=== 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: 399e6db9ee1a21058cc4319e934d6c8fef1bc356 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001.rc2, 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 (15.12.2022) # # \ 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 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001.rc2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/01 job id: 43545671 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/02 job id: 43545673 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/03 job id: 43545674 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/04 job id: 43545676 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/05 job id: 43545677 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/06 job id: 43545678 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/07 job id: 43545680 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/08 job id: 43545681 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/09 job id: 43545682 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/10 job id: 43545685 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/11 job id: 43545686 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/12 job id: 43545687 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/13 job id: 43545688 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/14 job id: 43545689 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/15 job id: 43545690 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/16 job id: 43545691 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/17 job id: 43545692 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/18 job id: 43545693 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/19 job id: 43545694 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/20 job id: 43545696 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/21 job id: 43545698 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/22 job id: 43545699 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/23 job id: 43545700 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/24 job id: 43545701 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/25 job id: 43545702 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/26 job id: 43545703 --- 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 134.137 134.138 farming_run 1 2.0 133.657 133.658 134.108 134.111 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.455354E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.023 116.373 116.374 qs_energies 1 2.0 0.000 0.000 116.171 116.173 mp2_main 1 3.0 0.000 0.000 114.132 114.135 mp2_gpw_main 1 4.0 0.020 0.026 113.243 113.246 mp2_ri_gpw_compute_in 1 5.0 0.173 0.206 93.956 94.435 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.424 55.903 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.172 41.728 47.179 get_2c_integrals 1 6.0 0.002 0.031 37.579 38.326 integrate_v_rspace 273 8.0 0.438 0.450 25.133 30.373 pw_transfer 6555 10.6 0.375 0.393 27.413 28.068 grid_integrate_task_list 273 9.0 20.947 26.664 20.947 26.664 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.070 26.643 fft_wrap_pw1pw2_100 2178 12.4 1.218 1.429 23.637 24.210 compute_2c_integrals 1 7.0 0.002 0.003 19.726 19.727 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.959 19.442 mp2_eri_2c_integrate_gpw 1 9.0 2.376 2.435 18.956 19.440 rpa_ri_compute_en 1 5.0 0.016 0.018 19.178 19.305 cp_fm_cholesky_decompose 12 8.2 17.885 18.562 17.885 18.562 cholesky_decomp 1 7.0 0.000 0.000 16.698 17.468 fft3d_s 5443 13.4 16.175 16.614 16.197 16.635 ao_to_mo_and_store_B_mult_1 272 7.0 10.861 15.595 10.861 15.595 calculate_wavefunction 272 8.0 5.402 5.559 12.519 13.120 rpa_num_int 1 6.0 0.000 0.006 10.658 10.667 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.565 10.647 calc_mat_Q 8 8.0 0.000 0.000 9.329 9.429 contract_S_to_Q 8 9.0 0.000 0.000 8.751 8.854 calc_potential_gpw 544 9.5 0.005 0.006 8.240 8.708 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.254 8.595 parallel_gemm_fm 14 9.1 0.000 0.000 8.322 8.428 parallel_gemm_fm_cosma 14 10.1 8.322 8.428 8.322 8.428 potential_pw2rs 545 10.0 0.107 0.108 7.731 8.379 create_integ_mat 1 6.0 0.014 0.027 7.928 7.938 collocate_single_gaussian 272 10.0 0.041 0.043 7.508 7.752 array2fm 1 7.0 0.000 0.000 6.665 7.265 pw_scatter_s 2720 13.7 4.417 4.622 4.417 4.622 pw_gather_s 2722 13.2 3.892 4.223 3.892 4.223 array2fm_buffer_send 1 8.0 2.939 3.146 2.939 3.146 pw_poisson_solve 545 10.5 1.131 1.177 2.180 2.408 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.245662, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.038 396.622 396.625 farming_run 1 2.0 395.853 395.857 396.588 396.590 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222660E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 28085220. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.041 210.147 210.148 qs_energies 1 2.0 0.000 0.000 209.643 209.928 scf_env_do_scf 1 3.0 0.000 0.000 106.232 106.233 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.317 105.324 rebuild_ks_matrix 4 6.0 0.000 0.000 105.315 105.323 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 105.315 105.323 hfx_ks_matrix 4 8.0 0.001 0.001 104.927 104.931 integrate_four_center 4 9.0 0.144 0.457 104.927 104.930 mp2_main 1 3.0 0.000 0.000 103.122 103.407 mp2_gpw_main 1 4.0 0.032 0.046 102.158 102.443 integrate_four_center_main 4 10.0 0.110 0.499 96.681 99.352 integrate_four_center_bin 265 11.0 96.571 99.315 96.571 99.315 init_scf_loop 1 4.0 0.000 0.000 92.048 92.049 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 75.093 76.270 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.555 55.732 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.204 47.428 integrate_v_rspace 95 8.0 0.397 0.578 28.582 33.590 pw_transfer 2240 10.6 0.145 0.204 29.962 30.381 fft_wrap_pw1pw2 1868 11.4 0.017 0.024 28.946 29.398 grid_integrate_task_list 95 9.0 23.860 29.055 23.860 29.055 mp2_ri_gpw_compute_en 1 5.0 0.223 0.507 26.917 28.940 fft_wrap_pw1pw2_100 730 12.4 1.314 1.498 26.647 27.012 ao_to_mo_and_store_B_mult_1 91 7.0 10.667 26.348 10.667 26.348 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.836 1.895 24.917 24.928 get_2c_integrals 1 6.0 0.000 0.000 20.443 20.475 compute_2c_integrals 1 7.0 0.003 0.005 19.423 19.427 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.967 19.260 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.902 18.966 19.259 fft3d_s 1823 13.4 18.419 18.715 18.432 18.728 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.182 14.182 calculate_wavefunction 91 8.0 2.020 2.058 9.745 9.985 potential_pw2rs 186 10.0 0.034 0.036 8.648 9.165 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.595 8.726 9.112 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.315 8.558 local_gemm 172 8.0 8.169 8.543 8.169 8.543 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.525 7.958 8.521 calc_potential_gpw 182 9.5 0.002 0.003 7.989 8.177 collocate_single_gaussian 91 10.0 0.017 0.025 7.918 8.116 mp_sync 38 10.4 2.901 6.681 2.901 6.681 mp_sendrecv_dm3 2068 8.0 5.996 6.554 5.996 6.554 mp2_ri_gpw_compute_en_ener 172 7.0 6.351 6.436 6.351 6.436 pw_gather_s 912 13.2 4.876 5.383 4.876 5.383 pw_scatter_s 910 13.7 3.945 4.237 3.945 4.237 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.440341, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 451.055616E+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 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.026 54.060 54.061 qs_mol_dyn_low 1 2.0 0.003 0.003 53.874 53.882 qs_forces 11 3.9 0.002 0.002 53.815 53.815 qs_energies 11 4.9 0.001 0.001 52.282 52.297 scf_env_do_scf 11 5.9 0.000 0.001 46.159 46.159 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.947 43.947 dbcsr_multiply_generic 2286 12.5 0.094 0.097 34.337 34.832 qs_scf_new_mos 108 7.5 0.000 0.001 33.331 33.641 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.331 33.640 ot_scf_mini 108 9.5 0.002 0.002 31.641 31.855 multiply_cannon 2286 13.5 0.183 0.190 26.204 27.871 velocity_verlet 10 3.0 0.001 0.001 27.254 27.254 multiply_cannon_loop 2286 14.5 1.500 1.583 25.368 27.009 ot_mini 108 10.5 0.001 0.001 19.961 20.206 qs_ot_get_derivative 108 11.5 0.001 0.001 16.860 17.065 mp_waitall_1 245248 16.5 9.048 14.936 9.048 14.936 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.172 13.203 multiply_cannon_multrec 54864 15.5 4.185 6.497 7.470 10.875 rebuild_ks_matrix 119 8.3 0.000 0.000 8.204 8.340 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.203 8.340 multiply_cannon_sync_h2d 54864 15.5 5.823 7.607 5.823 7.607 mp_sum_l 7207 12.9 5.766 7.514 5.766 7.514 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.214 7.335 qs_ot_get_p 119 10.4 0.001 0.001 6.834 7.151 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.714 6.172 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.590 5.705 init_scf_run 11 5.9 0.000 0.001 4.862 4.862 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.862 4.862 sum_up_and_integrate 119 10.3 0.012 0.015 4.793 4.799 integrate_v_rspace 119 11.3 0.002 0.002 4.781 4.789 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.508 4.743 calculate_rho_elec 119 8.7 0.011 0.016 4.507 4.742 dbcsr_mm_accdrv_process 76910 16.1 1.176 1.825 3.207 4.509 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.824 3.898 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.962 3.420 rs_pw_transfer 974 11.9 0.011 0.012 3.082 3.205 calculate_dm_sparse 119 9.5 0.000 0.000 2.998 3.183 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.957 3.172 apply_single 119 13.6 0.000 0.000 2.957 3.172 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.948 2.948 ot_diis_step 108 11.5 0.006 0.006 2.823 2.824 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.627 2.681 jit_kernel_multiply 13 15.8 1.968 2.681 1.968 2.681 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.676 2.676 cp_fm_redistribute_end 50 14.0 2.432 2.650 2.437 2.651 density_rs2pw 119 9.7 0.004 0.004 2.503 2.590 cp_fm_diag_elpa_base 50 14.0 0.211 2.552 0.212 2.561 calculate_first_density_matrix 1 7.0 0.000 0.000 2.509 2.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.281 2.283 wfi_extrapolate 11 7.9 0.001 0.001 2.259 2.259 init_scf_loop 11 6.9 0.000 0.000 2.194 2.194 acc_transpose_blocks 54864 15.5 0.231 0.250 1.741 2.168 potential_pw2rs 119 12.3 0.004 0.004 2.127 2.145 make_m2s 4572 13.5 0.055 0.058 2.056 2.125 grid_integrate_task_list 119 12.3 2.007 2.122 2.007 2.122 pw_transfer 1439 11.6 0.051 0.057 1.978 2.055 make_images 4572 14.5 0.133 0.138 1.971 2.040 mp_sum_d 4125 12.0 1.409 2.024 1.409 2.024 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.965 2.024 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.902 1.981 mp_alltoall_d11v 2130 13.8 1.551 1.815 1.551 1.815 fft3d_ps 1201 14.6 0.360 0.464 1.681 1.756 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.481 1.538 fft_wrap_pw1pw2_140 487 13.2 0.079 0.093 1.453 1.530 mp_waitany 12084 13.8 1.291 1.454 1.291 1.454 grid_collocate_task_list 119 9.7 1.292 1.363 1.292 1.363 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.800 1.220 prepare_preconditioner 11 7.9 0.000 0.000 1.104 1.133 make_preconditioner 11 8.9 0.000 0.000 1.104 1.132 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.061000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 486.494208E+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 1421557. