_cannon_sync_h2d 10108 15.6 11.749 12.201 11.749 12.201 init_scf_run 11 5.9 0.000 0.001 11.606 11.606 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.605 11.606 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.787 10.789 cp_fm_diag_elpa_base 83 14.4 10.520 10.608 10.772 10.773 dbcsr_mm_accdrv_process 20926 16.1 2.749 3.735 7.750 8.502 grid_integrate_task_list 129 12.3 7.887 8.165 7.887 8.165 multiply_cannon_metrocomm1 10108 15.6 0.028 0.029 4.552 8.109 cp_fm_cholesky_decompose 22 10.9 7.911 8.016 7.911 8.016 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.940 7.983 pw_transfer 1559 11.6 0.084 0.091 7.916 7.947 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.694 7.721 mp_allgather_i34 2527 14.6 2.905 7.532 2.905 7.532 wfi_extrapolate 11 7.9 0.001 0.001 7.316 7.316 calculate_dm_sparse 129 9.5 0.001 0.001 6.889 6.958 fft_wrap_pw1pw2_140 527 13.2 0.505 0.525 6.754 6.789 density_rs2pw 129 9.7 0.005 0.006 6.074 6.280 fft3d_ps 1301 14.7 2.745 2.817 5.845 5.880 dbcsr_complete_redistribute 395 12.7 2.113 2.240 5.380 5.769 grid_collocate_task_list 129 9.7 5.544 5.765 5.544 5.765 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.274 5.279 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.219 5.220 mp_alltoall_d11v 2423 14.1 4.800 5.146 4.800 5.146 rs_pw_transfer 1054 12.0 0.013 0.013 4.168 4.404 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.181 4.201 potential_pw2rs 129 12.3 0.027 0.028 4.156 4.173 calculate_first_density_matrix 1 7.0 0.000 0.000 4.094 4.097 multiply_cannon_metrocomm4 7581 15.6 0.023 0.026 1.871 3.933 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.532 3.866 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.554 3.860 mp_irecv_dv 29086 15.9 1.835 3.856 1.835 3.856 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.712 3.743 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.665 3.711 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.588 3.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.509000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1409.454545, yerr=39.199637 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_performance_tests/20/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 11.696233E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 2.964177E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59285. MP_Allreduce 10977 1514. MP_Sync 87 MP_Alltoall 1712 36974159. MP_ISendRecv 3584 218624. MP_Wait 11594 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.050 290.938 290.939 qs_mol_dyn_low 1 2.0 0.003 0.004 290.251 290.263 qs_forces 11 3.9 0.003 0.004 290.158 290.161 qs_energies 11 4.9 0.001 0.001 281.529 281.540 scf_env_do_scf 11 5.9 0.001 0.004 258.715 258.733 velocity_verlet 10 3.0 0.001 0.001 208.091 208.099 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.604 134.606 init_scf_loop 11 6.9 0.000 0.000 123.854 123.860 prepare_preconditioner 11 7.9 0.000 0.000 119.120 119.153 make_preconditioner 11 8.9 0.000 0.000 119.120 119.153 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.218 116.250 qs_scf_new_mos 117 7.6 0.001 0.001 92.635 92.765 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.634 92.764 ot_scf_mini 117 9.6 0.004 0.004 87.832 87.891 dbcsr_multiply_generic 2507 12.6 0.216 0.224 85.255 85.632 cp_fm_upper_to_full 105 14.8 52.411 75.233 52.411 75.233 multiply_cannon 2507 13.6 0.701 0.774 60.899 61.409 multiply_cannon_loop 2507 14.6 0.474 0.485 57.209 58.382 ot_mini 117 10.6 0.001 0.001 45.390 45.449 dbcsr_complete_redistribute 395 12.7 3.987 4.026 29.550 42.256 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.169 38.860 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.857 36.448 mp_alltoall_i22 716 14.1 21.691 34.571 21.691 34.571 rebuild_ks_matrix 