=== 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: 4083d9693461e026a35c103709c5fd193a3d5841 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/01 job id: 42584272 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/02 job id: 42584273 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/03 job id: 42584275 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/04 job id: 42584277 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/05 job id: 42584279 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/06 job id: 42584281 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/07 job id: 42584286 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/08 job id: 42584289 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/09 job id: 42584291 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/10 job id: 42584293 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/11 job id: 42584295 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/12 job id: 42584297 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/13 job id: 42584300 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/14 job id: 42584302 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/15 job id: 42584304 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/16 job id: 42584306 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/17 job id: 42584308 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/18 job id: 42584311 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/19 job id: 42584313 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/20 job id: 42584315 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/21 job id: 42584317 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/22 job id: 42584318 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/23 job id: 42584319 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/24 job id: 42584320 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/25 job id: 42584321 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/26 job id: 42584322 --- 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/4083d9693461e026a35c103709c5fd193a3d5841_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 0.022 0.034 135.198 135.199 farming_run 1 2.0 134.495 134.496 135.171 135.174 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.462448E+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 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.032 115.436 115.437 qs_energies 1 2.0 0.000 0.000 115.224 115.225 mp2_main 1 3.0 0.000 0.000 113.416 113.417 mp2_gpw_main 1 4.0 0.018 0.021 112.618 112.620 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 94.131 94.282 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.416 55.566 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.720 46.715 get_2c_integrals 1 6.0 0.000 0.000 37.169 38.543 integrate_v_rspace 273 8.0 0.436 0.449 25.061 29.653 pw_transfer 6555 10.6 0.373 0.392 27.330 27.687 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 25.985 26.132 grid_integrate_task_list 273 9.0 20.883 25.921 20.883 25.921 fft_wrap_pw1pw2_100 2178 12.4 1.212 1.420 23.545 23.698 compute_2c_integrals 1 7.0 0.002 0.002 19.705 19.706 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.920 19.442 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.434 18.917 19.439 cp_fm_cholesky_decompose 12 8.2 17.399 18.797 17.399 18.797 rpa_ri_compute_en 1 5.0 0.000 0.000 18.379 18.468 cholesky_decomp 1 7.0 0.000 0.000 16.314 17.714 fft3d_s 5443 13.4 16.107 16.285 16.129 16.307 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.551 10.853 15.551 calculate_wavefunction 272 8.0 5.476 5.553 12.575 13.258 rpa_num_int 1 6.0 0.000 0.000 10.500 10.500 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.458 10.497 calc_mat_Q 8 8.0 0.000 0.000 9.327 9.433 contract_S_to_Q 8 9.0 0.000 0.000 8.749 8.856 calc_potential_gpw 544 9.5 0.005 0.006 8.243 8.736 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.239 8.590 parallel_gemm_fm 14 9.1 0.000 0.000 8.335 8.404 parallel_gemm_fm_cosma 14 10.1 8.335 8.404 8.335 8.404 potential_pw2rs 545 10.0 0.108 0.110 7.715 8.336 collocate_single_gaussian 272 10.0 0.041 0.043 7.480 7.747 create_integ_mat 1 6.0 0.022 0.027 7.633 7.633 array2fm 1 7.0 0.000 0.000 6.701 7.105 pw_scatter_s 2720 13.7 4.436 4.567 4.436 4.567 pw_gather_s 2722 13.2 3.859 4.136 3.859 4.136 array2fm_buffer_send 1 8.0 2.933 3.176 2.933 3.176 mp_sync 27 8.8 1.384 2.567 1.384 2.567 pw_poisson_solve 545 10.5 1.145 1.180 2.197 2.484 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.617175, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 0.049 0.072 398.248 398.250 farming_run 1 2.0 397.254 397.261 398.169 398.192 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.225810E+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 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.049 209.789 209.790 qs_energies 1 2.0 0.000 0.000 209.486 209.523 scf_env_do_scf 1 3.0 0.000 0.000 106.105 106.105 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.228 105.236 rebuild_ks_matrix 4 6.0 0.000 0.000 105.190 105.197 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.190 105.197 hfx_ks_matrix 4 8.0 0.001 0.001 104.822 104.825 integrate_four_center 4 9.0 0.145 0.453 104.821 104.824 mp2_main 1 3.0 0.012 0.052 103.100 103.136 mp2_gpw_main 1 4.0 0.055 0.141 102.270 102.278 integrate_four_center_main 4 10.0 0.103 0.516 96.541 99.377 integrate_four_center_bin 264 11.0 96.438 99.323 96.438 99.323 init_scf_loop 1 4.0 0.000 0.000 91.898 91.898 mp2_ri_gpw_compute_in 1 5.0 0.067 0.093 75.174 76.275 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.590 55.682 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.159 42.060 47.498 integrate_v_rspace 95 8.0 0.397 0.578 28.406 33.649 pw_transfer 2240 10.6 0.145 0.167 29.862 30.282 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.893 29.338 grid_integrate_task_list 95 9.0 23.694 29.143 23.694 29.143 mp2_ri_gpw_compute_en 1 5.0 0.072 0.119 26.861 28.498 fft_wrap_pw1pw2_100 730 12.4 1.246 1.382 26.635 27.072 ao_to_mo_and_store_B_mult_1 91 7.0 10.845 25.987 10.845 25.987 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.930 2.266 25.046 25.057 get_2c_integrals 1 6.0 0.018 0.069 20.451 20.528 compute_2c_integrals 1 7.0 0.093 0.248 19.402 19.425 compute_2c_integrals_loop_lm 1 8.0 0.016 0.055 18.788 19.010 mp2_eri_2c_integrate_gpw 1 9.0 1.752 1.843 18.772 18.963 fft3d_s 1823 13.4 18.425 18.939 18.438 18.953 scf_env_do_scf_inner_loop 4 4.0 0.001 0.038 14.206 14.206 calculate_wavefunction 91 8.0 2.024 2.071 9.760 9.969 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.583 8.763 9.220 potential_pw2rs 186 10.0 0.033 0.035 8.601 9.125 local_gemm 172 8.0 8.206 8.652 8.206 8.652 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.181 8.495 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.529 7.964 8.290 calc_potential_gpw 182 9.5 0.002 0.002 7.893 8.098 collocate_single_gaussian 91 10.0 0.018 0.035 7.842 8.062 mp2_ri_gpw_compute_en_ener 172 7.0 6.344 6.423 6.344 6.423 mp_sendrecv_dm3 2068 8.0 5.982 6.322 5.982 6.322 mp_sync 38 10.4 2.977 5.573 2.977 5.573 pw_gather_s 912 13.2 4.927 5.330 4.927 5.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.276819, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.923968E+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 814967. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.177 52.225 52.227 qs_mol_dyn_low 1 2.0 0.004 0.009 49.947 51.632 qs_forces 11 3.9 0.002 0.003 49.063 49.066 qs_energies 11 4.9 0.002 0.009 47.634 47.644 scf_env_do_scf 11 5.9 0.001 0.002 41.224 41.225 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 39.090 39.094 dbcsr_multiply_generic 2286 12.5 0.092 0.098 31.813 32.296 qs_scf_new_mos 108 7.5 0.000 0.001 29.471 29.766 qs_scf_loop_do_ot 108 8.5 0.001 0.001 29.470 29.766 ot_scf_mini 108 9.5 0.002 0.003 27.909 28.101 multiply_cannon 2286 13.5 0.187 0.195 25.599 27.353 multiply_cannon_loop 2286 14.5 1.450 1.519 24.944 26.715 velocity_verlet 10 3.0 0.001 0.002 24.610 24.619 ot_mini 108 10.5 0.001 0.001 17.733 17.973 qs_ot_get_derivative 108 11.5 0.001 0.001 14.874 15.060 mp_waitall_1 267858 16.1 7.806 14.488 7.806 14.488 multiply_cannon_metrocomm3 54864 15.5 0.068 0.072 5.640 12.774 multiply_cannon_multrec 54864 15.5 4.372 6.777 7.953 11.649 rebuild_ks_matrix 119 8.3 0.000 0.000 7.676 7.818 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.014 7.676 7.818 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.771 6.898 multiply_cannon_sync_h2d 54864 15.5 6.066 6.861 6.066 6.861 qs_ot_get_p 119 10.4 0.002 0.012 5.957 6.223 mp_sum_l 7207 12.9 4.449 6.004 4.449 6.004 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.926 5.374 init_scf_run 11 5.9 0.000 0.001 5.161 5.162 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.161 5.161 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.039 5.137 dbcsr_mm_accdrv_process 76910 16.1 1.084 1.710 3.504 4.912 sum_up_and_integrate 119 10.3 0.013 0.015 4.377 4.385 integrate_v_rspace 119 11.3 0.002 0.002 4.364 4.373 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.867 3.938 calculate_rho_elec 119 8.7 0.011 0.016 3.867 3.937 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.324 3.360 jit_kernel_multiply 13 15.8 2.362 3.118 2.362 3.118 calculate_first_density_matrix 1 7.0 0.000 0.003 3.081 3.094 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.859 3.067 apply_single 119 13.6 0.000 0.000 2.859 3.067 calculate_dm_sparse 119 9.5 0.000 0.000 2.889 3.053 rs_pw_transfer 974 11.9 0.011 0.013 2.661 2.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.007 2.633 2.635 ot_diis_step 108 11.5 0.006 0.006 2.632 2.633 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.577 2.577 md_output 10 3.0 0.000 0.000 0.029 2.544 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.406 2.463 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.421 2.453 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.387 2.387 cp_fm_redistribute_end 50 14.0 2.168 2.368 2.172 2.369 md_write_output 11 3.9 0.009 0.839 0.026 2.355 cp_fm_diag_elpa_base 50 14.0 0.195 2.336 0.195 2.346 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.181 2.228 acc_transpose_blocks 54864 15.5 0.210 0.231 1.700 2.154 init_scf_loop 11 6.9 0.002 0.015 2.115 2.137 grid_integrate_task_list 119 12.3 2.025 2.128 2.025 2.128 density_rs2pw 119 9.7 0.004 0.004 2.038 2.112 wfi_extrapolate 11 7.9 0.001 0.001 2.014 2.014 mp_sum_d 4129 12.0 1.242 1.840 1.242 1.840 potential_pw2rs 119 12.3 0.004 0.004 1.757 1.768 mp_max_l 33 2.8 1.683 1.704 1.683 1.704 pw_transfer 1439 11.6 0.051 0.055 1.515 1.582 make_m2s 4572 13.5 0.053 0.055 1.508 1.558 write_trajectory 44 4.9 0.016 1.510 0.016 1.516 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.483 1.512 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.441 1.512 make_images 4572 14.5 0.132 0.137 1.426 1.475 mp_waitany 12084 13.8 1.253 1.472 1.253 1.472 grid_collocate_task_list 119 9.7 1.289 1.351 1.289 1.351 mp_alltoall_d11v 2130 13.8 1.185 1.334 1.185 1.334 fft3d_ps 1201 14.6 0.356 0.458 1.220 1.283 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.112 1.182 acc_transpose_blocks_kernels 54864 16.5 0.238 0.368 0.795 1.082 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.667 1.057 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.227000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.808448E+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 1032074. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.091 39.239 39.241 qs_mol_dyn_low 1 2.0 0.003 0.005 38.839 38.846 qs_forces 11 3.9 0.002 0.004 38.429 38.438 qs_energies 11 4.9 0.002 0.006 36.768 36.780 scf_env_do_scf 11 5.9 0.001 0.001 31.310 31.311 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.843 28.853 dbcsr_multiply_generic 2286 12.5 0.177 0.186 22.086 22.420 qs_scf_new_mos 108 7.5 0.001 0.001 20.106 20.338 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.106 20.338 ot_scf_mini 108 9.5 0.003 0.003 19.212 19.393 multiply_cannon 2286 13.5 0.207 0.218 16.975 18.560 velocity_verlet 10 3.0 0.001 0.002 18.161 18.169 multiply_cannon_loop 2286 14.5 0.897 0.975 15.850 17.229 ot_mini 108 10.5 0.002 0.013 12.260 12.505 mp_waitall_1 217478 16.2 6.380 11.772 6.380 11.772 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.427 10.153 qs_ot_get_derivative 108 11.5 0.001 0.001 9.762 9.948 multiply_cannon_multrec 27432 15.5 2.029 4.839 5.993 9.188 rebuild_ks_matrix 119 8.3 0.000 0.000 6.990 7.131 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.990 7.131 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.170 6.299 dbcsr_mm_accdrv_process 47894 16.0 3.014 5.266 3.895 5.933 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.797 4.655 qs_ot_get_p 119 10.4 0.001 0.001 4.131 4.376 init_scf_run 11 5.9 0.000 0.001 4.254 4.254 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.253 4.254 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.057 4.137 apply_single 119 13.6 0.000 0.000 3.056 4.137 sum_up_and_integrate 119 10.3 0.024 0.026 4.053 4.061 integrate_v_rspace 119 11.3 0.002 0.002 4.029 4.038 mp_sum_l 7207 12.9 2.064 3.965 2.064 3.965 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.498 3.535 calculate_rho_elec 119 8.7 0.021 0.024 3.497 3.534 multiply_cannon_sync_h2d 27432 15.5 2.192 2.798 2.192 2.798 make_m2s 4572 13.5 0.052 0.054 2.537 2.776 calculate_first_density_matrix 1 7.0 0.000 0.002 2.768 2.770 make_images 4572 14.5 0.200 0.237 2.449 2.685 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.648 2.666 init_scf_loop 11 6.9 0.001 0.003 2.445 2.477 rs_pw_transfer 974 11.9 0.010 0.011 2.384 2.476 ot_diis_step 108 11.5 0.015 0.048 2.447 2.450 calculate_dm_sparse 119 9.5 0.000 0.000 2.194 2.272 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.221 2.221 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.173 2.176 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.085 2.176 jit_kernel_multiply 10 16.1 0.829 1.960 0.829 1.960 density_rs2pw 119 9.7 0.004 0.004 1.858 1.952 grid_integrate_task_list 119 12.3 1.842 1.939 1.842 1.939 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.890 1.931 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.906 1.907 cp_fm_redistribute_end 50 14.0 1.570 1.884 1.573 1.885 cp_fm_diag_elpa_base 50 14.0 0.303 1.834 0.311 1.865 potential_pw2rs 119 12.3 0.006 0.006 1.800 1.811 pw_transfer 1439 11.6 0.063 0.066 1.714 1.746 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.625 1.660 make_images_data 4572 15.5 0.050 0.089 1.167 1.624 acc_transpose_blocks 27432 15.5 0.107 0.112 1.258 1.553 prepare_preconditioner 11 7.9 0.000 0.000 1.497 1.523 make_preconditioner 11 8.9 0.000 0.001 1.497 1.523 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.397 1.455 wfi_extrapolate 11 7.9 0.001 0.001 1.435 1.435 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.996 1.411 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.405 1.411 fft3d_ps 1201 14.6 0.497 0.549 1.331 1.362 mp_allgather_i34 2286 14.5 0.568 1.351 0.568 1.351 grid_collocate_task_list 119 9.7 1.243 1.330 1.243 1.330 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.276 1.311 mp_alltoall_d11v 2130 13.8 1.164 1.271 1.164 1.271 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.169 1.217 mp_sum_d 4129 12.0 0.603 1.064 0.603 1.064 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.941 0.950 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 0.730 0.939 multiply_cannon_metrocomm1 27432 15.5 0.031 0.036 0.512 0.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.241000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.272727, yerr=1.710444 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.354688E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.071 32.876 32.877 qs_mol_dyn_low 1 2.0 0.003 0.004 32.422 32.429 qs_forces 11 3.9 0.005 0.018 32.205 32.207 qs_energies 11 4.9 0.002 0.006 30.223 30.228 scf_env_do_scf 11 5.9 0.001 0.001 25.187 25.187 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.688 22.689 dbcsr_multiply_generic 2286 12.5 0.111 0.155 16.786 16.854 velocity_verlet 10 3.0 0.126 0.169 15.006 15.008 qs_scf_new_mos 108 7.5 0.001 0.001 14.902 14.918 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.901 14.917 multiply_cannon 2286 13.5 0.195 0.201 13.494 14.286 ot_scf_mini 108 9.5 0.002 0.003 14.183 14.195 multiply_cannon_loop 2286 14.5 0.637 0.665 12.712 13.543 ot_mini 108 10.5 0.001 0.001 8.996 9.009 qs_ot_get_derivative 108 11.5 0.001 0.001 7.504 7.517 multiply_cannon_multrec 18288 15.5 1.947 2.879 6.891 7.420 rebuild_ks_matrix 119 8.3 0.000 0.000 6.634 6.653 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.634 6.652 dbcsr_mm_accdrv_process 38222 16.0 4.086 5.707 4.858 5.863 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.491 5.509 sum_up_and_integrate 119 10.3 0.030 0.031 4.224 4.281 integrate_v_rspace 119 11.3 0.002 0.003 4.193 4.255 mp_waitall_1 169478 16.3 2.992 3.995 2.992 3.995 init_scf_run 11 5.9 0.000 0.001 3.865 3.865 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.865 3.865 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.023 3.670 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.158 3.166 calculate_rho_elec 119 8.7 0.030 0.031 3.157 3.165 qs_ot_get_p 119 10.4 0.001 0.001 3.048 3.073 calculate_first_density_matrix 1 7.0 0.001 0.004 2.676 2.677 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.651 2.505 init_scf_loop 11 6.9 0.001 0.003 2.482 2.486 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.061 2.366 apply_single 119 13.6 0.000 0.000 2.061 2.366 rs_pw_transfer 974 11.9 0.009 0.010 2.150 2.294 pw_transfer 1439 11.6 0.063 0.066 2.048 2.109 potential_pw2rs 119 12.3 0.007 0.008 2.010 2.067 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.958 2.019 calculate_dm_sparse 119 9.5 0.000 0.000 1.997 2.010 qs_ot_p2m_diag 50 11.0 0.012 0.013 1.999 2.004 make_m2s 4572 13.5 0.044 0.046 1.788 1.928 grid_integrate_task_list 119 12.3 1.810 1.905 1.810 1.905 jit_kernel_multiply 10 15.8 0.720 1.897 0.720 1.897 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.893 1.896 density_rs2pw 119 9.7 0.004 0.004 1.755 1.882 make_images 4572 14.5 0.191 0.207 1.703 1.842 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.736 1.737 fft3d_ps 1201 14.6 0.505 0.521 1.645 1.706 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.640 1.704 prepare_preconditioner 11 7.9 0.000 0.000 1.694 1.696 make_preconditioner 11 8.9 0.000 0.001 1.694 1.696 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.550 1.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.579 1.587 multiply_cannon_sync_h2d 18288 15.5 1.389 1.545 1.389 1.545 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.540 1.545 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.511 1.511 cp_fm_redistribute_end 50 14.0 1.120 1.491 1.121 1.492 cp_fm_diag_elpa_base 50 14.0 0.355 1.438 0.369 1.476 ot_diis_step 108 11.5 0.011 0.011 1.473 1.473 mp_sum_l 7207 12.9 1.105 1.425 1.105 1.425 acc_transpose_blocks 18288 15.5 0.074 0.076 1.387 1.414 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.370 1.374 grid_collocate_task_list 119 9.7 1.204 1.287 1.204 1.287 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.213 1.215 wfi_extrapolate 11 7.9 0.001 0.001 1.140 1.140 mp_alltoall_z22v 1201 16.6 0.987 1.093 0.987 1.093 acc_transpose_blocks_kernels 18288 16.5 0.211 0.222 0.948 0.963 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.946 0.948 make_images_data 4572 15.5 0.045 0.049 0.782 0.930 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.865 0.892 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.667 0.847 yz_to_x 487 15.3 0.048 0.052 0.742 0.831 mp_alltoall_d11v 2130 13.8 0.665 0.786 0.665 0.786 dbcsr_complete_redistribute 329 12.2 0.178 0.251 0.630 0.742 jit_kernel_transpose 5 15.6 0.737 0.741 0.737 0.741 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.647 0.722 cp_fm_cholesky_invert 11 10.9 0.700 0.704 0.700 0.704 mp_waitany 9880 13.7 0.524 0.688 0.524 0.688 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.518 0.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.877000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=2.374103 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 545.595392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.054 35.634 35.634 qs_mol_dyn_low 1 2.0 0.003 0.004 35.307 35.315 qs_forces 11 3.9 0.002 0.003 35.214 35.215 qs_energies 11 4.9 0.002 0.003 33.524 33.529 scf_env_do_scf 11 5.9 0.016 0.372 27.790 27.791 scf_env_do_scf_inner_loop 108 6.5 0.006 0.021 24.034 24.034 dbcsr_multiply_generic 2286 12.5 0.100 0.107 18.826 18.946 velocity_verlet 10 3.0 0.036 0.044 17.328 17.330 qs_scf_new_mos 108 7.5 0.001 0.001 16.080 16.123 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.080 16.122 multiply_cannon 2286 13.5 0.237 0.273 15.374 15.748 ot_scf_mini 108 9.5 0.003 0.003 15.153 15.198 multiply_cannon_loop 2286 14.5 0.936 0.965 14.454 14.765 ot_mini 108 10.5 0.001 0.001 9.450 9.511 multiply_cannon_multrec 27432 15.5 2.328 2.988 8.887 9.422 dbcsr_mm_accdrv_process 47916 15.9 5.215 7.061 6.460 7.759 qs_ot_get_derivative 108 11.5 0.001 0.001 7.655 7.701 rebuild_ks_matrix 119 8.3 0.000 0.000 6.358 6.403 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.357 6.403 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.644 5.685 init_scf_run 11 5.9 0.000 0.001 4.383 4.384 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.383 4.384 sum_up_and_integrate 119 10.3 0.034 0.038 3.657 3.665 integrate_v_rspace 119 11.3 0.002 0.002 3.622 3.632 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.072 3.462 init_scf_loop 11 6.9 0.000 0.001 3.366 3.366 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.174 3.204 calculate_rho_elec 119 8.7 0.040 0.046 3.173 3.203 qs_ot_get_p 119 10.4 0.001 0.002 3.119 3.190 calculate_first_density_matrix 1 7.0 0.001 0.003 3.057 3.060 mp_waitall_1 145218 16.4 2.038 2.744 2.038 2.744 prepare_preconditioner 11 7.9 0.000 0.000 2.541 2.548 make_preconditioner 11 8.9 0.000 0.001 2.541 2.548 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.143 2.470 calculate_dm_sparse 119 9.5 0.000 0.000 2.313 2.362 jit_kernel_multiply 10 15.8 1.185 2.312 1.185 2.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.261 2.262 make_m2s 4572 13.5 0.054 0.056 2.112 2.218 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.024 2.150 apply_single 119 13.6 0.000 0.000 2.024 2.150 make_images 4572 14.5 0.280 0.345 2.005 2.110 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.002 2.044 2.078 acc_transpose_blocks 27432 15.5 0.111 0.114 2.047 2.070 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.950 1.973 rs_pw_transfer 974 11.9 0.009 0.009 1.860 1.954 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.899 1.908 grid_integrate_task_list 119 12.3 1.819 1.898 1.819 1.898 density_rs2pw 119 9.7 0.004 0.004 1.668 1.757 ot_diis_step 108 11.5 0.012 0.012 1.753 1.754 pw_transfer 1439 11.6 0.062 0.066 1.712 1.742 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.899 1.702 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.623 1.656 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.580 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.504 1.512 potential_pw2rs 119 12.3 0.008 0.010 1.452 1.456 acc_transpose_blocks_kernels 27432 16.5 0.269 0.279 1.423 1.439 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.338 1.373 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.360 1.361 cp_fm_redistribute_end 50 14.0 0.895 1.338 0.896 1.339 grid_collocate_task_list 119 9.7 1.222 1.332 1.222 1.332 fft3d_ps 1201 14.6 0.532 0.585 1.304 1.326 cp_fm_diag_elpa_base 50 14.0 0.421 1.282 0.441 1.323 wfi_extrapolate 11 7.9 0.001 0.001 1.279 1.279 mp_sum_l 7207 12.9 0.921 1.181 0.921 1.181 jit_kernel_transpose 5 15.6 1.153 1.166 1.153 1.166 mp_alltoall_i22 627 13.8 0.855 1.155 0.855 1.155 cp_fm_upper_to_full 72 13.5 0.812 1.124 0.812 1.124 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.084 1.107 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.101 1.102 multiply_cannon_sync_h2d 27432 15.5 0.988 1.058 0.988 1.058 dbcsr_complete_redistribute 329 12.2 0.126 0.154 0.775 1.052 make_images_data 4572 15.5 0.045 0.049 0.815 0.919 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.870 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.586 0.857 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.709 0.855 mp_alltoall_d11v 2130 13.8 0.741 0.845 0.741 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.802 0.807 copy_dbcsr_to_fm 153 11.3 0.002 0.003 0.739 0.781 cp_fm_cholesky_invert 11 10.9 0.752 0.755 0.752 0.