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 39.763 39.763 qs_mol_dyn_low 1 2.0 0.003 0.003 39.503 39.511 qs_forces 11 3.9 0.001 0.002 39.411 39.412 qs_energies 11 4.9 0.001 0.006 37.719 37.722 scf_env_do_scf 11 5.9 0.000 0.001 32.560 32.560 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.865 29.865 dbcsr_multiply_generic 2286 12.5 0.100 0.102 21.705 22.030 qs_scf_new_mos 108 7.5 0.001 0.001 20.552 20.795 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.551 20.795 ot_scf_mini 108 9.5 0.002 0.003 19.618 19.791 velocity_verlet 10 3.0 0.001 0.001 19.006 19.008 multiply_cannon 2286 13.5 0.208 0.217 16.399 18.156 multiply_cannon_loop 2286 14.5 0.904 0.980 15.232 16.529 ot_mini 108 10.5 0.001 0.001 12.091 12.327 mp_waitall_1 200699 16.5 5.951 11.032 5.951 11.032 qs_ot_get_derivative 108 11.5 0.001 0.001 9.598 9.775 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.201 9.340 multiply_cannon_multrec 27432 15.5 1.970 4.334 5.846 8.576 rebuild_ks_matrix 119 8.3 0.000 0.000 7.291 7.437 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.291 7.437 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.458 6.591 dbcsr_mm_accdrv_process 47894 16.0 3.065 5.199 3.807 5.632 qs_ot_get_p 119 10.4 0.001 0.001 4.573 4.807 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.717 4.554 sum_up_and_integrate 119 10.3 0.024 0.027 4.315 4.319 integrate_v_rspace 119 11.3 0.002 0.002 4.290 4.298 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.052 4.134 apply_single 119 13.6 0.000 0.000 3.051 4.134 mp_sum_l 7207 12.9 2.127 4.010 2.127 4.010 init_scf_run 11 5.9 0.000 0.001 3.895 3.896 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.895 3.895 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.819 3.846 calculate_rho_elec 119 8.7 0.021 0.024 3.818 3.845 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.054 3.073 rs_pw_transfer 974 11.9 0.010 0.011 2.894 2.984 make_m2s 4572 13.5 0.053 0.056 2.729 2.947 multiply_cannon_sync_h2d 27432 15.5 2.187 2.859 2.187 2.859 make_images 4572 14.5 0.199 0.237 2.639 2.855 init_scf_loop 11 6.9 0.000 0.001 2.672 2.672 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.612 2.613 ot_diis_step 108 11.5 0.011 0.011 2.435 2.435 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.247 2.339 calculate_first_density_matrix 1 7.0 0.000 0.001 2.289 2.292 density_rs2pw 119 9.7 0.004 0.004 2.185 2.277 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.247 2.248 cp_fm_redistribute_end 50 14.0 1.857 2.218 1.861 2.220 cp_fm_diag_elpa_base 50 14.0 0.343 2.095 0.356 2.158 calculate_dm_sparse 119 9.5 0.000 0.000 2.079 2.157 potential_pw2rs 119 12.3 0.006 0.006 2.058 2.073 grid_integrate_task_list 119 12.3 1.828 1.954 1.828 1.954 jit_kernel_multiply 10 16.2 0.691 1.943 0.691 1.943 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.935 1.938 pw_transfer 1439 11.6 0.063 0.066 1.793 1.824 prepare_preconditioner 11 7.9 0.000 0.000 1.764 1.792 make_preconditioner 11 8.9 0.000 0.001 1.764 1.792 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.703 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.682 1.724 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.656 1.713 make_images_data 4572 15.5 0.045 0.051 1.204 1.637 wfi_extrapolate 11 7.9 0.001 0.001 1.552 1.552 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.048 1.524 acc_transpose_blocks 27432 15.5 0.108 0.113 1.192 1.508 fft3d_ps 1201 14.6 0.495 0.550 1.412 1.441 mp_alltoall_d11v 2130 13.8 1.276 1.384 1.276 1.384 mp_allgather_i34 2286 14.5 0.611 1.368 0.611 1.368 grid_collocate_task_list 119 9.7 1.240 1.319 1.240 1.319 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.271 1.306 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.266 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.207 1.257 mp_sum_d 4125 12.0 0.603 1.032 0.603 1.032 make_images_sizes 4572 15.5 0.005 0.005 0.707 0.996 mp_alltoall_i44 4572 16.5 0.702 0.991 0.702 0.991 qs_energies_init_hamiltonians 11 5.9 0.000 0.003 0.966 0.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.891 0.904 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.611 0.832 0.884 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.658 0.883 cp_fm_cholesky_invert 11 10.9 0.838 0.843 0.838 0.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.763000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.636364, yerr=1.149919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.681792E+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_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.072 33.492 33.493 qs_mol_dyn_low 1 2.0 0.003 0.004 33.181 33.188 qs_forces 11 3.9 0.002 0.002 33.120 33.121 qs_energies 11 4.9 0.001 0.004 31.503 31.506 scf_env_do_scf 11 5.9 0.001 0.005 26.767 26.767 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.933 23.934 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.879 17.072 velocity_verlet 10 3.0 0.001 0.002 16.115 16.116 qs_scf_new_mos 108 7.5 0.001 0.001 15.590 15.607 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.590 15.607 ot_scf_mini 108 9.5 0.002 0.003 14.838 14.850 multiply_cannon 2286 13.5 0.196 0.204 13.213 14.080 multiply_cannon_loop 2286 14.5 0.639 0.674 12.336 13.377 ot_mini 108 10.5 0.001 0.001 9.141 9.161 qs_ot_get_derivative 108 11.5 0.001 0.001 7.576 7.590 multiply_cannon_multrec 18288 15.5 1.944 2.931 6.743 7.123 rebuild_ks_matrix 119 8.3 0.000 0.000 6.603 6.638 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.603 6.638 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.817 5.844 dbcsr_mm_accdrv_process 38222 16.0 4.087 5.277 4.717 5.499 mp_waitall_1 158411 16.6 3.044 4.697 3.044 4.697 sum_up_and_integrate 119 10.3 0.030 0.031 4.193 4.200 integrate_v_rspace 119 11.3 0.002 0.003 4.163 4.173 init_scf_run 11 5.9 0.000 0.001 3.543 3.543 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.542 3.543 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.854 3.477 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.457 3.471 calculate_rho_elec 119 8.7 0.030 0.031 3.457 3.470 qs_ot_get_p 119 10.4 0.001 0.001 3.438 3.469 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.516 2.948 init_scf_loop 11 6.9 0.001 0.003 2.810 2.810 rs_pw_transfer 974 11.9 0.009 0.010 2.624 2.780 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.066 2.401 apply_single 119 13.6 0.000 0.000 2.066 2.401 calculate_first_density_matrix 1 7.0 0.000 0.003 2.272 2.273 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.240 2.253 make_m2s 4572 13.5 0.046 0.046 1.984 2.198 density_rs2pw 119 9.7 0.004 0.004 2.038 2.176 make_images 4572 14.5 0.190 0.201 1.897 2.111 prepare_preconditioner 11 7.9 0.000 0.000 1.974 1.977 make_preconditioner 11 8.9 0.000 0.001 1.974 1.977 potential_pw2rs 119 12.3 0.007 0.008 1.961 1.969 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.941 1.941 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.821 1.906 grid_integrate_task_list 119 12.3 1.796 1.877 1.796 1.877 calculate_dm_sparse 119 9.5 0.001 0.005 1.848 1.863 pw_transfer 1439 11.6 0.064 0.066 1.821 1.834 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.730 1.741 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.706 1.714 mp_sum_l 7207 12.9 1.331 1.708 1.331 1.708 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.700 1.702 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.663 1.665 cp_fm_redistribute_end 50 14.0 1.239 1.636 1.240 1.637 cp_fm_diag_elpa_base 50 14.0 0.380 1.541 0.394 1.582 multiply_cannon_sync_h2d 18288 15.5 1.386 1.558 1.386 1.558 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.535 1.548 ot_diis_step 108 11.5 0.011 0.011 1.532 1.532 jit_kernel_multiply 8 16.4 0.580 1.494 0.580 1.494 fft3d_ps 1201 14.6 0.505 0.521 1.417 1.430 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.307 1.321 grid_collocate_task_list 119 9.7 1.204 1.317 1.204 1.317 acc_transpose_blocks 18288 15.5 0.075 0.077 1.220 1.252 wfi_extrapolate 11 7.9 0.001 0.001 1.223 1.223 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.197 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.430 1.151 make_images_data 4572 15.5 0.044 0.048 0.884 1.120 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.751 1.007 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.945 0.962 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.958 0.959 mp_alltoall_d11v 2130 13.8 0.782 0.902 0.782 0.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.859 0.865 cp_fm_cholesky_invert 11 10.9 0.857 0.861 0.857 0.861 mp_alltoall_z22v 1201 16.6 0.740 0.803 0.740 0.803 acc_transpose_blocks_kernels 18288 16.5 0.210 0.221 0.777 0.795 mp_allgather_i34 2286 14.5 0.339 0.723 0.339 0.723 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.722 mp_waitany 9880 13.7 0.557 0.693 0.557 0.693 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.533 0.678 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.493000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.454545, yerr=2.641062 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 544.464896E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 36.509 36.509 qs_mol_dyn_low 1 2.0 0.003 0.003 36.203 36.210 qs_forces 11 3.9 0.001 0.002 36.140 36.144 qs_energies 11 4.9 0.001 0.001 34.378 34.387 scf_env_do_scf 11 5.9 0.000 0.001 29.300 29.302 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.574 25.574 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.656 18.767 velocity_verlet 10 3.0 0.001 0.001 18.757 18.759 qs_scf_new_mos 108 7.5 0.001 0.001 17.082 17.134 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.081 17.133 ot_scf_mini 108 9.5 0.002 0.003 16.117 16.172 multiply_cannon 2286 13.5 0.221 0.241 14.607 15.067 multiply_cannon_loop 2286 14.5 0.941 0.974 13.555 13.842 ot_mini 108 10.5 0.001 0.001 9.626 9.686 multiply_cannon_multrec 27432 15.5 2.311 2.944 8.569 8.959 qs_ot_get_derivative 108 11.5 0.001 0.001 7.770 7.819 dbcsr_mm_accdrv_process 47916 15.9 5.405 7.206 6.165 7.390 rebuild_ks_matrix 119 8.3 0.000 0.000 6.646 6.707 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.646 6.707 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.874 5.925 sum_up_and_integrate 119 10.3 0.035 0.037 3.906 3.913 integrate_v_rspace 119 11.3 0.002 0.003 3.871 3.878 qs_ot_get_p 119 10.4 0.001 0.001 3.771 3.855 init_scf_loop 11 6.9 0.000 0.000 3.700 3.700 init_scf_run 11 5.9 0.000 0.001 3.696 3.696 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.696 3.696 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.522 3.554 calculate_rho_elec 119 8.7 0.040 0.046 3.521 3.553 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.930 3.301 prepare_preconditioner 11 7.9 0.000 0.000 2.849 2.857 make_preconditioner 11 8.9 0.000 0.000 2.849 2.857 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.452 2.782 mp_waitall_1 137007 16.6 2.216 2.642 2.216 2.642 make_m2s 4572 13.5 0.055 0.057 2.498 2.589 make_images 4572 14.5 0.271 0.332 2.388 2.477 rs_pw_transfer 974 11.9 0.009 0.009 2.298 2.407 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.278 2.288 calculate_first_density_matrix 1 7.0 0.000 0.000 2.208 2.211 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.032 2.150 apply_single 119 13.6 0.000 0.000 2.031 2.150 calculate_dm_sparse 119 9.5 0.000 0.000 2.060 2.123 density_rs2pw 119 9.7 0.004 0.004 1.986 2.084 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.054 2.079 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.924 1.925 grid_integrate_task_list 119 12.3 1.829 1.914 1.829 1.914 jit_kernel_multiply 10 15.9 0.700 1.874 0.700 1.874 pw_transfer 1439 11.6 0.063 0.066 1.823 1.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.849 1.850 ot_diis_step 108 11.5 0.012 0.012 1.810 1.811 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.733 1.767 potential_pw2rs 119 12.3 0.008 0.009 1.677 1.683 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.635 1.650 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.608 1.610 cp_fm_redistribute_end 50 14.0 1.067 1.582 1.068 1.583 cp_fm_diag_elpa_base 50 14.0 0.491 1.504 0.512 1.544 acc_transpose_blocks 27432 15.5 0.113 0.116 1.456 1.479 mp_sum_l 7207 12.9 1.129 1.444 1.129 1.444 fft3d_ps 1201 14.6 0.531 0.587 1.415 1.440 wfi_extrapolate 11 7.9 0.001 0.001 1.432 1.433 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.366 1.402 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.868 1.354 grid_collocate_task_list 119 9.7 1.219 1.322 1.219 1.322 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.228 1.236 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.190 1.207 dbcsr_complete_redistribute 329 12.2 0.117 0.143 0.891 1.171 cp_fm_upper_to_full 72 13.5 0.811 1.126 0.811 1.126 make_images_data 4572 15.5 0.045 0.049 0.982 1.097 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.084 1.088 multiply_cannon_sync_h2d 27432 15.5 0.976 1.032 0.976 1.032 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.856 1.015 cp_fm_cholesky_invert 11 10.9 0.956 0.961 0.956 0.961 mp_alltoall_d11v 2130 13.8 0.877 0.959 0.877 0.959 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.626 0.898 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.865 0.873 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.869 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.828 0.838 mp_alltoall_z22v 1201 16.6 0.767 0.799 0.767 0.799 mp_alltoall_i22 627 13.8 0.461 0.763 0.461 0.