=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: b01cdf04a1de84b00fa31cfcfd63373d97607d6b ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # 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 (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/01 job id: 44348010 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/02 job id: 44348012 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/03 job id: 44348013 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/04 job id: 44348014 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/05 job id: 44348015 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/06 job id: 44348016 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/07 job id: 44348017 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/08 job id: 44348018 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/09 job id: 44348019 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/10 job id: 44348020 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/11 job id: 44348021 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/12 job id: 44348022 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/13 job id: 44348023 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/14 job id: 44348024 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/15 job id: 44348025 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/16 job id: 44348026 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/17 job id: 44348027 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/18 job id: 44348029 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/19 job id: 44348031 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/20 job id: 44348032 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/21 job id: 44348033 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/22 job id: 44348034 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/23 job id: 44348035 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/24 job id: 44348036 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/25 job id: 44348037 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/26 job id: 44348038 --- 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/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 0.023 0.038 135.719 135.720 farming_run 1 2.0 135.229 135.234 135.682 135.686 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459053E+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 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 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 0.014 0.028 116.163 116.163 qs_energies 1 2.0 0.000 0.000 115.932 115.934 mp2_main 1 3.0 0.000 0.000 113.800 113.802 mp2_gpw_main 1 4.0 0.021 0.028 112.816 112.818 mp2_ri_gpw_compute_in 1 5.0 0.173 0.183 93.890 94.326 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.423 55.859 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.172 41.748 47.145 get_2c_integrals 1 6.0 0.001 0.007 37.837 38.294 integrate_v_rspace 273 8.0 0.434 0.450 25.094 30.284 pw_transfer 6555 10.6 0.380 0.392 27.398 27.848 grid_integrate_task_list 273 9.0 20.920 26.584 20.920 26.584 fft_wrap_pw1pw2 5465 11.4 0.046 0.047 26.029 26.466 fft_wrap_pw1pw2_100 2178 12.4 1.211 1.450 23.545 23.987 compute_2c_integrals 1 7.0 0.003 0.003 19.954 19.954 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.936 19.602 mp2_eri_2c_integrate_gpw 1 9.0 2.367 2.421 18.933 19.601 rpa_ri_compute_en 1 5.0 0.010 0.012 18.820 19.034 cp_fm_cholesky_decompose 12 8.2 17.944 18.462 17.944 18.462 cholesky_decomp 1 7.0 0.000 0.000 16.734 17.229 fft3d_s 5443 13.4 16.127 16.437 16.149 16.458 ao_to_mo_and_store_B_mult_1 272 7.0 10.847 15.557 10.847 15.557 calculate_wavefunction 272 8.0 5.479 5.621 12.579 13.223 rpa_num_int 1 6.0 0.002 0.028 10.648 10.657 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.587 10.601 calc_mat_Q 8 8.0 0.000 0.000 9.329 9.430 contract_S_to_Q 8 9.0 0.000 0.000 8.752 8.851 calc_potential_gpw 544 9.5 0.005 0.005 8.257 8.738 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.248 8.651 parallel_gemm_fm 14 9.1 0.000 0.000 8.336 8.445 parallel_gemm_fm_cosma 14 10.1 8.336 8.445 8.336 8.445 potential_pw2rs 545 10.0 0.107 0.109 7.703 8.347 collocate_single_gaussian 272 10.0 0.039 0.042 7.499 7.858 create_integ_mat 1 6.0 0.015 0.027 7.632 7.641 array2fm 1 7.0 0.000 0.000 6.630 7.109 pw_scatter_s 2720 13.7 4.424 4.557 4.424 4.557 pw_gather_s 2722 13.2 3.852 4.167 3.852 4.167 array2fm_buffer_send 1 8.0 2.960 3.096 2.960 3.096 pw_poisson_solve 545 10.5 1.124 1.173 2.198 2.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.818441, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 0.041 0.083 397.574 397.575 farming_run 1 2.0 396.682 396.691 397.519 397.527 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227665E+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 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.054 210.229 210.230 qs_energies 1 2.0 0.000 0.000 209.957 209.970 scf_env_do_scf 1 3.0 0.000 0.000 106.608 106.609 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.685 105.693 rebuild_ks_matrix 4 6.0 0.000 0.000 105.684 105.692 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.684 105.692 hfx_ks_matrix 4 8.0 0.001 0.001 105.282 105.286 integrate_four_center 4 9.0 0.145 0.458 105.282 105.285 mp2_main 1 3.0 0.000 0.000 103.045 103.058 mp2_gpw_main 1 4.0 0.037 0.075 102.151 102.166 integrate_four_center_main 4 10.0 0.107 0.527 96.838 99.708 integrate_four_center_bin 267 11.0 96.731 99.236 96.731 99.236 init_scf_loop 1 4.0 0.000 0.000 92.561 92.561 mp2_ri_gpw_compute_in 1 5.0 0.075 0.097 74.937 75.991 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.004 54.444 55.496 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.161 42.212 47.296 integrate_v_rspace 95 8.0 0.398 0.567 28.586 33.497 pw_transfer 2240 10.6 0.146 0.163 29.926 30.320 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.926 29.313 grid_integrate_task_list 95 9.0 23.868 28.974 23.868 28.974 mp2_ri_gpw_compute_en 1 5.0 0.067 0.118 27.049 28.704 ao_to_mo_and_store_B_mult_1 91 7.0 10.553 27.310 10.553 27.310 fft_wrap_pw1pw2_100 730 12.4 1.294 1.456 26.605 27.044 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.887 25.281 25.293 get_2c_integrals 1 6.0 0.001 0.013 20.385 20.430 compute_2c_integrals 1 7.0 0.010 0.030 19.361 19.384 compute_2c_integrals_loop_lm 1 8.0 0.006 0.016 18.909 19.203 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.866 18.903 19.202 fft3d_s 1823 13.4 18.394 18.759 18.407 18.772 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.046 14.046 calculate_wavefunction 91 8.0 2.037 2.080 9.758 10.032 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.575 8.857 9.534 potential_pw2rs 186 10.0 0.034 0.035 8.634 9.249 local_gemm 172 8.0 8.298 8.959 8.298 8.959 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.517 8.210 8.746 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.278 8.599 calc_potential_gpw 182 9.5 0.002 0.002 7.958 8.175 collocate_single_gaussian 91 10.0 0.017 0.026 7.896 8.156 mp_sendrecv_dm3 2068 8.0 6.256 6.818 6.256 6.818 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.402 6.337 6.402 mp_sync 38 10.4 3.096 5.848 3.096 5.848 pw_gather_s 912 13.2 4.863 5.318 4.863 5.318 pw_scatter_s 910 13.7 3.969 4.262 3.969 4.262 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.157222, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.063232E+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 1066576. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 51.966 51.967 qs_mol_dyn_low 1 2.0 0.003 0.003 51.713 51.720 qs_forces 11 3.9 0.006 0.008 51.651 51.652 qs_energies 11 4.9 0.028 0.034 50.146 50.160 scf_env_do_scf 11 5.9 0.000 0.001 44.027 44.028 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.934 41.934 dbcsr_multiply_generic 2286 12.5 0.092 0.096 33.063 33.470 qs_scf_new_mos 108 7.5 0.000 0.001 31.607 31.934 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.606 31.933 ot_scf_mini 108 9.5 0.002 0.002 30.002 30.216 multiply_cannon 2286 13.5 0.187 0.199 25.833 27.347 multiply_cannon_loop 2286 14.5 1.520 1.586 25.104 26.669 velocity_verlet 10 3.0 0.003 0.006 25.316 25.317 ot_mini 108 10.5 0.001 0.001 18.901 19.148 qs_ot_get_derivative 108 11.5 0.001 0.002 15.950 16.172 mp_waitall_1 245248 16.5 8.273 14.657 8.273 14.657 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 5.866 13.002 multiply_cannon_multrec 54864 15.5 4.211 6.480 7.705 11.132 rebuild_ks_matrix 119 8.3 0.000 0.000 8.054 8.195 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 8.053 8.195 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.087 7.212 multiply_cannon_sync_h2d 54864 15.5 5.867 7.127 5.867 7.127 mp_sum_l 7207 12.9 5.157 6.852 5.157 6.852 qs_ot_get_p 119 10.4 0.001 0.001 6.490 6.757 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.302 5.741 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.441 5.546 init_scf_run 11 5.9 0.000 0.001 4.823 4.823 scf_env_initial_rho_setup 11 6.9 0.002 0.002 4.823 4.823 dbcsr_mm_accdrv_process 76910 16.1 1.186 1.821 3.416 4.754 sum_up_and_integrate 119 10.3 0.013 0.015 4.616 4.625 integrate_v_rspace 119 11.3 0.002 0.002 4.603 4.614 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.272 4.375 calculate_rho_elec 119 8.7 0.011 0.017 4.271 4.375 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.774 3.840 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.892 3.086 apply_single 119 13.6 0.000 0.000 2.892 3.085 calculate_dm_sparse 119 9.5 0.000 0.001 2.859 3.016 rs_pw_transfer 974 11.9 0.011 0.013 2.812 2.924 multiply_cannon_metrocomm1 54864 15.5 0.052 0.056 1.677 2.895 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.891 2.892 jit_kernel_multiply 13 15.8 2.168 2.843 2.168 2.843 ot_diis_step 108 11.5 0.006 0.006 2.713 2.714 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.662 2.663 cp_fm_redistribute_end 50 14.0 2.421 2.638 2.426 2.640 calculate_first_density_matrix 1 7.0 0.000 0.001 2.615 2.620 cp_fm_diag_elpa_base 50 14.0 0.211 2.548 0.212 2.560 density_rs2pw 119 9.7 0.004 0.005 2.310 2.453 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.392 2.447 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.430 2.432 acc_transpose_blocks 54864 15.5 0.235 0.256 1.771 2.196 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.115 2.164 wfi_extrapolate 11 7.9 0.001 0.001 2.145 2.145 grid_integrate_task_list 119 12.3 2.022 2.116 2.022 2.116 init_scf_loop 11 6.9 0.000 0.000 2.074 2.074 pw_transfer 1439 11.6 0.054 0.060 1.926 2.006 potential_pw2rs 119 12.3 0.004 0.004 1.964 1.992 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.848 1.933 mp_sum_d 4125 12.0 1.277 1.858 1.277 1.858 make_m2s 4572 13.5 0.054 0.055 1.765 1.821 make_images 4572 14.5 0.133 0.139 1.683 1.738 fft3d_ps 1201 14.6 0.371 0.476 1.620 1.696 fft_wrap_pw1pw2_140 487 13.2 0.081 0.096 1.424 1.509 mp_alltoall_d11v 2130 13.8 1.269 1.451 1.269 1.451 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.415 1.430 mp_waitany 12084 13.8 1.262 1.424 1.262 1.424 grid_collocate_task_list 119 9.7 1.290 1.348 1.290 1.348 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.709 1.079 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.967000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 487.759872E+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 1030216. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 39.077 39.077 qs_mol_dyn_low 1 2.0 0.003 0.003 38.791 38.798 qs_forces 11 3.9 0.002 0.002 38.731 38.732 qs_energies 11 4.9 0.003 0.014 37.039 37.043 scf_env_do_scf 11 5.9 0.001 0.002 31.956 31.957 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.441 29.441 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.308 21.699 qs_scf_new_mos 108 7.5 0.001 0.001 20.042 20.294 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.041 20.293 ot_scf_mini 108 9.5 0.002 0.003 19.153 19.327 velocity_verlet 10 3.0 0.001 0.002 18.523 18.524 multiply_cannon 2286 13.5 0.210 0.220 16.277 17.818 multiply_cannon_loop 2286 14.5 0.907 0.974 15.145 16.626 ot_mini 108 10.5 0.001 0.001 11.796 12.035 mp_waitall_1 200699 16.5 5.727 11.047 5.727 11.047 qs_ot_get_derivative 108 11.5 0.001 0.003 9.328 9.506 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.154 9.430 multiply_cannon_multrec 27432 15.5 1.971 4.558 5.844 8.640 rebuild_ks_matrix 119 8.3 0.000 0.000 7.324 7.472 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.324 7.472 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.481 6.616 dbcsr_mm_accdrv_process 47894 16.0 2.956 5.104 3.803 5.626 qs_ot_get_p 119 10.4 0.001 0.001 4.443 4.667 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.591 4.443 sum_up_and_integrate 119 10.3 0.024 0.026 4.293 4.299 integrate_v_rspace 119 11.3 0.002 0.003 4.269 4.276 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.050 4.131 apply_single 119 13.6 0.000 0.000 3.049 4.131 mp_sum_l 7207 12.9 2.085 4.067 2.085 4.067 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.912 3.953 calculate_rho_elec 119 8.7 0.021 0.024 3.912 3.953 init_scf_run 11 5.9 0.000 0.001 3.858 3.859 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.858 3.859 rs_pw_transfer 974 11.9 0.010 0.011 2.720 3.182 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.957 2.977 multiply_cannon_sync_h2d 27432 15.5 2.180 2.807 2.180 2.807 make_m2s 4572 13.5 0.052 0.054 2.502 2.734 density_rs2pw 119 9.7 0.004 0.004 2.236 2.715 make_images 4572 14.5 0.199 0.235 2.414 2.645 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.521 2.522 init_scf_loop 11 6.9 0.001 0.007 2.494 2.495 ot_diis_step 108 11.5 0.011 0.011 2.415 2.415 calculate_first_density_matrix 1 7.0 0.001 0.003 2.301 2.305 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.150 2.240 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.156 2.157 cp_fm_redistribute_end 50 14.0 1.787 2.133 1.790 2.134 jit_kernel_multiply 11 16.0 0.795 2.119 0.795 2.119 calculate_dm_sparse 119 9.5 0.000 0.000 2.026 2.106 pw_transfer 1439 11.6 0.066 0.073 2.032 2.077 cp_fm_diag_elpa_base 50 14.0 0.329 2.019 0.342 2.071 potential_pw2rs 119 12.3 0.006 0.006 2.004 2.018 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.939 1.988 grid_integrate_task_list 119 12.3 1.842 1.932 1.842 1.