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.634000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.000000, yerr=3.643175 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 595.804160E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 27.842 27.842 qs_mol_dyn_low 1 2.0 0.003 0.003 27.569 27.575 qs_forces 11 3.9 0.002 0.002 27.514 27.515 qs_energies 11 4.9 0.001 0.001 25.839 25.842 scf_env_do_scf 11 5.9 0.001 0.001 20.879 20.879 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.456 18.457 velocity_verlet 10 3.0 0.002 0.002 13.779 13.782 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.642 12.783 qs_scf_new_mos 108 7.5 0.001 0.001 11.103 11.130 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.102 11.129 multiply_cannon 2286 13.5 0.228 0.236 10.240 10.729 ot_scf_mini 108 9.5 0.002 0.002 10.440 10.470 multiply_cannon_loop 2286 14.5 0.332 0.344 9.354 9.615 multiply_cannon_multrec 9144 15.5 1.593 1.873 6.267 6.488 ot_mini 108 10.5 0.001 0.001 6.094 6.129 rebuild_ks_matrix 119 8.3 0.000 0.000 5.668 5.691 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.668 5.691 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.051 5.072 qs_ot_get_derivative 108 11.5 0.001 0.001 4.813 4.842 dbcsr_mm_accdrv_process 12550 15.8 3.546 4.596 4.572 4.643 init_scf_run 11 5.9 0.000 0.001 3.545 3.545 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.544 3.544 sum_up_and_integrate 119 10.3 0.038 0.042 3.447 3.452 integrate_v_rspace 119 11.3 0.003 0.003 3.409 3.414 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.093 3.098 calculate_rho_elec 119 8.7 0.059 0.061 3.093 3.097 qs_ot_get_p 119 10.4 0.001 0.001 2.602 2.657 calculate_first_density_matrix 1 7.0 0.000 0.000 2.463 2.463 init_scf_loop 11 6.9 0.000 0.000 2.403 2.404 mp_waitall_1 121218 16.5 1.562 2.097 1.562 2.097 calculate_dm_sparse 119 9.5 0.000 0.000 1.932 1.951 grid_integrate_task_list 119 12.3 1.853 1.924 1.853 1.924 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.901 1.916 make_m2s 4572 13.5 0.034 0.035 1.663 1.826 jit_kernel_multiply 8 15.7 0.988 1.798 0.988 1.798 make_images 4572 14.5 0.268 0.303 1.574 1.735 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.727 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.675 1.680 make_preconditioner 11 8.9 0.000 0.000 1.675 1.680 pw_transfer 1439 11.6 0.063 0.065 1.646 1.656 density_rs2pw 119 9.7 0.004 0.004 1.525 1.645 rs_pw_transfer 974 11.9 0.008 0.008 1.523 1.642 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.598 1.599 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.567 1.591 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.555 1.564 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.537 1.537 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.394 1.401 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.362 1.367 grid_collocate_task_list 119 9.7 1.272 1.354 1.272 1.354 fft_wrap_pw1pw2_140 487 13.2 0.082 0.087 1.282 1.290 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.272 1.273 ot_diis_step 108 11.5 0.012 0.013 1.270 1.270 cp_fm_redistribute_end 50 14.0 0.631 1.254 0.632 1.255 cp_fm_diag_elpa_base 50 14.0 0.578 1.188 0.622 1.244 potential_pw2rs 119 12.3 0.010 0.010 1.241 1.244 fft3d_ps 1201 14.6 0.537 0.550 1.223 1.231 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.213 1.214 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.151 1.167 apply_single 119 13.6 0.000 0.000 1.151 1.166 acc_transpose_blocks 9144 15.5 0.039 0.040 1.099 1.108 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.090 1.105 wfi_extrapolate 11 7.9 0.001 0.001 1.039 1.039 hybrid_alltoall_any 4725 16.4 0.062 0.177 0.759 0.998 make_images_data 4572 15.5 0.039 0.042 0.783 0.966 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.922 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.858 0.863 cp_fm_cholesky_invert 11 10.9 0.841 0.844 0.841 0.844 mp_alltoall_d11v 2130 13.8 0.739 0.838 0.739 0.838 multiply_cannon_sync_h2d 9144 15.5 0.712 0.801 0.712 0.801 jit_kernel_transpose 5 15.6 0.741 0.745 0.741 0.745 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.737 0.740 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.670 0.717 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.701 0.709 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.414 0.673 mp_allgather_i34 2286 14.5 0.218 0.628 0.218 0.628 mp_alltoall_z22v 1201 16.6 0.559 0.587 0.559 0.587 dbcsr_complete_redistribute 329 12.2 0.197 0.226 0.550 0.575 qs_create_task_list 11 7.9 0.006 0.006 0.550 0.572 generate_qs_task_list 11 8.9 0.191 0.215 0.544 0.566 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.842000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=565.363636, yerr=3.226953 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 756.047872E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 42.013 42.013 qs_mol_dyn_low 1 2.0 0.003 0.003 41.762 41.770 qs_forces 11 3.9 0.024 0.026 40.204 40.205 qs_energies 11 4.9 0.001 0.001 38.256 38.259 scf_env_do_scf 11 5.9 0.001 0.001 32.199 32.200 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.226 24.227 velocity_verlet 10 3.0 0.002 0.002 23.947 24.140 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.386 17.546 qs_scf_new_mos 108 7.5 0.001 0.001 15.520 15.614 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.519 15.613 multiply_cannon 2286 13.5 0.303 0.310 13.673 14.543 ot_scf_mini 108 9.5 0.002 0.002 14.441 14.540 multiply_cannon_loop 2286 14.5 0.344 0.351 12.410 13.289 ot_mini 108 10.5 0.001 0.001 8.656 8.771 multiply_cannon_multrec 9144 15.5 3.392 4.682 8.666 8.738 init_scf_loop 11 6.9 0.000 0.000 7.946 7.948 rebuild_ks_matrix 119 8.3 0.000 0.000 6.943 7.086 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.943 7.085 prepare_preconditioner 11 7.9 0.000 0.000 6.825 6.839 make_preconditioner 11 8.9 0.000 0.000 6.825 6.839 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.379 6.718 qs_ot_get_derivative 108 11.5 0.001 0.001 6.616 6.716 dbcsr_mm_accdrv_process 12550 15.8 4.197 5.761 5.149 6.480 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.291 6.420 cp_fm_upper_to_full 72 14.2 3.235 4.628 3.235 4.628 init_scf_run 11 5.9 0.000 0.001 4.055 4.055 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.054 4.054 sum_up_and_integrate 119 10.3 0.064 0.066 3.667 3.706 integrate_v_rspace 119 11.3 0.003 0.003 3.602 3.643 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.584 3.588 calculate_rho_elec 119 8.7 0.118 0.121 3.584 3.588 mp_waitall_1 97218 16.6 2.420 3.379 2.420 3.379 qs_ot_get_p 119 10.4 0.001 0.001 3.072 3.203 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.462 2.845 dbcsr_complete_redistribute 329 12.2 0.288 0.293 1.961 2.796 calculate_first_density_matrix 1 7.0 0.000 0.000 2.666 2.669 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.665 2.501 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.167 2.473 apply_single 119 13.6 0.000 0.000 2.167 2.473 make_m2s 4572 13.5 0.037 0.038 2.227 2.407 calculate_dm_sparse 119 9.5 0.000 0.000 2.313 2.329 mp_alltoall_i22 627 13.8 1.480 2.321 1.480 2.321 make_images 4572 14.5 0.354 0.383 2.107 2.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.441 2.270 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.301 2.147 pw_transfer 1439 11.6 0.066 0.066 2.046 2.087 ot_diis_step 108 11.5 0.014 0.014 2.017 2.017 grid_integrate_task_list 119 12.3 2.001 2.017 2.001 2.017 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.951 1.992 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.906 1.907 jit_kernel_multiply 10 15.5 0.926 1.878 0.926 1.878 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.842 1.843 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.741 1.744 density_rs2pw 119 9.7 0.003 0.003 1.697 1.720 fft_wrap_pw1pw2_140 487 13.2 0.088 0.089 1.654 1.699 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.639 1.689 mp_sum_l 7207 12.9 0.945 1.684 0.945 1.684 fft3d_ps 1201 14.6 0.565 0.576 1.585 1.626 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.591 1.617 mp_sum_dm 438 4.9 1.374 1.580 1.374 1.580 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.570 1.571 md_write_output 11 3.9 0.006 0.045 0.195 1.547 md_output 10 3.0 0.000 0.000 0.196 1.546 update_particle_set 20 4.0 0.000 0.000 1.350 1.543 write_trajectory 44 4.9 0.187 1.495 0.188 1.501 grid_collocate_task_list 119 9.7 1.445 1.455 1.445 1.455 cp_fm_cholesky_invert 11 10.9 1.411 1.414 1.411 1.414 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.380 1.387 hybrid_alltoall_any 4725 16.4 0.088 0.147 1.099 1.354 wfi_extrapolate 11 7.9 0.001 0.001 1.332 1.332 rs_pw_transfer 974 11.9 0.009 0.009 1.295 1.326 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.293 1.294 cp_fm_diag_elpa_base 50 14.0 1.146 1.199 1.292 1.292 make_images_data 4572 15.5 0.042 0.046 1.048 1.279 potential_pw2rs 119 12.3 0.014 0.014 1.223 1.265 mp_alltoall_d11v 2130 13.8 1.114 1.138 1.114 1.138 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.119 1.137 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.080 1.093 multiply_cannon_sync_h2d 9144 15.5 1.042 1.045 1.042 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.981 1.025 qs_create_task_list 11 7.9 0.000 0.001 0.938 0.950 generate_qs_task_list 11 8.9 0.374 0.394 0.938 0.949 mp_alltoall_z22v 1201 16.6 0.884 0.938 0.884 0.938 acc_transpose_blocks 9144 15.5 0.038 0.039 0.908 0.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.013000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.000000, yerr=13.810405 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.895168E+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 2470249. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.101 82.399 82.407 qs_mol_dyn_low 1 2.0 0.005 0.020 81.990 82.001 qs_forces 11 3.9 0.002 0.002 81.822 81.823 qs_energies 11 4.9 0.001 0.002 79.013 79.035 scf_env_do_scf 11 5.9 0.000 0.001 69.851 69.854 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.409 64.409 dbcsr_multiply_generic 2055 12.4 0.105 0.108 52.031 52.304 qs_scf_new_mos 99 7.5 0.000 0.001 47.736 47.839 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.735 47.839 ot_scf_mini 99 9.5 0.002 0.002 45.352 45.476 multiply_cannon 2055 13.4 0.180 0.187 42.593 43.315 multiply_cannon_loop 2055 14.4 1.554 1.594 41.655 42.423 velocity_verlet 10 3.0 0.001 0.002 41.558 41.559 ot_mini 99 10.5 0.001 0.001 27.671 27.806 qs_ot_get_derivative 99 11.5 0.001 0.001 20.803 20.903 multiply_cannon_multrec 49320 15.4 12.196 12.952 17.218 17.775 rebuild_ks_matrix 110 8.3 0.000 0.000 14.285 14.415 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.285 14.414 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.549 12.673 mp_waitall_1 241148 16.1 11.645 12.451 11.645 12.451 multiply_cannon_sync_h2d 49320 15.4 10.020 10.620 10.020 10.620 qs_ot_get_p 110 10.4 0.001 0.001 9.211 9.329 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.014 8.184 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.189 7.747 apply_single 110 13.6 0.000 0.000 7.188 7.746 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.142 7.666 init_scf_run 11 5.9 0.000 0.001 7.131 7.132 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.131 7.132 sum_up_and_integrate 110 10.3 0.037 0.044 6.903 6.917 integrate_v_rspace 110 11.3 0.003 0.003 6.866 6.887 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.638 6.718 ot_diis_step 99 11.5 0.006 0.006 6.574 6.574 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.058 6.200 calculate_rho_elec 110 8.6 0.020 0.024 6.058 6.199 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.035 6.069 mp_sum_l 6514 12.8 5.178 6.015 5.178 6.015 init_scf_loop 11 6.9 0.000 0.000 5.415 5.416 dbcsr_mm_accdrv_process 87628 16.1 2.066 2.173 4.901 5.201 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.039 5.040 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.568 4.569 cp_fm_redistribute_end 48 14.0 3.975 4.544 3.979 4.545 cp_fm_diag_elpa_base 48 14.0 0.560 4.486 0.564 4.511 wfi_extrapolate 11 7.9 0.001 0.001 3.975 3.975 make_m2s 4110 13.4 0.060 0.065 3.856 3.968 make_images 4110 14.4 0.178 0.191 3.761 3.877 calculate_dm_sparse 110 9.5 0.000 0.001 3.749 3.840 rs_pw_transfer 902 11.9 0.012 0.013 3.674 3.794 cp_dbcsr_sm_fm_multiply 37 9.5 0.004 0.021 3.697 3.701 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.603 3.657 grid_integrate_task_list 110 12.3 3.224 3.376 3.224 3.376 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.300 3.338 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.241 3.297 density_rs2pw 110 9.6 0.004 0.004 3.149 3.278 prepare_preconditioner 11 7.9 0.000 0.000 3.241 3.270 make_preconditioner 11 8.9 0.000 0.000 3.241 3.270 calculate_first_density_matrix 1 7.0 0.000 0.001 3.069 3.076 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.010 3.055 pw_transfer 1331 11.6 0.053 0.061 2.722 2.785 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.635 2.701 jit_kernel_multiply 13 15.9 2.556 2.570 2.556 2.570 potential_pw2rs 110 12.3 0.006 0.007 2.519 2.544 mp_alltoall_d11v 2046 13.8 2.105 2.544 2.105 2.544 fft_wrap_pw1pw2_140 451 13.1 0.168 0.187 2.229 2.296 acc_transpose_blocks 49320 15.4 0.212 0.222 2.178 2.248 fft3d_ps 1111 14.6 0.750 0.837 2.175 2.230 mp_sum_d 3883 11.9 1.504 2.187 1.504 2.187 grid_collocate_task_list 110 9.6 2.089 2.182 2.089 2.182 mp_waitany 14300 13.8 1.841 2.104 1.841 2.104 make_images_data 4110 15.4 0.043 0.046 1.766 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.875 1.889 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.530 1.788 cp_fm_cholesky_invert 11 10.9 1.755 1.759 1.755 1.759 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.643 1.674 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.407000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.489648 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 586.919936E+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 2759861. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.061 68.270 68.271 qs_mol_dyn_low 1 2.0 0.003 0.004 67.914 67.923 qs_forces 11 3.9 0.002 0.002 67.850 67.851 qs_energies 11 4.9 0.002 0.007 64.522 64.525 scf_env_do_scf 11 5.9 0.001 0.002 55.467 55.470 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 47.945 47.945 dbcsr_multiply_generic 2055 12.4 0.113 0.116 37.891 38.044 velocity_verlet 10 3.0 0.002 0.002 35.574 35.575 multiply_cannon 2055 13.4 0.221 0.245 31.423 32.588 qs_scf_new_mos 99 7.5 0.001 0.001 32.257 32.410 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.256 32.409 multiply_cannon_loop 2055 14.4 0.929 0.954 30.123 31.000 ot_scf_mini 99 9.5 0.003 0.004 30.611 30.769 ot_mini 99 10.5 0.001 0.001 18.059 18.212 multiply_cannon_multrec 24660 15.4 7.665 9.756 14.155 16.140 rebuild_ks_matrix 110 8.3 0.000 0.001 13.555 13.684 qs_ks_build_kohn_sham_matrix 110 9.3 0.017 0.049 13.554 13.684 qs_ot_get_derivative 99 11.5 0.001 0.001 12.226 12.382 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.921 12.038 mp_waitall_1 186928 16.3 7.842 10.275 7.842 10.275 multiply_cannon_sync_h2d 24660 15.4 7.059 7.918 7.059 7.918 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.198 7.615 init_scf_loop 11 6.9 0.001 0.003 7.485 7.485 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.555 7.182 apply_single 110 13.6 0.000 0.001 6.555 7.182 init_scf_run 11 5.9 0.000 0.001 6.647 6.648 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.647 6.648 dbcsr_mm_accdrv_process 52282 16.1 4.709 5.782 6.330 6.633 sum_up_and_integrate 110 10.3 0.052 0.060 6.446 6.456 integrate_v_rspace 110 11.3 0.002 0.002 6.393 6.408 qs_ot_get_p 110 10.4 0.001 0.001 5.936 6.126 ot_diis_step 99 11.5 0.010 0.010 5.786 5.786 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.685 5.698 calculate_rho_elec 110 8.6 0.039 0.047 5.685 5.697 prepare_preconditioner 11 7.9 0.000 0.000 5.470 5.494 make_preconditioner 11 8.9 0.000 0.001 5.470 5.494 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.598 5.308 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.035 5.191 make_m2s 4110 13.4 0.056 0.059 4.124 4.583 make_images 4110 14.4 0.398 0.443 4.016 4.474 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.113 4.133 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.687 3.688 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.544 3.545 wfi_extrapolate 11 7.9 0.001 0.001 3.451 3.451 pw_transfer 1331 11.6 0.065 0.070 3.219 3.354 grid_integrate_task_list 110 12.3 3.161 3.301 3.161 3.301 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.113 3.246 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.137 3.218 calculate_dm_sparse 110 9.5 0.001 0.001 3.174 3.209 density_rs2pw 110 9.6 0.004 0.004 2.977 3.198 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.142 3.188 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.129 3.130 rs_pw_transfer 902 11.9 0.012 0.014 2.925 3.121 calculate_first_density_matrix 1 7.0 0.003 0.013 3.110 3.113 cp_fm_redistribute_end 48 14.0 2.331 3.105 2.332 3.106 cp_fm_diag_elpa_base 48 14.0 0.739 2.996 0.770 3.083 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.047 2.805 make_images_data 4110 15.4 0.047 0.051 2.326 2.805 fft_wrap_pw1pw2_140 451 13.1 0.200 0.219 2.643 2.777 mp_sum_l 6514 12.8 1.734 2.607 1.734 2.607 fft3d_ps 1111 14.6 1.069 1.263 2.471 2.587 cp_fm_cholesky_invert 11 10.9 2.522 2.528 2.522 2.528 jit_kernel_multiply 11 16.4 1.273 2.399 1.273 2.399 potential_pw2rs 110 12.3 0.008 0.009 2.306 2.323 grid_collocate_task_list 110 9.6 2.121 2.304 2.121 2.304 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.948 1.974 mp_alltoall_d11v 2046 13.8 1.785 1.971 1.785 1.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.888 1.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.797 1.809 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.758 1.768 mp_allgather_i34 2055 14.4 0.610 1.664 0.610 1.664 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.776 1.638 acc_transpose_blocks 24660 15.4 0.106 0.110 1.509 1.541 mp_irecv_dv 57340 16.2 0.653 1.512 0.653 1.512 dbcsr_complete_redistribute 325 12.2 0.280 0.353 1.209 1.482 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.476 mp_waitany 10164 13.8 1.231 1.415 1.231 1.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.271000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.454545, yerr=6.624522 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 657.727488E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.180 63.245 63.252 qs_mol_dyn_low 1 2.0 0.003 0.004 61.990 62.016 qs_forces 11 3.9 0.055 0.424 60.375 60.375 qs_energies 11 4.9 0.014 0.017 56.810 56.815 scf_env_do_scf 11 5.9 0.001 0.001 47.734 47.734 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 38.948 38.948 velocity_verlet 10 3.0 0.002 0.002 33.408 33.458 dbcsr_multiply_generic 2055 12.4 0.110 0.124 28.536 28.786 qs_scf_new_mos 99 7.5 0.001 0.001 24.617 24.712 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.616 24.711 ot_scf_mini 99 9.5 0.002 0.003 23.409 23.515 multiply_cannon 2055 13.4 0.212 0.222 22.323 23.513 multiply_cannon_loop 2055 14.4 0.619 0.632 21.111 21.980 ot_mini 99 10.5 0.001 0.001 13.328 13.438 rebuild_ks_matrix 110 8.3 0.000 0.000 12.096 12.234 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.096 12.234 multiply_cannon_multrec 16440 15.4 3.958 4.938 10.094 11.010 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.674 10.795 mp_waitall_1 146766 16.3 6.836 9.590 6.836 9.590 qs_ot_get_derivative 99 11.5 0.001 0.001 8.941 9.048 init_scf_loop 11 6.9 0.000 0.002 8.752 8.753 prepare_preconditioner 11 7.9 0.000 0.000 6.980 6.999 make_preconditioner 11 8.9 0.036 0.290 6.980 6.999 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.147 6.771 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.259 6.674 sum_up_and_integrate 110 10.3 0.060 0.061 6.358 6.380 integrate_v_rspace 110 11.3 0.002 0.003 6.298 6.320 dbcsr_mm_accdrv_process 34862 16.1 5.018 5.564 5.989 6.187 init_scf_run 11 5.9 0.000 0.001 6.033 6.033 scf_env_initial_rho_setup 11 6.9 0.025 0.030 6.032 6.033 qs_ot_get_p 110 10.4 0.001 0.001 5.326 5.468 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.450 5.459 calculate_rho_elec 110 8.6 0.058 0.058 5.450 5.459 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.922 5.326 apply_single 110 13.6 0.000 0.000 4.922 5.326 make_m2s 4110 13.4 0.049 0.051 4.130 4.524 multiply_cannon_sync_h2d 16440 15.4 3.705 4.418 3.705 4.418 make_images 4110 14.4 0.395 0.508 4.016 4.409 ot_diis_step 99 11.5 0.011 0.011 4.359 4.359 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.742 3.747 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.053 3.698 rs_pw_transfer 902 11.9 0.010 0.012 3.146 3.463 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.411 3.412 grid_integrate_task_list 110 12.3 3.214 3.356 3.214 3.356 pw_transfer 1331 11.6 0.064 0.070 3.141 3.158 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.035 3.053 calculate_first_density_matrix 1 7.0 0.000 0.000 3.025 3.050 density_rs2pw 110 9.6 0.004 0.004 2.729 2.981 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.914 2.916 wfi_extrapolate 11 7.9 0.001 0.001 2.904 2.904 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.899 2.901 calculate_dm_sparse 110 9.5 0.001 0.001 2.863 2.894 cp_fm_redistribute_end 48 14.0 1.801 2.876 1.803 2.876 make_images_data 4110 15.4 0.046 0.068 2.347 2.863 cp_fm_diag_elpa_base 48 14.0 1.012 2.745 1.068 2.853 hybrid_alltoall_any 4261 16.3 0.105 0.378 2.142 2.800 fft_wrap_pw1pw2_140 451 13.1 0.209 0.212 2.611 2.628 cp_fm_cholesky_invert 11 10.9 2.600 2.607 2.600 2.607 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.605 2.605 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.554 2.604 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.859 2.392 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.301 2.367 fft3d_ps 1111 14.6 1.065 1.072 2.350 2.365 grid_collocate_task_list 110 9.6 2.178 2.340 2.178 2.340 mp_irecv_dv 48980 15.7 0.789 2.272 0.789 2.272 potential_pw2rs 110 12.3 0.011 0.011 2.177 2.192 mp_sum_l 6514 12.8 1.550 2.167 1.550 2.167 mp_sum_dm 438 4.9 1.951 2.071 1.951 2.071 mp_alltoall_d11v 2046 13.8 1.746 2.017 1.746 2.017 mp_waitany 17072 13.8 1.679 1.972 1.679 1.972 dbcsr_complete_redistribute 325 12.2 0.332 0.367 1.468 1.917 jit_kernel_multiply 8 16.8 0.586 1.881 0.586 1.881 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.837 1.850 cp_fm_upper_to_full 70 13.6 1.368 1.835 1.368 1.835 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.638 1.789 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.177 1.395 1.731 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.639 md_output 10 3.0 0.000 0.000 0.054 1.619 md_write_output 11 3.9 0.050 1.579 0.052 1.618 update_particle_set 20 4.0 0.000 0.000 1.545 1.596 mp_allgather_i34 2055 14.4 0.535 1.582 0.535 1.582 cp_fm_cholesky_decompose 22 10.9 1.557 1.579 1.557 1.579 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.355 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.010 1.454 build_core_hamiltonian_matrix 11 6.9 0.001 0.004 0.763 1.290 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.252000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.363636, yerr=7.946099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 732.065792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 66.387 66.388 qs_mol_dyn_low 1 2.0 0.003 0.003 65.833 65.842 qs_forces 11 3.9 0.002 0.002 65.053 65.059 qs_energies 11 4.9 0.001 0.002 61.704 61.712 scf_env_do_scf 11 5.9 0.000 0.001 52.464 52.467 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.720 40.721 velocity_verlet 10 3.0 0.002 0.002 37.110 37.142 dbcsr_multiply_generic 2055 12.4 0.118 0.120 30.053 30.310 qs_scf_new_mos 99 7.5 0.001 0.001 26.549 26.648 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.548 26.647 ot_scf_mini 99 9.5 0.003 0.003 24.932 25.034 multiply_cannon 2055 13.4 0.243 0.261 23.091 24.546 multiply_cannon_loop 2055 14.4 0.883 0.902 21.735 22.425 ot_mini 99 10.5 0.001 0.001 14.330 14.448 multiply_cannon_multrec 24660 15.4 4.227 6.750 13.165 14.408 rebuild_ks_matrix 110 8.3 0.000 0.000 11.765 11.875 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.764 11.874 init_scf_loop 11 6.9 0.000 0.000 11.702 11.703 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.415 10.511 qs_ot_get_derivative 99 11.5 0.001 0.001 10.190 10.288 prepare_preconditioner 11 7.9 0.000 0.000 9.985 10.000 make_preconditioner 11 8.9 0.000 0.000 9.985 10.000 dbcsr_mm_accdrv_process 52304 16.0 6.813 8.811 8.791 9.683 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.162 9.666 init_scf_run 11 5.9 0.000 0.001 6.531 6.532 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.531 6.531 mp_waitall_1 126806 16.4 4.594 6.378 4.594 6.378 sum_up_and_integrate 110 10.3 0.068 0.072 6.226 6.239 integrate_v_rspace 110 11.3 0.003 0.003 6.157 6.169 qs_ot_get_p 110 10.4 0.001 0.001 5.644 5.791 make_m2s 4110 13.4 0.058 0.060 5.327 5.602 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.479 5.487 calculate_rho_elec 110 8.6 0.077 0.081 5.478 5.486 make_images 4110 14.4 0.575 0.695 5.187 5.458 cp_fm_upper_to_full 70 13.8 3.347 4.772 3.347 4.772 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.012 4.117 apply_single 110 13.6 0.000 0.000 4.012 4.117 ot_diis_step 99 11.5 0.011 0.011 4.104 4.104 dbcsr_complete_redistribute 325 12.2 0.490 0.531 2.793 3.945 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.857 3.871 calculate_first_density_matrix 1 7.0 0.000 0.000 3.524 3.527 grid_integrate_task_list 110 12.3 3.270 3.425 3.270 3.425 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.250 3.397 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.391 3.392 multiply_cannon_sync_h2d 24660 15.4 3.186 3.385 3.186 3.385 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.383 3.384 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.284 3.331 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.594 3.259 pw_transfer 1331 11.6 0.064 0.072 3.119 3.145 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.015 3.043 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.973 3.024 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.957 2.997 make_images_data 4110 15.4 0.046 0.050 2.671 2.979 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.243 2.947 calculate_dm_sparse 110 9.5 0.001 0.001 2.903 2.941 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.814 2.940 wfi_extrapolate 11 7.9 0.001 0.001 2.920 2.920 mp_alltoall_i22 605 13.7 1.684 2.894 1.684 2.894 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.845 2.846 cp_fm_redistribute_end 48 14.0 1.415 2.816 1.416 2.817 cp_fm_diag_elpa_base 48 14.0 1.318 2.686 1.398 2.798 density_rs2pw 110 9.6 0.004 0.004 2.669 2.783 cp_fm_cholesky_invert 11 10.9 2.679 2.688 2.679 2.688 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.559 2.589 rs_pw_transfer 902 11.9 0.010 0.011 2.352 2.518 jit_kernel_multiply 10 15.7 1.650 2.364 1.650 2.364 grid_collocate_task_list 110 9.6 2.221 2.359 2.221 2.359 fft3d_ps 1111 14.6 1.058 1.095 2.318 2.331 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.225 2.231 potential_pw2rs 110 12.3 0.012 0.013 2.034 2.045 mp_alltoall_d11v 2046 13.8 1.751 2.021 1.751 2.021 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.689 1.719 cp_fm_cholesky_decompose 22 10.9 1.659 1.708 1.659 1.708 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.603 1.702 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.593 1.605 mp_allgather_i34 2055 14.4 0.451 1.557 0.451 1.557 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.858 1.547 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.523 1.538 acc_transpose_blocks 24660 15.4 0.104 0.106 1.493 1.510 mp_irecv_dv 62702 16.1 0.760 1.473 0.760 1.473 mp_sum_l 6514 12.8 0.943 1.449 0.943 1.