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.509000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.363636, yerr=2.637931 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 598.908928E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 30.073 30.074 qs_mol_dyn_low 1 2.0 0.003 0.004 29.900 29.907 qs_forces 11 3.9 0.002 0.002 29.836 29.837 qs_energies 11 4.9 0.001 0.001 28.107 28.111 scf_env_do_scf 11 5.9 0.000 0.001 23.273 23.273 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.473 20.473 velocity_verlet 10 3.0 0.001 0.002 15.275 15.278 dbcsr_multiply_generic 2286 12.5 0.092 0.095 13.039 13.123 qs_scf_new_mos 108 7.5 0.001 0.001 12.286 12.328 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.285 12.327 ot_scf_mini 108 9.5 0.002 0.002 11.562 11.599 multiply_cannon 2286 13.5 0.230 0.239 10.044 10.630 multiply_cannon_loop 2286 14.5 0.331 0.344 9.004 9.224 ot_mini 108 10.5 0.001 0.001 6.370 6.415 rebuild_ks_matrix 119 8.3 0.000 0.000 6.170 6.196 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.169 6.196 multiply_cannon_multrec 9144 15.5 1.644 1.870 5.895 6.113 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.531 5.555 qs_ot_get_derivative 108 11.5 0.001 0.001 4.992 5.028 dbcsr_mm_accdrv_process 12550 15.8 3.199 4.010 4.149 4.236 sum_up_and_integrate 119 10.3 0.037 0.041 3.848 3.854 integrate_v_rspace 119 11.3 0.002 0.003 3.811 3.817 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.567 3.582 calculate_rho_elec 119 8.7 0.059 0.061 3.566 3.581 init_scf_run 11 5.9 0.000 0.001 3.386 3.386 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.386 3.386 qs_ot_get_p 119 10.4 0.001 0.001 3.280 3.334 init_scf_loop 11 6.9 0.000 0.000 2.776 2.776 mp_waitall_1 115863 16.7 2.050 2.656 2.050 2.656 make_m2s 4572 13.5 0.036 0.037 2.063 2.322 make_images 4572 14.5 0.267 0.303 1.972 2.230 rs_pw_transfer 974 11.9 0.008 0.008 2.110 2.194 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.155 2.158 calculate_first_density_matrix 1 7.0 0.000 0.000 2.111 2.113 density_rs2pw 119 9.7 0.003 0.004 2.005 2.090 prepare_preconditioner 11 7.9 0.000 0.000 2.004 2.010 make_preconditioner 11 8.9 0.000 0.000 2.004 2.010 pw_transfer 1439 11.6 0.064 0.066 1.930 1.941 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.929 1.929 grid_integrate_task_list 119 12.3 1.846 1.929 1.846 1.929 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.886 1.917 jit_kernel_multiply 10 15.8 0.913 1.852 0.913 1.852 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.838 1.849 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.771 1.787 calculate_dm_sparse 119 9.5 0.000 0.000 1.752 1.774 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.620 1.623 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.604 1.606 potential_pw2rs 119 12.3 0.010 0.010 1.586 1.596 cp_fm_redistribute_end 50 14.0 0.808 1.585 0.809 1.586 cp_fm_diag_elpa_base 50 14.0 0.727 1.490 0.773 1.551 fft3d_ps 1201 14.6 0.542 0.553 1.507 1.520 fft_wrap_pw1pw2_140 487 13.2 0.081 0.085 1.428 1.440 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.421 1.430 grid_collocate_task_list 119 9.7 1.275 1.386 1.275 1.386 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.356 1.376 ot_diis_step 108 11.5 0.012 0.012 1.355 1.355 make_images_data 4572 15.5 0.039 0.042 0.987 1.324 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.887 1.241 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.210 1.237 apply_single 119 13.6 0.000 0.000 1.210 1.237 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.224 1.225 wfi_extrapolate 11 7.9 0.001 0.001 1.218 1.218 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.130 1.137 cp_fm_cholesky_invert 11 10.9 1.038 1.042 1.038 1.042 mp_alltoall_d11v 2130 13.8 0.845 0.971 0.845 0.971 acc_transpose_blocks 9144 15.5 0.038 0.040 0.924 0.931 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.919 mp_alltoall_z22v 1201 16.6 0.838 0.865 0.838 0.865 mp_allgather_i34 2286 14.5 0.374 0.853 0.374 0.853 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.367 0.820 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.804 0.815 multiply_cannon_sync_h2d 9144 15.5 0.711 0.784 0.711 0.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.764 0.768 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.677 0.729 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.682 0.686 dbcsr_complete_redistribute 329 12.2 0.158 0.169 0.607 0.638 mp_sum_l 7207 12.9 0.508 0.630 0.508 0.630 make_images_sizes 4572 15.5 0.005 0.005 0.413 0.623 mp_alltoall_i44 4572 16.5 0.409 0.618 0.409 0.618 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.507 0.609 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.074000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.363636, yerr=3.937529 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 764.022784E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 44.194 44.194 qs_mol_dyn_low 1 2.0 0.003 0.004 43.861 43.869 qs_forces 11 3.9 0.001 0.002 43.797 43.799 qs_energies 11 4.9 0.001 0.001 41.810 41.816 scf_env_do_scf 11 5.9 0.001 0.001 35.865 35.865 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.492 27.494 velocity_verlet 10 3.0 0.001 0.001 24.592 24.597 dbcsr_multiply_generic 2286 12.5 0.100 0.101 18.767 19.009 qs_scf_new_mos 108 7.5 0.001 0.001 17.863 17.958 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.862 17.957 ot_scf_mini 108 9.5 0.002 0.002 16.722 16.826 multiply_cannon 2286 13.5 0.307 0.318 14.189 15.249 multiply_cannon_loop 2286 14.5 0.344 0.348 12.813 13.815 ot_mini 108 10.5 0.001 0.001 9.732 9.860 multiply_cannon_multrec 9144 15.5 3.421 4.768 8.622 8.705 init_scf_loop 11 6.9 0.000 0.000 8.334 8.338 qs_ot_get_derivative 108 11.5 0.001 0.001 7.549 7.653 rebuild_ks_matrix 119 8.3 0.000 0.000 7.275 7.429 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.274 7.429 prepare_preconditioner 11 7.9 0.000 0.000 7.338 7.352 make_preconditioner 11 8.9 0.000 0.000 7.338 7.352 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.906 7.211 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.592 6.733 dbcsr_mm_accdrv_process 12550 15.8 4.119 5.661 5.078 6.432 cp_fm_upper_to_full 72 14.2 3.190 4.545 3.190 4.545 qs_ot_get_p 119 10.4 0.001 0.001 4.097 4.250 mp_waitall_1 94719 16.7 3.111 4.121 3.111 4.121 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.096 4.102 calculate_rho_elec 119 8.7 0.118 0.121 4.095 4.101 sum_up_and_integrate 119 10.3 0.064 0.066 4.009 4.015 integrate_v_rspace 119 11.3 0.003 0.003 3.944 3.950 init_scf_run 11 5.9 0.000 0.001 3.860 3.860 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.859 3.860 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.693 3.102 make_m2s 4572 13.5 0.039 0.040 2.745 2.972 dbcsr_complete_redistribute 329 12.2 0.285 0.286 2.091 2.922 make_images 4572 14.5 0.352 0.384 2.622 2.849 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.642 2.585 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.559 2.560 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.282 2.554 apply_single 119 13.6 0.000 0.000 2.282 2.554 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.702 2.517 calculate_dm_sparse 119 9.5 0.000 0.000 2.313 2.334 pw_transfer 1439 11.6 0.066 0.067 2.313 2.319 calculate_first_density_matrix 1 7.0 0.000 0.000 2.293 2.295 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.249 2.249 mp_alltoall_i22 627 13.8 1.436 2.240 1.436 2.240 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.427 2.237 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.217 2.224 density_rs2pw 119 9.7 0.003 0.003 2.127 2.152 ot_diis_step 108 11.5 0.014 0.014 2.139 2.140 mp_sum_l 7207 12.9 1.300 2.087 1.300 2.087 grid_integrate_task_list 119 12.3 2.045 2.068 2.045 2.068 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.997 2.052 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.997 1.998 cp_fm_cholesky_invert 11 10.9 1.854 1.859 1.854 1.859 fft3d_ps 1201 14.6 0.569 0.579 1.852 1.857 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.856 1.857 cp_fm_diag_elpa_base 50 14.0 1.704 1.755 1.853 1.854 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.794 1.802 rs_pw_transfer 974 11.9 0.009 0.009 1.767 1.791 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.779 1.782 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.671 1.690 make_images_data 4572 15.5 0.043 0.046 1.357 1.680 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.378 1.675 jit_kernel_multiply 8 15.4 0.932 1.508 0.932 1.508 grid_collocate_task_list 119 9.7 1.479 1.502 1.479 1.502 wfi_extrapolate 11 7.9 0.001 0.001 1.499 1.499 potential_pw2rs 119 12.3 0.014 0.014 1.462 1.468 mp_alltoall_d11v 2130 13.8 1.385 1.411 1.385 1.411 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.345 1.354 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.228 1.257 mp_alltoall_z22v 1201 16.6 1.146 1.168 1.146 1.168 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.116 1.131 multiply_cannon_sync_h2d 9144 15.5 1.044 1.047 1.044 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.979 1.020 acc_transpose_blocks 9144 15.5 0.038 0.038 1.001 1.004 qs_create_task_list 11 7.9 0.001 0.001 0.952 0.964 generate_qs_task_list 11 8.9 0.372 0.391 0.952 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.896 0.910 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.194000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.727273, yerr=14.997520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 501.301248E+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 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 85.555 85.556 qs_mol_dyn_low 1 2.0 0.003 0.004 85.316 85.325 qs_forces 11 3.9 0.002 0.002 85.238 85.242 qs_energies 11 4.9 0.001 0.001 82.309 82.327 scf_env_do_scf 11 5.9 0.000 0.001 73.359 73.361 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 67.213 67.214 dbcsr_multiply_generic 2055 12.4 0.107 0.112 52.880 53.140 qs_scf_new_mos 99 7.5 0.000 0.001 49.451 49.580 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.451 49.580 ot_scf_mini 99 9.5 0.002 0.002 47.001 47.090 multiply_cannon 2055 13.4 0.178 0.184 42.784 43.757 velocity_verlet 10 3.0 0.001 0.001 43.493 43.494 multiply_cannon_loop 2055 14.4 1.546 1.588 41.638 42.545 ot_mini 99 10.5 0.001 0.001 28.327 28.420 qs_ot_get_derivative 99 11.5 0.001 0.001 21.489 21.572 multiply_cannon_multrec 49320 15.4 12.117 12.812 17.133 17.826 rebuild_ks_matrix 110 8.3 0.000 0.001 14.909 15.038 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.909 15.038 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.099 13.210 mp_waitall_1 220248 16.4 12.123 13.158 12.123 13.158 multiply_cannon_sync_h2d 49320 15.4 9.927 10.729 9.927 10.729 qs_ot_get_p 110 10.4 0.001 0.001 10.057 10.187 multiply_cannon_metrocomm3 49320 15.4 0.078 0.084 6.963 8.118 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.330 7.883 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.200 7.740 apply_single 110 13.6 0.000 0.001 7.200 7.739 sum_up_and_integrate 110 10.3 0.036 0.042 7.434 7.447 integrate_v_rspace 110 11.3 0.002 0.003 7.398 7.420 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.994 7.060 init_scf_run 11 5.9 0.000 0.001 6.826 6.826 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.826 6.826 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.688 6.818 calculate_rho_elec 110 8.6 0.020 0.024 6.688 6.817 ot_diis_step 99 11.5 0.005 0.006 6.572 6.573 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.528 6.552 mp_sum_l 6514 12.8 5.234 6.099 5.234 6.099 init_scf_loop 11 6.9 0.000 0.000 6.094 6.095 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.515 5.515 dbcsr_mm_accdrv_process 87628 16.1 2.054 2.161 4.896 5.227 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.909 4.912 cp_fm_redistribute_end 48 14.0 4.271 4.868 4.276 4.870 cp_fm_diag_elpa_base 48 14.0 0.586 4.700 0.589 4.732 make_m2s 4110 13.4 0.063 0.067 4.404 4.513 make_images 4110 14.4 0.178 0.191 4.306 4.415 rs_pw_transfer 902 11.9 0.012 0.013 4.228 4.392 wfi_extrapolate 11 7.9 0.001 0.001 4.159 4.160 calculate_dm_sparse 110 9.5 0.000 0.001 3.828 3.940 density_rs2pw 110 9.6 0.004 0.004 3.732 3.894 prepare_preconditioner 11 7.9 0.000 0.000 3.838 3.855 make_preconditioner 11 8.9 0.000 0.000 3.838 3.855 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.838 3.848 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.568 3.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.586 3.590 grid_integrate_task_list 110 12.3 3.244 3.455 3.244 3.455 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.329 3.384 pw_transfer 1331 11.6 0.054 0.062 3.269 3.347 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.182 3.263 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.208 3.244 potential_pw2rs 110 12.3 0.006 0.007 2.980 3.021 fft3d_ps 1111 14.6 0.753 0.839 2.721 2.784 mp_alltoall_d11v 2046 13.8 2.364 2.745 2.364 2.745 fft_wrap_pw1pw2_140 451 13.1 0.168 0.186 2.565 2.650 jit_kernel_multiply 13 15.9 2.558 2.609 2.558 2.609 calculate_first_density_matrix 1 7.0 0.000 0.000 2.576 2.583 acc_transpose_blocks 49320 15.4 0.222 0.233 2.159 2.240 grid_collocate_task_list 110 9.6 2.086 2.215 2.086 2.215 make_images_data 4110 15.4 0.042 0.045 1.988 2.138 mp_waitany 14300 13.8 1.911 2.118 1.911 2.118 cp_fm_cholesky_invert 11 10.9 2.033 2.038 2.033 2.038 hybrid_alltoall_any 4261 16.3 0.080 0.474 1.723 2.004 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.950 1.973 mp_sum_d 3879 11.9 1.491 1.927 1.491 1.927 mp_alltoall_z22v 1111 16.6 1.606 1.760 1.606 1.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.556000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.822577 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.423744E+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 1765220. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.060 70.022 70.024 qs_mol_dyn_low 1 2.0 0.004 0.009 69.749 69.757 qs_forces 11 3.9 0.002 0.002 69.673 69.674 qs_energies 11 4.9 0.001 0.001 66.260 66.264 scf_env_do_scf 11 5.9 0.000 0.001 57.664 57.667 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.543 49.543 dbcsr_multiply_generic 2055 12.4 0.114 0.119 37.929 38.106 velocity_verlet 10 3.0 0.001 0.002 36.896 36.898 qs_scf_new_mos 99 7.5 0.001 0.001 33.195 33.345 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.194 33.345 multiply_cannon 2055 13.4 0.222 0.245 31.016 32.153 ot_scf_mini 99 9.5 0.003 0.003 31.511 31.639 multiply_cannon_loop 2055 14.4 0.930 0.959 29.643 30.559 ot_mini 99 10.5 0.001 0.001 18.483 18.627 multiply_cannon_multrec 24660 15.4 7.641 9.105 13.848 15.120 rebuild_ks_matrix 110 8.3 0.000 0.001 13.881 13.975 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.881 13.975 qs_ot_get_derivative 99 11.5 0.001 0.001 12.670 12.798 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.251 12.340 mp_waitall_1 176588 16.5 7.882 10.594 7.882 10.594 multiply_cannon_sync_h2d 24660 15.4 6.980 8.133 6.980 8.133 init_scf_loop 11 6.9 0.000 0.000 8.082 8.082 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.120 7.565 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.528 7.246 apply_single 110 13.6 0.000 0.001 6.528 7.246 sum_up_and_integrate 110 10.3 0.053 0.059 6.786 6.797 integrate_v_rspace 110 11.3 0.002 0.002 6.732 6.744 qs_ot_get_p 110 10.4 0.001 0.001 6.354 6.507 dbcsr_mm_accdrv_process 52282 16.1 4.728 5.834 6.046 6.300 init_scf_run 11 5.9 0.000 0.001 6.208 6.208 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.207 6.208 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.047 6.061 calculate_rho_elec 110 8.6 0.039 0.047 6.047 6.060 prepare_preconditioner 11 7.9 0.000 0.000 6.026 6.048 make_preconditioner 11 8.9 0.000 0.000 6.026 6.048 ot_diis_step 99 11.5 0.010 0.010 5.764 5.764 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.577 5.735 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.710 5.435 make_m2s 4110 13.4 0.058 0.062 4.455 4.928 make_images 4110 14.4 0.396 0.439 4.344 4.812 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.347 4.366 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.894 3.895 pw_transfer 1331 11.6 0.066 0.072 3.474 3.617 wfi_extrapolate 11 7.9 0.001 0.001 3.599 3.599 rs_pw_transfer 902 11.9 0.012 0.013 3.335 3.576 density_rs2pw 110 9.6 0.004 0.004 3.321 3.541 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.367 3.509 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.313 3.385 grid_integrate_task_list 110 12.3 3.136 3.301 3.136 3.301 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.293 3.294 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.277 3.279 cp_fm_redistribute_end 48 14.0 2.447 3.245 2.449 3.246 cp_fm_diag_elpa_base 48 14.0 0.761 3.094 0.792 3.182 calculate_dm_sparse 110 9.5 0.001 0.001 2.993 3.032 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.895 2.932 make_images_data 4110 15.4 0.046 0.050 2.424 2.883 fft_wrap_pw1pw2_140 451 13.1 0.199 0.219 2.717 2.862 fft3d_ps 1111 14.6 1.074 1.277 2.725 2.855 cp_fm_cholesky_invert 11 10.9 2.813 2.822 2.813 2.822 hybrid_alltoall_any 4261 16.3 0.101 0.444 2.119 2.796 potential_pw2rs 110 12.3 0.008 0.009 2.654 2.676 mp_sum_l 6514 12.8 1.869 2.552 1.869 2.552 calculate_first_density_matrix 1 7.0 0.000 0.001 2.514 2.517 grid_collocate_task_list 110 9.6 2.118 2.266 2.118 2.266 mp_alltoall_d11v 2046 13.8 1.876 2.203 1.876 2.203 jit_kernel_multiply 11 16.3 0.966 2.160 0.966 2.160 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.988 2.010 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.856 1.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.795 1.806 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.781 1.799 mp_irecv_dv 57340 16.2 0.660 1.682 0.660 1.682 mp_allgather_i34 2055 14.4 0.702 1.663 0.702 1.663 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.552 1.563 acc_transpose_blocks 24660 15.4 0.109 0.113 1.526 1.550 dbcsr_complete_redistribute 325 12.2 0.250 0.307 1.277 1.550 cp_fm_cholesky_decompose 22 10.9 1.542 1.550 1.542 1.550 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.380 1.475 mp_alltoall_z22v 1111 16.6 1.372 1.455 1.372 1.455 mp_waitany 10164 13.8 1.217 1.447 1.217 1.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.024000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.272727, yerr=6.981676 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 661.446656E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.115 62.046 62.047 qs_mol_dyn_low 1 2.0 0.003 0.004 61.675 61.684 qs_forces 11 3.9 0.002 0.003 61.558 61.559 qs_energies 11 4.9 0.003 0.012 58.260 58.272 scf_env_do_scf 11 5.9 0.001 0.003 50.099 50.099 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.860 40.863 velocity_verlet 10 3.0 0.005 0.007 33.547 33.550 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.849 29.102 qs_scf_new_mos 99 7.5 0.001 0.001 25.820 25.924 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.819 25.923 ot_scf_mini 99 9.5 0.003 0.006 24.565 24.679 multiply_cannon 2055 13.4 0.211 0.221 22.154 23.329 multiply_cannon_loop 2055 14.4 0.618 0.630 20.887 21.963 ot_mini 99 10.5 0.001 0.001 13.823 13.934 rebuild_ks_matrix 110 8.3 0.000 0.001 12.584 12.720 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.584 12.720 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.062 11.192 multiply_cannon_multrec 16440 15.4 3.939 5.329 9.667 10.987 mp_waitall_1 139946 16.5 7.318 10.686 7.318 10.686 qs_ot_get_derivative 99 11.5 0.001 0.001 9.393 9.504 init_scf_loop 11 6.9 0.001 0.002 9.202 9.207 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.235 7.431 prepare_preconditioner 11 7.9 0.000 0.000 7.325 7.341 make_preconditioner 11 8.9 0.000 0.001 7.325 7.341 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.610 6.994 sum_up_and_integrate 110 10.3 0.061 0.061 6.760 6.784 integrate_v_rspace 110 11.3 0.002 0.003 6.699 6.723 qs_ot_get_p 110 10.4 0.005 0.032 5.923 6.053 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.815 5.824 calculate_rho_elec 110 8.6 0.058 0.058 5.814 5.823 init_scf_run 11 5.9 0.000 0.001 5.720 5.720 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.720 5.720 dbcsr_mm_accdrv_process 34862 16.1 4.554 5.237 5.583 5.693 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.950 5.396 apply_single 110 13.6 0.000 0.000 4.950 5.395 make_m2s 4110 13.4 0.051 0.053 4.494 4.916 make_images 4110 14.4 0.390 0.508 4.376 4.797 multiply_cannon_sync_h2d 16440 15.4 3.690 4.483 3.690 4.483 ot_diis_step 99 11.5 0.010 0.011 4.399 4.400 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.115 4.120 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.210 3.846 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.737 3.738 pw_transfer 1331 11.6 0.065 0.071 3.427 3.445 grid_integrate_task_list 110 12.3 3.182 3.371 3.182 3.371 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.320 3.341 density_rs2pw 110 9.6 0.004 0.004 3.039 3.186 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.166 3.169 rs_pw_transfer 902 11.9 0.011 0.012 3.007 3.155 cp_fm_redistribute_end 48 14.0 1.977 3.136 1.979 3.137 make_images_data 4110 15.4 0.043 0.047 2.560 3.122 wfi_extrapolate 11 7.9 0.001 0.001 3.103 3.103 cp_fm_diag_elpa_base 48 14.0 1.089 2.971 1.151 3.077 cp_fm_cholesky_invert 11 10.9 2.987 2.993 2.987 2.993 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.246 2.895 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.802 2.804 fft_wrap_pw1pw2_140 451 13.1 0.209 0.215 2.780 2.800 fft3d_ps 1111 14.6 1.068 1.075 2.636 2.655 calculate_dm_sparse 110 9.5 0.001 0.002 2.605 2.651 potential_pw2rs 110 12.3 0.010 0.011 2.547 2.564 calculate_first_density_matrix 1 7.0 0.022 0.090 2.530 2.531 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.434 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.460 2.504 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.875 2.418 grid_collocate_task_list 110 9.6 2.176 2.364 2.176 2.364 mp_alltoall_d11v 2046 13.8 1.944 2.306 1.944 2.306 mp_irecv_dv 48980 15.7 0.805 2.294 0.805 2.294 mp_sum_l 6514 12.8 1.619 2.188 1.619 2.188 dbcsr_complete_redistribute 325 12.2 0.320 0.345 1.647 2.140 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.972 1.973 cp_fm_upper_to_full 70 13.6 1.419 1.894 1.419 1.894 cp_fm_cholesky_decompose 22 10.9 1.734 1.756 1.734 1.756 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.723 1.741 mp_allgather_i34 2055 14.4 0.593 1.714 0.593 1.714 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.105 1.596 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.569 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.506 jit_kernel_multiply 8 16.6 0.641 1.473 0.641 1.473 mp_alltoall_z22v 1111 16.6 1.353 1.391 1.353 1.391 mp_waitany 17072 13.8 1.183 1.358 1.183 1.358 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.292 1.303 rs_gather_matrices 110 12.3 0.138 0.150 0.917 1.291 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.047000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.090909, yerr=10.334888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 736.710656E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.078 65.777 65.778 qs_mol_dyn_low 1 2.0 0.003 0.004 65.383 65.392 qs_forces 11 3.9 0.008 0.045 65.290 65.291 qs_energies 11 4.9 0.002 0.005 61.908 61.912 scf_env_do_scf 11 5.9 0.001 0.001 53.556 53.559 scf_env_do_scf_inner_loop 99 6.5 0.006 0.014 41.530 41.531 velocity_verlet 10 3.0 0.001 0.002 37.193 37.199 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.609 29.772 qs_scf_new_mos 99 7.5 0.001 0.001 27.001 27.105 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.000 27.104 ot_scf_mini 99 9.5 0.003 0.003 25.362 25.473 multiply_cannon 2055 13.4 0.242 0.256 22.421 23.620 multiply_cannon_loop 2055 14.4 0.890 0.908 20.938 21.470 ot_mini 99 10.5 0.001 0.002 14.231 14.354 multiply_cannon_multrec 24660 15.4 4.204 6.817 12.632 13.747 rebuild_ks_matrix 110 8.3 0.000 0.000 11.984 12.086 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.983 12.085 init_scf_loop 11 6.9 0.001 0.002 11.984 11.986 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.624 10.714 prepare_preconditioner 11 7.9 0.000 0.000 10.257 10.272 make_preconditioner 11 8.9 0.000 0.001 10.257 10.272 qs_ot_get_derivative 99 11.5 0.001 0.001 10.047 10.154 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.494 9.932 dbcsr_mm_accdrv_process 52304 16.0 6.894 8.082 8.281 9.247 sum_up_and_integrate 110 10.3 0.068 0.071 6.403 6.418 integrate_v_rspace 110 11.3 0.002 0.003 6.335 6.351 qs_ot_get_p 110 10.4 0.001 0.001 6.092 6.243 mp_waitall_1 121746 16.5 4.429 6.192 4.429 6.192 make_m2s 4110 13.4 0.061 0.064 5.538 5.899 make_images 4110 14.4 0.573 0.697 5.394 5.752 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.658 5.672 calculate_rho_elec 110 8.6 0.077 0.081 5.657 5.671 init_scf_run 11 5.9 0.000 0.001 5.511 5.512 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.511 5.512 cp_fm_upper_to_full 70 13.8 3.274 4.650 3.274 4.650 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.169 4.184 ot_diis_step 99 11.5 0.011 0.016 4.140 4.141 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.028 4.104 apply_single 110 13.6 0.000 0.000 4.028 4.103 dbcsr_complete_redistribute 325 12.2 0.412 0.452 2.657 3.778 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.688 3.688 grid_integrate_task_list 110 12.3 3.283 3.406 3.283 3.406 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.303 3.360 multiply_cannon_sync_h2d 24660 15.4 3.165 3.295 3.165 3.295 pw_transfer 1331 11.6 0.065 0.073 3.242 3.269 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.120 3.221 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.125 3.185 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.137 3.167 hybrid_alltoall_any 4261 16.3 0.119 0.456 2.317 3.166 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.368 3.150 make_images_data 4110 15.4 0.046 0.050 2.782 3.147 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.085 3.087 cp_fm_redistribute_end 48 14.0 1.545 3.050 1.547 3.051 wfi_extrapolate 11 7.9 0.001 0.001 3.043 3.043 cp_fm_diag_elpa_base 48 14.0 1.417 2.897 1.500 3.011 density_rs2pw 110 9.6 0.004 0.004 2.828 2.987 cp_fm_cholesky_invert 11 10.9 2.956 2.965 2.956 2.965 calculate_dm_sparse 110 9.5 0.001 0.001 2.923 2.948 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.755 2.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.809 2.811 mp_alltoall_i22 605 13.7 1.647 2.803 1.647 2.803 rs_pw_transfer 902 11.9 0.010 0.011 2.579 2.758 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.588 2.