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.876 1.878 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.671 1.713 fft3d_ps 1201 14.6 0.513 0.567 1.641 1.683 prepare_preconditioner 11 7.9 0.000 0.000 1.575 1.602 make_preconditioner 11 8.9 0.000 0.002 1.575 1.602 make_images_data 4572 15.5 0.045 0.051 1.159 1.568 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.518 1.566 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.473 1.526 wfi_extrapolate 11 7.9 0.001 0.001 1.502 1.502 acc_transpose_blocks 27432 15.5 0.108 0.112 1.184 1.482 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.011 1.461 grid_collocate_task_list 119 9.7 1.226 1.359 1.226 1.359 mp_allgather_i34 2286 14.5 0.571 1.353 0.571 1.353 mp_alltoall_d11v 2130 13.8 1.206 1.287 1.206 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.183 1.229 mp_sum_d 4125 12.0 0.607 1.069 0.607 1.069 mp_waitany 5720 13.7 0.530 1.023 0.530 1.023 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.554 1.021 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.953 0.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.902 0.917 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.653 0.859 mp_alltoall_z22v 1201 16.6 0.732 0.832 0.732 0.832 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.607 0.766 0.820 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.077000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 523.235328E+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 538098. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.020 0.047 33.058 33.060 qs_mol_dyn_low 1 2.0 0.003 0.005 32.763 32.770 qs_forces 11 3.9 0.002 0.005 32.671 32.672 qs_energies 11 4.9 0.003 0.012 31.069 31.072 scf_env_do_scf 11 5.9 0.001 0.002 26.044 26.044 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 23.426 23.431 dbcsr_multiply_generic 2286 12.5 0.092 0.095 16.692 16.811 velocity_verlet 10 3.0 0.002 0.002 15.586 15.588 qs_scf_new_mos 108 7.5 0.001 0.001 15.001 15.015 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.001 15.014 ot_scf_mini 108 9.5 0.002 0.003 14.268 14.282 multiply_cannon 2286 13.5 0.198 0.206 13.294 14.212 multiply_cannon_loop 2286 14.5 0.639 0.667 12.472 13.428 ot_mini 108 10.5 0.001 0.001 8.839 8.856 qs_ot_get_derivative 108 11.5 0.001 0.001 7.312 7.327 multiply_cannon_multrec 18288 15.5 1.968 3.004 6.920 7.278 rebuild_ks_matrix 119 8.3 0.000 0.001 6.537 6.552 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.537 6.552 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.767 5.782 dbcsr_mm_accdrv_process 38222 16.0 4.113 5.497 4.867 5.717 sum_up_and_integrate 119 10.3 0.031 0.032 4.061 4.066 integrate_v_rspace 119 11.3 0.002 0.003 4.029 4.037 mp_waitall_1 158411 16.6 2.718 3.832 2.718 3.832 init_scf_run 11 5.9 0.000 0.001 3.772 3.773 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.772 3.772 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.578 3.589 calculate_rho_elec 119 8.7 0.030 0.031 3.578 3.588 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.737 3.384 qs_ot_get_p 119 10.4 0.001 0.001 3.255 3.287 rs_pw_transfer 974 11.9 0.009 0.010 2.398 2.690 init_scf_loop 11 6.9 0.001 0.005 2.597 2.603 calculate_first_density_matrix 1 7.0 0.001 0.004 2.520 2.521 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.064 2.380 apply_single 119 13.6 0.000 0.000 2.064 2.380 density_rs2pw 119 9.7 0.004 0.004 2.080 2.359 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.493 2.312 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.144 2.151 pw_transfer 1439 11.6 0.066 0.072 2.037 2.052 jit_kernel_multiply 10 16.1 0.704 2.036 0.704 2.036 make_m2s 4572 13.5 0.044 0.046 1.844 1.989 calculate_dm_sparse 119 9.5 0.000 0.000 1.979 1.989 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.943 1.961 make_images 4572 14.5 0.189 0.201 1.759 1.902 grid_integrate_task_list 119 12.3 1.795 1.885 1.795 1.885 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.869 1.870 potential_pw2rs 119 12.3 0.007 0.008 1.818 1.826 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.781 1.782 prepare_preconditioner 11 7.9 0.000 0.000 1.767 1.769 make_preconditioner 11 8.9 0.000 0.001 1.767 1.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.620 1.703 fft3d_ps 1201 14.6 0.524 0.542 1.622 1.642 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.624 1.625 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.602 1.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.599 1.606 cp_fm_redistribute_end 50 14.0 1.210 1.602 1.211 1.602 fft_wrap_pw1pw2_140 487 13.2 0.090 0.093 1.554 1.570 cp_fm_diag_elpa_base 50 14.0 0.375 1.521 0.389 1.565 mp_sum_l 7207 12.9 1.174 1.553 1.174 1.553 multiply_cannon_sync_h2d 18288 15.5 1.375 1.508 1.375 1.508 ot_diis_step 108 11.5 0.011 0.011 1.501 1.501 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.340 1.345 grid_collocate_task_list 119 9.7 1.215 1.338 1.215 1.338 acc_transpose_blocks 18288 15.5 0.075 0.076 1.304 1.328 wfi_extrapolate 11 7.9 0.001 0.001 1.201 1.201 make_images_data 4572 15.5 0.045 0.049 0.844 1.005 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.003 1.005 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.732 0.940 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.885 0.908 acc_transpose_blocks_kernels 18288 16.5 0.208 0.218 0.864 0.875 mp_waitany 9880 13.7 0.543 0.846 0.543 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.845 mp_alltoall_d11v 2130 13.8 0.718 0.836 0.718 0.836 rs_pw_transfer_RS2PW_140 130 11.5 0.125 0.140 0.537 0.824 mp_alltoall_z22v 1201 16.6 0.702 0.819 0.702 0.819 cp_fm_cholesky_invert 11 10.9 0.753 0.757 0.753 0.757 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.733 mp_allgather_i34 2286 14.5 0.282 0.663 0.282 0.663 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.060000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.909091, yerr=1.831767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 551.878656E+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_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.088 38.520 38.521 qs_mol_dyn_low 1 2.0 0.003 0.005 38.129 38.136 qs_forces 11 3.9 0.004 0.022 38.040 38.041 qs_energies 11 4.9 0.011 0.055 36.201 36.207 scf_env_do_scf 11 5.9 0.004 0.015 30.441 30.443 scf_env_do_scf_inner_loop 108 6.5 0.003 0.015 26.558 26.560 velocity_verlet 10 3.0 0.028 0.036 19.455 19.457 dbcsr_multiply_generic 2286 12.5 0.099 0.101 19.171 19.277 qs_scf_new_mos 108 7.5 0.001 0.001 17.354 17.411 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.353 17.410 ot_scf_mini 108 9.5 0.003 0.004 16.377 16.429 multiply_cannon 2286 13.5 0.227 0.260 14.940 15.429 multiply_cannon_loop 2286 14.5 0.943 0.974 13.836 14.206 ot_mini 108 10.5 0.001 0.001 9.990 10.051 multiply_cannon_multrec 27432 15.5 2.393 3.109 8.944 9.337 qs_ot_get_derivative 108 11.5 0.001 0.001 8.114 8.169 dbcsr_mm_accdrv_process 47916 15.9 5.288 7.171 6.457 7.771 rebuild_ks_matrix 119 8.3 0.000 0.000 7.064 7.117 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.064 7.117 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.253 6.299 sum_up_and_integrate 119 10.3 0.036 0.039 4.205 4.223 integrate_v_rspace 119 11.3 0.002 0.003 4.169 4.187 init_scf_run 11 5.9 0.000 0.001 4.170 4.171 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.170 4.171 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.973 4.000 calculate_rho_elec 119 8.7 0.040 0.046 3.973 4.000 init_scf_loop 11 6.9 0.001 0.006 3.837 3.842 qs_ot_get_p 119 10.4 0.001 0.003 3.687 3.774 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.980 3.478 prepare_preconditioner 11 7.9 0.000 0.000 2.898 2.906 make_preconditioner 11 8.9 0.000 0.002 2.898 2.906 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.495 2.821 mp_waitall_1 137007 16.6 2.172 2.787 2.172 2.787 rs_pw_transfer 974 11.9 0.009 0.009 2.536 2.720 make_m2s 4572 13.5 0.054 0.056 2.536 2.676 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.177 2.657 apply_single 119 13.6 0.000 0.000 2.177 2.656 calculate_first_density_matrix 1 7.0 0.001 0.003 2.584 2.586 make_images 4572 14.5 0.269 0.329 2.428 2.566 density_rs2pw 119 9.7 0.004 0.004 2.359 2.544 pw_transfer 1439 11.6 0.067 0.073 2.340 2.392 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.246 2.302 calculate_dm_sparse 119 9.5 0.000 0.000 2.223 2.273 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.191 2.200 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.110 2.140 jit_kernel_multiply 10 15.6 1.110 2.136 1.110 2.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.999 2.000 fft3d_ps 1201 14.6 0.563 0.612 1.921 1.971 potential_pw2rs 119 12.3 0.008 0.009 1.944 1.961 grid_integrate_task_list 119 12.3 1.805 1.884 1.805 1.884 mp_sum_l 7207 12.9 1.307 1.849 1.307 1.849 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.847 1.847 fft_wrap_pw1pw2_140 487 13.2 0.087 0.094 1.774 1.833 ot_diis_step 108 11.5 0.012 0.012 1.829 1.829 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.777 1.788 acc_transpose_blocks 27432 15.5 0.112 0.115 1.505 1.538 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.531 1.532 wfi_extrapolate 11 7.9 0.001 0.001 1.529 1.529 cp_fm_redistribute_end 50 14.0 1.010 1.497 1.011 1.498 cp_fm_diag_elpa_base 50 14.0 0.465 1.426 0.484 1.464 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.382 1.391 grid_collocate_task_list 119 9.7 1.220 1.336 1.220 1.336 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.786 1.295 dbcsr_complete_redistribute 329 12.2 0.119 0.148 0.932 1.209 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.177 1.192 make_images_data 4572 15.5 0.045 0.049 1.001 1.183 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.147 1.148 cp_fm_upper_to_full 72 13.5 0.807 1.119 0.807 1.119 mp_alltoall_d11v 2130 13.8 0.946 1.093 0.946 1.093 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.859 1.050 multiply_cannon_sync_h2d 27432 15.5 0.969 1.024 0.969 1.024 cp_fm_cholesky_invert 11 10.9 0.996 1.001 0.996 1.001 mp_alltoall_z22v 1201 16.6 0.942 0.994 0.942 0.994 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.913 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.639 0.908 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.880 0.906 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.521000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.090909, yerr=2.574510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 612.962304E+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_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.018 0.051 31.382 31.383 qs_mol_dyn_low 1 2.0 0.003 0.003 31.091 31.099 qs_forces 11 3.9 0.002 0.002 31.014 31.016 qs_energies 11 4.9 0.002 0.002 29.229 29.234 scf_env_do_scf 11 5.9 0.000 0.001 24.007 24.007 scf_env_do_scf_inner_loop 108 6.5 0.003 0.014 21.003 21.004 velocity_verlet 10 3.0 0.019 0.019 16.298 16.301 dbcsr_multiply_generic 2286 12.5 0.092 0.094 13.281 13.372 qs_scf_new_mos 108 7.5 0.001 0.001 12.289 12.309 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.289 12.309 ot_scf_mini 108 9.5 0.002 0.002 11.569 11.588 multiply_cannon 2286 13.5 0.233 0.238 10.230 10.752 multiply_cannon_loop 2286 14.5 0.334 0.346 9.162 9.374 ot_mini 108 10.5 0.001 0.001 6.451 6.477 rebuild_ks_matrix 119 8.3 0.000 0.000 6.414 6.441 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.414 6.441 multiply_cannon_multrec 9144 15.5 1.715 1.960 6.059 6.279 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.726 5.750 qs_ot_get_derivative 108 11.5 0.001 0.001 5.075 5.094 dbcsr_mm_accdrv_process 12550 15.8 3.166 4.025 4.244 4.311 sum_up_and_integrate 119 10.3 0.038 0.041 3.992 3.997 integrate_v_rspace 119 11.3 0.003 0.003 3.954 3.959 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.935 3.946 calculate_rho_elec 119 8.7 0.060 0.061 3.935 3.946 init_scf_run 11 5.9 0.000 0.001 3.707 3.707 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.706 3.707 qs_ot_get_p 119 10.4 0.001 0.001 3.208 3.231 init_scf_loop 11 6.9 0.000 0.000 2.974 2.976 mp_waitall_1 115863 16.7 1.961 2.553 1.961 2.553 density_rs2pw 119 9.7 0.004 0.004 2.222 2.359 make_m2s 4572 13.5 0.034 0.036 2.092 2.350 pw_transfer 1439 11.6 0.067 0.070 2.326 2.338 calculate_first_density_matrix 1 7.0 0.001 0.001 2.314 2.314 rs_pw_transfer 974 11.9 0.008 0.008 2.148 2.301 make_images 4572 14.5 0.267 0.297 2.003 2.258 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.231 2.244 prepare_preconditioner 11 7.9 0.000 0.000 2.176 2.180 make_preconditioner 11 8.9 0.000 0.000 2.176 2.180 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.093 2.096 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.056 2.085 grid_integrate_task_list 119 12.3 1.860 1.969 1.860 1.969 jit_kernel_multiply 10 15.8 1.041 1.936 1.041 1.936 calculate_dm_sparse 119 9.5 0.000 0.000 1.875 1.900 fft3d_ps 1201 14.6 0.564 0.575 1.891 1.897 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.873 1.873 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.799 1.809 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.769 1.