128 8.3 0.001 0.001 33.547 33.592 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 33.546 33.591 cp_fm_cholesky_invert 11 10.9 33.529 33.535 33.529 33.535 mp_waitall_1 105466 16.7 29.108 32.817 29.108 32.817 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.303 31.347 qs_ot_get_p 128 10.4 0.001 0.001 26.836 26.890 qs_ot_get_derivative 117 11.6 0.002 0.002 25.086 25.143 qs_ot_p2m_diag 83 11.4 0.878 0.883 22.774 22.806 cp_dbcsr_syevd 83 12.4 0.006 0.006 21.036 21.039 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 19.721 21.013 make_m2s 5014 13.6 0.076 0.079 19.723 20.680 ot_diis_step 117 11.6 0.022 0.023 20.273 20.273 make_images 5014 14.6 3.786 4.016 19.244 20.204 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.512 19.715 apply_single 128 13.6 0.001 0.001 19.511 19.715 multiply_cannon_multrec 10028 15.6 10.547 12.439 18.852 19.356 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.819 17.820 cp_fm_diag_elpa_base 83 14.4 13.499 15.053 17.814 17.814 sum_up_and_integrate 128 10.3 0.322 0.323 15.692 15.790 multiply_cannon_sync_h2d 10028 15.6 15.648 15.660 15.648 15.660 integrate_v_rspace 128 11.3 0.004 0.004 15.370 15.468 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.949 14.968 calculate_rho_elec 128 8.7 0.482 0.483 14.949 14.968 init_scf_run 11 5.9 0.000 0.001 12.871 12.871 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.870 12.871 make_images_data 5014 15.6 0.060 0.063 10.462 12.278 hybrid_alltoall_any 5200 16.5 1.300 3.024 10.534 12.229 dbcsr_mm_accdrv_process 20762 16.1 3.870 5.612 8.069 10.055 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.489 9.541 cp_fm_cholesky_decompose 22 10.9 8.988 9.029 8.988 9.029 grid_integrate_task_list 128 12.3 8.537 8.731 8.537 8.731 wfi_extrapolate 11 7.9 0.001 0.001 8.675 8.675 pw_transfer 1547 11.6 0.090 0.091 8.338 8.346 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.104 8.112 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 7.942 7.943 calculate_dm_sparse 128 9.5 0.001 0.001 7.492 7.594 fft_wrap_pw1pw2_140 523 13.2 0.540 0.542 7.185 7.203 mp_alltoall_d11v 2415 14.1 6.903 7.051 6.903 7.051 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.517 6.621 grid_collocate_task_list 128 9.7 6.336 6.389 6.336 6.389 fft3d_ps 1291 14.7 2.763 2.785 6.146 6.159 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.039 6.158 density_rs2pw 128 9.7 0.005 0.005 6.013 6.052 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.939000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2636.363636, yerr=170.695515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 1.261474E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.073 85.695 85.697 qs_energies 1 2.0 0.000 0.000 85.062 85.073 ls_scf 1 3.0 0.038 0.304 84.147 84.157 dbcsr_multiply_generic 111 6.7 0.015 0.015 73.157 73.363 multiply_cannon 111 7.7 0.018 0.021 55.947 57.049 multiply_cannon_loop 111 8.7 0.212 0.227 52.531 53.790 ls_scf_main 1 4.0 0.009 0.092 52.453 52.454 density_matrix_trs4 2 5.0 0.002 0.003 46.949 47.048 ls_scf_init_scf 1 4.0 0.001 0.003 28.328 28.329 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.214 27.279 mp_waitall_1 11316 10.9 22.740 25.770 22.740 25.770 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.088 25.109 multiply_cannon_multrec 2664 9.7 8.167 8.940 15.488 17.303 multiply_cannon_sync_h2d 2664 9.7 13.573 14.773 13.573 14.773 make_m2s 222 7.7 0.008 0.011 13.207 13.687 make_images 222 8.7 0.099 0.109 13.185 13.667 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.789 12.710 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.483 9.233 make_images_data 222 9.7 0.004 0.005 7.788 8.388 hybrid_alltoall_any 227 10.6 0.216 1.845 6.676 8.148 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.630 6.939 