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.432 2.471 fft3d_ps 1111 14.6 1.062 1.093 2.440 2.458 calculate_first_density_matrix 1 7.0 0.001 0.002 2.382 2.385 grid_collocate_task_list 110 9.6 2.224 2.358 2.224 2.358 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.318 2.319 potential_pw2rs 110 12.3 0.012 0.013 2.209 2.218 mp_alltoall_d11v 2046 13.8 1.791 1.975 1.791 1.975 cp_fm_cholesky_decompose 22 10.9 1.826 1.882 1.826 1.882 jit_kernel_multiply 10 15.6 1.058 1.845 1.058 1.845 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.727 1.760 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.602 1.717 mp_allgather_i34 2055 14.4 0.583 1.632 0.583 1.632 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.606 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.538 acc_transpose_blocks 24660 15.4 0.106 0.108 1.507 1.528 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.842 1.512 mp_sum_l 6514 12.8 0.980 1.507 0.980 1.507 mp_irecv_dv 62702 16.1 0.744 1.434 0.744 1.434 qs_env_update_s_mstruct 11 6.9 0.003 0.022 1.268 1.368 mp_waitany 13376 13.8 1.086 1.322 1.086 1.322 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.778000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.636364, yerr=11.105058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 818.053120E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.048 55.819 55.820 qs_mol_dyn_low 1 2.0 0.003 0.004 55.573 55.582 qs_forces 11 3.9 0.002 0.003 55.507 55.508 qs_energies 11 4.9 0.001 0.001 51.901 51.906 scf_env_do_scf 11 5.9 0.000 0.001 43.468 43.468 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.420 35.421 velocity_verlet 10 3.0 0.001 0.001 31.220 31.223 dbcsr_multiply_generic 2055 12.4 0.109 0.110 23.060 23.190 qs_scf_new_mos 99 7.5 0.001 0.001 21.249 21.316 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.248 21.315 ot_scf_mini 99 9.5 0.002 0.002 19.994 20.027 multiply_cannon 2055 13.4 0.247 0.259 17.412 18.787 multiply_cannon_loop 2055 14.4 0.321 0.333 15.992 16.282 rebuild_ks_matrix 110 8.3 0.000 0.000 11.499 11.524 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.498 11.523 ot_mini 99 10.5 0.001 0.001 10.695 10.720 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.253 10.276 multiply_cannon_multrec 8220 15.4 3.299 4.608 7.579 8.614 init_scf_loop 11 6.9 0.000 0.000 7.999 8.000 mp_waitall_1 103326 16.6 6.034 7.835 6.034 7.835 qs_ot_get_derivative 99 11.5 0.001 0.001 6.836 6.868 prepare_preconditioner 11 7.9 0.000 0.000 6.364 6.368 make_preconditioner 11 8.9 0.000 0.000 6.364 6.368 sum_up_and_integrate 110 10.3 0.080 0.081 6.244 6.258 integrate_v_rspace 110 11.3 0.003 0.003 6.164 6.177 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.931 6.011 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.648 5.662 calculate_rho_elec 110 8.6 0.115 0.115 5.647 5.662 qs_ot_get_p 110 10.4 0.001 0.001 5.372 5.387 dbcsr_mm_accdrv_process 17442 15.9 2.844 3.739 4.151 5.147 init_scf_run 11 5.9 0.000 0.001 5.105 5.105 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.104 5.105 make_m2s 4110 13.4 0.040 0.041 4.309 4.596 make_images 4110 14.4 0.632 0.686 4.177 4.461 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.876 4.360 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.934 3.938 ot_diis_step 99 11.5 0.012 0.012 3.833 3.833 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.741 3.754 apply_single 110 13.6 0.000 0.000 3.741 3.754 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.603 3.604 grid_integrate_task_list 110 12.3 3.351 3.491 3.351 3.491 pw_transfer 1331 11.6 0.065 0.069 3.303 3.312 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.196 3.208 cp_fm_cholesky_invert 11 10.9 3.128 3.132 3.128 3.132 multiply_cannon_sync_h2d 8220 15.4 2.910 3.071 2.910 3.071 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.042 3.043 cp_fm_redistribute_end 48 14.0 0.775 3.007 0.781 3.008 cp_fm_diag_elpa_base 48 14.0 2.035 2.799 2.215 2.967 make_images_data 4110 15.4 0.038 0.043 2.441 2.884 density_rs2pw 110 9.6 0.004 0.004 2.697 2.860 qs_energies_init_hamiltonians 11 5.9 0.006 0.042 2.852 2.853 wfi_extrapolate 11 7.9 0.001 0.001 2.818 2.819 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.719 2.732 hybrid_alltoall_any 4261 16.3 0.199 0.860 2.328 2.717 calculate_dm_sparse 110 9.5 0.001 0.001 2.479 2.524 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.502 2.503 fft3d_ps 1111 14.6 1.112 1.123 2.456 2.471 grid_collocate_task_list 110 9.6 2.328 2.402 2.328 2.402 rs_pw_transfer 902 11.9 0.010 0.011 2.275 2.394 calculate_first_density_matrix 1 7.0 0.000 0.000 2.199 2.200 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.137 2.146 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.008 2.019 potential_pw2rs 110 12.3 0.015 0.015 1.995 2.003 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.988 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.923 1.944 mp_alltoall_d11v 2046 13.8 1.678 1.908 1.678 1.908 cp_fm_cholesky_decompose 22 10.9 1.800 1.818 1.800 1.818 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.652 1.770 mp_allgather_i34 2055 14.4 0.609 1.700 0.609 1.700 dbcsr_complete_redistribute 325 12.2 0.566 0.586 1.500 1.609 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.581 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.429 1.441 qs_create_task_list 11 7.9 0.001 0.003 1.237 1.333 generate_qs_task_list 11 8.9 0.379 0.445 1.235 1.332 jit_kernel_multiply 7 15.8 1.000 1.304 1.000 1.304 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.481 1.273 mp_irecv_dv 24056 15.7 0.457 1.231 0.457 1.231 mp_waitany 9240 13.8 1.097 1.213 1.097 1.213 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.839 1.182 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.147 1.166 mp_alltoall_z22v 1111 16.6 1.110 1.124 1.110 1.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.820000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.000000, yerr=4.729021 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 1.384550E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.053 87.858 87.858 qs_mol_dyn_low 1 2.0 0.003 0.004 87.489 87.498 qs_forces 11 3.9 0.007 0.007 87.421 87.423 qs_energies 11 4.9 0.001 0.001 83.327 83.333 scf_env_do_scf 11 5.9 0.001 0.001 73.270 73.270 velocity_verlet 10 3.0 0.001 0.002 56.053 56.059 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.293 44.294 dbcsr_multiply_generic 2055 12.4 0.120 0.123 29.322 29.425 init_scf_loop 11 6.9 0.000 0.000 28.900 28.903 qs_scf_new_mos 99 7.5 0.001 0.001 27.524 27.592 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.523 27.591 prepare_preconditioner 11 7.9 0.000 0.000 26.957 26.968 make_preconditioner 11 8.9 0.000 0.000 26.957 26.968 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.143 26.436 ot_scf_mini 99 9.5 0.002 0.002 25.704 25.759 multiply_cannon 2055 13.4 0.342 0.361 21.864 22.638 multiply_cannon_loop 2055 14.4 0.341 0.344 20.043 20.342 cp_fm_upper_to_full 70 14.2 12.578 18.118 12.578 18.118 ot_mini 99 10.5 0.001 0.001 14.075 14.133 rebuild_ks_matrix 110 8.3 0.001 0.001 13.348 13.402 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.348 13.402 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.148 12.195 dbcsr_complete_redistribute 325 12.2 1.021 1.041 7.472 10.617 multiply_cannon_multrec 8220 15.4 4.432 4.629 9.694 9.815 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.404 9.554 qs_ot_get_derivative 99 11.5 0.001 0.001 9.422 9.476 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.799 8.929 mp_waitall_1 84994 16.7 7.960 8.880 7.960 8.880 mp_alltoall_i22 605 13.7 5.456 8.595 5.456 8.595 sum_up_and_integrate 110 10.3 0.150 0.152 6.769 6.784 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.650 6.685 calculate_rho_elec 110 8.6 0.227 0.227 6.650 6.684 integrate_v_rspace 110 11.3 0.003 0.003 6.619 6.633 qs_ot_get_p 110 10.4 0.001 0.001 6.335 6.401 make_m2s 4110 13.4 0.045 0.045 5.524 6.040 make_images 4110 14.4 0.873 0.907 5.334 5.850 cp_fm_cholesky_invert 11 10.9 5.820 5.825 5.820 5.825 init_scf_run 11 5.9 0.000 0.001 5.793 5.793 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.792 5.793 dbcsr_mm_accdrv_process 11614 15.7 3.278 3.644 5.120 5.378 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.999 5.263 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.739 5.206 apply_single 110 13.6 0.000 0.000 4.739 5.206 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.674 4.681 ot_diis_step 99 11.5 0.015 0.015 4.620 4.620 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.201 4.202 multiply_cannon_sync_h2d 8220 15.4 3.950 3.955 3.950 3.955 pw_transfer 1331 11.6 0.074 0.074 3.858 3.869 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.742 3.752 hybrid_alltoall_any 4261 16.3 0.255 0.551 2.987 3.732 grid_integrate_task_list 110 12.3 3.661 3.711 3.661 3.711 make_images_data 4110 15.4 0.042 0.044 3.021 3.694 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.644 3.647 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.533 3.533 cp_fm_diag_elpa_base 48 14.0 2.999 3.193 3.531 3.531 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.999 3.464 wfi_extrapolate 11 7.9 0.001 0.001 3.338 3.338 calculate_dm_sparse 110 9.5 0.001 0.001 3.317 3.334 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.211 3.219 density_rs2pw 110 9.6 0.004 0.004 3.120 3.138 fft3d_ps 1111 14.6 1.266 1.272 2.951 2.961 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.916 2.919 grid_collocate_task_list 110 9.6 2.629 2.646 2.629 2.646 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.515 2.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.375 2.399 calculate_first_density_matrix 1 7.0 0.000 0.000 2.348 2.349 rs_pw_transfer 902 11.9 0.010 0.010 2.310 2.346 cp_fm_cholesky_decompose 22 10.9 2.224 2.251 2.224 2.251 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.184 2.239 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.221 mp_alltoall_d11v 2046 13.8 2.168 2.218 2.168 2.218 potential_pw2rs 110 12.3 0.021 0.021 2.175 2.177 qs_create_task_list 11 7.9 0.000 0.001 1.911 1.957 generate_qs_task_list 11 8.9 0.739 0.796 1.911 1.956 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.830 1.862 jit_kernel_multiply 10 15.3 1.643 1.849 1.643 1.849 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.799 1.815 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.806 1.813 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.858000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1244.363636, yerr=65.459596 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 625.868800E+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 1835144. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.071 208.175 208.177 qs_mol_dyn_low 1 2.0 0.006 0.025 207.654 207.667 qs_forces 11 3.9 0.003 0.003 207.535 207.536 qs_energies 11 4.9 0.001 0.001 201.902 201.917 scf_env_do_scf 11 5.9 0.001 0.001 185.073 185.077 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.361 163.362 dbcsr_multiply_generic 2507 12.6 0.179 0.186 124.881 125.675 velocity_verlet 10 3.0 0.001 0.002 125.379 125.381 qs_scf_new_mos 117 7.6 0.001 0.001 123.368 123.666 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.367 123.666 ot_scf_mini 117 9.6 0.003 0.003 116.740 117.013 multiply_cannon 2507 13.6 0.237 0.245 100.866 103.409 multiply_cannon_loop 2507 14.6 2.119 2.171 98.446 101.172 ot_mini 117 10.6 0.001 0.001 66.109 66.442 multiply_cannon_multrec 60168 15.6 33.123 35.141 41.480 43.414 qs_ot_get_derivative 117 11.6 0.001 0.002 41.306 41.608 rebuild_ks_matrix 128 8.3 0.001 0.001 34.022 34.465 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.022 34.465 mp_waitall_1 267128 16.5 28.884 32.287 28.884 32.287 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.576 30.980 multiply_cannon_sync_h2d 60168 15.6 27.232 29.220 27.232 29.220 qs_ot_get_p 128 10.4 0.001 0.001 28.601 28.873 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.255 25.094 apply_single 128 13.6 0.001 0.001 24.255 25.094 ot_diis_step 117 11.6 0.007 0.008 24.477 24.479 qs_ot_p2m_diag 83 11.4 0.077 0.091 21.740 21.877 init_scf_loop 11 6.9 0.000 0.000 21.623 21.625 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.414 19.674 multiply_cannon_metrocomm3 60168 15.6 0.115 0.121 15.452 19.656 cp_dbcsr_syevd 83 12.4 0.004 0.005 19.069 19.070 prepare_preconditioner 11 7.9 0.000 0.000 17.013 17.069 make_preconditioner 11 8.9 0.000 0.000 17.013 17.069 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.229 16.417 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.837 15.846 cp_fm_redistribute_end 83 14.4 12.508 15.746 12.523 15.750 cp_fm_diag_elpa_base 83 14.4 3.176 15.335 3.211 15.444 make_m2s 5014 13.6 0.106 0.114 14.519 14.998 make_images 5014 14.6 0.403 0.421 14.335 14.826 sum_up_and_integrate 128 10.3 0.089 0.107 14.627 14.643 integrate_v_rspace 128 11.3 0.003 0.004 14.538 14.557 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.354 13.492 calculate_rho_elec 128 8.7 0.045 0.063 13.353 13.492 init_scf_run 11 5.9 0.000 0.001 12.605 12.605 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.605 12.605 cp_fm_cholesky_invert 11 10.9 9.854 9.862 9.854 9.862 mp_sum_l 7870 13.0 8.276 9.576 8.276 9.576 wfi_extrapolate 11 7.9 0.001 0.001 9.317 9.318 calculate_dm_sparse 128 9.5 0.001 0.001 8.477 8.563 multiply_cannon_metrocomm1 60168 15.6 0.087 0.092 6.182 8.501 dbcsr_mm_accdrv_process 124484 16.2 3.297 3.447 7.923 8.460 make_images_data 5014 15.6 0.066 0.071 7.107 8.132 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.987 8.118 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.880 7.972 pw_transfer 1547 11.6 0.076 0.108 7.343 7.621 density_rs2pw 128 9.7 0.006 0.007 7.048 7.512 grid_integrate_task_list 128 12.3 7.076 7.506 7.076 7.506 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 7.138 7.394 hybrid_alltoall_any 5200 16.5 0.288 2.258 6.143 7.302 rs_pw_transfer 1046 11.9 0.016 0.019 6.324 6.