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.742 1.744 potential_pw2rs 119 12.3 0.010 0.011 1.702 1.708 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.532 1.533 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.505 1.522 cp_fm_redistribute_end 50 14.0 0.763 1.500 0.765 1.501 cp_fm_diag_elpa_base 50 14.0 0.690 1.418 0.734 1.476 grid_collocate_task_list 119 9.7 1.281 1.397 1.281 1.397 ot_diis_step 108 11.5 0.013 0.013 1.355 1.356 wfi_extrapolate 11 7.9 0.001 0.001 1.333 1.333 make_images_data 4572 15.5 0.039 0.042 0.988 1.313 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.300 1.311 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.936 1.298 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.265 1.271 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.251 1.253 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.218 1.239 apply_single 119 13.6 0.000 0.000 1.218 1.238 cp_fm_cholesky_invert 11 10.9 1.190 1.193 1.190 1.193 mp_alltoall_d11v 2130 13.8 0.919 1.030 0.919 1.030 acc_transpose_blocks 9144 15.5 0.038 0.039 1.003 1.008 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.865 0.916 mp_allgather_i34 2286 14.5 0.398 0.913 0.398 0.913 mp_alltoall_z22v 1201 16.6 0.776 0.825 0.776 0.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.815 0.817 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.808 0.816 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.761 0.763 multiply_cannon_sync_h2d 9144 15.5 0.710 0.763 0.710 0.763 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.708 0.760 dbcsr_complete_redistribute 329 12.2 0.165 0.173 0.667 0.706 mp_sum_l 7207 12.9 0.556 0.697 0.556 0.697 yz_to_x 606 15.1 0.264 0.275 0.666 0.682 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.341 0.674 x_to_yz 595 16.2 0.275 0.287 0.649 0.656 jit_kernel_transpose 5 15.6 0.644 0.647 0.644 0.647 mp_waitany 5200 13.7 0.495 0.634 0.495 0.634 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.383000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=581.454545, yerr=4.793246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 752.832512E+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_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.049 43.816 43.816 qs_mol_dyn_low 1 2.0 0.003 0.003 43.493 43.502 qs_forces 11 3.9 0.002 0.002 43.424 43.425 qs_energies 11 4.9 0.001 0.001 41.379 41.383 scf_env_do_scf 11 5.9 0.001 0.001 35.447 35.447 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.138 27.140 velocity_verlet 10 3.0 0.002 0.006 24.763 24.769 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.067 18.326 qs_scf_new_mos 108 7.5 0.001 0.001 16.860 16.950 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.859 16.949 ot_scf_mini 108 9.5 0.002 0.002 15.739 15.836 multiply_cannon 2286 13.5 0.308 0.321 13.807 14.821 multiply_cannon_loop 2286 14.5 0.347 0.351 12.483 13.462 ot_mini 108 10.5 0.001 0.001 9.198 9.318 multiply_cannon_multrec 9144 15.5 3.339 4.835 8.635 8.784 init_scf_loop 11 6.9 0.000 0.000 8.273 8.277 rebuild_ks_matrix 119 8.3 0.000 0.001 7.644 7.791 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.644 7.791 prepare_preconditioner 11 7.9 0.000 0.000 7.262 7.277 make_preconditioner 11 8.9 0.000 0.000 7.262 7.277 qs_ot_get_derivative 108 11.5 0.001 0.001 7.153 7.251 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.830 7.148 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.891 7.025 dbcsr_mm_accdrv_process 12550 15.8 4.246 5.705 5.173 6.496 cp_fm_upper_to_full 72 14.2 3.151 4.519 3.151 4.519 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.493 4.501 calculate_rho_elec 119 8.7 0.118 0.121 4.493 4.500 sum_up_and_integrate 119 10.3 0.066 0.067 4.264 4.270 integrate_v_rspace 119 11.3 0.003 0.003 4.198 4.206 qs_ot_get_p 119 10.4 0.001 0.001 3.729 3.875 init_scf_run 11 5.9 0.000 0.001 3.856 3.856 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.856 3.856 mp_waitall_1 94719 16.7 2.670 3.706 2.670 3.706 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.638 3.044 pw_transfer 1439 11.6 0.069 0.069 2.956 2.961 dbcsr_complete_redistribute 329 12.2 0.283 0.287 2.092 2.906 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.857 2.865 make_m2s 4572 13.5 0.038 0.039 2.513 2.699 make_images 4572 14.5 0.348 0.377 2.392 2.579 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.710 2.520 fft3d_ps 1201 14.6 0.595 0.602 2.480 2.490 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.154 2.412 apply_single 119 13.6 0.000 0.000 2.154 2.412 density_rs2pw 119 9.7 0.004 0.004 2.386 2.408 fft_wrap_pw1pw2_140 487 13.2 0.094 0.097 2.371 2.378 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.298 2.300 calculate_dm_sparse 119 9.5 0.000 0.000 2.270 2.293 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.378 2.292 mp_alltoall_i22 627 13.8 1.431 2.260 1.431 2.260 calculate_first_density_matrix 1 7.0 0.001 0.001 2.256 2.257 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.427 2.226 grid_integrate_task_list 119 12.3 2.083 2.112 2.083 2.112 ot_diis_step 108 11.5 0.014 0.014 2.013 2.013 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.010 2.011 mp_sum_l 7207 12.9 1.227 1.980 1.227 1.980 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.974 1.974 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.797 1.854 cp_fm_cholesky_invert 11 10.9 1.802 1.806 1.802 1.806 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.780 1.781 rs_pw_transfer 974 11.9 0.009 0.010 1.672 1.697 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.644 1.670 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.659 1.660 cp_fm_diag_elpa_base 50 14.0 1.513 1.566 1.657 1.657 potential_pw2rs 119 12.3 0.014 0.015 1.607 1.610 jit_kernel_multiply 8 15.2 0.901 1.554 0.901 1.554 grid_collocate_task_list 119 9.7 1.508 1.541 1.508 1.541 wfi_extrapolate 11 7.9 0.001 0.001 1.533 1.533 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.257 1.508 make_images_data 4572 15.5 0.043 0.045 1.212 1.454 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.305 1.316 mp_alltoall_d11v 2130 13.8 1.228 1.267 1.228 1.267 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.188 1.215 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.116 1.135 multiply_cannon_sync_h2d 9144 15.5 1.044 1.048 1.044 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.978 1.014 yz_to_x 606 15.1 0.460 0.469 0.993 1.003 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.969 0.983 mp_alltoall_z22v 1201 16.6 0.933 0.968 0.933 0.968 qs_create_task_list 11 7.9 0.001 0.002 0.950 0.963 generate_qs_task_list 11 8.9 0.372 0.392 0.949 0.963 acc_transpose_blocks 9144 15.5 0.038 0.038 0.915 0.921 x_to_yz 595 16.2 0.480 0.493 0.879 0.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.816000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=706.545455, yerr=16.283643 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 503.955456E+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 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 85.080 85.083 qs_mol_dyn_low 1 2.0 0.003 0.004 84.762 84.772 qs_forces 11 3.9 0.002 0.002 84.679 84.683 qs_energies 11 4.9 0.002 0.006 81.728 81.747 scf_env_do_scf 11 5.9 0.000 0.001 72.482 72.487 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.542 66.545 dbcsr_multiply_generic 2055 12.4 0.105 0.109 52.245 52.545 qs_scf_new_mos 99 7.5 0.000 0.001 48.423 48.541 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.423 48.540 ot_scf_mini 99 9.5 0.002 0.002 45.976 46.055 multiply_cannon 2055 13.4 0.182 0.189 42.588 43.457 velocity_verlet 10 3.0 0.002 0.002 42.745 42.747 multiply_cannon_loop 2055 14.4 1.541 1.585 41.504 42.197 ot_mini 99 10.5 0.001 0.001 27.590 27.672 qs_ot_get_derivative 99 11.5 0.001 0.001 20.715 20.790 multiply_cannon_multrec 49320 15.4 12.278 13.063 17.470 18.403 rebuild_ks_matrix 110 8.3 0.000 0.001 15.012 15.134 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 15.011 15.133 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.158 13.265 mp_waitall_1 220248 16.4 11.244 11.914 11.244 11.914 multiply_cannon_sync_h2d 49320 15.4 10.096 10.654 10.096 10.654 qs_ot_get_p 110 10.4 0.001 0.001 9.851 9.966 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.201 7.712 apply_single 110 13.6 0.000 0.001 7.201 7.712 sum_up_and_integrate 110 10.3 0.037 0.043 7.450 7.465 integrate_v_rspace 110 11.3 0.003 0.003 7.412 7.437 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.861 7.386 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 6.552 7.381 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.003 7.145 calculate_rho_elec 110 8.6 0.021 0.026 7.003 7.144 init_scf_run 11 5.9 0.000 0.001 7.102 7.102 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.101 7.102 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.871 6.920 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.571 6.606 ot_diis_step 99 11.5 0.006 0.006 6.591 6.592 mp_sum_l 6514 12.8 5.004 5.920 5.004 5.920 init_scf_loop 11 6.9 0.000 0.002 5.901 5.903 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.535 5.535 dbcsr_mm_accdrv_process 87628 16.1 1.956 2.067 5.070 5.381 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.957 4.959 cp_fm_redistribute_end 48 14.0 4.319 4.921 4.323 4.923 cp_fm_diag_elpa_base 48 14.0 0.592 4.759 0.596 4.787 rs_pw_transfer 902 11.9 0.012 0.013 4.073 4.619 density_rs2pw 110 9.6 0.004 0.005 3.828 4.360 make_m2s 4110 13.4 0.061 0.066 4.207 4.291 make_images 4110 14.4 0.178 0.192 4.111 4.200 wfi_extrapolate 11 7.9 0.001 0.001 4.154 4.154 calculate_dm_sparse 110 9.5 0.001 0.001 3.959 4.049 pw_transfer 1331 11.6 0.056 0.066 3.658 3.733 multiply_cannon_metrocomm1 49320 15.4 0.061 0.063 2.634 3.729 prepare_preconditioner 11 7.9 0.000 0.000 3.687 3.703 make_preconditioner 11 8.9 0.000 0.000 3.687 3.703 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.569 3.645 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.637 3.641 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.463 3.505 grid_integrate_task_list 110 12.3 3.261 3.398 3.261 3.398 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.274 3.342 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.284 3.337 fft3d_ps 1111 14.6 0.786 0.902 3.100 3.160 potential_pw2rs 110 12.3 0.006 0.007 2.981 3.011 fft_wrap_pw1pw2_140 451 13.1 0.169 0.192 2.929 3.009 jit_kernel_multiply 13 15.9 2.838 2.883 2.838 2.883 calculate_first_density_matrix 1 7.0 0.001 0.002 2.857 2.859 mp_alltoall_d11v 2046 13.8 2.092 2.527 2.092 2.527 mp_waitany 14300 13.8 1.807 2.399 1.807 2.399 grid_collocate_task_list 110 9.6 2.083 2.317 2.083 2.317 acc_transpose_blocks 49320 15.4 0.216 0.230 2.127 2.217 mp_sum_d 3879 11.9 1.508 2.082 1.508 2.082 cp_fm_cholesky_invert 11 10.9 2.054 2.058 2.054 2.058 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.997 2.015 make_images_data 4110 15.4 0.042 0.046 1.799 1.916 mp_alltoall_z22v 1111 16.6 1.595 1.875 1.595 1.875 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.802 1.829 hybrid_alltoall_any 4261 16.3 0.081 0.481 1.557 1.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.083000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=3.077511 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 586.866688E+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 3431240. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.048 72.749 72.751 qs_mol_dyn_low 1 2.0 0.003 0.004 72.378 72.388 qs_forces 11 3.9 0.002 0.004 72.305 72.306 qs_energies 11 4.9 0.002 0.007 68.893 68.898 scf_env_do_scf 11 5.9 0.001 0.001 60.152 60.155 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 51.628 51.630 dbcsr_multiply_generic 2055 12.4 0.114 0.120 38.488 38.748 velocity_verlet 10 3.0 0.001 0.002 38.348 38.350 qs_scf_new_mos 99 7.5 0.001 0.001 34.470 34.628 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.469 34.628 ot_scf_mini 99 9.5 0.003 0.004 32.775 32.911 multiply_cannon 2055 13.4 0.223 0.243 31.292 32.570 multiply_cannon_loop 2055 14.4 0.934 0.962 29.820 30.628 ot_mini 99 10.5 0.001 0.001 18.937 19.073 multiply_cannon_multrec 24660 15.4 7.580 9.408 13.736 15.808 rebuild_ks_matrix 110 8.3 0.000 0.001 14.278 14.403 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.277 14.403 qs_ot_get_derivative 99 11.5 0.001 0.002 13.055 13.191 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.588 12.693 mp_waitall_1 176588 16.5 8.125 11.251 8.125 11.251 init_scf_loop 11 6.9 0.001 0.006 8.480 8.485 multiply_cannon_metrocomm3 24660 15.4 0.070 0.071 5.370 8.257 multiply_cannon_sync_h2d 24660 15.4 7.052 8.190 7.052 8.190 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.581 7.246 apply_single 110 13.6 0.000 0.001 6.581 7.246 qs_ot_get_p 110 10.4 0.001 0.002 7.079 7.221 sum_up_and_integrate 110 10.3 0.052 0.060 7.022 7.032 integrate_v_rspace 110 11.3 0.002 0.002 6.970 6.983 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.624 6.636 calculate_rho_elec 110 8.6 0.040 0.047 6.623 6.635 prepare_preconditioner 11 7.9 0.000 0.000 6.370 6.389 make_preconditioner 11 8.9 0.000 0.001 6.370 6.389 dbcsr_mm_accdrv_process 52282 16.1 4.684 5.428 6.000 6.283 init_scf_run 11 5.9 0.000 0.001 6.280 6.281 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.280 6.280 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.920 6.082 ot_diis_step 99 11.5 0.010 0.010 5.827 5.828 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.867 5.594 make_m2s 4110 13.4 0.057 0.061 4.518 5.027 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.903 4.926 make_images 4110 14.4 0.395 0.437 4.408 4.914 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.424 4.424 density_rs2pw 110 9.6 0.004 0.004 3.745 4.255 pw_transfer 1331 11.6 0.067 0.074 4.054 4.201 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.946 4.097 rs_pw_transfer 902 11.9 0.012 0.013 3.473 4.019 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.744 3.746 cp_fm_redistribute_end 48 14.0 2.785 3.693 2.787 3.694 wfi_extrapolate 11 7.9 0.001 0.001 3.684 3.684 cp_fm_diag_elpa_base 48 14.0 0.868 3.520 0.901 3.620 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.418 3.488 fft3d_ps 1111 14.6 1.104 1.324 3.296 3.430 grid_integrate_task_list 110 12.3 3.150 3.336 3.150 3.336 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.302 3.304 fft_wrap_pw1pw2_140 451 13.1 0.200 0.217 3.144 3.301 cp_fm_cholesky_invert 11 10.9 3.097 3.106 3.097 3.106 calculate_dm_sparse 110 9.5 0.001 0.001 3.002 3.034 make_images_data 4110 15.4 0.047 0.050 2.437 2.956 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.907 2.955 potential_pw2rs 110 12.3 0.008 0.009 2.830 2.852 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.102 2.839 mp_sum_l 6514 12.8 2.059 2.809 2.059 2.809 calculate_first_density_matrix 1 7.0 0.001 0.004 2.493 2.495 grid_collocate_task_list 110 9.6 2.049 2.487 2.049 2.487 mp_alltoall_d11v 2046 13.8 1.854 2.169 1.854 2.169 jit_kernel_multiply 11 16.2 0.966 2.115 0.966 2.115 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.042 2.058 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.894 1.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.858 1.875 mp_allgather_i34 2055 14.4 0.799 1.837 0.799 1.837 mp_waitany 10164 13.8 1.267 1.769 1.267 1.769 cp_fm_cholesky_decompose 22 10.9 1.619 1.625 1.619 1.625 dbcsr_complete_redistribute 325 12.2 0.240 0.307 1.336 1.605 mp_alltoall_z22v 1111 16.6 1.472 1.579 1.472 1.579 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.561 acc_transpose_blocks 24660 15.4 0.109 0.111 1.515 1.544 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.780 1.536 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.982 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.472 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.751000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.545455, yerr=7.088578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 659.218432E+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_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.014 0.049 62.975 62.976 qs_mol_dyn_low 1 2.0 0.003 0.004 62.684 62.693 qs_forces 11 3.9 0.002 0.002 62.615 62.616 qs_energies 11 4.9 0.001 0.002 59.321 59.326 scf_env_do_scf 11 5.9 0.000 0.001 51.259 51.259 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.926 41.928 velocity_verlet 10 3.0 0.002 0.002 34.248 34.250 dbcsr_multiply_generic 2055 12.4 0.108 0.111 29.070 29.303 qs_scf_new_mos 99 7.5 0.001 0.001 26.387 26.477 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.386 26.476 ot_scf_mini 99 9.5 0.002 0.003 25.138 25.245 multiply_cannon 2055 13.4 0.212 0.222 22.355 23.632 multiply_cannon_loop 2055 14.4 0.622 0.639 21.015 22.093 ot_mini 99 10.5 0.001 0.001 14.130 14.238 rebuild_ks_matrix 110 8.3 0.000 0.001 12.709 12.814 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.023 12.709 12.814 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.183 11.275 multiply_cannon_multrec 16440 15.4 4.000 5.104 9.748 10.876 mp_waitall_1 139946 16.5 7.196 10.429 7.196 10.429 qs_ot_get_derivative 99 11.5 0.001 0.001 9.610 9.714 init_scf_loop 11 6.9 0.000 0.000 9.289 9.291 prepare_preconditioner 11 7.9 0.000 0.000 7.472 7.486 make_preconditioner 11 8.9 0.000 0.000 7.472 7.486 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.336 7.352 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.775 7.139 sum_up_and_integrate 110 10.3 0.059 0.060 6.830 6.847 integrate_v_rspace 110 11.3 0.002 0.003 6.770 6.788 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.236 6.244 calculate_rho_elec 110 8.6 0.058 0.059 6.236 6.243 qs_ot_get_p 110 10.4 0.001 0.001 6.077 6.213 dbcsr_mm_accdrv_process 34862 16.1 4.538 5.231 5.605 5.792 init_scf_run 11 5.9 0.000 0.001 5.624 5.625 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.624 5.624 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.073 5.545 apply_single 110 13.6 0.000 0.000 5.073 5.545 make_m2s 4110 13.4 0.050 0.052 4.335 4.774 make_images 4110 14.4 0.388 0.503 4.220 4.659 density_rs2pw 110 9.6 0.004 0.005 3.328 4.530 ot_diis_step 99 11.5 0.011 0.011 4.482 4.482 multiply_cannon_sync_h2d 16440 15.4 3.694 4.336 3.694 4.336 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.290 4.296 rs_pw_transfer 902 11.9 0.010 0.011 3.042 4.239 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.275 3.959 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.898 3.899 pw_transfer 1331 11.6 0.066 0.075 3.781 3.793 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.673 3.688 grid_integrate_task_list 110 12.3 3.174 3.389 3.174 3.389 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.273 3.275 cp_fm_redistribute_end 48 14.0 2.037 3.231 2.040 3.232 cp_fm_diag_elpa_base 48 14.0 1.126 3.064 1.186 3.170 fft_wrap_pw1pw2_140 451 13.1 0.212 0.217 3.095 3.109 cp_fm_cholesky_invert 11 10.9 3.094 3.103 3.094 3.103 wfi_extrapolate 11 7.9 0.001 0.001 3.100 3.100 fft3d_ps 1111 14.6 1.089 1.100 2.973 2.991 make_images_data 4110 15.4 0.043 0.048 2.431 2.949 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.142 2.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.820 2.821 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.519 2.577 potential_pw2rs 110 12.3 0.011 0.011 2.543 2.551 calculate_dm_sparse 110 9.5 0.001 0.001 2.519 2.546 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.459 2.497 mp_waitany 17072 13.8 1.253 2.495 1.253 2.495 grid_collocate_task_list 110 9.6 2.079 2.447 2.079 2.447 mp_sum_l 6514 12.8 1.787 2.439 1.787 2.439 calculate_first_density_matrix 1 7.0 0.001 0.001 2.432 2.434 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.848 2.421 mp_irecv_dv 48980 15.7 0.779 2.293 0.779 2.293 mp_alltoall_d11v 2046 13.8 1.924 2.277 1.924 2.277 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.179 0.899 2.083 dbcsr_complete_redistribute 325 12.2 0.327 0.371 1.600 2.071 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.956 1.956 jit_kernel_multiply 10 16.6 0.679 1.897 0.679 1.897 cp_fm_upper_to_full 70 13.6 1.414 1.858 1.414 1.858 cp_fm_cholesky_decompose 22 10.9 1.798 1.818 1.798 1.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.731 1.746 mp_allgather_i34 2055 14.4 0.665 1.693 0.665 1.693 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.050 1.514 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.470 1.485 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.366 1.480 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.382 1.392 rs_gather_matrices 110 12.3 0.232 0.259 0.970 1.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.976000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.545455, yerr=7.499862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 724.983808E+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_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.053 0.102 69.425 69.426 qs_mol_dyn_low 1 2.0 0.003 0.004 68.985 68.996 qs_forces 11 3.9 0.003 0.014 68.912 68.913 qs_energies 11 4.9 0.008 0.055 65.324 65.329 scf_env_do_scf 11 5.9 0.001 0.002 56.479 56.482 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.867 43.871 velocity_verlet 10 3.0 0.001 0.002 38.929 38.932 dbcsr_multiply_generic 2055 12.4 0.115 0.117 30.659 30.895 qs_scf_new_mos 99 7.5 0.001 0.001 28.229 28.329 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.229 28.328 ot_scf_mini 99 9.5 0.003 0.004 26.571 26.681 multiply_cannon 2055 13.4 0.242 0.258 23.122 24.222 multiply_cannon_loop 2055 14.4 0.891 0.905 21.575 22.094 ot_mini 99 10.5 0.001 0.001 15.047 15.160 multiply_cannon_multrec 24660 15.4 4.241 6.961 12.989 14.266 rebuild_ks_matrix 110 8.3 0.000 0.000 12.646 12.761 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.646 12.760 init_scf_loop 11 6.9 0.001 0.009 12.557 12.561 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.101 11.204 qs_ot_get_derivative 99 11.5 0.001 0.001 10.812 10.918 prepare_preconditioner 11 7.9 0.000 0.000 10.720 10.736 make_preconditioner 11 8.9 0.000 0.001 10.720 10.736 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.971 10.412 dbcsr_mm_accdrv_process 52304 16.0 6.991 8.371 8.601 9.503 sum_up_and_integrate 110 10.3 0.068 0.071 6.884 6.911 integrate_v_rspace 110 11.3 0.003 0.003 6.816 6.841 mp_waitall_1 121746 16.5 4.702 6.804 4.702 6.804 qs_ot_get_p 110 10.4 0.001 0.004 6.484 6.636 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.444 6.454 calculate_rho_elec 110 8.6 0.077 0.081 6.443 6.454 make_m2s 4110 13.4 0.060 0.061 5.714 6.071 init_scf_run 11 5.9 0.000 0.001 5.930 5.930 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.930 5.930 make_images 4110 14.4 0.571 0.690 5.573 5.926 cp_fm_upper_to_full 70 13.8 3.264 4.640 3.264 4.640 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.474 4.489 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.099 4.259 apply_single 110 13.6 0.000 0.000 4.098 4.258 ot_diis_step 99 11.5 0.011 0.011 4.188 4.189 pw_transfer 1331 11.6 0.066 0.075 4.064 4.105 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.958 4.002 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.964 3.965 dbcsr_complete_redistribute 325 12.2 0.417 0.459 2.768 3.900 density_rs2pw 110 9.6 0.004 0.004 3.353 3.865 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.731 3.803 grid_integrate_task_list 110 12.3 3.261 3.475 3.261 3.475 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.588 3.444 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.301 3.352 rs_pw_transfer 902 11.9 0.010 0.011 2.793 3.326 multiply_cannon_sync_h2d 24660 15.4 3.171 3.324 3.171 3.324 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.291 3.293 fft3d_ps 1111 14.6 1.091 1.127 3.251 3.287 cp_fm_cholesky_invert 11 10.9 3.266 3.275 3.266 3.275 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.219 3.274 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.145 3.261 make_images_data 4110 15.4 0.046 0.051 2.841 3.256 cp_fm_redistribute_end 48 14.0 1.641 3.240 1.643 3.242 cp_fm_diag_elpa_base 48 14.0 1.514 3.088 1.595 3.200 wfi_extrapolate 11 7.9 0.001 0.001 3.136 3.136 calculate_dm_sparse 110 9.5 0.001 0.001 3.098 3.119 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.409 3.119 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.953 2.954 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.741 2.838 mp_alltoall_i22 605 13.7 1.655 2.831 1.655 2.831 calculate_first_density_matrix 1 7.0 0.001 0.004 2.698 2.700 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.557 2.589 potential_pw2rs 110 12.3 0.012 0.013 2.530 2.553 grid_collocate_task_list 110 9.6 2.180 2.468 2.180 2.468 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.279 2.279 mp_alltoall_d11v 2046 13.8 1.870 2.152 1.870 2.152 jit_kernel_multiply 10 15.5 1.282 2.147 1.282 2.147 cp_fm_cholesky_decompose 22 10.9 1.964 2.003 1.964 2.003 acc_transpose_blocks 24660 15.4 0.105 0.106 1.612 1.812 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.797 1.810 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.773 1.809 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.685 1.699 