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.388000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.727273, yerr=11.338518 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 843.083776E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.047 55.230 55.230 qs_mol_dyn_low 1 2.0 0.003 0.003 54.954 54.962 qs_forces 11 3.9 0.002 0.002 54.861 54.861 qs_energies 11 4.9 0.019 0.044 51.258 51.263 scf_env_do_scf 11 5.9 0.000 0.001 42.477 42.478 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.828 34.829 velocity_verlet 10 3.0 0.002 0.002 30.267 30.272 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.966 23.119 qs_scf_new_mos 99 7.5 0.001 0.001 20.819 20.868 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.818 20.868 ot_scf_mini 99 9.5 0.002 0.002 19.586 19.627 multiply_cannon 2055 13.4 0.250 0.266 17.601 18.803 multiply_cannon_loop 2055 14.4 0.322 0.333 16.258 16.579 rebuild_ks_matrix 110 8.3 0.000 0.000 11.387 11.426 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.387 11.426 ot_mini 99 10.5 0.001 0.001 10.423 10.463 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.145 10.180 multiply_cannon_multrec 8220 15.4 3.324 4.932 7.596 8.738 mp_waitall_1 106626 16.5 6.061 7.665 6.061 7.665 init_scf_loop 11 6.9 0.000 0.000 7.600 7.601 qs_ot_get_derivative 99 11.5 0.001 0.001 6.577 6.620 sum_up_and_integrate 110 10.3 0.079 0.081 6.147 6.158 integrate_v_rspace 110 11.3 0.003 0.003 6.068 6.078 prepare_preconditioner 11 7.9 0.000 0.000 5.989 5.996 make_preconditioner 11 8.9 0.000 0.000 5.989 5.995 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.575 5.646 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.562 5.576 calculate_rho_elec 110 8.6 0.115 0.115 5.562 5.576 qs_ot_get_p 110 10.4 0.001 0.001 5.253 5.310 init_scf_run 11 5.9 0.000 0.001 5.298 5.298 scf_env_initial_rho_setup 11 6.9 0.018 0.050 5.297 5.297 dbcsr_mm_accdrv_process 17442 15.9 2.753 3.637 4.141 5.179 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.045 4.485 make_m2s 4110 13.4 0.038 0.040 4.114 4.352 make_images 4110 14.4 0.636 0.690 3.984 4.222 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.887 3.891 ot_diis_step 99 11.5 0.012 0.012 3.824 3.824 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.731 3.784 apply_single 110 13.6 0.000 0.000 3.730 3.784 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.579 3.580 grid_integrate_task_list 110 12.3 3.370 3.468 3.370 3.468 pw_transfer 1331 11.6 0.065 0.069 3.212 3.224 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.105 3.121 multiply_cannon_sync_h2d 8220 15.4 2.908 3.018 2.908 3.018 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.002 3.003 qs_energies_init_hamiltonians 11 5.9 0.343 0.349 2.986 2.987 cp_fm_redistribute_end 48 14.0 0.758 2.966 0.763 2.967 cp_fm_diag_elpa_base 48 14.0 2.021 2.765 2.197 2.933 cp_fm_cholesky_invert 11 10.9 2.896 2.900 2.896 2.900 density_rs2pw 110 9.6 0.004 0.004 2.617 2.796 make_images_data 4110 15.4 0.038 0.043 2.334 2.726 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.713 2.715 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.659 2.678 wfi_extrapolate 11 7.9 0.001 0.001 2.647 2.647 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.273 2.626 calculate_first_density_matrix 1 7.0 0.000 0.000 2.548 2.567 calculate_dm_sparse 110 9.5 0.001 0.001 2.455 2.501 grid_collocate_task_list 110 9.6 2.329 2.459 2.329 2.459 fft3d_ps 1111 14.6 1.118 1.137 2.362 2.373 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.326 2.341 rs_pw_transfer 902 11.9 0.010 0.010 2.139 2.296 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.775 1.998 potential_pw2rs 110 12.3 0.015 0.016 1.899 1.903 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.875 1.894 mp_alltoall_d11v 2046 13.8 1.631 1.853 1.631 1.853 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.821 1.845 jit_kernel_multiply 10 15.8 1.081 1.751 1.081 1.751 cp_fm_cholesky_decompose 22 10.9 1.662 1.677 1.662 1.677 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.491 1.610 mp_allgather_i34 2055 14.4 0.527 1.585 0.527 1.585 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.574 1.578 dbcsr_complete_redistribute 325 12.2 0.582 0.615 1.482 1.574 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.424 1.431 qs_create_task_list 11 7.9 0.000 0.001 1.219 1.317 generate_qs_task_list 11 8.9 0.381 0.450 1.218 1.317 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.477 1.256 mp_irecv_dv 24056 15.7 0.453 1.216 0.453 1.216 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.745 1.209 mp_waitany 9240 13.8 1.044 1.203 1.044 1.203 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.142 1.172 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.230000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=791.454545, yerr=10.849077 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 1.376612E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.039 86.216 86.217 qs_mol_dyn_low 1 2.0 0.003 0.003 85.925 85.933 qs_forces 11 3.9 0.002 0.002 85.854 85.855 qs_energies 11 4.9 0.001 0.002 81.814 81.816 scf_env_do_scf 11 5.9 0.001 0.001 71.541 71.541 velocity_verlet 10 3.0 0.005 0.006 54.792 54.798 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.866 42.868 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.711 28.762 init_scf_loop 11 6.9 0.000 0.000 28.602 28.605 qs_scf_new_mos 99 7.5 0.001 0.001 26.520 26.558 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.520 26.557 prepare_preconditioner 11 7.9 0.000 0.000 26.464 26.470 make_preconditioner 11 8.9 0.000 0.000 26.464 26.470 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.622 25.934 ot_scf_mini 99 9.5 0.002 0.002 24.757 24.783 multiply_cannon 2055 13.4 0.336 0.353 21.608 22.322 multiply_cannon_loop 2055 14.4 0.342 0.343 19.813 20.126 cp_fm_upper_to_full 70 14.2 12.761 18.318 12.761 18.318 ot_mini 99 10.5 0.001 0.001 13.646 13.673 rebuild_ks_matrix 110 8.3 0.000 0.001 13.318 13.361 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.318 13.361 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.133 12.172 dbcsr_complete_redistribute 325 12.2 1.030 1.063 7.454 10.633 multiply_cannon_multrec 8220 15.4 4.371 4.557 9.728 9.906 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.404 9.585 qs_ot_get_derivative 99 11.5 0.001 0.001 9.065 9.090 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.826 8.972 mp_waitall_1 87304 16.6 7.874 8.809 7.874 8.809 mp_alltoall_i22 605 13.7 5.440 8.631 5.440 8.631 sum_up_and_integrate 110 10.3 0.151 0.152 6.611 6.634 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.451 6.486 calculate_rho_elec 110 8.6 0.227 0.227 6.450 6.486 integrate_v_rspace 110 11.3 0.003 0.004 6.460 6.483 init_scf_run 11 5.9 0.000 0.001 6.049 6.049 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.048 6.048 qs_ot_get_p 110 10.4 0.001 0.001 5.908 5.930 make_m2s 4110 13.4 0.043 0.043 5.384 5.899 make_images 4110 14.4 0.876 0.927 5.196 5.711 dbcsr_mm_accdrv_process 11614 15.7 3.336 3.782 5.215 5.540 cp_fm_cholesky_invert 11 10.9 5.338 5.343 5.338 5.343 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.744 5.204 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.666 5.089 apply_single 110 13.6 0.000 0.000 4.666 5.089 ot_diis_step 99 11.5 0.015 0.015 4.558 4.558 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.344 4.351 multiply_cannon_sync_h2d 8220 15.4 3.949 3.954 3.949 3.954 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.903 3.903 pw_transfer 1331 11.6 0.073 0.074 3.772 3.789 hybrid_alltoall_any 4261 16.3 0.258 0.551 2.999 3.751 grid_integrate_task_list 110 12.3 3.653 3.707 3.653 3.707 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.656 3.673 make_images_data 4110 15.4 0.041 0.044 3.011 3.658 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.625 3.627 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.010 3.432 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.290 3.291 calculate_dm_sparse 110 9.5 0.001 0.001 3.272 3.290 cp_fm_diag_elpa_base 48 14.0 2.750 2.947 3.288 3.289 wfi_extrapolate 11 7.9 0.001 0.001 3.231 3.231 fft_wrap_pw1pw2_140 451 13.1 0.215 0.221 3.167 3.187 density_rs2pw 110 9.6 0.004 0.004 2.953 2.972 fft3d_ps 1111 14.6 1.272 1.292 2.865 2.885 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.844 2.846 calculate_first_density_matrix 1 7.0 0.000 0.000 2.710 2.711 grid_collocate_task_list 110 9.6 2.623 2.649 2.623 2.649 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.402 2.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.336 2.349 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.162 2.219 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.194 rs_pw_transfer 902 11.9 0.010 0.011 2.083 2.121 mp_alltoall_d11v 2046 13.8 2.045 2.092 2.045 2.092 potential_pw2rs 110 12.3 0.021 0.021 2.058 2.079 cp_fm_cholesky_decompose 22 10.9 1.988 2.014 1.988 2.014 jit_kernel_multiply 10 15.3 1.681 1.989 1.681 1.989 qs_create_task_list 11 7.9 0.000 0.000 1.902 1.945 generate_qs_task_list 11 8.9 0.743 0.796 1.902 1.944 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.806 1.822 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.784 1.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.217000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1222.909091, yerr=71.670273 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.347456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 1586820. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.086 203.875 203.876 qs_mol_dyn_low 1 2.0 0.007 0.032 203.297 203.310 qs_forces 11 3.9 0.004 0.006 202.356 202.359 qs_energies 11 4.9 0.023 0.025 196.831 196.843 scf_env_do_scf 11 5.9 0.001 0.002 179.519 179.523 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 158.531 158.533 dbcsr_multiply_generic 2507 12.6 0.176 0.180 123.016 123.777 velocity_verlet 10 3.0 0.001 0.003 122.348 122.357 qs_scf_new_mos 117 7.6 0.001 0.001 120.023 120.385 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.022 120.385 ot_scf_mini 117 9.6 0.003 0.003 113.432 113.856 multiply_cannon 2507 13.6 0.239 0.247 100.569 102.197 multiply_cannon_loop 2507 14.6 2.040 2.118 98.415 99.970 ot_mini 117 10.6 0.001 0.001 64.779 65.146 multiply_cannon_multrec 60168 15.6 33.620 35.825 41.827 43.970 qs_ot_get_derivative 117 11.6 0.001 0.001 39.927 40.307 rebuild_ks_matrix 128 8.3 0.001 0.001 33.192 33.631 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.053 33.191 33.630 mp_waitall_1 291448 16.2 27.573 30.851 27.573 30.851 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.829 30.238 multiply_cannon_sync_h2d 60168 15.6 27.872 30.134 27.872 30.134 qs_ot_get_p 128 10.4 0.001 0.001 27.309 27.764 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.261 24.928 apply_single 128 13.6 0.001 0.001 24.261 24.928 ot_diis_step 117 11.6 0.007 0.008 24.536 24.537 init_scf_loop 11 6.9 0.000 0.000 20.910 20.911 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.373 20.410 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.214 18.497 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.004 18.005 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 14.979 17.296 prepare_preconditioner 11 7.9 0.000 0.000 16.023 16.075 make_preconditioner 11 8.9 0.000 0.000 16.023 16.075 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.258 15.465 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.182 15.188 cp_fm_redistribute_end 83 14.4 11.961 15.105 11.974 15.110 cp_fm_diag_elpa_base 83 14.4 3.088 14.873 3.122 15.000 make_m2s 5014 13.6 0.103 0.111 13.737 14.182 make_images 5014 14.6 0.405 0.422 13.557 14.012 sum_up_and_integrate 128 10.3 0.091 0.108 13.898 13.940 integrate_v_rspace 128 11.3 0.003 0.004 13.807 13.853 init_scf_run 11 5.9 0.000 0.001 13.075 13.075 scf_env_initial_rho_setup 11 6.9 0.005 0.033 13.075 13.075 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.538 12.647 calculate_rho_elec 128 8.7 0.045 0.063 12.538 12.646 cp_fm_cholesky_invert 11 10.9 9.229 9.237 9.229 9.237 wfi_extrapolate 11 7.9 0.001 0.001 8.988 8.988 mp_sum_l 7870 13.0 7.446 8.931 7.446 8.931 calculate_dm_sparse 128 9.5 0.001 0.001 8.642 8.720 multiply_cannon_metrocomm1 60168 15.6 0.086 0.091 5.777 8.432 dbcsr_mm_accdrv_process 124484 16.2 3.045 3.302 7.753 8.358 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.924 8.086 make_images_data 5014 15.6 0.069 0.074 6.704 7.718 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.592 7.698 grid_integrate_task_list 128 12.3 7.057 7.440 7.057 7.440 pw_transfer 1547 11.6 0.074 0.106 6.786 7.083 hybrid_alltoall_any 5200 16.5 0.291 2.260 5.859 7.068 density_rs2pw 128 9.7 0.006 0.007 6.398 6.962 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.584 6.