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.798 6.807 fft3d_ps 1291 14.7 2.092 2.565 5.934 6.128 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.900 6.126 fft_wrap_pw1pw2_140 523 13.2 0.442 0.509 5.895 6.109 mp_alltoall_d11v 2415 14.1 4.619 5.946 4.619 5.946 grid_collocate_task_list 128 9.7 4.727 5.092 4.727 5.092 potential_pw2rs 128 12.3 0.009 0.010 4.991 5.025 cp_fm_cholesky_decompose 22 10.9 4.935 4.950 4.935 4.950 mp_sum_d 4469 12.2 3.582 4.246 3.582 4.246 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.177000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.818182, yerr=6.589498 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 2.166474E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 830.455808E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843365392 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57200. MP_Allreduce 11252 947. MP_Sync 168 MP_Alltoall 1955 4804669. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 197.236 197.237 qs_mol_dyn_low 1 2.0 0.003 0.003 196.858 196.871 qs_forces 11 3.9 0.003 0.004 196.771 196.772 qs_energies 11 4.9 0.001 0.001 189.971 189.983 scf_env_do_scf 11 5.9 0.001 0.001 173.211 173.221 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 137.881 137.884 velocity_verlet 10 3.0 0.001 0.001 123.419 123.420 dbcsr_multiply_generic 2485 12.5 0.184 0.190 99.109 100.500 qs_scf_new_mos 116 7.6 0.001 0.001 98.590 99.114 qs_scf_loop_do_ot 116 8.6 0.001 0.001 98.590 99.113 ot_scf_mini 116 9.6 0.004 0.004 93.744 94.339 multiply_cannon 2485 13.5 0.474 0.527 77.912 82.388 multiply_cannon_loop 2485 14.5 1.245 1.288 74.254 77.123 ot_mini 116 10.6 0.001 0.001 51.137 51.689 mp_waitall_1 212858 16.6 26.649 41.285 26.649 41.285 multiply_cannon_multrec 29820 15.5 21.415 26.013 30.951 35.776 init_scf_loop 11 6.9 0.000 0.000 35.220 35.221 rebuild_ks_matrix 127 8.3 0.001 0.001 33.125 33.705 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 33.124 33.705 prepare_preconditioner 11 7.9 0.000 0.000 30.762 30.845 make_preconditioner 11 8.9 0.000 0.000 30.762 30.845 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.844 30.373 multiply_cannon_metrocomm3 29820 15.5 0.093 0.100 17.061 30.105 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.414 29.974 qs_ot_get_derivative 116 11.6 0.001 0.002 29.118 29.693 qs_ot_get_p 127 10.4 0.001 0.001 24.139 24.861 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.042 23.184 apply_single 127 13.6 0.001 0.001 22.041 23.184 ot_diis_step 116 11.6 0.014 0.015 21.840 21.844 multiply_cannon_sync_h2d 29820 15.5 18.698 20.674 18.698 20.674 qs_ot_p2m_diag 82 11.4 0.186 0.214 18.989 19.032 cp_fm_cholesky_invert 11 10.9 18.209 18.222 18.209 18.222 cp_dbcsr_syevd 82 12.4 0.005 0.005 17.690 17.692 make_m2s 4970 13.5 0.091 0.096 14.991 16.711 make_images 4970 14.5 1.138 1.332 14.780 16.500 sum_up_and_integrate 127 10.3 0.115 0.133 14.648 14.677 integrate_v_rspace 127 11.3 0.003 0.003 14.533 14.565 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.328 14.338 cp_fm_redistribute_end 82 14.4 8.366 14.246 8.380 14.247 cp_fm_diag_elpa_base 82 14.4 5.611 13.679 5.843 14.061 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.340 13.393 calculate_rho_elec 127 8.7 0.087 0.104 13.339 13.393 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.606 12.025 init_scf_run 11 5.9 0.000 0.001 11.847 11.849 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.847 11.848 make_images_data 4970 15.5 0.064 0.071 9.124 11.163 multiply_cannon_metrocomm4 27335 15.5 0.092 0.108 3.807 10.649 hybrid_alltoall_any 5155 16.4 0.339 1.505 7.700 10.437 mp_irecv_dv 68888 16.3 3.617 10.264 3.617 10.264 dbcsr_mm_accdrv_process 61726 16.2 4.493 5.345 9.027 9.593 wfi_extrapolate 11 7.9 0.001 0.001 8.712 8.712 pw_transfer 1535 11.6 0.085 0.102 7.943 8.007 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.720 7.784 cp_fm_cholesky_decompose 22 10.9 7.654 7.737 7.654 7.737 density_rs2pw 127 9.7 0.006 0.006 7.181 7.618 grid_integrate_task_list 127 12.3 7.100 7.527 7.100 7.527 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.385 7.131 mp_sum_l 7804 13.0 4.471 6.791 4.471 6.791 rs_pw_transfer 1038 11.9 0.014 0.016 6.216 6.691 calculate_dm_sparse 127 9.5 0.001 0.001 6.506 6.662 fft_wrap_pw1pw2_140 519 13.2 0.464 0.511 6.566 6.632 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.299 6.308 fft3d_ps 1281 14.7 2.752 2.900 6.104 6.139 mp_allgather_i34 2485 14.5 2.246 5.680 2.246 5.680 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.451 5.526 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.355 5.522 grid_collocate_task_list 127 9.7 4.876 5.243 4.876 5.243 potential_pw2rs 127 12.3 0.015 0.018 5.149 5.172 mp_alltoall_d11v 2401 14.1 4.341 5.081 4.341 5.081 dbcsr_complete_redistribute 393 12.7 0.768 0.845 3.366 4.255 mp_sum_d 4442 12.2 2.783 4.140 2.783 4.140 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=197.237000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.545455, yerr=1.233151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/17/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 2.950886E+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 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 937.885696E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56898. MP_Allreduce 11306 983. MP_Sync 170 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_comm_split 83 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.056 182.301 182.302 qs_mol_dyn_low 1 2.0 0.003 0.004 181.862 181.875 qs_forces 11 3.9 0.003 0.004 181.767 181.775 qs_energies 11 4.9 0.009 0.063 175.155 175.163 scf_env_do_scf 11 5.9 0.001 0.001 159.278 159.279 scf_env_do_scf_inner_loop 118 6.6 0.008 0.046 122.737 122.738 velocity_verlet 10 3.0 0.001 0.002 116.677 116.679 qs_scf_new_mos 118 7.6 0.001 0.001 84.824 85.225 qs_scf_loop_do_ot 118 8.6 0.001 0.001 84.824 85.224 dbcsr_multiply_generic 2527 12.6 0.183 0.187 83.706 84.721 ot_scf_mini 118 9.6 0.004 0.004 80.504 80.934 multiply_cannon 2527 13.6 0.504 0.526 62.621 67.410 multiply_cannon_loop 2527 14.6 0.863 0.887 59.298 61.772 ot_mini 118 10.6 0.001 0.001 43.777 44.190 mp_waitall_1 171878 16.6 26.523 37.162 26.523 37.162 init_scf_loop 11 6.9 0.000 0.002 36.430 36.431 prepare_preconditioner 11 7.9 0.000 0.000 32.325 32.373 make_preconditioner 11 8.9 0.000 0.001 32.325 32.373 rebuild_ks_matrix 129 8.3 0.001 0.001 30.964 31.385 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.964 31.385 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.911 31.301 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.012 28.405 multiply_cannon_metrocomm3 20216 15.6 0.060 0.065 15.601 25.943 multiply_cannon_multrec 20216 15.6 13.502 16.732 22.194 25.514 qs_ot_get_derivative 118 11.6 0.001 0.002 23.717 24.133 qs_ot_get_p 129 10.4 0.001 0.001 21.727 22.194 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.138 21.192 apply_single 129 13.6 0.001 0.001 20.138 21.192 ot_diis_step 118 11.6 0.018 0.018 19.947 19.949 qs_ot_p2m_diag 83 11.4 0.266 0.272 17.238 17.263 make_m2s 5054 13.6 0.083 0.089 15.676 16.658 make_images 5054 14.6 1.186 1.296 15.439 16.419 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.116 16.117 multiply_cannon_sync_h2d 20216 15.6 14.340 15.996 14.340 15.996 cp_fm_cholesky_invert 11 10.9 15.339 15.349 15.339 15.349 sum_up_and_integrate 129 10.3 0.134 0.146 14.833 14.860 integrate_v_rspace 129 11.3 0.003 0.004 14.699 14.724 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.658 13.709 calculate_rho_elec 129 8.7 0.133 0.147 13.657 13.708 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.840 12.844 cp_fm_redistribute_end 83 14.4 4.852 12.750 4.870 12.756 cp_fm_diag_elpa_base 83 14.4 7.434 12.116 7.857 12.588 make_images_data 5054 15.6 0.061 0.069 9.739 11.164 init_scf_run 11 5.9 0.000 0.001 10.669 10.669 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.669 10.669 hybrid_alltoall_any 5240 16.5 0.433 1.985 8.293 10.201 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.347 9.642 multiply_cannon_metrocomm4 17689 15.6 0.062 0.072 3.490 9.231 mp_irecv_dv 50625 16.2 3.368 8.978 3.368 8.978 dbcsr_mm_accdrv_process 41830 16.2 4.469 5.346 8.148 8.311 pw_transfer 1559 11.6 0.086 0.106 8.195 8.303 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.970 8.084 grid_integrate_task_list 129 12.3 7.373 7.865 7.373 7.865 cp_fm_cholesky_decompose 22 10.9 7.744 7.795 7.744 7.795 wfi_extrapolate 11 7.9 0.001 0.001 7.645 7.645 cp_fm_upper_to_full 105 14.5 5.775 7.605 5.775 7.605 density_rs2pw 129 9.7 0.006 0.006 7.007 7.368 fft_wrap_pw1pw2_140 527 13.2 0.480 0.535 6.839 6.955 dbcsr_complete_redistribute 395 12.7 1.166 1.197 4.696 6.516 fft3d_ps 1301 14.7 2.717 2.923 6.230 6.304 rs_pw_transfer 1054 12.0 0.014 0.014 5.747 6.084 calculate_dm_sparse 129 9.5 0.001 0.002 5.938 6.033 grid_collocate_task_list 129 9.7 5.114 5.608 5.114 5.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.532 5.538 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.780 5.427 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.417 5.233 potential_pw2rs 129 12.3 0.020 0.022 5.054 5.080 mp_alltoall_d11v 2423 14.1 4.505 5.051 4.505 5.051 mp_allgather_i34 2527 14.6 1.782 4.910 1.782 4.910 mp_sum_l 7930 13.1 3.596 4.891 3.596 4.891 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.613 4.760 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.228 4.259 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.395 4.149 mp_alltoall_i22 718 14.1 1.953 3.899 1.953 3.899 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.749 3.750 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=182.302000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.272727, yerr=9.106351 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.142800E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 1067. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 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 0.016 0.064 192.894 192.895 qs_mol_dyn_low 1 2.0 0.003 0.003 192.501 192.514 qs_forces 11 3.9 0.003 0.003 192.382 192.393 qs_energies 11 4.9 0.001 0.001 185.324 185.333 scf_env_do_scf 11 5.9 0.001 0.001 168.448 168.460 velocity_verlet 10 3.0 0.001 0.001 127.715 127.718 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.893 119.894 qs_scf_new_mos 117 7.6 0.001 0.001 83.466 83.823 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.465 83.822 dbcsr_multiply_generic 2507 12.6 0.190 0.196 80.068 80.821 ot_scf_mini 117 9.6 0.003 0.003 78.959 79.289 multiply_cannon 2507 13.6 0.553 0.584 55.119 58.021 multiply_cannon_loop 2507 14.6 1.186 1.218 51.185 52.847 init_scf_loop 11 6.9 0.000 0.000 48.426 48.427 prepare_preconditioner 11 7.9 0.000 0.000 44.335 44.362 make_preconditioner 11 8.9 0.000 0.000 44.335 44.362 ot_mini 117 10.6 0.001 0.001 42.971 43.295 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.942 42.944 multiply_cannon_multrec 30084 15.6 13.976 19.429 25.999 30.859 rebuild_ks_matrix 128 8.3 0.001 0.001 29.462 29.790 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.461 29.790 mp_waitall_1 147882 16.7 17.766 27.979 17.766 27.979 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.643 26.934 qs_ot_get_derivative 117 11.6 0.001 0.002 23.246 23.579 qs_ot_get_p 128 10.4 0.001 0.001 22.162 22.504 make_m2s 5014 13.6 0.098 0.103 20.454 21.803 make_images 5014 14.6 1.952 2.262 20.144 21.491 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.096 19.661 apply_single 128 13.6 0.001 0.001 19.096 19.661 ot_diis_step 117 11.6 0.017 0.018 19.599 19.601 qs_ot_p2m_diag 83 11.4 0.343 0.389 17.757 17.810 cp_fm_cholesky_invert 11 10.9 17.421 17.431 17.421 17.431 cp_fm_upper_to_full 105 14.7 11.210 16.581 11.210 16.581 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.448 16.449 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.619 15.200 sum_up_and_integrate 128 10.3 0.140 0.151 14.493 14.516 integrate_v_rspace 128 11.3 0.003 0.003 14.353 14.382 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.642 13.686 calculate_rho_elec 128 8.7 0.175 0.190 13.642 13.685 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.153 13.157 cp_fm_redistribute_end 83 14.4 2.270 13.008 2.291 13.013 dbcsr_complete_redistribute 395 12.7 1.504 1.626 9.097 12.841 make_images_data 5014 15.6 0.064 0.068 10.912 12.821 cp_fm_diag_elpa_base 83 14.4 10.049 12.330 10.662 12.811 multiply_cannon_sync_h2d 30084 15.6 11.686 12.561 11.686 12.561 dbcsr_mm_accdrv_process 62264 16.2 7.319 8.392 11.596 12.044 hybrid_alltoall_any 5200 16.5 0.524 2.211 9.767 11.608 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.640 11.397 init_scf_run 11 5.9 0.000 0.001 10.940 10.941 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.939 10.941 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.370 10.033 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.681 9.920 mp_alltoall_i22 716 14.1 5.506 9.203 5.506 9.203 cp_fm_cholesky_decompose 22 10.9 8.250 8.333 8.250 8.333 pw_transfer 1547 11.6 0.085 0.099 8.202 8.292 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.978 8.074 grid_integrate_task_list 128 12.3 7.519 7.933 7.519 7.933 wfi_extrapolate 11 7.9 0.001 0.001 7.717 7.718 multiply_cannon_metrocomm4 25070 15.6 0.075 0.086 2.822 7.166 density_rs2pw 128 9.7 0.006 0.006 6.723 7.097 fft_wrap_pw1pw2_140 523 13.2 0.479 0.490 6.944 7.042 mp_irecv_dv 76098 16.2 2.678 6.901 2.678 6.901 calculate_dm_sparse 128 9.5 0.001 0.001 6.355 6.467 fft3d_ps 1291 14.7 2.800 2.917 6.209 6.308 mp_alltoall_d11v 2415 14.1 5.175 5.700 5.175 5.700 grid_collocate_task_list 128 9.7 5.206 5.616 5.206 5.616 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.454 5.488 rs_pw_transfer 1046 11.9 0.013 0.014 5.108 5.452 potential_pw2rs 128 12.3 0.022 0.023 4.588 4.611 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.425 4.