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.591 1.694 mp_sum_l 6514 12.8 1.186 1.674 1.186 1.674 mp_allgather_i34 2055 14.4 0.647 1.667 0.647 1.667 mp_waitany 13376 13.8 1.128 1.600 1.128 1.600 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.829 1.547 mp_irecv_dv 62702 16.1 0.731 1.473 0.731 1.473 mp_alltoall_z22v 1111 16.6 1.297 1.413 1.297 1.413 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.426000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.272727, yerr=6.607034 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 822.652928E+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_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.016 0.043 58.572 58.572 qs_mol_dyn_low 1 2.0 0.003 0.003 58.310 58.320 qs_forces 11 3.9 0.002 0.002 58.239 58.240 qs_energies 11 4.9 0.004 0.004 54.492 54.497 scf_env_do_scf 11 5.9 0.000 0.001 46.063 46.063 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 37.529 37.529 velocity_verlet 10 3.0 0.002 0.002 32.816 32.820 dbcsr_multiply_generic 2055 12.4 0.105 0.106 23.278 23.401 qs_scf_new_mos 99 7.5 0.001 0.001 22.009 22.050 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.008 22.049 ot_scf_mini 99 9.5 0.002 0.002 20.736 20.767 multiply_cannon 2055 13.4 0.246 0.269 17.496 18.870 multiply_cannon_loop 2055 14.4 0.323 0.337 16.025 16.333 rebuild_ks_matrix 110 8.3 0.000 0.000 12.146 12.176 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.145 12.175 ot_mini 99 10.5 0.001 0.001 10.848 10.870 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.789 10.814 init_scf_loop 11 6.9 0.000 0.000 8.480 8.480 multiply_cannon_multrec 8220 15.4 3.228 4.595 7.539 8.473 mp_waitall_1 103326 16.6 6.123 7.693 6.123 7.693 qs_ot_get_derivative 99 11.5 0.001 0.001 6.994 7.022 prepare_preconditioner 11 7.9 0.000 0.000 6.783 6.786 make_preconditioner 11 8.9 0.000 0.000 6.783 6.786 sum_up_and_integrate 110 10.3 0.079 0.081 6.691 6.703 integrate_v_rspace 110 11.3 0.003 0.003 6.612 6.625 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.612 6.623 calculate_rho_elec 110 8.6 0.116 0.116 6.611 6.622 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.353 6.433 qs_ot_get_p 110 10.4 0.001 0.001 5.900 5.921 init_scf_run 11 5.9 0.000 0.001 5.215 5.215 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.215 5.215 dbcsr_mm_accdrv_process 17442 15.9 2.772 3.654 4.181 5.132 make_m2s 4110 13.4 0.039 0.040 4.354 4.600 make_images 4110 14.4 0.634 0.686 4.224 4.469 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.377 4.382 pw_transfer 1331 11.6 0.067 0.073 4.268 4.283 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.159 4.179 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.906 4.149 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.046 4.047 ot_diis_step 99 11.5 0.012 0.012 3.827 3.828 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.726 3.754 apply_single 110 13.6 0.000 0.000 3.726 3.754 density_rs2pw 110 9.6 0.004 0.005 3.351 3.733 fft_wrap_pw1pw2_140 451 13.1 0.214 0.218 3.539 3.563 grid_integrate_task_list 110 12.3 3.350 3.538 3.350 3.538 cp_fm_cholesky_invert 11 10.9 3.489 3.493 3.489 3.493 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.456 3.457 fft3d_ps 1111 14.6 1.149 1.178 3.404 3.417 cp_fm_redistribute_end 48 14.0 0.884 3.412 0.889 3.414 cp_fm_diag_elpa_base 48 14.0 2.341 3.210 2.515 3.361 multiply_cannon_sync_h2d 8220 15.4 2.918 3.054 2.918 3.054 wfi_extrapolate 11 7.9 0.001 0.001 2.891 2.891 make_images_data 4110 15.4 0.038 0.043 2.425 2.820 rs_pw_transfer 902 11.9 0.010 0.010 2.415 2.816 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.339 2.757 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 2.702 2.703 grid_collocate_task_list 110 9.6 2.280 2.543 2.280 2.543 calculate_dm_sparse 110 9.5 0.001 0.001 2.507 2.542 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.528 2.529 potential_pw2rs 110 12.3 0.015 0.015 2.246 2.252 calculate_first_density_matrix 1 7.0 0.001 0.001 2.221 2.222 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.164 2.180 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.143 2.151 mp_alltoall_d11v 2046 13.8 1.740 2.044 1.740 2.044 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.765 1.984 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.933 1.946 cp_fm_cholesky_decompose 22 10.9 1.834 1.850 1.834 1.850 mp_allgather_i34 2055 14.4 0.656 1.730 0.656 1.730 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.695 1.700 jit_kernel_multiply 9 15.9 1.102 1.700 1.102 1.700 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.540 1.665 dbcsr_complete_redistribute 325 12.2 0.551 0.583 1.539 1.642 mp_waitany 9240 13.8 1.102 1.534 1.102 1.534 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.462 1.473 qs_create_task_list 11 7.9 0.001 0.001 1.245 1.348 generate_qs_task_list 11 8.9 0.380 0.448 1.244 1.347 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.879 1.244 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.216 1.236 rs_gather_matrices 110 12.3 0.322 0.364 0.929 1.222 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.173 0.809 1.210 mp_alltoall_z22v 1111 16.6 1.146 1.184 1.146 1.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.572000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=775.090909, yerr=9.746370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 1.377825E+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_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 91.542 91.543 qs_mol_dyn_low 1 2.0 0.003 0.003 91.175 91.186 qs_forces 11 3.9 0.002 0.002 91.099 91.100 qs_energies 11 4.9 0.001 0.002 86.784 86.788 scf_env_do_scf 11 5.9 0.001 0.001 76.391 76.391 velocity_verlet 10 3.0 0.002 0.002 57.907 57.913 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 46.771 46.772 init_scf_loop 11 6.9 0.000 0.000 29.532 29.534 dbcsr_multiply_generic 2055 12.4 0.119 0.120 29.158 29.293 qs_scf_new_mos 99 7.5 0.001 0.001 27.863 27.904 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.862 27.903 prepare_preconditioner 11 7.9 0.000 0.000 27.464 27.473 make_preconditioner 11 8.9 0.000 0.000 27.464 27.473 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.750 26.933 ot_scf_mini 99 9.5 0.002 0.002 26.071 26.107 multiply_cannon 2055 13.4 0.350 0.383 21.726 22.557 multiply_cannon_loop 2055 14.4 0.343 0.344 19.850 20.187 cp_fm_upper_to_full 70 14.2 12.684 18.108 12.684 18.108 rebuild_ks_matrix 110 8.3 0.001 0.001 14.533 14.578 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.533 14.578 ot_mini 99 10.5 0.001 0.001 13.979 14.023 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.143 13.182 dbcsr_complete_redistribute 325 12.2 1.023 1.045 7.452 10.542 multiply_cannon_multrec 8220 15.4 4.371 4.559 9.747 9.872 qs_ot_get_derivative 99 11.5 0.001 0.001 9.404 9.443 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.312 9.406 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.699 8.770 mp_waitall_1 84994 16.7 7.750 8.732 7.750 8.732 mp_alltoall_i22 605 13.7 5.337 8.407 5.337 8.407 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.069 8.105 calculate_rho_elec 110 8.6 0.227 0.227 8.068 8.105 sum_up_and_integrate 110 10.3 0.150 0.151 7.661 7.675 integrate_v_rspace 110 11.3 0.003 0.003 7.510 7.526 qs_ot_get_p 110 10.4 0.001 0.001 6.835 6.903 cp_fm_cholesky_invert 11 10.9 6.232 6.238 6.232 6.238 init_scf_run 11 5.9 0.000 0.001 6.090 6.090 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.090 6.090 make_m2s 4110 13.4 0.043 0.044 5.518 6.047 make_images 4110 14.4 0.876 0.923 5.329 5.857 pw_transfer 1331 11.6 0.076 0.077 5.541 5.546 dbcsr_mm_accdrv_process 11614 15.7 3.320 3.648 5.234 5.495 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.423 5.428 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.645 5.144 apply_single 110 13.6 0.000 0.000 4.645 5.144 qs_ot_p2m_diag 48 11.0 0.151 0.155 5.087 5.095 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.705 4.935 fft_wrap_pw1pw2_140 451 13.1 0.221 0.224 4.650 4.663 fft3d_ps 1111 14.6 1.296 1.315 4.617 4.623 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.601 4.602 ot_diis_step 99 11.5 0.015 0.016 4.547 4.547 density_rs2pw 110 9.6 0.004 0.004 4.059 4.083 multiply_cannon_sync_h2d 8220 15.4 3.949 3.954 3.949 3.954 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.895 3.896 cp_fm_diag_elpa_base 48 14.0 3.358 3.548 3.892 3.893 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.959 3.746 make_images_data 4110 15.4 0.042 0.044 3.029 3.720 grid_integrate_task_list 110 12.3 3.674 3.717 3.674 3.717 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.687 3.688 wfi_extrapolate 11 7.9 0.001 0.001 3.437 3.437 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.963 3.409 calculate_dm_sparse 110 9.5 0.001 0.001 3.297 3.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.007 3.012 potential_pw2rs 110 12.3 0.021 0.021 2.746 2.751 grid_collocate_task_list 110 9.6 2.639 2.663 2.639 2.663 rs_pw_transfer 902 11.9 0.011 0.011 2.546 2.582 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.537 2.556 calculate_first_density_matrix 1 7.0 0.001 0.001 2.530 2.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.480 2.498 cp_fm_cholesky_decompose 22 10.9 2.281 2.295 2.281 2.295 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.219 2.276 mp_alltoall_d11v 2046 13.8 2.115 2.211 2.115 2.211 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.171 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.997 2.003 jit_kernel_multiply 10 15.2 1.716 1.959 1.716 1.959 qs_create_task_list 11 7.9 0.001 0.001 1.907 1.953 generate_qs_task_list 11 8.9 0.738 0.790 1.906 1.953 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.895 1.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.543000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.181818, yerr=74.171568 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.727808E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 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 0.012 0.030 206.197 206.198 qs_mol_dyn_low 1 2.0 0.003 0.003 205.827 205.840 qs_forces 11 3.9 0.003 0.003 205.738 205.739 qs_energies 11 4.9 0.001 0.002 200.129 200.145 scf_env_do_scf 11 5.9 0.001 0.001 183.271 183.275 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 162.114 162.116 dbcsr_multiply_generic 2507 12.6 0.176 0.183 124.827 125.617 velocity_verlet 10 3.0 0.001 0.002 125.497 125.498 qs_scf_new_mos 117 7.6 0.001 0.001 121.561 121.817 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.560 121.816 ot_scf_mini 117 9.6 0.003 0.003 114.985 115.235 multiply_cannon 2507 13.6 0.235 0.246 101.399 103.137 multiply_cannon_loop 2507 14.6 2.112 2.199 99.230 100.912 ot_mini 117 10.6 0.001 0.001 66.341 66.576 multiply_cannon_multrec 60168 15.6 33.411 35.420 41.883 43.999 qs_ot_get_derivative 117 11.6 0.001 0.002 41.290 41.511 rebuild_ks_matrix 128 8.3 0.001 0.001 34.107 34.461 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 34.107 34.460 mp_waitall_1 267128 16.5 28.831 31.859 28.831 31.859 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.668 31.016 multiply_cannon_sync_h2d 60168 15.6 27.303 29.816 27.303 29.816 qs_ot_get_p 128 10.4 0.001 0.001 26.657 26.988 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.459 25.030 apply_single 128 13.6 0.001 0.001 24.458 25.030 ot_diis_step 117 11.6 0.008 0.008 24.691 24.692 init_scf_loop 11 6.9 0.000 0.000 21.079 21.081 qs_ot_p2m_diag 83 11.4 0.077 0.092 19.902 19.966 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.774 18.966 multiply_cannon_metrocomm3 60168 15.6 0.118 0.123 15.729 18.051 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.435 17.436 prepare_preconditioner 11 7.9 0.000 0.000 16.422 16.470 make_preconditioner 11 8.9 0.000 0.000 16.422 16.470 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.647 15.844 sum_up_and_integrate 128 10.3 0.090 0.107 14.515 14.530 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.454 14.460 integrate_v_rspace 128 11.3 0.003 0.004 14.425 14.441 make_m2s 5014 13.6 0.105 0.112 14.016 14.439 cp_fm_redistribute_end 83 14.4 11.422 14.373 11.434 14.376 make_images 5014 14.6 0.403 0.420 13.833 14.261 cp_fm_diag_elpa_base 83 14.4 2.898 13.991 2.929 14.094 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.948 14.067 calculate_rho_elec 128 8.7 0.045 0.065 13.948 14.067 init_scf_run 11 5.9 0.000 0.001 12.649 12.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.649 12.650 density_rs2pw 128 9.7 0.006 0.007 7.349 10.512 cp_fm_cholesky_invert 11 10.9 9.352 9.360 9.352 9.360 wfi_extrapolate 11 7.9 0.001 0.001 9.300 9.300 mp_sum_l 7870 13.0 8.141 9.245 8.141 9.245 rs_pw_transfer 1046 11.9 0.016 0.019 6.034 9.185 multiply_cannon_metrocomm1 60168 15.6 0.090 0.095 6.341 8.920 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.430 8.617 dbcsr_mm_accdrv_process 124484 16.2 3.184 3.421 8.028 8.579 calculate_dm_sparse 128 9.5 0.001 0.001 8.400 8.481 pw_transfer 1547 11.6 0.077 0.103 8.046 8.281 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.840 8.080 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.813 7.892 make_images_data 5014 15.6 0.066 0.073 6.885 7.735 grid_integrate_task_list 128 12.3 7.017 7.500 7.017 7.500 hybrid_alltoall_any 5200 16.5 0.289 2.245 6.055 7.235 fft_wrap_pw1pw2_140 523 13.2 0.442 0.510 6.686 6.971 fft3d_ps 1291 14.7 2.104 2.879 6.630 6.