7.936 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.102 6.208 7.102 calculate_norms 4752 9.8 5.542 6.237 5.542 6.237 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.369 5.540 mp_sum_l 807 5.4 3.496 4.757 3.496 4.757 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.047 3.972 mp_irecv_dv 6231 10.9 2.029 3.936 2.029 3.936 make_images_sizes 222 9.7 0.000 0.000 0.736 3.556 mp_alltoall_i44 222 10.7 0.736 3.555 0.736 3.555 ls_scf_post 1 4.0 0.000 0.001 3.328 3.375 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.353 3.362 arnoldi_extremal 4 6.8 0.000 0.000 3.212 3.241 arnoldi_normal_ev 4 7.8 0.001 0.003 3.212 3.241 ls_scf_store_result 1 5.0 0.000 0.000 3.145 3.221 build_subspace 16 8.4 0.009 0.012 3.111 3.113 dbcsr_special_finalize 555 9.7 0.005 0.006 2.329 2.800 dbcsr_merge_single_wm 555 10.7 0.455 0.580 2.321 2.791 make_images_pack 222 9.7 2.208 2.629 2.210 2.631 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.311 2.554 dbcsr_sort_data 658 11.4 2.125 2.529 2.125 2.529 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.461 2.070 2.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.221 2.295 buffer_matrices_ensure_size 222 8.7 1.745 2.111 1.745 2.111 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.784 1.787 rebuild_ks_matrix 3 7.3 0.000 0.000 1.774 1.777 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.774 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.697000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 2.108568E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.064 91.660 91.662 qs_energies 1 2.0 0.002 0.011 90.996 91.020 ls_scf 1 3.0 0.000 0.002 89.680 89.700 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.122 75.509 multiply_cannon 111 7.7 0.034 0.058 53.226 56.607 ls_scf_main 1 4.0 0.000 0.001 54.659 54.666 multiply_cannon_loop 111 8.7 0.117 0.122 49.940 53.063 density_matrix_trs4 2 5.0 0.007 0.031 49.019 49.170 ls_scf_init_scf 1 4.0 0.000 0.002 31.487 31.492 ls_scf_init_matrix_S 1 5.0 0.002 0.005 30.145 30.264 mp_waitall_1 9246 10.9 21.210 29.977 21.210 29.977 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.681 27.697 multiply_cannon_multrec 1332 9.7 13.084 16.702 22.300 27.014 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.718 20.602 make_m2s 222 7.7 0.006 0.007 15.331 15.924 make_images 222 8.7 1.569 1.961 15.301 15.896 dbcsr_mm_accdrv_process 4041 10.4 0.287 0.490 8.819 10.384 dbcsr_mm_accdrv_process_sort 4041 11.4 8.389 9.894 8.389 9.894 make_images_data 222 9.7 0.004 0.004 8.841 9.705 hybrid_alltoall_any 227 10.6 0.520 2.476 8.298 9.345 mp_sum_l 807 5.4 5.276 8.445 5.276 8.445 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.233 7.958 mp_irecv_dv 3311 11.0 3.213 7.905 3.213 7.905 calculate_norms 2376 9.8 6.020 6.790 6.020 6.790 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.036 6.632 multiply_cannon_sync_h2d 1332 9.7 4.801 6.154 4.801 6.154 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.051 5.262 arnoldi_extremal 4 6.8 0.000 0.000 5.087 5.130 arnoldi_normal_ev 4 7.8 0.048 0.381 5.087 5.130 build_subspace 16 8.4 0.014 0.021 4.389 4.393 ls_scf_post 1 4.0 0.005 0.020 3.533 3.553 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.141 3.385 ls_scf_store_result 1 5.0 0.000 0.000 3.220 3.349 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.217 2.741 3.219 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.273 3.083 ls_scf_dm_to_ks 2 5.0 0.001 0.015 2.547 2.653 make_images_pack 222 9.7 2.027 2.372 2.029 2.374 mp_allgather_i34 111 8.7 0.975 2.306 0.975 2.306 dbcsr_sort_data 436 11.2 1.808 2.055 1.808 2.055 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.016 2.020 rebuild_ks_matrix 3 7.3 0.000 0.000 2.002 2.007 qs_ks_build_kohn_sham_matrix 3 8.3 0.013 0.065 2.002 2.007 buffer_matrices_ensure_size 222 8.7 1.011 1.862 1.011 1.862 dbcsr_data_new 4174 10.1 1.614 1.837 1.614 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.662000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1781.