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.361 4.444 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.414 4.415 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.234 4.276 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.895000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1071.000000, yerr=16.503443 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.509511E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57953. MP_Allreduce 11047 1173. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.057 183.800 183.801 qs_mol_dyn_low 1 2.0 0.003 0.003 183.263 183.275 qs_forces 11 3.9 0.002 0.003 182.956 182.960 qs_energies 11 4.9 0.001 0.001 175.492 175.498 scf_env_do_scf 11 5.9 0.001 0.001 157.824 157.840 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.679 119.681 velocity_verlet 10 3.0 0.205 0.207 119.227 119.231 qs_scf_new_mos 117 7.6 0.001 0.001 82.749 82.858 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.748 82.858 ot_scf_mini 117 9.6 0.003 0.004 78.123 78.196 dbcsr_multiply_generic 2507 12.6 0.183 0.189 77.369 78.067 multiply_cannon 2507 13.6 0.584 0.623 56.058 60.321 multiply_cannon_loop 2507 14.6 0.446 0.456 51.187 52.702 ot_mini 117 10.6 0.001 0.001 41.677 41.775 init_scf_loop 11 6.9 0.000 0.000 37.980 37.982 mp_waitall_1 125778 16.7 28.082 36.119 28.082 36.119 prepare_preconditioner 11 7.9 0.000 0.000 34.056 34.087 make_preconditioner 11 8.9 0.000 0.000 34.056 34.087 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.959 32.244 rebuild_ks_matrix 128 8.3 0.001 0.001 29.266 29.409 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.265 29.409 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.693 26.831 qs_ot_get_p 128 10.4 0.001 0.001 23.031 23.132 qs_ot_get_derivative 117 11.6 0.001 0.002 21.343 21.424 multiply_cannon_multrec 10028 15.6 10.414 14.325 17.979 21.403 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.908 20.531 ot_diis_step 117 11.6 0.020 0.020 20.247 20.248 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.701 20.135 apply_single 128 13.6 0.001 0.001 19.700 20.134 cp_fm_cholesky_invert 11 10.9 19.678 19.685 19.678 19.685 make_m2s 5014 13.6 0.068 0.072 17.241 19.571 make_images 5014 14.6 2.304 2.838 16.934 19.262 qs_ot_p2m_diag 83 11.4 0.496 0.501 18.706 18.719 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.452 17.454 sum_up_and_integrate 128 10.3 0.181 0.192 14.608 14.658 integrate_v_rspace 128 11.3 0.003 0.004 14.427 14.484 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.065 14.074 cp_fm_diag_elpa_base 83 14.4 13.794 13.891 14.059 14.071 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.873 13.915 calculate_rho_elec 128 8.7 0.258 0.269 13.873 13.915 make_images_data 5014 15.6 0.053 0.060 10.556 13.280 hybrid_alltoall_any 5200 16.5 0.838 3.779 10.146 13.030 multiply_cannon_sync_h2d 10028 15.6 11.528 12.029 11.528 12.029 init_scf_run 11 5.9 0.000 0.001 10.811 10.812 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.811 10.812 cp_fm_cholesky_decompose 22 10.9 8.752 8.858 8.752 8.858 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.678 8.731 pw_transfer 1547 11.6 0.084 0.090 8.287 8.302 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 8.065 8.085 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 5.119 8.063 grid_integrate_task_list 128 12.3 7.792 8.025 7.792 8.025 dbcsr_mm_accdrv_process 20762 16.1 2.657 3.429 7.200 7.792 mp_allgather_i34 2507 14.6 3.084 7.624 3.084 7.624 wfi_extrapolate 11 7.9 0.001 0.001 7.548 7.548 density_rs2pw 128 9.7 0.005 0.006 6.550 6.965 fft_wrap_pw1pw2_140 523 13.2 0.500 0.523 6.909 6.933 calculate_dm_sparse 128 9.5 0.001 0.001 6.499 6.565 fft3d_ps 1291 14.7 2.721 2.789 6.233 6.261 grid_collocate_task_list 128 9.7 5.495 5.890 5.495 5.890 dbcsr_complete_redistribute 395 12.7 2.107 2.162 5.414 5.808 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.429 5.435 mp_alltoall_d11v 2415 14.1 5.051 5.350 5.051 5.350 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 5.261 5.262 rs_pw_transfer 1046 11.9 0.012 0.013 4.728 5.177 potential_pw2rs 128 12.3 0.026 0.027 4.585 4.603 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.866 4.395 mp_irecv_dv 28860 15.9 1.829 4.317 1.829 4.317 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.259 4.287 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.948 3.999 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.563 3.872 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.786 3.858 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.754 3.781 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.450 3.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=183.801000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1412.909091, yerr=50.193181 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.993213E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59299. MP_Allreduce 10975 1515. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.133 300.475 300.475 qs_mol_dyn_low 1 2.0 0.003 0.003 299.842 299.853 qs_forces 11 3.9 0.003 0.003 299.556 299.559 qs_energies 11 4.9 0.001 0.001 290.823 290.829 scf_env_do_scf 11 5.9 0.001 0.001 268.520 268.530 velocity_verlet 10 3.0 0.002 0.002 215.373 215.403 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 141.488 141.490 init_scf_loop 11 6.9 0.000 0.000 126.765 126.768 prepare_preconditioner 11 7.9 0.000 0.000 121.884 121.910 make_preconditioner 11 8.9 0.000 0.000 121.884 121.910 make_full_inverse_cholesky 11 9.9 0.000 0.000 98.348 118.968 qs_scf_new_mos 116 7.6 0.001 0.001 97.900 98.057 qs_scf_loop_do_ot 116 8.6 0.001 0.001 97.899 98.057 ot_scf_mini 116 9.6 0.004 0.004 92.999 93.115 dbcsr_multiply_generic 2485 12.5 0.211 0.218 86.968 87.854 cp_fm_upper_to_full 104 14.8 52.008 74.442 52.008 74.442 multiply_cannon 2485 13.5 0.707 0.746 60.943 61.891 multiply_cannon_loop 2485 14.5 0.465 0.474 57.046 59.088 ot_mini 116 10.6 0.001 0.001 47.094 47.240 dbcsr_complete_redistribute 393 12.7 3.976 4.017 29.480 41.932 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.956 38.343 cp_fm_cholesky_invert 11 10.9 36.130 36.136 36.130 36.136 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.490 35.759 mp_waitall_1 102768 16.8 31.457 35.672 31.457 35.672 rebuild_ks_matrix 127 8.3 0.001 0.001 34.688 34.844 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.688 34.844 mp_alltoall_i22 712 14.1 21.291 33.732 21.291 33.732 qs_ks_update_qs_env 127 7.6 0.001 0.001 32.292 32.442 qs_ot_get_p 127 10.4 0.001 0.001 29.869 30.004 qs_ot_get_derivative 116 11.6 0.002 0.002 25.870 25.987 qs_ot_p2m_diag 82 11.4 0.868 0.873 25.297 25.331 cp_dbcsr_syevd 82 12.4 0.006 0.006 23.518 23.520 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 21.093 23.030 make_m2s 4970 13.5 0.078 0.080 20.873 21.959 make_images 4970 14.5 3.724 3.843 20.394 21.479 ot_diis_step 116 11.6 0.022 0.022 21.174 21.175 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.356 20.743 apply_single 127 13.6 0.001 0.001 20.356 20.743 cp_fm_diag_elpa 82 13.4 0.000 0.000 20.110 20.112 cp_fm_diag_elpa_base 82 14.4 15.790 17.360 20.105 20.105 multiply_cannon_multrec 9940 15.5 10.487 12.209 17.837 17.905 sum_up_and_integrate 127 10.3 0.318 0.319 16.260 16.365 integrate_v_rspace 127 11.3 0.004 0.004 15.941 16.047 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.776 15.793 calculate_rho_elec 127 8.7 0.479 0.480 15.776 15.792 multiply_cannon_sync_h2d 9940 15.5 15.541 15.549 15.541 15.549 make_images_data 4970 15.5 0.060 0.063 11.462 13.410 hybrid_alltoall_any 5155 16.4 1.288 3.002 11.453 13.394 init_scf_run 11 5.9 0.000 0.001 12.155 12.155 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.154 12.155 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.154 10.248 cp_fm_cholesky_decompose 22 10.9 9.867 9.892 9.867 9.892 pw_transfer 1535 11.6 0.090 0.091 9.222 9.240 wfi_extrapolate 11 7.9 0.001 0.001 9.078 9.079 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.989 9.008 dbcsr_mm_accdrv_process 20590 16.0 3.828 5.713 7.113 8.939 grid_integrate_task_list 127 12.3 8.481 8.665 8.481 8.665 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.085 8.086 mp_alltoall_d11v 2401 14.1 7.745 7.904 7.745 7.904 fft_wrap_pw1pw2_140 519 13.2 0.535 0.539 7.851 7.870 fft3d_ps 1281 14.7 2.737 2.748 7.048 7.065 calculate_dm_sparse 127 9.5 0.001 0.001 6.748 6.802 density_rs2pw 127 9.7 0.005 0.005 6.690 6.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.556 6.611 grid_collocate_task_list 127 9.7 6.289 6.328 6.289 6.328 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.241 6.302 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.475000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2680.909091, yerr=172.247947 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 1.261511E+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 15561534. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 85.228 85.229 qs_energies 1 2.0 0.000 0.000 84.823 84.830 ls_scf 1 3.0 0.000 0.000 83.870 83.877 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.857 73.026 multiply_cannon 111 7.7 0.018 0.021 55.993 57.348 multiply_cannon_loop 111 8.7 0.210 0.225 52.531 53.973 ls_scf_main 1 4.0 0.000 0.000 52.434 52.435 density_matrix_trs4 2 5.0 0.002 0.003 46.924 46.995 ls_scf_init_scf 1 4.0 0.000 0.000 28.360 28.361 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.314 27.358 mp_waitall_1 11031 10.9 22.563 25.402 22.563 25.402 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.202 25.223 multiply_cannon_multrec 2664 9.7 8.171 8.805 15.481 17.104 multiply_cannon_sync_h2d 2664 9.7 13.721 15.979 13.721 15.979 make_m2s 222 7.7 0.008 0.011 13.162 13.708 make_images 222 8.7 0.098 0.112 13.140 13.689 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.730 12.292 make_images_data 222 9.7 0.004 0.005 7.739 8.431 hybrid_alltoall_any 227 10.6 0.216 1.838 6.508 7.887 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.639 6.929 7.879 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.417 7.700 dbcsr_mm_accdrv_process_sort 4760 11.4 6.218 7.112 6.218 7.112 calculate_norms 4752 9.8 5.528 6.189 5.528 6.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.096 5.248 mp_sum_l 807 5.4 3.159 4.576 3.159 4.576 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.060 3.699 mp_irecv_dv 6231 10.9 2.042 3.668 2.042 3.668 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.338 3.497 arnoldi_extremal 4 6.8 0.000 0.000 3.419 3.452 arnoldi_normal_ev 4 7.8 0.001 0.002 3.419 3.451 make_images_sizes 222 9.7 0.000 0.000 0.856 3.376 mp_alltoall_i44 222 10.7 0.856 3.376 0.856 3.376 build_subspace 16 8.4 0.009 0.012 3.317 3.319 ls_scf_post 1 4.0 0.000 0.000 3.076 3.083 ls_scf_store_result 1 5.0 0.000 0.000 2.893 2.935 dbcsr_special_finalize 555 9.7 0.005 0.005 2.218 2.758 dbcsr_merge_single_wm 555 10.7 0.458 0.602 2.210 2.750 make_images_pack 222 9.7 2.204 2.629 2.206 2.631 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.363 2.610 dbcsr_sort_data 658 11.4 2.012 2.486 2.012 2.486 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.457 2.068 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.235 2.348 buffer_matrices_ensure_size 222 8.7 1.768 2.097 1.768 2.097 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.731 1.732 rebuild_ks_matrix 3 7.3 0.000 0.000 1.722 1.723 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.722 1.