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.686 6.695 rs_pw_transfer 1046 11.9 0.017 0.019 5.651 6.211 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.809 5.920 fft_wrap_pw1pw2_140 523 13.2 0.444 0.508 5.646 5.836 fft3d_ps 1291 14.7 2.093 2.595 5.378 5.592 mp_alltoall_d11v 2415 14.1 4.130 5.375 4.130 5.375 grid_collocate_task_list 128 9.7 4.726 5.133 4.726 5.133 cp_fm_cholesky_decompose 22 10.9 4.615 4.629 4.615 4.629 mp_sum_d 4473 12.2 3.857 4.535 3.857 4.535 potential_pw2rs 128 12.3 0.009 0.010 4.425 4.479 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.876000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.909091, yerr=6.625770 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 828.796928E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 5621117. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.078 190.761 190.763 qs_mol_dyn_low 1 2.0 0.003 0.003 190.093 190.106 qs_forces 11 3.9 0.003 0.003 189.963 189.964 qs_energies 11 4.9 0.005 0.033 183.286 183.296 scf_env_do_scf 11 5.9 0.001 0.002 166.148 166.159 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.584 133.586 velocity_verlet 10 3.0 0.010 0.023 119.342 119.345 dbcsr_multiply_generic 2527 12.6 0.186 0.190 97.879 99.017 qs_scf_new_mos 118 7.6 0.001 0.001 94.940 95.368 qs_scf_loop_do_ot 118 8.6 0.001 0.002 94.939 95.367 ot_scf_mini 118 9.6 0.004 0.005 90.142 90.639 multiply_cannon 2527 13.6 0.481 0.531 78.468 82.306 multiply_cannon_loop 2527 14.6 1.265 1.300 75.216 77.417 ot_mini 118 10.6 0.001 0.001 49.909 50.339 mp_waitall_1 228564 16.4 23.960 37.030 23.960 37.030 multiply_cannon_multrec 30324 15.6 22.441 26.415 32.343 36.618 rebuild_ks_matrix 129 8.3 0.001 0.001 32.720 33.182 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 32.720 33.182 init_scf_loop 11 6.9 0.001 0.003 32.476 32.477 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.515 29.941 qs_ot_get_derivative 118 11.6 0.001 0.002 27.860 28.366 prepare_preconditioner 11 7.9 0.000 0.000 28.151 28.214 make_preconditioner 11 8.9 0.000 0.002 28.151 28.214 multiply_cannon_metrocomm3 30324 15.6 0.098 0.103 15.393 28.199 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.823 27.333 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.071 23.172 apply_single 129 13.6 0.001 0.001 22.071 23.172 qs_ot_get_p 129 10.4 0.001 0.001 21.935 22.475 ot_diis_step 118 11.6 0.014 0.015 21.876 21.878 multiply_cannon_sync_h2d 30324 15.6 19.665 21.663 19.665 21.663 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.205 17.241 cp_fm_cholesky_invert 11 10.9 16.408 16.420 16.408 16.420 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.066 16.067 make_m2s 5054 13.6 0.089 0.094 13.794 15.329 make_images 5054 14.6 1.161 1.352 13.585 15.122 sum_up_and_integrate 129 10.3 0.118 0.133 14.290 14.320 integrate_v_rspace 129 11.3 0.003 0.004 14.172 14.204 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.962 12.971 cp_fm_redistribute_end 83 14.4 7.536 12.894 7.546 12.895 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.828 12.859 calculate_rho_elec 129 8.7 0.088 0.106 12.828 12.858 cp_fm_diag_elpa_base 83 14.4 5.114 12.434 5.331 12.805 init_scf_run 11 5.9 0.000 0.001 12.247 12.249 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.247 12.249 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.800 11.152 multiply_cannon_metrocomm4 27797 15.6 0.098 0.114 3.745 10.236 dbcsr_mm_accdrv_process 62734 16.2 4.695 5.579 9.350 9.995 make_images_data 5054 15.6 0.067 0.074 8.150 9.928 mp_irecv_dv 70031 16.3 3.550 9.849 3.550 9.849 hybrid_alltoall_any 5240 16.5 0.345 1.529 7.047 9.481 wfi_extrapolate 11 7.9 0.001 0.001 8.294 8.294 pw_transfer 1559 11.6 0.085 0.102 7.674 7.745 grid_integrate_task_list 129 12.3 7.230 7.593 7.230 7.593 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.449 7.517 density_rs2pw 129 9.7 0.006 0.007 6.636 7.192 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.243 6.980 calculate_dm_sparse 129 9.5 0.001 0.001 6.801 6.931 cp_fm_cholesky_decompose 22 10.9 6.806 6.887 6.806 6.887 fft_wrap_pw1pw2_140 527 13.2 0.473 0.516 6.512 6.593 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.222 6.229 rs_pw_transfer 1054 12.0 0.014 0.016 5.530 6.089 fft3d_ps 1301 14.7 2.796 2.950 5.808 5.853 mp_sum_l 7930 13.1 3.855 5.496 3.855 5.496 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.323 5.454 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.354 5.411 grid_collocate_task_list 129 9.7 4.947 5.294 4.947 5.294 potential_pw2rs 129 12.3 0.015 0.018 4.716 4.739 mp_alltoall_d11v 2423 14.1 4.029 4.562 4.029 4.562 mp_allgather_i34 2527 14.6 1.814 4.457 1.814 4.457 dbcsr_complete_redistribute 395 12.7 0.784 0.894 3.242 4.081 mp_sum_d 4493 12.2 2.660 3.974 2.660 3.974 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.763000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.727273, yerr=1.911254 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 943.202304E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57139. MP_Allreduce 11261 985. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.207 173.706 173.708 qs_mol_dyn_low 1 2.0 0.003 0.004 172.711 172.725 qs_forces 11 3.9 0.004 0.010 172.611 172.621 qs_energies 11 4.9 0.009 0.042 166.151 166.161 scf_env_do_scf 11 5.9 0.001 0.002 149.989 149.995 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.490 115.491 velocity_verlet 10 3.0 0.002 0.002 109.973 109.975 dbcsr_multiply_generic 2485 12.5 0.180 0.185 81.298 82.384 qs_scf_new_mos 116 7.6 0.001 0.001 79.617 79.988 qs_scf_loop_do_ot 116 8.6 0.001 0.002 79.616 79.987 ot_scf_mini 116 9.6 0.004 0.004 75.517 75.949 multiply_cannon 2485 13.5 0.495 0.517 61.646 65.202 multiply_cannon_loop 2485 14.5 0.852 0.875 58.681 61.334 ot_mini 116 10.6 0.001 0.001 42.277 42.728 mp_waitall_1 176908 16.5 25.218 35.090 25.218 35.090 init_scf_loop 11 6.9 0.001 0.004 34.400 34.402 prepare_preconditioner 11 7.9 0.000 0.000 30.423 30.487 make_preconditioner 11 8.9 0.000 0.002 30.423 30.487 rebuild_ks_matrix 127 8.3 0.001 0.001 29.662 30.180 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.662 30.179 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.068 29.407 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.742 27.219 multiply_cannon_multrec 19880 15.5 13.301 16.350 22.324 25.366 multiply_cannon_metrocomm3 19880 15.5 0.058 0.062 15.323 24.957 qs_ot_get_derivative 116 11.6 0.001 0.002 22.590 23.029 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.947 21.117 apply_single 127 13.6 0.001 0.001 19.947 21.117 ot_diis_step 116 11.6 0.018 0.018 19.587 19.588 qs_ot_get_p 127 10.4 0.001 0.001 18.743 19.329 multiply_cannon_sync_h2d 19880 15.5 14.112 15.711 14.112 15.711 make_m2s 4970 13.5 0.079 0.083 14.695 15.491 make_images 4970 14.5 1.162 1.256 14.465 15.260 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.663 14.671 cp_fm_cholesky_invert 11 10.9 14.156 14.164 14.156 14.164 sum_up_and_integrate 127 10.3 0.133 0.145 13.878 13.901 integrate_v_rspace 127 11.3 0.004 0.004 13.745 13.768 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.685 13.687 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.738 12.779 calculate_rho_elec 127 8.7 0.131 0.145 12.737 12.778 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.764 10.765 cp_fm_redistribute_end 82 14.4 4.042 10.713 4.055 10.714 init_scf_run 11 5.9 0.000 0.001 10.693 10.693 scf_env_initial_rho_setup 11 6.9 0.005 0.034 10.693 10.693 cp_fm_diag_elpa_base 82 14.4 6.244 10.152 6.641 10.637 make_images_data 4970 15.5 0.060 0.067 9.173 10.395 hybrid_alltoall_any 5155 16.4 0.428 1.972 7.925 9.554 multiply_cannon_metrocomm4 17395 15.5 0.063 0.074 3.466 9.405 mp_irecv_dv 49801 16.2 3.341 9.143 3.341 9.143 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.617 8.942 dbcsr_mm_accdrv_process 41158 16.2 4.611 5.492 8.479 8.711 pw_transfer 1535 11.6 0.083 0.103 7.583 7.694 grid_integrate_task_list 127 12.3 7.267 7.626 7.267 7.626 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.363 7.480 cp_fm_upper_to_full 104 14.5 5.666 7.346 5.666 7.346 wfi_extrapolate 11 7.9 0.001 0.001 7.326 7.326 cp_fm_cholesky_decompose 22 10.9 7.172 7.209 7.172 7.209 density_rs2pw 127 9.7 0.006 0.006 6.256 6.674 fft_wrap_pw1pw2_140 519 13.2 0.475 0.521 6.508 6.632 dbcsr_complete_redistribute 393 12.7 1.170 1.200 4.447 6.206 calculate_dm_sparse 127 9.5 0.001 0.001 5.806 5.899 fft3d_ps 1281 14.7 2.674 2.878 5.642 5.712 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.634 5.642 grid_collocate_task_list 127 9.7 5.039 5.483 5.039 5.483 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.603 5.396 rs_pw_transfer 1038 11.9 0.013 0.014 4.811 5.324 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.274 5.030 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.743 4.898 mp_alltoall_d11v 2401 14.1 4.123 4.713 4.123 4.713 mp_sum_l 7804 13.0 3.171 4.667 3.171 4.667 mp_allgather_i34 2485 14.5 1.497 4.540 1.497 4.540 potential_pw2rs 127 12.3 0.020 0.022 4.340 4.352 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.336 4.019 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.981 4.008 qs_energies_init_hamiltonians 11 5.9 0.024 0.031 3.732 3.733 mp_alltoall_i22 712 14.1 1.863 3.705 1.863 3.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.708000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.727273, yerr=9.096362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.143689E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.098 185.613 185.614 qs_mol_dyn_low 1 2.0 0.003 0.003 185.022 185.034 qs_forces 11 3.9 0.003 0.007 184.859 184.877 qs_energies 11 4.9 0.003 0.007 177.853 177.868 scf_env_do_scf 11 5.9 0.001 0.017 161.214 161.227 velocity_verlet 10 3.0 0.029 0.065 123.111 123.114 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.873 114.875 qs_scf_new_mos 117 7.6 0.001 0.001 79.588 79.870 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.587 79.870 dbcsr_multiply_generic 2507 12.6 0.190 0.196 79.170 79.712 ot_scf_mini 117 9.6 0.004 0.004 75.073 75.351 multiply_cannon 2507 13.6 0.559 0.598 54.564 57.015 multiply_cannon_loop 2507 14.6 1.181 1.213 50.896 52.192 init_scf_loop 11 6.9 0.001 0.003 46.202 46.203 ot_mini 117 10.6 0.001 0.001 42.061 42.325 prepare_preconditioner 11 7.9 0.000 0.000 42.141 42.163 make_preconditioner 11 8.9 0.000 0.002 42.141 42.163 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.722 40.830 multiply_cannon_multrec 30084 15.6 14.321 19.472 26.194 30.807 rebuild_ks_matrix 128 8.3 0.001 0.001 28.866 29.106 qs_ks_build_kohn_sham_matrix 128 9.3 0.022 0.055 28.865 29.106 mp_waitall_1 153770 16.5 17.345 27.196 17.345 27.196 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.082 26.303 qs_ot_get_derivative 117 11.6 0.001 0.002 22.579 22.855 make_m2s 5014 13.6 0.093 0.098 20.184 21.301 make_images 5014 14.6 1.950 2.288 19.880 20.997 qs_ot_get_p 128 10.4 0.001 0.001 19.267 19.560 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.899 19.361 apply_single 128 13.6 0.001 0.001 18.898 19.360 ot_diis_step 117 11.6 0.017 0.018 19.342 19.343 cp_fm_upper_to_full 105 14.7 11.242 16.672 11.242 16.672 cp_fm_cholesky_invert 11 10.9 15.966 15.975 15.966 15.975 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.325 15.814 qs_ot_p2m_diag 83 11.4 0.342 0.390 15.166 15.218 sum_up_and_integrate 128 10.3 0.142 0.154 13.946 13.972 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.901 13.902 integrate_v_rspace 128 11.3 0.004 0.004 13.804 13.837 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.040 13.084 calculate_rho_elec 128 8.7 0.175 0.190 13.039 13.083 dbcsr_complete_redistribute 395 12.7 1.513 1.667 8.992 12.797 multiply_cannon_sync_h2d 30084 15.6 11.690 12.767 11.690 12.767 make_images_data 5014 15.6 0.063 0.067 10.894 12.645 dbcsr_mm_accdrv_process 62264 16.2 7.310 8.043 11.448 11.959 hybrid_alltoall_any 5200 16.5 0.527 2.200 9.711 11.586 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.627 11.442 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.875 10.877 