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 2.714063E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.077 92.892 92.893 qs_energies 1 2.0 0.000 0.000 91.896 91.927 ls_scf 1 3.0 0.000 0.000 90.509 90.540 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.081 75.360 ls_scf_main 1 4.0 0.000 0.000 56.217 56.231 multiply_cannon 111 7.7 0.039 0.084 51.811 55.982 multiply_cannon_loop 111 8.7 0.100 0.106 48.190 51.854 density_matrix_trs4 2 5.0 0.002 0.003 50.426 50.605 mp_waitall_1 7374 11.0 22.098 31.929 22.098 31.929 ls_scf_init_scf 1 4.0 0.000 0.000 30.597 30.601 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.363 29.441 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.918 26.926 multiply_cannon_multrec 888 9.7 12.753 15.375 21.292 24.641 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 9.962 21.828 make_m2s 222 7.7 0.006 0.007 16.640 17.894 make_images 222 8.7 1.973 2.298 16.601 17.857 hybrid_alltoall_any 227 10.6 0.624 2.867 8.801 10.273 make_images_data 222 9.7 0.003 0.004 9.175 10.229 dbcsr_mm_accdrv_process 3754 10.4 0.274 0.511 8.067 9.360 dbcsr_mm_accdrv_process_sort 3754 11.4 7.675 8.849 7.675 8.849 mp_sum_l 807 5.4 5.082 8.427 5.082 8.427 multiply_cannon_sync_h2d 888 9.7 6.053 7.587 6.053 7.587 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.412 7.077 mp_irecv_dv 2335 11.1 2.396 7.015 2.396 7.015 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.893 6.716 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.586 6.656 arnoldi_extremal 4 6.8 0.000 0.000 5.121 5.129 arnoldi_normal_ev 4 7.8 0.002 0.006 5.121 5.129 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.920 5.073 build_subspace 16 8.4 0.014 0.020 4.809 4.816 calculate_norms 1584 9.8 4.275 4.644 4.275 4.644 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.452 3.798 mp_allgather_i34 111 8.7 1.453 3.788 1.453 3.788 ls_scf_post 1 4.0 0.000 0.000 3.695 3.728 dbcsr_matrix_vector_mult_local 304 10.0 3.022 3.611 3.024 3.613 ls_scf_store_result 1 5.0 0.000 0.000 3.251 3.338 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.708 2.817 make_images_pack 222 9.7 1.829 2.137 1.831 2.139 dbcsr_sort_data 325 11.1 1.862 2.124 1.862 2.124 make_images_sizes 222 9.7 0.000 0.000 1.146 2.088 mp_alltoall_i44 222 10.7 1.146 2.088 1.146 2.088 dbcsr_data_release 9322 10.9 1.312 2.005 1.312 2.005 dbcsr_finalize 304 7.8 0.026 0.032 1.609 1.913 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.898 1.901 rebuild_ks_matrix 3 7.3 0.000 0.000 1.881 1.884 qs_ks_build_kohn_sham_matrix 3 8.3 0.013 0.015 1.881 1.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.893000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2181.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 3.325415E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.084 99.326 99.334 qs_energies 1 2.0 0.000 0.000 98.181 98.199 ls_scf 1 3.0 0.010 0.043 96.518 96.537 dbcsr_multiply_generic 111 6.7 0.024 0.076 79.505 79.809 ls_scf_main 1 4.0 0.000 0.000 59.332 59.340 multiply_cannon 111 7.7 0.079 0.159 52.192 56.000 density_matrix_trs4 2 5.0 0.002 0.003 53.278 53.396 multiply_cannon_loop 111 8.7 0.115 0.126 47.198 50.881 ls_scf_init_scf 1 4.0 0.000 0.002 33.861 33.870 ls_scf_init_matrix_S 1 5.0 0.002 0.006 31.868 31.959 mp_waitall_1 6438 11.0 23.297 29.525 23.297 29.525 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.388 29.401 multiply_cannon_multrec 1332 9.7 14.141 17.022 22.099 24.959 make_m2s 222 7.7 0.007 0.009 21.325 22.652 make_images 222 8.7 3.143 3.602 21.275 22.605 