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.229000, 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/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 2.109903E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.041 92.084 92.085 qs_energies 1 2.0 0.000 0.000 91.619 91.623 ls_scf 1 3.0 0.000 0.000 90.317 90.320 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.993 76.422 multiply_cannon 111 7.7 0.028 0.041 53.773 57.966 ls_scf_main 1 4.0 0.000 0.000 55.574 55.581 multiply_cannon_loop 111 8.7 0.116 0.123 50.512 53.908 density_matrix_trs4 2 5.0 0.002 0.003 49.857 50.051 ls_scf_init_scf 1 4.0 0.000 0.000 31.101 31.102 mp_waitall_1 9105 10.9 21.827 30.161 21.827 30.161 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.914 30.030 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.527 27.540 multiply_cannon_multrec 1332 9.7 12.988 16.566 22.178 26.897 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.411 22.028 make_m2s 222 7.7 0.006 0.007 15.302 15.899 make_images 222 8.7 1.567 1.911 15.272 15.868 dbcsr_mm_accdrv_process 4041 10.4 0.256 0.450 8.799 10.319 dbcsr_mm_accdrv_process_sort 4041 11.4 8.391 9.893 8.391 9.893 make_images_data 222 9.7 0.004 0.004 8.750 9.595 hybrid_alltoall_any 227 10.6 0.520 2.444 8.056 9.060 mp_sum_l 807 5.4 5.627 9.024 5.627 9.024 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.217 7.701 mp_irecv_dv 3311 11.0 3.197 7.629 3.197 7.629 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.366 7.189 calculate_norms 2376 9.8 5.985 6.702 5.985 6.702 multiply_cannon_sync_h2d 1332 9.7 4.844 5.847 4.844 5.847 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.094 5.278 arnoldi_extremal 4 6.8 0.000 0.000 4.840 4.862 arnoldi_normal_ev 4 7.8 0.001 0.005 4.840 4.862 build_subspace 16 8.4 0.014 0.021 4.561 4.565 ls_scf_post 1 4.0 0.000 0.000 3.641 3.646 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.227 3.439 ls_scf_store_result 1 5.0 0.000 0.000 3.246 3.356 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.228 2.739 3.230 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.295 2.878 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.544 2.657 make_images_pack 222 9.7 2.018 2.414 2.020 2.416 mp_allgather_i34 111 8.7 0.959 2.349 0.959 2.349 dbcsr_sort_data 436 11.2 1.819 2.057 1.819 2.057 dbcsr_data_new 4174 10.1 1.615 1.842 1.615 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.085000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1791.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 2.700591E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.135 94.587 94.589 qs_energies 1 2.0 0.000 0.000 93.940 93.943 ls_scf 1 3.0 0.000 0.001 92.427 92.430 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.124 77.432 ls_scf_main 1 4.0 0.000 0.000 58.110 58.115 multiply_cannon 111 7.7 0.043 0.119 52.998 56.769 multiply_cannon_loop 111 8.7 0.100 0.105 49.452 53.739 density_matrix_trs4 2 5.0 0.002 0.003 52.216 52.322 mp_waitall_1 7281 11.0 24.227 34.588 24.227 34.588 ls_scf_init_scf 1 4.0 0.000 0.002 30.775 30.777 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.570 29.666 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.188 27.200 multiply_cannon_multrec 888 9.7 12.564 15.337 21.082 24.450 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.178 23.638 make_m2s 222 7.7 0.006 0.007 17.174 18.402 make_images 222 8.7 1.971 2.282 17.136 18.363 hybrid_alltoall_any 227 10.6 0.621 2.867 9.525 10.918 make_images_data 222 9.7 0.003 0.004 9.904 10.904 dbcsr_mm_accdrv_process 3754 10.4 0.239 0.412 8.047 9.242 mp_sum_l 807 5.4 5.438 9.156 5.438 9.156 dbcsr_mm_accdrv_process_sort 3754 11.4 7.679 8.830 7.679 8.830 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.265 7.346 multiply_cannon_sync_h2d 888 9.7 6.088 7.268 6.088 7.268 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.765 7.111 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.448 6.781 mp_irecv_dv 2335 11.1 2.432 6.724 2.432 6.724 arnoldi_extremal 4 6.8 0.000 0.000 5.178 5.194 arnoldi_normal_ev 4 7.8 0.001 0.005 5.178 5.194 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.916 5.106 build_subspace 16 8.4 0.014 0.020 4.868 4.875 calculate_norms 1584 9.8 4.269 4.641 4.269 4.641 mp_allgather_i34 111 8.7 1.390 3.862 1.390 3.862 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.468 3.792 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.577 3.015 3.579 ls_scf_post 1 4.0 0.000 0.001 3.541 3.545 ls_scf_store_result 1 5.0 0.000 0.000 3.279 3.364 ls_scf_dm_to_ks 2 5.0 0.000 0.004 2.764 2.859 dbcsr_sort_data 325 11.1 1.901 2.191 1.901 2.191 make_images_pack 222 9.7 1.813 2.136 1.816 2.138 make_images_sizes 222 9.7 0.000 0.000 0.941 1.975 mp_alltoall_i44 222 10.7 0.941 1.975 0.941 1.975 dbcsr_data_release 9322 10.9 1.304 1.953 1.304 1.953 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.916 1.918 rebuild_ks_matrix 3 7.3 0.000 0.000 1.898 1.900 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.898 1.900 dbcsr_finalize 304 7.8 0.026 0.032 1.616 1.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.589000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2180.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 3.360379E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.075 97.316 97.317 qs_energies 1 2.0 0.000 0.000 96.619 96.639 ls_scf 1 3.0 0.001 0.011 94.981 95.002 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.659 78.967 ls_scf_main 1 4.0 0.000 0.000 58.944 58.946 multiply_cannon 111 7.7 0.056 0.146 51.592 55.665 density_matrix_trs4 2 5.0 0.002 0.003 52.915 53.034 multiply_cannon_loop 111 8.7 0.115 0.126 46.601 49.219 ls_scf_init_scf 1 4.0 0.000 0.001 32.785 32.787 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.483 31.543 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.028 29.044 mp_waitall_1 6369 11.0 22.980 28.894 22.980 28.894 multiply_cannon_multrec 1332 9.7 14.210 17.339 22.050 24.590 make_m2s 222 7.7 0.007 0.008 21.392 22.746 make_images 222 8.7 3.122 3.587 21.341 22.699 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.243 17.362 make_images_data 222 9.7 0.004 0.004 12.000 13.691 hybrid_alltoall_any 227 10.6 0.796 3.830 11.405 13.227 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.403 7.481 8.995 dbcsr_mm_accdrv_process_sort 3641 11.4 7.108 8.576 7.108 8.576 mp_sum_l 807 5.4 4.100 8.107 4.100 8.107 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.146 6.464 multiply_cannon_sync_h2d 1332 9.7 5.539 6.345 5.539 6.345 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.079 6.150 mp_irecv_dv 3229 10.9 2.057 6.073 2.057 6.073 arnoldi_extremal 4 6.8 0.000 0.000 5.279 5.294 arnoldi_normal_ev 4 7.8 0.001 0.005 5.279 5.294 build_subspace 16 8.4 0.014 0.021 4.943 4.951 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.533 4.715 calculate_norms 2376 9.8 4.193 4.511 4.193 4.511 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.562 4.452 mp_allgather_i34 111 8.7 2.115 4.289 2.115 4.289 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.591 3.895 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.674 3.186 3.676 dbcsr_sort_data 658 11.4 3.083 3.413 3.083 3.413 ls_scf_post 1 4.0 0.003 0.023 3.251 3.272 dbcsr_special_finalize 555 9.7 0.006 0.007 2.825 3.142 dbcsr_merge_single_wm 555 10.7 0.536 0.661 2.816 3.134 ls_scf_store_result 1 5.0 0.000 0.000 3.002 3.050 ls_scf_dm_to_ks 2 5.0 0.001 0.014 2.914 2.945 dbcsr_data_release 10477 10.7 1.581 2.496 1.581 2.496 dbcsr_finalize 304 7.8 0.049 0.061 1.802 2.001 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.985 1.986 rebuild_ks_matrix 3 7.3 0.000 0.000 1.962 1.963 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.020 1.962 1.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.317000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2693.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 4.629434E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.059 99.011 99.017 qs_energies 1 2.0 0.000 0.000 98.244 98.250 ls_scf 1 3.0 0.000 0.000 96.307 96.311 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.881 78.069 ls_scf_main 1 4.0 0.000 0.000 62.031 62.032 multiply_cannon 111 7.7 0.080 0.124 55.174 60.684 density_matrix_trs4 2 5.0 0.002 0.003 55.031 55.134 multiply_cannon_loop 111 8.7 0.069 0.077 50.637 52.401 mp_waitall_1 5436 11.0 26.611 33.020 26.611 33.020 ls_scf_init_scf 1 4.0 0.000 0.000 30.650 30.655 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.481 29.522 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.314 27.328 multiply_cannon_multrec 444 9.7 13.973 16.403 20.993 23.697 make_m2s 222 7.7 0.005 0.005 17.866 20.390 make_images 222 8.7 3.715 4.405 17.804 20.331 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.086 16.203 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.230 15.122 make_images_data 222 9.7 0.003 0.004 10.095 12.649 hybrid_alltoall_any 227 10.6 0.790 3.761 9.886 12.439 multiply_cannon_sync_h2d 444 9.7 6.558 8.055 6.558 8.055 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.345 6.722 7.835 dbcsr_mm_accdrv_process_sort 3003 11.4 6.407 7.491 6.407 7.491 mp_allgather_i34 111 8.7 2.773 6.954 2.773 6.954 arnoldi_extremal 4 6.8 0.000 0.000 6.022 6.039 arnoldi_normal_ev 4 7.8 0.001 0.004 6.022 6.039 build_subspace 16 8.4 0.015 0.020 5.631 5.640 mp_sum_l 807 5.4 2.879 4.809 2.879 4.809 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.583 4.759 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.265 4.454 dbcsr_matrix_vector_mult_local 304 10.0 3.692 4.173 3.694 4.175 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.582 3.998 mp_irecv_dv 1241 11.2 1.560 3.973 1.560 3.973 make_images_sizes 222 9.7 0.000 0.000 1.132 3.714 mp_alltoall_i44 222 10.7 1.132 3.714 1.132 3.714 calculate_norms 792 9.8 3.555 3.676 3.555 3.676 ls_scf_post 1 4.0 0.000 0.000 3.625 3.630 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.970 3.533 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.396 3.482 ls_scf_store_result 1 5.0 0.000 0.000 3.407 3.453 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.250 dbcsr_merge_all 275 8.9 0.473 0.519 2.049 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.983 1.984 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.017000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3625.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/399e6db9ee1a21058cc4319e934d6c8fef1bc356_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 8.727192E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.099 108.212 108.213 qs_energies 1 2.0 0.000 0.000 106.726 106.740 ls_scf 1 3.0 0.000 0.000 103.875 103.889 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.558 77.679 ls_scf_main 1 4.0 0.000 0.000 66.019 66.020 density_matrix_trs4 2 5.0 0.002 0.003 57.056 57.115 multiply_cannon 111 7.7 0.131 0.203 49.948 51.670 multiply_cannon_loop 111 8.7 0.067 0.069 46.679 47.602 ls_scf_init_scf 1 4.0 0.000 0.000 34.189 34.190 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.830 32.841 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.987 29.997 mp_waitall_1 4527 11.1 22.583 25.727 22.583 25.727 make_m2s 222 7.7 0.005 0.005 23.954 24.742 make_images 222 8.7 4.577 5.005 23.848 24.635 multiply_cannon_multrec 444 9.7 17.808 18.587 22.412 23.058 hybrid_alltoall_any 227 10.6 1.659 3.619 13.005 15.557 make_images_data 222 9.7 0.003 0.003 13.184 15.389 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.813 11.388 multiply_cannon_sync_h2d 444 9.7 8.854 8.891 8.854 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.698 7.709 arnoldi_normal_ev 4 7.8 0.002 0.008 7.698 7.708 build_subspace 16 8.4 0.026 0.036 7.154 7.169 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.635 5.786 dbcsr_matrix_vector_mult_local 304 10.0 5.040 5.357 5.043 5.360 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.060 5.312 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.870 4.969 dbcsr_mm_accdrv_process 1814 10.4 0.212 0.359 4.437 4.567 dbcsr_mm_accdrv_process_sort 1814 11.4 4.118 4.253 4.118 4.253 ls_scf_post 1 4.0 0.000 0.000 3.667 3.680 make_images_sizes 222 9.7 0.000 0.000 1.447 3.573 mp_alltoall_i44 222 10.7 1.446 3.573 1.446 3.573 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.430 calculate_norms 792 9.8 3.239 3.273 3.239 3.273 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.145 dbcsr_complete_redistribute 5 7.6 1.432 1.469 2.791 2.936 dbcsr_merge_all 275 8.9 0.889 0.915 2.866 2.923 dbcsr_data_release 12724 10.6 2.331 2.876 2.331 2.876 mp_allgather_i34 111 8.7 0.903 2.849 0.903 2.849 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.820 2.820 matrix_ls_to_qs 2 6.0 0.000 0.000 2.431 2.588 dbcsr_sort_data 325 11.1 2.443 2.522 2.443 2.522 dbcsr_new_transposed 4 7.5 0.246 0.259 2.341 2.361 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.249 2.251 dbcsr_frobenius_norm 74 6.6 2.056 2.132 2.203 2.238 mp_alltoall_d11v 48 9.2 2.149 2.211 2.149 2.211 dbcsr_add_d 103 6.2 0.000 0.000 2.135 2.206 dbcsr_add_anytype 103 7.2 0.859 0.892 2.134 2.205 rebuild_ks_matrix 3 7.3 0.000 0.000 2.179 2.181 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.179 2.181 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.213000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6865.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 399e6db9ee1a21058cc4319e934d6c8fef1bc356 Summary: empty Status: OK