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.817 6.825 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.976 6.124 mp_waitany 16020 13.9 2.673 5.850 2.673 5.850 grid_collocate_task_list 128 9.7 4.554 5.782 4.554 5.782 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.296 2.107 5.295 mp_alltoall_d11v 2415 14.1 4.190 5.256 4.190 5.256 potential_pw2rs 128 12.3 0.009 0.011 4.918 4.949 cp_fm_cholesky_decompose 22 10.9 4.880 4.892 4.880 4.892 mp_sum_d 4455 12.2 3.661 4.417 3.661 4.417 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.198000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.818182, yerr=6.965084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 826.920960E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57145. MP_Allreduce 11262 946. MP_Sync 168 MP_Alltoall 1955 5616425. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.066 192.653 192.655 qs_mol_dyn_low 1 2.0 0.003 0.007 192.154 192.170 qs_forces 11 3.9 0.008 0.038 192.020 192.033 qs_energies 11 4.9 0.003 0.008 185.098 185.120 scf_env_do_scf 11 5.9 0.001 0.001 168.566 168.577 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 134.845 134.848 velocity_verlet 10 3.0 0.002 0.010 120.626 120.628 dbcsr_multiply_generic 2485 12.5 0.184 0.191 96.532 97.765 qs_scf_new_mos 116 7.6 0.001 0.001 94.517 95.120 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.516 95.120 ot_scf_mini 116 9.6 0.004 0.005 89.741 90.418 multiply_cannon 2485 13.5 0.476 0.529 76.580 81.251 multiply_cannon_loop 2485 14.5 1.249 1.284 73.577 76.343 ot_mini 116 10.6 0.001 0.002 49.479 50.082 mp_waitall_1 212858 16.6 24.447 37.415 24.447 37.415 multiply_cannon_multrec 29820 15.5 21.811 25.850 31.330 35.526 rebuild_ks_matrix 127 8.3 0.001 0.001 33.479 33.926 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 33.478 33.925 init_scf_loop 11 6.9 0.003 0.012 33.626 33.627 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.081 30.495 prepare_preconditioner 11 7.9 0.000 0.000 29.145 29.218 make_preconditioner 11 8.9 0.001 0.002 29.145 29.218 qs_ot_get_derivative 116 11.6 0.001 0.003 27.741 28.419 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.810 28.376 multiply_cannon_metrocomm3 29820 15.5 0.093 0.099 15.624 28.050 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.867 22.998 apply_single 127 13.6 0.001 0.001 21.867 22.998 qs_ot_get_p 127 10.4 0.001 0.001 22.246 22.954 multiply_cannon_sync_h2d 29820 15.5 19.207 21.619 19.207 21.619 ot_diis_step 116 11.6 0.017 0.037 21.564 21.578 qs_ot_p2m_diag 82 11.4 0.185 0.213 17.456 17.498 cp_fm_cholesky_invert 11 10.9 16.952 16.965 16.952 16.965 cp_dbcsr_syevd 82 12.4 0.005 0.005 16.291 16.291 make_m2s 4970 13.5 0.090 0.097 14.199 15.400 make_images 4970 14.5 1.132 1.317 13.989 15.190 sum_up_and_integrate 127 10.3 0.115 0.130 15.010 15.040 integrate_v_rspace 127 11.3 0.003 0.003 14.895 14.930 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.341 14.363 calculate_rho_elec 127 8.7 0.087 0.106 14.340 14.362 cp_fm_diag_elpa 82 13.4 0.000 0.001 13.153 13.163 cp_fm_redistribute_end 82 14.4 7.680 13.082 7.691 13.084 cp_fm_diag_elpa_base 82 14.4 5.152 12.547 5.372 12.929 init_scf_run 11 5.9 0.000 0.001 11.682 11.683 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.681 11.683 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.845 11.274 multiply_cannon_metrocomm4 27335 15.5 0.093 0.108 3.743 10.391 density_rs2pw 127 9.7 0.006 0.006 7.742 10.219 make_images_data 4970 15.5 0.063 0.070 8.587 10.060 mp_irecv_dv 68888 16.3 3.552 10.006 3.552 10.006 dbcsr_mm_accdrv_process 61726 16.2 4.477 5.241 8.983 9.576 hybrid_alltoall_any 5155 16.4 0.340 1.511 7.325 9.350 pw_transfer 1535 11.6 0.088 0.105 9.172 9.253 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.945 9.034 wfi_extrapolate 11 7.9 0.001 0.001 8.497 8.497 rs_pw_transfer 1038 11.9 0.014 0.015 6.002 8.458 fft_wrap_pw1pw2_140 519 13.2 0.469 0.525 7.833 7.934 grid_integrate_task_list 127 12.3 7.133 7.536 7.133 7.536 fft3d_ps 1281 14.7 2.732 2.908 7.320 7.375 cp_fm_cholesky_decompose 22 10.9 7.277 7.363 7.277 7.363 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.242 7.009 calculate_dm_sparse 127 9.5 0.001 0.001 6.434 6.561 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.178 6.189 mp_sum_l 7804 13.0 4.014 5.967 4.014 5.967 grid_collocate_task_list 127 9.7 4.662 5.821 4.662 5.821 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.243 5.380 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.312 5.372 potential_pw2rs 127 12.3 0.015 0.017 5.292 5.321 mp_alltoall_d11v 2401 14.1 4.263 5.128 4.263 5.128 mp_waitany 11660 13.9 2.500 5.049 2.500 5.049 mp_allgather_i34 2485 14.5 1.581 4.719 1.581 4.719 rs_pw_transfer_RS2PW_140 138 11.5 0.349 0.384 2.091 4.536 dbcsr_complete_redistribute 393 12.7 0.790 0.937 3.296 4.147 mp_sum_d 4444 12.2 2.667 4.106 2.667 4.106 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.655000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.727273, yerr=3.413633 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 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 936.419328E+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 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57304. MP_Allreduce 11231 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 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 0.031 0.044 178.284 178.285 qs_mol_dyn_low 1 2.0 0.003 0.004 177.737 177.753 qs_forces 11 3.9 0.003 0.005 177.627 177.632 qs_energies 11 4.9 0.002 0.008 170.981 170.993 scf_env_do_scf 11 5.9 0.001 0.002 155.397 155.398 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 119.459 119.460 velocity_verlet 10 3.0 0.001 0.002 114.035 114.037 dbcsr_multiply_generic 2485 12.5 0.180 0.183 80.747 81.819 qs_scf_new_mos 116 7.6 0.001 0.001 81.284 81.585 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.283 81.584 ot_scf_mini 116 9.6 0.004 0.005 77.195 77.547 multiply_cannon 2485 13.5 0.494 0.514 61.044 65.470 multiply_cannon_loop 2485 14.5 0.852 0.879 57.999 60.339 ot_mini 116 10.6 0.001 0.001 42.379 42.752 init_scf_loop 11 6.9 0.001 0.010 35.832 35.833 mp_waitall_1 169034 16.6 24.555 33.686 24.555 33.686 prepare_preconditioner 11 7.9 0.000 0.000 31.764 31.809 make_preconditioner 11 8.9 0.000 0.001 31.764 31.809 rebuild_ks_matrix 127 8.3 0.001 0.001 30.662 31.093 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.661 31.092 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.440 30.847 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.598 27.983 multiply_cannon_multrec 19880 15.5 13.477 16.545 22.055 25.030 multiply_cannon_metrocomm3 19880 15.5 0.060 0.063 14.825 23.532 qs_ot_get_derivative 116 11.6 0.001 0.002 22.865 23.225 qs_ot_get_p 127 10.4 0.001 0.001 20.387 20.821 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.620 20.516 apply_single 127 13.6 0.001 0.001 19.620 20.516 ot_diis_step 116 11.6 0.018 0.019 19.411 19.412 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.123 16.139 multiply_cannon_sync_h2d 19880 15.5 14.316 15.864 14.316 15.864 make_m2s 4970 13.5 0.081 0.085 14.634 15.472 make_images 4970 14.5 1.141 1.226 14.402 15.232 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.091 15.092 cp_fm_cholesky_invert 11 10.9 15.022 15.031 15.022 15.031 sum_up_and_integrate 127 10.3 0.133 0.143 14.744 14.770 integrate_v_rspace 127 11.3 0.003 0.004 14.611 14.637 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.404 14.432 calculate_rho_elec 127 8.7 0.131 0.146 14.403 14.431 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.881 11.883 cp_fm_redistribute_end 82 14.4 4.486 11.804 4.504 11.806 cp_fm_diag_elpa_base 82 14.4 6.872 11.213 7.278 11.674 init_scf_run 11 5.9 0.000 0.001 10.558 10.558 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.558 10.558 make_images_data 4970 15.5 0.060 0.068 8.989 10.150 hybrid_alltoall_any 5155 16.4 0.430 1.962 7.786 9.514 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.949 9.204 pw_transfer 1535 11.6 0.087 0.106 9.062 9.188 density_rs2pw 127 9.7 0.006 0.006 7.396 9.090 multiply_cannon_metrocomm4 17395 15.5 0.061 0.071 3.362 8.973 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.838 8.968 mp_irecv_dv 49801 16.2 3.240 8.735 3.240 8.735 dbcsr_mm_accdrv_process 41158 16.2 4.321 5.193 8.032 8.133 fft_wrap_pw1pw2_140 519 13.2 0.474 0.529 7.724 7.859 grid_integrate_task_list 127 12.3 7.222 7.714 7.222 7.714 cp_fm_cholesky_decompose 22 10.9 7.661 7.698 7.661 7.698 wfi_extrapolate 11 7.9 0.001 0.001 7.578 7.578 cp_fm_upper_to_full 104 14.5 5.747 7.462 5.747 7.462 fft3d_ps 1281 14.7 2.647 2.878 7.113 7.205 rs_pw_transfer 1038 11.9 0.014 0.015 5.405 7.141 dbcsr_complete_redistribute 393 12.7 1.167 1.201 4.603 6.382 calculate_dm_sparse 127 9.5 0.001 0.001 5.689 5.791 grid_collocate_task_list 127 9.7 4.876 5.696 4.876 5.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.414 5.424 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.654 5.334 mp_alltoall_d11v 2401 14.1 4.231 5.190 4.231 5.190 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.393 5.166 potential_pw2rs 127 12.3 0.020 0.022 4.953 4.975 mp_sum_l 7804 13.0 3.269 4.714 3.269 4.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.518 4.634 mp_allgather_i34 2485 14.5 1.579 4.536 1.579 4.536 mp_waitany 11660 13.9 2.344 4.103 2.344 4.103 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.038 4.071 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.305 4.034 mp_alltoall_i22 712 14.1 1.940 3.813 1.940 3.813 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 3.766 3.768 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.575 3.622 rs_pw_transfer_RS2PW_140 138 11.5 0.329 0.348 1.907 3.606 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.285000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.727273, yerr=10.163132 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.131803E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669921808880 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509632. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 189.613 189.615 qs_mol_dyn_low 1 2.0 0.003 0.004 189.215 189.229 qs_forces 11 3.9 0.004 0.015 189.098 189.102 qs_energies 11 4.9 0.009 0.034 181.869 181.877 scf_env_do_scf 11 5.9 0.004 0.028 164.920 164.927 velocity_verlet 10 3.0 0.002 0.002 126.060 126.063 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 118.587 118.589 dbcsr_multiply_generic 2529 12.6 0.190 0.196 80.189 80.849 qs_scf_new_mos 118 7.6 0.001 0.001 80.194 80.531 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.193 80.530 ot_scf_mini 118 9.6 0.004 0.005 75.695 76.038 multiply_cannon 2529 13.6 0.560 0.592 55.215 59.101 multiply_cannon_loop 2529 14.6 1.199 1.225 51.502 53.240 init_scf_loop 11 6.9 0.001 0.003 46.184 46.185 ot_mini 118 10.6 0.001 0.001 42.780 43.125 prepare_preconditioner 11 7.9 0.000 0.000 42.005 42.028 make_preconditioner 11 8.9 0.000 0.004 42.005 42.028 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.577 40.580 multiply_cannon_multrec 30348 15.6 14.178 19.200 26.163 30.955 rebuild_ks_matrix 129 8.3 0.001 0.001 30.597 30.873 qs_ks_build_kohn_sham_matrix 129 9.3 0.022 0.031 30.596 30.872 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.583 27.818 mp_waitall_1 149172 16.7 17.892 27.630 17.892 27.630 qs_ot_get_derivative 118 11.6 0.002 0.007 22.985 23.337 make_m2s 5058 13.6 0.096 0.099 20.578 21.654 make_images 5058 14.6 1.952 2.237 20.272 21.347 ot_diis_step 118 11.6 0.017 0.018 19.673 19.675 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.211 19.637 apply_single 129 13.6 0.001 0.001 19.210 19.637 qs_ot_get_p 129 10.4 0.001 0.001 18.988 19.358 cp_fm_upper_to_full 106 14.7 11.046 16.378 11.046 16.378 cp_fm_cholesky_invert 11 10.9 15.972 15.982 15.972 15.982 multiply_cannon_metrocomm3 30348 15.6 0.048 0.050 6.649 15.736 sum_up_and_integrate 129 10.3 0.142 0.156 15.044 15.066 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.900 14.946 calculate_rho_elec 129 8.7 0.176 0.192 14.899 14.945 integrate_v_rspace 129 11.3 0.003 0.004 14.902 14.930 qs_ot_p2m_diag 84 11.4 0.347 0.394 14.807 14.858 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.522 13.523 make_images_data 5058 15.6 0.064 0.068 11.242 13.084 multiply_cannon_sync_h2d 30348 15.6 11.856 13.058 11.856 13.058 dbcsr_complete_redistribute 397 12.7 1.508 1.617 8.979 12.733 hybrid_alltoall_any 5245 16.5 0.527 2.233 9.873 12.192 dbcsr_mm_accdrv_process 62780 16.2 7.354 8.276 11.558 12.029 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.624 11.359 init_scf_run 11 5.9 0.000 0.001 10.794 10.795 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.794 10.795 cp_fm_diag_elpa 84 13.4 0.000 0.001 10.425 10.428 cp_fm_redistribute_end 84 14.4 1.802 10.357 1.818 10.363 cp_fm_diag_elpa_base 84 14.4 7.920 9.740 8.511 10.229 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.404 10.041 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.587 9.835 pw_transfer 1559 11.6 0.090 0.109 9.687 9.760 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.456 9.537 mp_alltoall_i22 720 14.1 5.593 9.324 5.593 9.324 fft_wrap_pw1pw2_140 527 13.2 0.488 0.498 8.389 8.480 density_rs2pw 129 9.7 0.006 0.006 7.238 8.416 grid_integrate_task_list 129 12.3 7.565 8.108 7.565 8.108 fft3d_ps 1301 14.7 2.807 2.889 7.665 7.725 wfi_extrapolate 11 7.9 0.001 0.001 7.687 7.687 cp_fm_cholesky_decompose 22 10.9 7.506 7.614 7.506 7.614 multiply_cannon_metrocomm4 25290 15.6 0.076 0.086 2.789 7.314 mp_irecv_dv 76751 16.2 2.643 7.054 2.643 7.054 mp_alltoall_d11v 2429 14.1 4.980 6.392 4.980 6.392 calculate_dm_sparse 129 9.5 0.001 0.001 6.181 6.278 rs_pw_transfer 1054 12.0 0.013 0.014 4.834 6.125 grid_collocate_task_list 129 9.7 5.126 5.848 5.126 5.848 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.376 5.405 potential_pw2rs 129 12.3 0.023 0.023 4.734 4.752 qs_energies_init_hamiltonians 11 5.9 0.003 0.020 4.610 4.611 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.414 4.518 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.332 4.401 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.195 4.247 mp_allgather_i34 2529 14.6 1.805 3.897 1.805 3.