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.734 17.594 make_images_data 222 9.7 0.004 0.004 11.803 13.327 hybrid_alltoall_any 227 10.6 0.798 3.827 11.222 12.859 dbcsr_mm_accdrv_process 3641 10.4 0.315 0.785 7.589 8.962 dbcsr_mm_accdrv_process_sort 3641 11.4 7.097 8.544 7.097 8.544 mp_sum_l 807 5.4 4.435 8.089 4.435 8.089 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.432 6.374 multiply_cannon_sync_h2d 1332 9.7 5.556 6.083 5.556 6.083 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 6.058 mp_irecv_dv 3229 10.9 2.059 5.984 2.059 5.984 arnoldi_extremal 4 6.8 0.000 0.000 5.223 5.242 arnoldi_normal_ev 4 7.8 0.002 0.006 5.223 5.242 build_subspace 16 8.4 0.014 0.021 4.872 4.880 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.575 4.767 calculate_norms 2376 9.8 4.194 4.548 4.194 4.548 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.584 4.502 mp_allgather_i34 111 8.7 2.093 4.326 2.093 4.326 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.582 3.880 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.691 3.192 3.693 dbcsr_sort_data 658 11.4 3.136 3.572 3.136 3.572 ls_scf_post 1 4.0 0.008 0.022 3.314 3.333 dbcsr_special_finalize 555 9.7 0.007 0.029 2.888 3.254 dbcsr_merge_single_wm 555 10.7 0.542 0.667 2.879 3.246 ls_scf_store_result 1 5.0 0.000 0.000 3.016 3.110 ls_scf_dm_to_ks 2 5.0 0.010 0.050 2.907 2.950 dbcsr_data_release 10477 10.7 1.590 2.447 1.590 2.447 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.211 2.221 rebuild_ks_matrix 3 7.3 0.000 0.000 2.188 2.197 qs_ks_build_kohn_sham_matrix 3 8.3 0.044 0.060 2.187 2.197 ls_scf_initial_guess 1 5.0 0.000 0.000 1.992 2.120 ls_scf_qs_atomic_guess 1 6.0 0.002 0.019 1.992 2.120 dbcsr_finalize 304 7.8 0.049 0.061 1.800 2.054 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.334000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2742.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 4.624282E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.057 99.449 99.450 qs_energies 1 2.0 0.000 0.000 98.692 98.696 ls_scf 1 3.0 0.000 0.000 96.472 96.478 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.101 78.273 ls_scf_main 1 4.0 0.000 0.000 62.040 62.045 multiply_cannon 111 7.7 0.096 0.175 55.471 60.708 density_matrix_trs4 2 5.0 0.002 0.003 55.025 55.126 multiply_cannon_loop 111 8.7 0.069 0.076 51.076 52.347 mp_waitall_1 5481 11.0 26.854 31.906 26.854 31.906 ls_scf_init_scf 1 4.0 0.000 0.000 30.795 30.818 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.518 29.566 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.288 27.300 multiply_cannon_multrec 444 9.7 13.972 16.483 21.004 22.882 make_m2s 222 7.7 0.004 0.005 17.864 20.258 make_images 222 8.7 3.728 4.433 17.802 20.198 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.985 16.652 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.598 13.954 make_images_data 222 9.7 0.003 0.004 10.070 12.424 hybrid_alltoall_any 227 10.6 0.788 3.757 9.966 12.312 dbcsr_mm_accdrv_process 3003 10.4 0.196 0.383 6.734 7.883 multiply_cannon_sync_h2d 444 9.7 6.590 7.676 6.590 7.676 dbcsr_mm_accdrv_process_sort 3003 11.4 6.408 7.502 6.408 7.502 mp_allgather_i34 111 8.7 2.628 7.002 2.628 7.002 arnoldi_extremal 4 6.8 0.000 0.000 5.803 5.817 arnoldi_normal_ev 4 7.8 0.002 0.006 5.803 5.817 build_subspace 16 8.4 0.015 0.020 5.407 5.421 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.630 4.758 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.679 4.392 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.156 4.363 mp_irecv_dv 1241 11.2 1.659 4.361 1.659 4.361 mp_sum_l 807 5.4 2.798 4.240 2.798 4.240 dbcsr_matrix_vector_mult_local 304 10.0 3.693 4.168 3.695 4.170 ls_scf_post 1 4.0 0.000 0.000 3.637 3.641 calculate_norms 792 9.8 3.557 3.636 3.557 3.636 