cp_fm_redistribute_end 83 14.4 1.852 10.811 1.865 10.817 cp_fm_diag_elpa_base 83 14.4 8.333 10.224 8.922 10.714 init_scf_run 11 5.9 0.000 0.001 10.712 10.713 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.711 10.713 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.395 10.110 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.332 9.535 mp_alltoall_i22 716 14.1 5.697 9.479 5.697 9.479 grid_integrate_task_list 128 12.3 7.505 7.927 7.505 7.927 pw_transfer 1547 11.6 0.084 0.099 7.778 7.855 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.554 7.637 wfi_extrapolate 11 7.9 0.001 0.001 7.557 7.557 cp_fm_cholesky_decompose 22 10.9 7.433 7.537 7.433 7.537 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.722 7.024 fft_wrap_pw1pw2_140 523 13.2 0.481 0.489 6.698 6.792 mp_irecv_dv 76098 16.2 2.578 6.751 2.578 6.751 density_rs2pw 128 9.7 0.006 0.006 6.142 6.479 calculate_dm_sparse 128 9.5 0.001 0.001 6.185 6.287 fft3d_ps 1291 14.7 2.796 2.895 5.784 5.872 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.527 5.576 grid_collocate_task_list 128 9.7 5.206 5.529 5.206 5.529 mp_alltoall_d11v 2415 14.1 4.768 5.155 4.768 5.155 rs_pw_transfer 1046 11.9 0.013 0.015 4.497 4.902 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.453 4.537 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 4.464 4.465 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.360 4.450 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.164 4.233 potential_pw2rs 128 12.3 0.023 0.023 4.171 4.186 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.614000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.000000, yerr=18.935297 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 1.523188E+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 3983 58480. MP_Allreduce 10955 1177. MP_Sync 86 MP_Alltoall 1700 18828162. MP_ISendRecv 7620 122880. MP_Wait 19810 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 0.025 0.076 169.540 169.541 qs_mol_dyn_low 1 2.0 0.003 0.003 169.033 169.122 qs_forces 11 3.9 0.003 0.003 168.256 168.262 qs_energies 11 4.9 0.002 0.002 160.977 160.986 scf_env_do_scf 11 5.9 0.003 0.034 143.926 143.935 velocity_verlet 10 3.0 0.113 0.219 110.930 110.962 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.461 108.462 dbcsr_multiply_generic 2485 12.5 0.179 0.185 73.767 74.380 qs_scf_new_mos 116 7.6 0.001 0.001 73.515 73.691 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.514 73.691 ot_scf_mini 116 9.6 0.003 0.004 69.104 69.273 multiply_cannon 2485 13.5 0.585 0.627 53.970 58.134 multiply_cannon_loop 2485 14.5 0.445 0.455 49.409 50.477 ot_mini 116 10.6 0.001 0.001 39.202 39.363 init_scf_loop 11 6.9 0.000 0.000 35.283 35.284 mp_waitall_1 128490 16.6 26.002 33.288 26.002 33.288 prepare_preconditioner 11 7.9 0.000 0.000 31.475 31.501 make_preconditioner 11 8.9 0.000 0.000 31.475 31.501 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.393 29.655 rebuild_ks_matrix 127 8.3 0.001 0.001 28.105 28.222 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.104 28.221 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.634 25.742 multiply_cannon_multrec 9940 15.5 10.296 13.744 17.617 20.635 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.240 19.709 apply_single 127 13.6 0.001 0.001 19.240 19.709 ot_diis_step 116 11.6 0.019 0.020 19.696 19.696 qs_ot_get_derivative 116 11.6 0.002 0.002 19.442 19.600 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 11.978 19.114 make_m2s 4970 13.5 0.065 0.069 16.223 18.627 make_images 4970 14.5 2.295 2.879 15.919 18.321 cp_fm_cholesky_invert 11 10.9 17.966 17.971 17.966 17.971 qs_ot_get_p 127 10.4 0.001 0.001 17.075 17.252 sum_up_and_integrate 127 10.3 0.180 0.188 13.723 13.777 integrate_v_rspace 127 11.3 0.004 0.004 13.543 13.604 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.325 13.342 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.875 12.908 calculate_rho_elec 127 8.7 0.256 0.267 12.874 12.908 make_images_data 4970 15.5 0.052 0.060 9.955 12.678 hybrid_alltoall_any 5155 16.4 0.800 3.517 9.734 12.452 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.211 12.212 multiply_cannon_sync_h2d 9940 15.5 11.477 11.985 11.477 11.985 init_scf_run 11 5.9 0.000 0.001 10.305 10.305 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.305 10.305 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.285 9.286 cp_fm_diag_elpa_base 82 14.4 9.050 9.118 9.282 9.282 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.644 8.456 grid_integrate_task_list 127 12.3 7.715 7.987 7.715 7.987 cp_fm_cholesky_decompose 22 10.9 7.879 7.970 7.879 7.970 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.777 7.892 dbcsr_mm_accdrv_process 20590 16.1 2.679 3.456 6.952 7.607 pw_transfer 1535 11.6 0.082 0.089 7.578 7.589 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.359 7.375 mp_allgather_i34 2485 14.5 2.775 7.267 2.775 7.267 wfi_extrapolate 11 7.9 0.001 0.001 7.214 7.214 fft_wrap_pw1pw2_140 519 13.2 0.501 0.522 6.492 6.512 calculate_dm_sparse 127 9.5 0.001 0.001 6.175 6.263 density_rs2pw 127 9.7 0.005 0.006 5.771 5.998 grid_collocate_task_list 127 9.7 5.450 5.663 5.450 5.663 fft3d_ps 1281 14.7 2.704 2.775 5.531 5.562 dbcsr_complete_redistribute 393 12.7 2.119 2.171 5.135 5.521 qs_energies_init_hamiltonians 11 5.9 0.003 0.010 5.233 5.235 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.222 5.233 mp_alltoall_d11v 2401 14.1 4.537 4.936 4.537 4.936 multiply_cannon_metrocomm4 7455 15.5 0.023 0.027 1.879 4.200 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.132 4.164 rs_pw_transfer 1038 11.9 0.012 0.013 3.883 4.145 mp_irecv_dv 28618 15.9 1.843 4.129 1.843 4.129 potential_pw2rs 127 12.3 0.026 0.026 3.891 3.901 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.572 3.855 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.644 3.680 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.348 3.676 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.556 3.638 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.431 3.488 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.541000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1414.000000, yerr=58.868111 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 3.077870E+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_ISendRecv 3612 218624. MP_Wait 11682 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 0.083 0.104 289.656 289.656 qs_mol_dyn_low 1 2.0 0.003 0.003 288.782 288.793 qs_forces 11 3.9 0.003 0.003 288.665 288.669 qs_energies 11 4.9 0.002 0.002 280.078 280.092 scf_env_do_scf 11 5.9 0.050 0.394 258.445 258.457 velocity_verlet 10 3.0 0.018 0.019 209.838 209.848 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.540 132.541 init_scf_loop 11 6.9 0.000 0.000 125.257 125.258 prepare_preconditioner 11 7.9 0.000 0.000 120.614 120.642 make_preconditioner 11 8.9 0.000 0.000 120.614 120.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.477 117.785 qs_scf_new_mos 118 7.6 0.001 0.001 90.527 90.630 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.526 90.629 ot_scf_mini 118 9.6 0.004 0.004 85.718 85.752 dbcsr_multiply_generic 2535 12.6 0.215 0.222 83.350 83.928 cp_fm_upper_to_full 106 14.8 53.661 77.893 53.661 77.893 multiply_cannon 2535 13.6 0.710 0.782 59.124 59.728 multiply_cannon_loop 2535 14.6 0.478 0.483 55.516 56.876 ot_mini 118 10.6 0.001 0.001 44.638 44.670 dbcsr_complete_redistribute 397 12.7 3.988 4.027 30.789 44.393 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.413 41.040 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 25.092 38.642 mp_alltoall_i22 720 14.1 23.239 36.903 23.239 36.903 rebuild_ks_matrix 129 8.3 0.001 0.001 33.546 33.588 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.545 33.587 cp_fm_cholesky_invert 11 10.9 32.979 32.985 32.979 32.985 mp_waitall_1 106626 16.7 28.196 32.083 28.196 32.083 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.351 31.389 qs_ot_get_p 129 10.4 0.001 0.001 25.607 25.616 qs_ot_get_derivative 118 11.6 0.002 0.002 24.558 24.594 qs_ot_p2m_diag 84 11.4 0.889 0.894 21.603 21.634 make_m2s 5070 13.6 0.076 0.079 19.720 20.650 make_images 5070 14.6 3.786 3.919 19.238 20.170 multiply_cannon_metrocomm3 10140 15.6 0.023 0.023 18.914 20.137 ot_diis_step 118 11.6 0.022 0.023 20.055 20.056 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.882 19.884 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.252 19.453 apply_single 129 13.6 0.001 0.001 19.252 19.453 multiply_cannon_multrec 10140 15.6 10.647 12.467 18.279 18.594 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.708 16.709 cp_fm_diag_elpa_base 84 14.4 12.187 13.864 16.704 16.704 multiply_cannon_sync_h2d 10140 15.6 15.782 15.796 15.782 15.796 sum_up_and_integrate 129 10.3 0.324 0.326 15.687 15.780 integrate_v_rspace 129 11.3 0.004 0.004 15.363 15.456 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.879 14.901 calculate_rho_elec 129 8.7 0.487 0.487 14.878 14.900 hybrid_alltoall_any 5257 16.5 1.307 3.048 10.648 12.438 make_images_data 5070 15.6 0.061 0.065 10.477 12.056 init_scf_run 11 5.9 0.000 0.001 11.697 11.697 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.696 11.697 dbcsr_mm_accdrv_process 20958 16.1 4.098 6.310 7.393 9.556 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.439 9.470 grid_integrate_task_list 129 12.3 8.603 8.783 8.603 8.783 cp_fm_cholesky_decompose 22 10.9 8.727 8.742 8.727 8.742 wfi_extrapolate 11 7.9 0.001 0.001 8.607 8.607 pw_transfer 1559 11.6 0.090 0.091 8.173 8.188 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 7.938 7.953 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.942 7.943 fft_wrap_pw1pw2_140 527 13.2 0.544 0.546 7.066 7.086 mp_alltoall_d11v 2429 14.1 6.868 7.007 6.868 7.007 calculate_dm_sparse 129 9.5 0.001 0.001 6.644 6.734 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.344 6.464 grid_collocate_task_list 129 9.7 6.392 6.424 6.392 6.424 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.273 6.370 fft3d_ps 1301 14.7 2.772 2.780 5.968 5.978 density_rs2pw 129 9.7 0.005 0.005 5.899 5.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.656000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2713.818182, yerr=189.265479 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 1.261306E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.082 84.686 84.723 qs_energies 1 2.0 0.000 0.000 84.207 84.222 ls_scf 1 3.0 0.000 0.000 83.282 83.297 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.473 72.643 multiply_cannon 111 7.7 0.017 0.020 55.894 57.136 multiply_cannon_loop 111 8.7 0.211 0.226 52.462 53.910 ls_scf_main 1 4.0 0.000 0.000 52.108 52.108 density_matrix_trs4 2 5.0 0.002 0.003 46.672 46.791 ls_scf_init_scf 1 4.0 0.000 0.000 28.145 28.146 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.034 27.085 mp_waitall_1 11316 10.9 22.584 25.477 22.584 25.477 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.955 24.976 multiply_cannon_multrec 2664 9.7 8.164 8.904 15.463 17.099 multiply_cannon_sync_h2d 2664 9.7 13.481 15.659 13.481 15.659 make_m2s 222 7.7 0.008 0.011 12.898 13.489 make_images 222 8.7 0.099 0.109 12.876 13.469 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.807 13.383 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.562 9.325 make_images_data 222 9.7 0.004 0.005 7.542 8.174 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.607 6.919 7.868 hybrid_alltoall_any 227 10.6 0.215 1.822 6.464 7.459 dbcsr_mm_accdrv_process_sort 4760 11.4 6.209 7.073 6.209 7.073 calculate_norms 4752 9.8 5.513 6.092 5.513 6.092 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.045 5.159 mp_sum_l 807 5.4 3.146 4.600 3.146 4.600 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.045 3.598 mp_irecv_dv 6231 10.9 2.028 3.569 2.028 3.569 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.351 3.355 make_images_sizes 222 9.7 0.000 0.000 0.668 3.236 mp_alltoall_i44 222 10.7 0.668 3.236 0.668 3.236 arnoldi_extremal 4 6.8 0.000 0.000 3.180 3.205 arnoldi_normal_ev 4 7.8 0.001 0.003 3.179 3.205 build_subspace 16 8.4 0.009 0.012 3.086 3.088 ls_scf_post 1 4.0 0.000 0.000 3.030 3.044 ls_scf_store_result 1 5.0 0.000 0.000 2.849 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.331 2.771 dbcsr_merge_single_wm 555 10.7 0.458 0.595 2.324 2.763 make_images_pack 222 9.7 2.209 2.634 2.211 2.635 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.292 2.533 dbcsr_sort_data 658 11.4 2.123 2.487 2.123 2.487 dbcsr_matrix_vector_mult_local 304 10.0 2.060 2.449 2.062 2.451 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.182 2.245 buffer_matrices_ensure_size 222 8.7 1.752 2.034 1.752 2.034 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.739 1.740 rebuild_ks_matrix 3 7.3 0.000 0.000 1.730 1.731 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.017 1.730 1.