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.615000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1071.454545, yerr=18.529405 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.520349E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3943 59063. MP_Allreduce 10855 1180. MP_Sync 86 MP_Alltoall 1700 18828150. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 175.808 175.809 qs_mol_dyn_low 1 2.0 0.003 0.003 175.423 175.437 qs_forces 11 3.9 0.003 0.003 175.300 175.304 qs_energies 11 4.9 0.001 0.002 167.690 167.696 scf_env_do_scf 11 5.9 0.001 0.001 150.317 150.327 velocity_verlet 10 3.0 0.014 0.015 114.898 114.903 scf_env_do_scf_inner_loop 116 6.6 0.003 0.018 114.092 114.093 qs_scf_new_mos 116 7.6 0.001 0.001 75.142 75.230 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.141 75.229 dbcsr_multiply_generic 2485 12.5 0.180 0.186 74.542 75.148 ot_scf_mini 116 9.6 0.003 0.004 70.728 70.806 multiply_cannon 2485 13.5 0.585 0.619 54.727 58.654 multiply_cannon_loop 2485 14.5 0.444 0.460 49.856 51.089 ot_mini 116 10.6 0.001 0.001 39.764 39.851 init_scf_loop 11 6.9 0.000 0.000 36.068 36.070 mp_waitall_1 124680 16.7 26.467 34.606 26.467 34.606 prepare_preconditioner 11 7.9 0.000 0.000 32.026 32.045 make_preconditioner 11 8.9 0.000 0.000 32.026 32.045 rebuild_ks_matrix 127 8.3 0.001 0.001 30.166 30.319 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.165 30.319 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.925 30.197 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.361 27.503 multiply_cannon_multrec 9940 15.5 10.331 15.485 17.544 21.591 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.728 19.991 qs_ot_get_derivative 116 11.6 0.001 0.002 19.827 19.904 ot_diis_step 116 11.6 0.020 0.020 19.864 19.864 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.416 19.777 apply_single 127 13.6 0.001 0.001 19.415 19.777 make_m2s 4970 13.5 0.065 0.070 16.151 18.700 cp_fm_cholesky_invert 11 10.9 18.448 18.454 18.448 18.454 make_images 4970 14.5 2.240 2.780 15.848 18.397 qs_ot_get_p 127 10.4 0.001 0.001 17.960 18.079 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.541 15.579 calculate_rho_elec 127 8.7 0.256 0.266 15.540 15.578 sum_up_and_integrate 127 10.3 0.179 0.188 15.173 15.231 integrate_v_rspace 127 11.3 0.004 0.004 14.993 15.046 qs_ot_p2m_diag 82 11.4 0.490 0.495 14.078 14.098 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.937 12.939 make_images_data 4970 15.5 0.052 0.060 9.892 12.574 hybrid_alltoall_any 5155 16.4 0.811 3.565 9.736 12.177 multiply_cannon_sync_h2d 9940 15.5 11.431 12.015 11.431 12.015 init_scf_run 11 5.9 0.000 0.001 10.497 10.498 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.497 10.498 pw_transfer 1535 11.6 0.086 0.092 10.027 10.065 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.913 9.922 cp_fm_diag_elpa_base 82 14.4 9.676 9.754 9.909 9.919 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.805 9.851 fft_wrap_pw1pw2_140 519 13.2 0.492 0.511 8.620 8.662 cp_fm_cholesky_decompose 22 10.9 7.967 8.149 7.967 8.149 grid_integrate_task_list 127 12.3 7.705 8.129 7.705 8.129 density_rs2pw 127 9.7 0.005 0.005 7.142 8.026 fft3d_ps 1281 14.7 2.689 2.764 7.975 8.004 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.942 8.002 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.551 7.839 wfi_extrapolate 11 7.9 0.001 0.001 7.546 7.546 dbcsr_mm_accdrv_process 20590 16.1 2.654 3.413 6.849 7.472 mp_allgather_i34 2485 14.5 3.088 7.352 3.088 7.352 calculate_dm_sparse 127 9.5 0.001 0.001 6.090 6.167 grid_collocate_task_list 127 9.7 5.355 6.052 5.355 6.052 mp_alltoall_d11v 2401 14.1 5.147 5.720 5.147 5.720 dbcsr_complete_redistribute 393 12.7 2.120 2.178 5.172 5.499 multiply_cannon_metrocomm4 7455 15.5 0.023 0.027 1.871 5.360 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.344 5.345 mp_irecv_dv 28618 15.9 1.835 5.278 1.835 5.278 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.222 5.237 rs_pw_transfer 1038 11.9 0.012 0.013 4.150 5.088 potential_pw2rs 127 12.3 0.026 0.026 4.627 4.641 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.139 4.177 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.546 3.848 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.741 3.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.702 3.713 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.380 3.700 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.561 3.649 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.809000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1421.545455, yerr=46.531248 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 2.961158E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.061 303.792 303.793 qs_mol_dyn_low 1 2.0 0.003 0.003 303.195 303.207 qs_forces 11 3.9 0.003 0.003 303.092 303.095 qs_energies 11 4.9 0.001 0.001 293.758 293.764 scf_env_do_scf 11 5.9 0.001 0.001 271.154 271.163 velocity_verlet 10 3.0 0.002 0.002 218.867 218.876 scf_env_do_scf_inner_loop 118 6.6 0.005 0.022 145.394 145.397 init_scf_loop 11 6.9 0.000 0.000 125.500 125.504 prepare_preconditioner 11 7.9 0.000 0.000 120.206 120.242 make_preconditioner 11 8.9 0.000 0.000 120.206 120.242 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.281 117.347 qs_scf_new_mos 118 7.6 0.001 0.001 93.530 93.667 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.529 93.666 ot_scf_mini 118 9.6 0.004 0.004 88.715 88.803 dbcsr_multiply_generic 2535 12.6 0.214 0.220 84.218 84.970 cp_fm_upper_to_full 106 14.8 52.627 75.558 52.627 75.558 multiply_cannon 2535 13.6 0.712 0.746 58.809 59.473 multiply_cannon_loop 2535 14.6 0.479 0.490 54.698 56.463 ot_mini 118 10.6 0.001 0.001 45.553 45.638 dbcsr_complete_redistribute 397 12.7 3.977 4.029 29.709 42.575 rebuild_ks_matrix 129 8.3 0.001 0.001 39.137 39.244 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 39.137 39.244 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.255 39.095 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.880 36.603 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.128 36.236 mp_alltoall_i22 720 14.1 21.735 34.419 21.735 34.419 cp_fm_cholesky_invert 11 10.9 34.263 34.269 34.263 34.269 mp_waitall_1 104820 16.8 28.119 32.354 28.119 32.354 qs_ot_get_p 129 10.4 0.001 0.001 27.548 27.626 qs_ot_get_derivative 118 11.6 0.002 0.002 24.629 24.719 qs_ot_p2m_diag 84 11.4 0.890 0.895 23.394 23.425 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.624 21.627 make_m2s 5070 13.6 0.077 0.079 20.379 21.601 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.117 21.141 calculate_rho_elec 129 8.7 0.486 0.487 21.117 21.141 make_images 5070 14.6 3.760 3.941 19.897 21.124 ot_diis_step 118 11.6 0.022 0.022 20.878 20.878 sum_up_and_integrate 129 10.3 0.323 0.326 20.296 20.373 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.799 20.164 apply_single 129 13.6 0.001 0.001 19.798 20.164 integrate_v_rspace 129 11.3 0.004 0.004 19.972 20.052 multiply_cannon_metrocomm3 10140 15.6 0.024 0.025 18.276 19.843 multiply_cannon_multrec 10140 15.6 10.561 12.224 18.064 18.460 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.282 18.284 cp_fm_diag_elpa_base 84 14.4 13.861 15.463 18.277 18.278 multiply_cannon_sync_h2d 10140 15.6 15.778 15.791 15.778 15.791 hybrid_alltoall_any 5257 16.5 1.309 3.057 11.117 13.195 make_images_data 5070 15.6 0.061 0.065 11.031 13.054 pw_transfer 1559 11.6 0.094 0.095 12.653 12.665 init_scf_run 11 5.9 0.000 0.001 12.590 12.591 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.590 12.591 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.415 12.427 fft_wrap_pw1pw2_140 527 13.2 0.546 0.548 10.975 10.989 fft3d_ps 1301 14.7 2.762 2.769 10.436 10.451 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.626 9.692 wfi_extrapolate 11 7.9 0.001 0.001 9.444 9.444 mp_alltoall_d11v 2429 14.1 8.406 9.243 8.406 9.243 cp_fm_cholesky_decompose 22 10.9 9.161 9.187 9.161 9.187 dbcsr_mm_accdrv_process 20958 16.1 3.788 5.634 7.263 9.045 grid_integrate_task_list 129 12.3 8.652 8.819 8.652 8.819 density_rs2pw 129 9.7 0.005 0.005 8.629 8.761 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.046 8.047 calculate_dm_sparse 129 9.5 0.001 0.001 6.755 6.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.415 6.503 grid_collocate_task_list 129 9.7 6.412 6.443 6.412 6.443 rs_scatter_matrices 140 9.7 3.662 4.486 6.036 6.284 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.190 6.263 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=303.793000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2651.636364, yerr=163.866505 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 1.262174E+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 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.069 85.717 85.721 qs_energies 1 2.0 0.000 0.001 85.020 85.030 ls_scf 1 3.0 0.000 0.000 84.104 84.116 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.855 72.991 multiply_cannon 111 7.7 0.017 0.020 55.924 56.878 multiply_cannon_loop 111 8.7 0.212 0.227 52.549 53.779 ls_scf_main 1 4.0 0.000 0.000 52.633 52.633 density_matrix_trs4 2 5.0 0.002 0.003 47.022 47.129 ls_scf_init_scf 1 4.0 0.000 0.002 28.422 28.423 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.303 27.348 mp_waitall_1 11031 10.9 22.702 26.484 22.702 26.484 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.149 25.160 multiply_cannon_multrec 2664 9.7 8.176 8.910 15.462 17.189 multiply_cannon_sync_h2d 2664 9.7 13.715 15.520 13.715 15.520 make_m2s 222 7.7 0.008 0.011 13.324 13.741 make_images 222 8.7 0.099 0.110 13.302 13.721 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.798 12.704 make_images_data 222 9.7 0.004 0.005 7.835 8.412 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.387 8.262 hybrid_alltoall_any 227 10.6 0.215 1.835 6.752 7.986 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.640 6.902 7.877 dbcsr_mm_accdrv_process_sort 4760 11.4 6.193 7.094 6.193 7.094 calculate_norms 4752 9.8 5.521 6.228 5.521 6.228 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.032 5.208 mp_sum_l 807 5.4 3.097 4.415 3.097 4.415 make_images_sizes 222 9.7 0.000 0.000 0.714 3.748 mp_alltoall_i44 222 10.7 0.714 3.748 0.714 3.748 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.067 3.494 mp_irecv_dv 6231 10.9 2.049 3.465 2.049 3.465 arnoldi_extremal 4 6.8 0.000 0.000 3.380 3.403 arnoldi_normal_ev 4 7.8 0.001 0.002 3.380 3.403 build_subspace 16 8.4 0.009 0.012 3.285 3.288 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.304 3.135 ls_scf_post 1 4.0 0.000 0.000 3.049 3.061 ls_scf_store_result 1 5.0 0.000 0.000 2.866 2.908 dbcsr_special_finalize 555 9.7 0.005 0.006 2.422 2.814 dbcsr_merge_single_wm 555 10.7 0.453 0.587 2.414 2.806 make_images_pack 222 9.7 2.206 2.631 2.208 2.633 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.367 2.610 dbcsr_sort_data 658 11.4 2.220 2.575 2.220 2.575 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.357 2.475 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.460 2.070 2.462 buffer_matrices_ensure_size 222 8.7 1.745 2.093 1.745 2.093 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.818 1.819 rebuild_ks_matrix 3 7.3 0.000 0.000 1.809 1.810 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.809 1.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.721000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1143.