make_images_sizes 222 9.7 0.000 0.000 1.143 3.489 mp_alltoall_i44 222 10.7 1.142 3.489 1.142 3.489 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.365 3.462 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.460 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.882 3.022 dbcsr_finalize 304 7.8 0.062 0.078 2.206 2.330 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 2.204 2.204 dbcsr_merge_all 275 8.9 0.473 0.531 2.056 2.164 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.050 2.073 rebuild_ks_matrix 3 7.3 0.000 0.000 2.017 2.040 qs_ks_build_kohn_sham_matrix 3 8.3 0.050 0.073 2.017 2.040 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.450000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3602.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/682b821dde03a8c13c17ad5faba6833238f81c82_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 8.717906E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.111 106.453 106.454 qs_energies 1 2.0 0.000 0.000 104.969 104.984 ls_scf 1 3.0 0.000 0.000 102.038 102.052 dbcsr_multiply_generic 111 6.7 0.023 0.028 76.192 76.371 ls_scf_main 1 4.0 0.000 0.000 64.684 64.684 density_matrix_trs4 2 5.0 0.002 0.003 55.962 56.018 multiply_cannon 111 7.7 0.125 0.168 48.894 51.214 multiply_cannon_loop 111 8.7 0.068 0.071 45.501 46.233 ls_scf_init_scf 1 4.0 0.000 0.000 33.737 33.738 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.356 32.368 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.540 29.547 make_m2s 222 7.7 0.005 0.005 23.586 24.475 make_images 222 8.7 4.598 5.025 23.480 24.366 mp_waitall_1 4569 11.1 21.031 24.255 21.031 24.255 multiply_cannon_multrec 444 9.7 17.845 18.636 22.433 23.101 hybrid_alltoall_any 227 10.6 1.664 3.650 12.623 15.309 make_images_data 222 9.7 0.003 0.003 12.827 15.149 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.634 10.256 multiply_cannon_sync_h2d 444 9.7 8.850 8.896 8.850 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.433 7.445 arnoldi_normal_ev 4 7.8 0.003 0.009 7.433 7.445 build_subspace 16 8.4 0.026 0.036 6.875 6.884 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.461 5.625 dbcsr_matrix_vector_mult_local 304 10.0 5.021 5.328 5.024 5.331 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.948 5.176 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.729 4.822 dbcsr_mm_accdrv_process 1814 10.4 0.257 0.326 4.418 4.541 dbcsr_mm_accdrv_process_sort 1814 11.4 4.115 4.246 4.115 4.246 ls_scf_post 1 4.0 0.000 0.000 3.617 3.631 make_images_sizes 222 9.7 0.000 0.000 1.422 3.569 mp_alltoall_i44 222 10.7 1.422 3.569 1.422 3.569 ls_scf_store_result 1 5.0 0.000 0.000 3.365 3.380 calculate_norms 792 9.8 3.238 3.275 3.238 3.275 mp_allgather_i34 111 8.7 1.020 3.173 1.020 3.173 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.146 dbcsr_merge_all 275 8.9 0.892 0.917 2.870 2.923 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.901 2.901 dbcsr_complete_redistribute 5 7.6 1.431 1.470 2.755 2.858 dbcsr_data_release 12724 10.6 2.330 2.854 2.330 2.854 dbcsr_sort_data 325 11.1 2.443 2.518 2.443 2.518 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.499 dbcsr_new_transposed 4 7.5 0.243 0.252 2.301 2.325 dbcsr_frobenius_norm 74 6.6 2.056 2.133 2.191 2.243 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.234 2.236 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.203 dbcsr_add_anytype 103 7.2 0.860 0.891 2.133 2.203 rebuild_ks_matrix 3 7.3 0.000 0.000 2.168 2.170 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.168 2.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.454000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6823.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 682b821dde03a8c13c17ad5faba6833238f81c82 Summary: empty Status: OK