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.723000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 2.060517E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.156 0.497 91.759 91.759 qs_energies 1 2.0 0.002 0.010 90.811 90.834 ls_scf 1 3.0 0.000 0.001 89.456 89.472 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.382 75.653 multiply_cannon 111 7.7 0.029 0.044 53.092 57.285 ls_scf_main 1 4.0 0.000 0.002 54.791 54.792 multiply_cannon_loop 111 8.7 0.116 0.123 49.825 53.284 density_matrix_trs4 2 5.0 0.002 0.004 49.113 49.305 ls_scf_init_scf 1 4.0 0.000 0.002 31.080 31.082 mp_waitall_1 9246 10.9 21.139 29.864 21.139 29.864 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.709 29.793 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.223 27.240 multiply_cannon_multrec 1332 9.7 13.183 16.726 22.392 27.045 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.697 20.645 make_m2s 222 7.7 0.006 0.007 15.501 16.163 make_images 222 8.7 1.571 1.930 15.471 16.134 dbcsr_mm_accdrv_process 4041 10.4 0.305 0.452 8.806 10.387 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.934 8.399 9.934 make_images_data 222 9.7 0.004 0.005 8.902 9.807 hybrid_alltoall_any 227 10.6 0.520 2.460 8.300 9.310 mp_sum_l 807 5.4 5.783 8.945 5.783 8.945 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.225 7.689 mp_irecv_dv 3311 11.0 3.205 7.633 3.205 7.633 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.269 6.760 calculate_norms 2376 9.8 6.006 6.678 6.006 6.678 multiply_cannon_sync_h2d 1332 9.7 4.774 5.954 4.774 5.954 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.121 5.326 arnoldi_extremal 4 6.8 0.000 0.000 4.633 4.659 arnoldi_normal_ev 4 7.8 0.001 0.004 4.633 4.659 build_subspace 16 8.4 0.014 0.022 4.369 4.372 ls_scf_post 1 4.0 0.003 0.017 3.585 3.601 ls_scf_store_result 1 5.0 0.000 0.000 3.254 3.401 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.131 3.350 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.220 2.746 3.222 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.164 2.646 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.524 2.602 make_images_pack 222 9.7 2.055 2.465 2.057 2.467 mp_allgather_i34 111 8.7 0.967 2.464 0.967 2.464 dbcsr_sort_data 436 11.2 1.838 2.076 1.838 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.908 1.910 rebuild_ks_matrix 3 7.3 0.000 0.000 1.895 1.897 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.895 1.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.759000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 2.758947E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.096 0.132 94.937 94.939 qs_energies 1 2.0 0.000 0.001 93.896 93.900 ls_scf 1 3.0 0.012 0.033 92.193 92.198 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.704 76.970 ls_scf_main 1 4.0 0.000 0.002 57.795 57.813 multiply_cannon 111 7.7 0.044 0.102 52.829 57.430 multiply_cannon_loop 111 8.7 0.100 0.106 49.247 52.664 density_matrix_trs4 2 5.0 0.057 0.247 51.865 52.029 mp_waitall_1 7374 11.0 23.965 33.802 23.965 33.802 ls_scf_init_scf 1 4.0 0.000 0.002 30.779 30.783 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.404 29.463 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.989 27.005 multiply_cannon_multrec 888 9.7 12.544 15.130 21.068 24.259 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.215 22.851 make_m2s 222 7.7 0.006 0.007 17.287 18.503 make_images 222 8.7 1.973 2.268 17.249 18.466 hybrid_alltoall_any 227 10.6 0.622 2.840 9.465 10.951 make_images_data 222 9.7 0.003 0.004 9.844 10.862 dbcsr_mm_accdrv_process 3754 10.4 0.244 0.437 8.042 9.271 dbcsr_mm_accdrv_process_sort 3754 11.4 7.666 8.834 7.666 8.834 mp_sum_l 807 5.4 5.025 8.450 5.025 8.450 multiply_cannon_sync_h2d 888 9.7 6.105 7.331 6.105 7.331 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.436 7.165 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.526 7.161 mp_irecv_dv 2335 11.1 2.420 7.117 2.420 7.117 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.668 6.341 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.986 5.218 arnoldi_extremal 4 6.8 0.000 0.000 5.134 5.152 arnoldi_normal_ev 4 7.8 0.001 0.005 5.134 5.152 build_subspace 16 8.4 0.014 0.020 4.828 4.836 calculate_norms 1584 9.8 4.277 4.625 4.277 4.625 mp_allgather_i34 111 8.7 1.424 3.899 1.424 3.899 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.484 3.829 dbcsr_matrix_vector_mult_local 304 10.0 3.043 3.624 3.045 3.626 ls_scf_post 1 4.0 0.000 0.001 3.608 3.613 ls_scf_store_result 1 5.0 0.000 0.000 3.341 3.416 ls_scf_dm_to_ks 2 5.0 0.001 0.018 2.756 2.879 make_images_sizes 222 9.7 0.000 0.000 1.098 2.193 mp_alltoall_i44 222 10.7 1.097 2.193 1.097 2.193 make_images_pack 222 9.7 1.818 2.141 1.821 2.144 dbcsr_sort_data 325 11.1 1.901 2.138 1.901 2.138 dbcsr_data_release 9322 10.9 1.323 2.050 1.323 2.050 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.940 1.942 rebuild_ks_matrix 3 7.3 0.000 0.000 1.922 1.924 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.922 1.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.939000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2181.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 3.373343E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.068 97.193 97.196 qs_energies 1 2.0 0.000 0.000 96.522 96.527 ls_scf 1 3.0 0.000 0.001 94.836 94.843 dbcsr_multiply_generic 111 6.7 0.018 0.022 78.663 78.926 ls_scf_main 1 4.0 0.000 0.000 59.079 59.080 multiply_cannon 111 7.7 0.049 0.117 51.621 56.552 density_matrix_trs4 2 5.0 0.002 0.003 53.031 53.163 multiply_cannon_loop 111 8.7 0.114 0.126 46.548 50.071 ls_scf_init_scf 1 4.0 0.001 0.002 32.557 32.559 ls_scf_init_matrix_S 1 5.0 0.001 0.002 31.379 31.448 mp_waitall_1 6438 11.0 22.764 29.242 22.764 29.242 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.926 28.938 multiply_cannon_multrec 1332 9.7 14.217 17.642 22.054 25.263 make_m2s 222 7.7 0.007 0.009 21.185 22.651 make_images 222 8.7 3.137 3.601 21.135 22.603 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.190 17.704 make_images_data 222 9.7 0.004 0.005 11.848 13.599 hybrid_alltoall_any 227 10.6 0.796 3.778 11.171 12.990 dbcsr_mm_accdrv_process 3641 10.4 0.203 0.398 7.471 8.992 dbcsr_mm_accdrv_process_sort 3641 11.4 7.099 8.576 7.099 8.576 mp_sum_l 807 5.4 4.282 7.768 4.282 7.768 multiply_cannon_sync_h2d 1332 9.7 5.530 6.157 5.530 6.157 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.320 6.039 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.057 5.936 mp_irecv_dv 3229 10.9 2.035 5.850 2.035 5.850 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.552 5.377 arnoldi_extremal 4 6.8 0.000 0.000 5.209 5.222 arnoldi_normal_ev 4 7.8 0.001 0.005 5.208 5.222 build_subspace 16 8.4 0.015 0.021 4.870 4.879 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.559 4.749 calculate_norms 2376 9.8 4.221 4.645 4.221 4.645 mp_allgather_i34 111 8.7 2.198 4.545 2.198 4.545 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.596 3.889 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.711 3.206 3.712 dbcsr_sort_data 658 11.4 3.020 3.437 3.020 3.437 dbcsr_special_finalize 555 9.7 0.006 0.008 2.761 3.240 dbcsr_merge_single_wm 555 10.7 0.530 0.665 2.753 3.232 ls_scf_post 1 4.0 0.001 0.018 3.200 3.205 ls_scf_store_result 1 5.0 0.000 0.000 2.943 2.999 ls_scf_dm_to_ks 2 5.0 0.001 0.011 2.903 2.984 dbcsr_data_release 10477 10.7 1.581 2.376 1.581 2.376 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.196000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 4.631450E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.130 0.167 99.539 99.541 qs_energies 1 2.0 0.000 0.000 98.664 98.669 ls_scf 1 3.0 0.000 0.000 96.695 96.699 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.387 78.598 ls_scf_main 1 4.0 0.000 0.000 61.989 61.989 multiply_cannon 111 7.7 0.087 0.179 55.272 61.121 density_matrix_trs4 2 5.0 0.002 0.003 55.029 55.118 multiply_cannon_loop 111 8.7 0.069 0.075 50.837 52.513 mp_waitall_1 5481 11.0 26.990 33.019 26.990 33.019 ls_scf_init_scf 1 4.0 0.000 0.000 31.065 31.070 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.880 29.933 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.338 27.348 multiply_cannon_multrec 444 9.7 14.013 16.899 21.035 22.577 make_m2s 222 7.7 0.004 0.005 18.369 20.771 make_images 222 8.7 3.718 4.437 18.307 20.711 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.061 16.103 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.315 14.798 make_images_data 222 9.7 0.003 0.004 10.413 12.935 hybrid_alltoall_any 227 10.6 0.789 3.768 10.092 12.466 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.342 6.728 7.879 multiply_cannon_sync_h2d 444 9.7 6.542 7.666 6.542 7.666 dbcsr_mm_accdrv_process_sort 3003 11.4 6.413 7.534 6.413 7.534 mp_allgather_i34 111 8.7 2.672 7.429 2.672 7.429 arnoldi_extremal 4 6.8 0.000 0.000 5.860 5.873 arnoldi_normal_ev 4 7.8 0.002 0.005 5.860 5.873 build_subspace 16 8.4 0.015 0.020 5.440 5.452 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.938 5.080 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.204 4.394 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.715 4.360 mp_irecv_dv 1241 11.2 1.695 4.337 1.695 4.337 dbcsr_matrix_vector_mult_local 304 10.0 3.727 4.202 3.729 4.205 mp_sum_l 807 5.4 2.788 4.124 2.788 4.124 ls_scf_post 1 4.0 0.000 0.000 3.640 3.645 make_images_sizes 222 9.7 0.000 0.000 1.174 3.644 mp_alltoall_i44 222 10.7 1.174 3.644 1.174 3.644 calculate_norms 792 9.8 3.534 3.632 3.534 3.632 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.470 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.390 3.465 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.937 3.105 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.332 dbcsr_merge_all 275 8.9 0.473 0.540 2.048 2.166 dbcsr_data_release 10123 10.8 1.330 2.015 1.330 2.015 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.541000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3633.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4083d9693461e026a35c103709c5fd193a3d5841_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 8.785437E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.124 0.471 107.524 107.848 qs_energies 1 2.0 0.000 0.000 106.065 106.086 ls_scf 1 3.0 0.000 0.000 103.197 103.218 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.254 77.417 ls_scf_main 1 4.0 0.000 0.000 65.435 65.435 density_matrix_trs4 2 5.0 0.002 0.003 56.621 56.688 multiply_cannon 111 7.7 0.175 0.234 49.112 50.260 multiply_cannon_loop 111 8.7 0.067 0.070 45.661 47.482 ls_scf_init_scf 1 4.0 0.000 0.000 34.117 34.118 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.788 32.809 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.712 29.724 mp_waitall_1 4569 11.1 21.375 25.688 21.375 25.688 make_m2s 222 7.7 0.005 0.005 23.852 24.824 make_images 222 8.7 4.582 4.971 23.746 24.714 multiply_cannon_multrec 444 9.7 17.881 18.513 22.506 23.101 make_images_data 222 9.7 0.003 0.003 13.077 15.533 hybrid_alltoall_any 227 10.6 1.658 3.624 12.822 15.501 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.713 10.818 multiply_cannon_sync_h2d 444 9.7 8.823 8.861 8.823 8.861 arnoldi_extremal 4 6.8 0.000 0.000 7.474 7.484 arnoldi_normal_ev 4 7.8 0.003 0.009 7.474 7.484 build_subspace 16 8.4 0.026 0.037 6.918 6.931 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.545 5.683 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.280 5.547 dbcsr_matrix_vector_mult_local 304 10.0 5.096 5.412 5.099 5.415 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.775 4.864 dbcsr_mm_accdrv_process 1814 10.4 0.207 0.317 4.459 4.641 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.249 4.124 4.249 ls_scf_post 1 4.0 0.000 0.000 3.645 3.666 make_images_sizes 222 9.7 0.000 0.000 1.446 3.522 mp_alltoall_i44 222 10.7 1.445 3.521 1.445 3.521 mp_allgather_i34 111 8.7 1.098 3.424 1.098 3.424 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.408 mp_sum_l 807 5.4 1.964 3.359 1.964 3.359 calculate_norms 792 9.8 3.268 3.317 3.268 3.317 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.122 dbcsr_merge_all 275 8.9 0.889 0.907 2.862 2.913 dbcsr_data_release 12724 10.6 2.339 2.885 2.339 2.885 dbcsr_complete_redistribute 5 7.6 1.435 1.471 2.764 2.874 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.838 2.838 matrix_ls_to_qs 2 6.0 0.000 0.000 2.414 2.534 dbcsr_sort_data 325 11.1 2.439 2.503 2.439 2.503 dbcsr_new_transposed 4 7.5 0.242 0.250 2.304 2.323 dbcsr_frobenius_norm 74 6.6 2.056 2.138 2.225 2.283 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.205 dbcsr_add_anytype 103 7.2 0.860 0.895 2.131 2.205 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.187 2.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.848000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6862.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 4083d9693461e026a35c103709c5fd193a3d5841 Summary: empty Status: OK