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 2.099577E+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_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.066 0.109 92.574 92.576 qs_energies 1 2.0 0.000 0.000 91.975 91.991 ls_scf 1 3.0 0.000 0.001 90.661 90.677 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.240 76.555 multiply_cannon 111 7.7 0.028 0.041 53.531 57.783 ls_scf_main 1 4.0 0.000 0.000 55.882 55.888 multiply_cannon_loop 111 8.7 0.116 0.122 50.258 53.950 density_matrix_trs4 2 5.0 0.002 0.003 50.108 50.290 ls_scf_init_scf 1 4.0 0.000 0.003 31.248 31.250 ls_scf_init_matrix_S 1 5.0 0.000 0.002 30.002 30.115 mp_waitall_1 9105 10.9 22.078 30.115 22.078 30.115 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.500 27.512 multiply_cannon_multrec 1332 9.7 13.034 16.749 22.210 27.030 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.290 21.446 make_m2s 222 7.7 0.006 0.007 15.739 16.404 make_images 222 8.7 1.570 1.978 15.709 16.375 dbcsr_mm_accdrv_process 4041 10.4 0.269 0.449 8.777 10.339 make_images_data 222 9.7 0.004 0.004 9.178 10.129 dbcsr_mm_accdrv_process_sort 4041 11.4 8.373 9.890 8.373 9.890 hybrid_alltoall_any 227 10.6 0.520 2.509 8.586 9.556 mp_sum_l 807 5.4 5.662 8.566 5.662 8.566 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.234 7.662 mp_irecv_dv 3311 11.0 3.215 7.611 3.215 7.611 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.418 7.064 calculate_norms 2376 9.8 5.979 6.649 5.979 6.649 multiply_cannon_sync_h2d 1332 9.7 4.741 6.147 4.741 6.147 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.051 5.225 arnoldi_extremal 4 6.8 0.000 0.000 4.832 4.853 arnoldi_normal_ev 4 7.8 0.001 0.005 4.832 4.853 build_subspace 16 8.4 0.014 0.021 4.579 4.582 ls_scf_post 1 4.0 0.005 0.019 3.530 3.547 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.260 3.448 ls_scf_store_result 1 5.0 0.000 0.000 3.230 3.350 dbcsr_matrix_vector_mult_local 304 10.0 2.772 3.261 2.774 3.263 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.235 3.071 ls_scf_dm_to_ks 2 5.0 0.001 0.016 2.653 2.744 mp_allgather_i34 111 8.7 0.976 2.474 0.976 2.474 make_images_pack 222 9.7 2.021 2.429 2.023 2.431 dbcsr_sort_data 436 11.2 1.826 2.062 1.826 2.062 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.928 1.930 rebuild_ks_matrix 3 7.3 0.000 0.000 1.915 1.917 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.915 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.576000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1791.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 2.686988E+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_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.070 92.970 92.972 qs_energies 1 2.0 0.000 0.000 92.334 92.337 ls_scf 1 3.0 0.000 0.001 90.936 90.940 dbcsr_multiply_generic 111 6.7 0.015 0.017 75.481 75.761 ls_scf_main 1 4.0 0.000 0.000 56.816 56.821 multiply_cannon 111 7.7 0.042 0.101 52.160 56.429 multiply_cannon_loop 111 8.7 0.101 0.106 48.510 52.377 density_matrix_trs4 2 5.0 0.002 0.003 50.902 51.073 mp_waitall_1 7281 11.0 22.558 32.641 22.558 32.641 ls_scf_init_scf 1 4.0 0.000 0.002 30.626 30.629 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.437 29.508 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.074 27.085 multiply_cannon_multrec 888 9.7 12.718 15.275 21.218 24.517 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.468 22.289 make_m2s 222 7.7 0.006 0.007 16.640 17.847 make_images 222 8.7 1.968 2.287 16.602 17.809 hybrid_alltoall_any 227 10.6 0.622 2.867 8.903 10.326 make_images_data 222 9.7 0.003 0.004 9.264 10.293 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.413 8.029 9.254 dbcsr_mm_accdrv_process_sort 3754 11.4 7.659 8.841 7.659 8.841 mp_sum_l 807 5.4 5.123 8.582 5.123 8.582 multiply_cannon_sync_h2d 888 9.7 6.068 7.464 6.068 7.464 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.421 6.889 mp_irecv_dv 2335 11.1 2.407 6.847 2.407 6.847 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.941 6.806 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.451 6.769 arnoldi_extremal 4 6.8 0.000 0.000 5.221 5.236 arnoldi_normal_ev 4 7.8 0.001 0.005 5.221 5.236 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.819 5.017 build_subspace 16 8.4 0.014 0.020 4.913 4.917 calculate_norms 1584 9.8 4.271 4.599 4.271 4.599 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.502 3.823 mp_allgather_i34 111 8.7 1.477 3.728 1.477 3.728 dbcsr_matrix_vector_mult_local 304 10.0 3.041 3.605 3.043 3.606 ls_scf_post 1 4.0 0.000 0.001 3.494 3.499 ls_scf_store_result 1 5.0 0.000 0.000 3.218 3.319 ls_scf_dm_to_ks 2 5.0 0.000 0.002 2.835 2.934 make_images_pack 222 9.7 1.809 2.139 1.812 2.142 dbcsr_sort_data 325 11.1 1.886 2.131 1.886 2.131 make_images_sizes 222 9.7 0.000 0.000 1.071 2.054 mp_alltoall_i44 222 10.7 1.071 2.054 1.071 2.054 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.921 1.923 rebuild_ks_matrix 3 7.3 0.000 0.000 1.903 1.905 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.903 1.905 dbcsr_data_release 9322 10.9 1.307 1.904 1.307 1.904 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.972000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 3.342696E+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_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.075 98.030 98.033 qs_energies 1 2.0 0.000 0.000 97.235 97.240 ls_scf 1 3.0 0.000 0.001 95.474 95.477 dbcsr_multiply_generic 111 6.7 0.016 0.018 79.025 79.321 ls_scf_main 1 4.0 0.000 0.000 59.465 59.466 multiply_cannon 111 7.7 0.056 0.142 51.704 55.944 density_matrix_trs4 2 5.0 0.002 0.003 53.276 53.428 multiply_cannon_loop 111 8.7 0.115 0.127 46.706 50.170 ls_scf_init_scf 1 4.0 0.000 0.002 32.666 32.668 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.417 31.485 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.955 28.966 mp_waitall_1 6369 11.0 23.027 28.669 23.027 28.669 multiply_cannon_multrec 1332 9.7 14.151 17.332 21.975 24.484 make_m2s 222 7.7 0.006 0.008 21.307 22.701 make_images 222 8.7 3.131 3.593 21.257 22.654 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.269 17.339 make_images_data 222 9.7 0.003 0.004 11.896 13.435 hybrid_alltoall_any 227 10.6 0.799 3.788 11.265 13.066 dbcsr_mm_accdrv_process 3641 10.4 0.229 0.411 7.460 8.981 dbcsr_mm_accdrv_process_sort 3641 11.4 7.086 8.562 7.086 8.562 mp_sum_l 807 5.4 4.437 8.394 4.437 8.394 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.389 6.650 multiply_cannon_sync_h2d 1332 9.7 5.559 6.159 5.559 6.159 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.099 6.084 mp_irecv_dv 3229 10.9 2.075 5.995 2.075 5.995 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.694 5.481 arnoldi_extremal 4 6.8 0.000 0.002 5.359 5.371 arnoldi_normal_ev 4 7.8 0.001 0.005 5.359 5.371 build_subspace 16 8.4 0.014 0.021 5.024 5.032 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.566 4.732 calculate_norms 2376 9.8 4.177 4.559 4.177 4.559 mp_allgather_i34 111 8.7 2.126 4.399 2.126 4.399 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.642 3.948 dbcsr_matrix_vector_mult_local 304 10.0 3.213 3.720 3.215 3.721 dbcsr_sort_data 658 11.4 3.116 3.473 3.116 3.473 ls_scf_post 1 4.0 0.000 0.001 3.342 3.347 dbcsr_special_finalize 555 9.7 0.006 0.007 2.863 3.248 dbcsr_merge_single_wm 555 10.7 0.539 0.658 2.855 3.240 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.069 3.146 ls_scf_store_result 1 5.0 0.000 0.000 3.025 3.084 dbcsr_data_release 10477 10.7 1.568 2.371 1.568 2.371 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.996 1.997 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.987 rebuild_ks_matrix 3 7.3 0.000 0.000 1.973 1.974 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.973 1.974 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.033000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 4.623725E+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_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.075 100.132 100.133 qs_energies 1 2.0 0.000 0.000 99.271 99.277 ls_scf 1 3.0 0.000 0.000 97.326 97.337 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.318 78.494 ls_scf_main 1 4.0 0.000 0.000 62.594 62.595 multiply_cannon 111 7.7 0.082 0.148 55.538 60.741 density_matrix_trs4 2 5.0 0.002 0.003 55.136 55.220 multiply_cannon_loop 111 8.7 0.069 0.078 50.972 52.421 mp_waitall_1 5436 11.0 27.123 32.417 27.123 32.417 ls_scf_init_scf 1 4.0 0.000 0.000 31.048 31.053 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.658 29.700 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.501 27.507 multiply_cannon_multrec 444 9.7 14.045 16.693 21.050 24.656 make_m2s 222 7.7 0.004 0.005 17.998 20.480 make_images 222 8.7 3.711 4.417 17.936 20.421 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.576 16.586 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.051 14.002 make_images_data 222 9.7 0.003 0.004 10.286 12.702 hybrid_alltoall_any 227 10.6 0.790 3.759 10.012 12.557 multiply_cannon_sync_h2d 444 9.7 6.507 8.777 6.507 8.777 dbcsr_mm_accdrv_process 3003 10.4 0.191 0.360 6.711 7.850 dbcsr_mm_accdrv_process_sort 3003 11.4 6.392 7.504 6.392 7.504 mp_allgather_i34 111 8.7 2.796 7.075 2.796 7.075 arnoldi_extremal 4 6.8 0.000 0.000 5.956 5.963 arnoldi_normal_ev 4 7.8 0.001 0.005 5.956 5.963 build_subspace 16 8.4 0.015 0.019 5.567 5.574 mp_sum_l 807 5.4 2.832 4.959 2.832 4.959 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.678 4.853 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.222 4.439 dbcsr_matrix_vector_mult_local 304 10.0 3.681 4.123 3.684 4.125 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.609 4.007 mp_irecv_dv 1241 11.2 1.589 3.983 1.589 3.983 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.785 3.897 make_images_sizes 222 9.7 0.000 0.000 1.075 3.739 mp_alltoall_i44 222 10.7 1.074 3.738 1.074 3.738 ls_scf_post 1 4.0 0.000 0.000 3.685 3.690 calculate_norms 792 9.8 3.526 3.670 3.526 3.670 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.932 3.658 ls_scf_store_result 1 5.0 0.000 0.000 3.477 3.510 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.270 2.271 dbcsr_finalize 304 7.8 0.062 0.078 2.193 2.264 rebuild_ks_matrix 3 7.3 0.000 0.000 2.237 2.239 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.237 2.239 dbcsr_merge_all 275 8.9 0.474 0.524 2.044 2.102 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.133000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3635.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b01cdf04a1de84b00fa31cfcfd63373d97607d6b_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 8.828092E+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_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.099 109.989 109.989 qs_energies 1 2.0 0.000 0.000 108.497 108.511 ls_scf 1 3.0 0.000 0.000 105.565 105.579 dbcsr_multiply_generic 111 6.7 0.023 0.027 78.814 78.933 ls_scf_main 1 4.0 0.000 0.000 67.172 67.173 density_matrix_trs4 2 5.0 0.002 0.003 57.763 57.839 multiply_cannon 111 7.7 0.144 0.209 51.011 52.896 multiply_cannon_loop 111 8.7 0.068 0.072 47.613 48.304 ls_scf_init_scf 1 4.0 0.000 0.000 34.705 34.706 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.026 33.033 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.159 30.166 mp_waitall_1 4527 11.1 23.633 27.237 23.633 27.237 make_m2s 222 7.7 0.005 0.005 24.179 25.107 make_images 222 8.7 4.581 4.964 24.073 24.999 multiply_cannon_multrec 444 9.7 17.921 18.792 22.504 23.151 hybrid_alltoall_any 227 10.6 1.661 3.628 13.145 16.156 make_images_data 222 9.7 0.003 0.003 13.403 15.803 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.634 12.262 multiply_cannon_sync_h2d 444 9.7 8.849 8.896 8.849 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.438 7.449 arnoldi_normal_ev 4 7.8 0.002 0.009 7.438 7.449 build_subspace 16 8.4 0.026 0.036 6.896 6.907 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.474 5.621 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.325 5.418 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.077 5.303 dbcsr_matrix_vector_mult_local 304 10.0 4.960 5.251 4.962 5.253 dbcsr_mm_accdrv_process 1814 10.4 0.234 0.324 4.418 4.550 dbcsr_mm_accdrv_process_sort 1814 11.4 4.116 4.240 4.116 4.240 ls_scf_post 1 4.0 0.000 0.000 3.688 3.702 make_images_sizes 222 9.7 0.000 0.000 1.458 3.628 mp_alltoall_i44 222 10.7 1.458 3.628 1.458 3.628 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.446 mp_allgather_i34 111 8.7 1.052 3.369 1.052 3.369 calculate_norms 792 9.8 3.239 3.280 3.239 3.280 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.127 dbcsr_merge_all 275 8.9 0.891 0.918 2.873 2.907 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 2.902 2.902 dbcsr_complete_redistribute 5 7.6 1.429 1.468 2.790 2.902 dbcsr_data_release 12724 10.6 2.334 2.878 2.334 2.878 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.624 2.625 rebuild_ks_matrix 3 7.3 0.000 0.000 2.554 2.556 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.005 2.554 2.556 matrix_ls_to_qs 2 6.0 0.000 0.000 2.415 2.534 dbcsr_sort_data 325 11.1 2.445 2.499 2.445 2.499 dbcsr_new_transposed 4 7.5 0.243 0.252 2.298 2.309 dbcsr_frobenius_norm 74 6.6 2.056 2.134 2.186 2.225 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.206 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.989000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6853.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b01cdf04a1de84b00fa31cfcfd63373d97607d6b Summary: empty Status: OK